WORLD TRADE WT/DS26/R/USA
18 August 1997
ORGANIZATION
(97-3368)
Original: English
The report of the Panel on EC Measures Concerning Meat and Meat Products (Hormones) - Complaint by the United States - is being circulated to all Members, pursuant to the DSU. The report is being circulated as an unrestricted document from 18 August 1997 pursuant to the Procedures for the Circulation and Derestriction of WTO Documents (WT/L/160/Rev.1). Members are reminded that in accordance with the DSU only parties to the dispute may appeal a panel report, an appeal shall be limited to issues of law covered in the panel report and legal interpretations developed by the panel, and that there shall be no ex parte communications with the panel or Appellate Body concerning matters under consideration by the panel or Appellate Body.
Note by the Secretariat: This Panel Report shall be adopted by the Dispute Settlement Body (DSB) within 60 days after the date of its circulation unless a party to the dispute decides to appeal or the DSB decides by consensus not to adopt the report. If the Panel Report is appealed to the Appellate Body, it shall not be considered for adoption by the DSB until after the completion of the appeal. Information on the current status of the Panel Report is available from the WTO Secretariat.
- i -
INTRODUCTION 1
FACTUAL ASPECTS 2
CLAIMS OF THE PARTIES 12
ARGUMENTS OF THE PARTIES 14
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
THIRD PARTIES SUBMISSION 104
PANEL'S CONSULTATION WITH SCIENTIFIC EXPERTS 112
INTERIM REVIEW
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
- ii-
FINDINGS 162
CLAIMS OF THE PARTIES 162
ORGANIZATIONAL ISSUES 163
Scientific evidence 163
Parallel panel requested by Canada 165
GENERAL INTERPRETATIVE ISSUES 166
Scope of the measures in dispute 166
Application of the SPSAgreement, the TBT Agreement and
GATT 167
Relationship between the SPS Agreement and GATT 169
THE SPS AGREEMENT 172
Overview of the provisions in dispute 172
Burden of proof 172
Article 3.1: sanitary measures based on international standards 174
Codex standards 175
Sanitary measures based on Codex standards 178
The meaning of based on 179
Comparison of levels of sanitary protection 180
Article 3.3: sanitary measures not based on international
standards 181
Requirements for justification . . . . . . . . . _ 181
Burden of proof 182
Article 5: "Assessment of Risk and Determination of the Appropriate Level of Sanitary or Phytosanitary
Protection" 184
Risk assessment and risk management 184
Articles 5.1to 5.3: risk assessment 185
Techniques and factors to be taken into account . . 186
The existence of a risk assessment 187
Sanitary measures to be based on a risk assessment 189
Articles 5.4 to 5.6: risk management 202
Article 5.4: minimizing trade effects 203
Article 5.5: distinctions in levels of protection . . . 204
Article 5.6: measures not more trade restrictive than required to achieve the appropriate level of
protection 221
Article 5.7: provisional sanitary measures 221
Sanitary measures where no international standards exist: melengestrol acetate ("MGA") 222
Burden of proof 222
Articles 5.1to 5.3: risk assessment 223
Article 5.5: distinctions in levels of protection 224
MGA for growth promotion compared to the natural hormones occurring endogenously in meat
and other foods 225
MGA for growth promotion compared to
carbadox 225
Article 2: "Basic Rights and Obligations" 226
- iii -
ARTICLES I AND III OF GATT 226
CONCLUDING REMARKS 227
CONCLUSIONS 227
ANNEX 229
WT/DS26/R/USA
Page 1
INTRODUCTION
On 26 January 1996, the United States requested consultations with the European Communities, pursuant to Article 4 of the Understanding on Rules and Procedures Governing the Settlement of Disputes ("DSU"), Article 11 of the Agreement on the Application of Sanitary and Phytosanitary Measures ("SPS Agreement"), Article 14 of the Agreement on Technical Barriers to Trade ("TBT Agreement"), Article 19 of the Agreement on Agriculture and Article XXII of the General Agreement on Tariffs and Trade 1994 ("GATT"), regarding the Council Directive Prohibiting the Use in Livestock Farming of Certain Substances Having a Hormonal Action and related measures (WT/DS26/1).
On 2 February 1997, pursuant to Article 4.11 of the DSU, Australia (WT/DS26/3) and New Zealand (WT/DS26/2), followed on 8 February by Canada (WT/DS26/4), requested to be joined in these consultations. The European Communities accepted these requests on 19 March 1996 (WT/DS26/5).
On 27 March 1996, the United States, Australia, Canada and New Zealand held joint consultations with the European Communities but failed to reach a mutually satisfactory solution.
On 25 April 1996, pursuant to Article 11 of the SPS Agreement, Article 14 of the TBT Agreement, Article 19 of the Agreement on Agriculture, Article XXIII:2 of the GATT, and Article 6 of the DSU, the United States requested the Dispute Settlement Body ("DSB") to establish a panel with standard terms of reference (WT/DS/26/6). The United States claimed that the EC measures:
"... adversely affect imports of meat and meat products and appear to be inconsistent with the obligations of the European Communities under the General Agreement on Tariffs and Trade 1994, the Agreement on the Application of Sanitary and Phytosanitary Measures, the Agreement on Technical Barriers to Trade, and the Agreement on Agriculture. The provisions of these agreements with which these measures appear to be inconsistent include, but are not limited to, the following:
General Agreement on Tariffs and Trade 1994, Article III or Article XI;
Agreement on the Application of Sanitary and Phytosanitary Measures, Articles 2, 3 and 5;
Agreement on Technical Barriers to Trade, Article 2; and
Agreement on Agriculture, Article 4.
These measures also appear to nullify or impair the benefits accruing to the United Statesdirectly or indirectly under the cited agreements".
On 20 May 1996, the DSB established a Panel in accordance with the request made by the United States. The agreed standard terms of reference of the Panel were (WT/DS26/7):
"To examine, in the light of the relevant provisions of the covered agreements cited by the United States in document WT/DS26/6, the matter referred to the DSB by the United States in that document and to make such findings as will assist the DSB in making the recommendations or in giving the rulings provided for in those agreements".
Australia, Canada, New Zealand and Norway reserved their rights to participate in the Panel proceedings as third parties.
WT/DS26/R/USA
Page 2
On 2 July 1996, the Panel was constituted with the following composition: Chairman: Mr. Thomas Cottier
Panellists: Mr. Jun Yokota
Mr. Peter Palecka
The Panel met with the parties on 10 October 1996 and 11 November 1996. It met with third parties on 10 October 1996. The Panel consulted scientific and technical experts on 17-18 February 1997 in a meeting held jointly with the panel proceeding brought by Canada on the same EC measures.1
On 28 November 1996, the Chairman of the Panel informed the DSB that the Panel would not be able to issue its report within six months. The reasons for that delay are stated in document WT/DS/26/8.
The Panel issued its interim report to the parties on 7 May 1997. Following a request made by the European Communities pursuant to Article 15.2 of the DSU, the Panel held a further meeting with the parties on 4 June 1997. The Panel issued its final report to the parties to the dispute on 30 June 1997.
FACTUAL ASPECTS
This dispute concerns EC measures, in particular Council Directive 81/602/EEC ("Directive 81/602/EEC"), Council Directive 88/146/EEC ("Directive 88/146/EEC") and Council Directive 88/299/EEC ("Directive 88/299/EEC").2
Directive 81/602/EEC prohibits the administering to farm animals of substances having a thyrostatic action or substances having an oestrogenic, androgenic or gestagenic action; the placing on the market or slaughtering of farm animals to which these substances have been administered; the placing on the market of meat from such animals; the processing of meat from such animals and the placing on the market of meat products prepared from or with such meat. The Directive provides two exceptions to the prohibition: one exception is provided for substances with an oestrogenic, androgenic or gestagenic action when they are used for therapeutic or zootechnical purposes and administered by a veterinarian or under a veterinarian's responsibility. The other exception was for oestradiol-17, progesterone, testosterone, trenbolone acetate (or TBA) and zeranol - when they were used for growth promotion purposes and their use was governed according to the individual regulatory schemes maintained by EC member States. This exception was made pending an examination of the effects of these hormones on the health of consumers and the adoption of an EC rule. EC member States are obliged to apply their regulatory schemes to imports from third countries in a manner not more favourable than that applied to intra-EC trade.
Directive 88/146/EEC extends the prohibition imposed by Directive 81/602/EEC to the administration to farm animals of trenbolone acetate and zeranol for any purpose, and , testosterone and progesterone for fattening purposes. However, the Directive maintains the permission
1WT/DS48/6.
2Other measures relevant to the dispute are contained in Directives 72/462/EEC, 81/602/EEC, 81/851/EEC, 81/852/EEC, 85/358/EEC, referenced in Directive 88/146/EEC; the decisions, control programme and derogations referred to in Article 6(2), Article 6(7) and Article 7, respectively, of Directive 88/146/EEC; and any amendments or modifications, including Directives 96/22/EC and 96/23/EC.
WT/DS26/R/USA
Page 3
to administer these three natural hormones to animals for therapeutic and zootechnical purposes under prescribed conditions; in particular, therapeutic treatment is defined to mean the administering to an individual animal of any of the substances which are authorized to treat a fertility problem diagnosed on examination by a veterinarian. The products which are used for therapeutic treatment may be administered only by a veterinarian, in the form of an injection (to the exclusion of implantation) to farm animals which have been clearly identified. Such treatment must be registered by the veterinarian and these animals may not be slaughtered before expiry of the period fixed. In the case of animals at the end of their reproductive career, the treatments are prohibited from being administered during the fattening period following the end of their breeding life. Article 4 of directive 88/146/EEC explicitly requires that undertakings in the EC member Statesproducing theprohibited hormones, those companies authorized to market these hormones for whatever purposes and undertakings producing pharmaceutical and veterinary products based on those substances, must keep a detailed register recording (in chronological order) the quantities produced or acquired and those sold or used for the production of pharmaceutical and veterinary products. The importation from third countries of animals and meat from animals to which have been administered substances with thyrostatic, oestrogenic, androgenic or gestagenic action is prohibited.3 However, under certain conditions, Article 7 of Directive 88/146/EEC allows trade in those animals and meat from those animals treated for therapeutic or zootechnical purposes, including imports from third countries.4
Directive 88/299/EEC lays down the conditions for applying the derogations, provided for in Article 7 of Directive 88/146/EEC, from the prohibition on trade in certain categories of animals and their meat. The first derogation of the Directive requires EC member States to authorize trade in animals intended for reproduction and reproductive animals at the end of their career (and of meat of such animals) which, during their reproductive career, have undergone one of two categories of treatments: The first category is therapeutic treatment with one of the following substances: oestradiol-
17, testosterone and progesterone; and those derivatives which readily yield the parent compound on hydrolysis after absorption at the site of application which appear in a list of approved products. The second category is the administration of substanceshaving an oestrogenic, androgenic or gestagenic action for synchronization of oestrus, termination of unwanted gestation, the improvement of fertility and the preparation of donors and recipients for the implantation of embryos, provided that the products in which they are contained appear on a list of approved products and with the respect of strict conditions of use concerning, in particular, the respect of the withdrawal period, the monitoring of those conditions of use and of the means of identification of the animals. In addition, Articles 3 and 4 of this Directive provide that trade between the EC member States of the European Communities in animals intended for reproduction and reproductive animals and meat from such animals is allowed only if all the conditions laid down in the Directive are respected, in particular as regards the waiting period and the requirement that animals have not received any of the above treatments with any of the above substances during the fattening period following the end of their breeding life. The EC stamp may be affixed to the meat only if the waiting time ended before the animals are slaughtered. The second
3Article 6(7) of Directive 88/146/EEC requires the establishment of a control programme as regards imports from third countries to ensure that imports do not receive more favourable treatment than EC products. This control programme also provides for rules on the frequency of controls on imports from each third country and on guarantees offered by the inspection regulation of third countries. Such checks on imports are now carried out in accordance with Directives 91/496/EEC and 90/675/EEC.
4Article 7 of Directive 88/146/EEC allows derogations in respect to trade in animals intended for reproduction and reproductive animals at the end of their career (and in respect of meat from these various animals, taking into account the guarantees given), which in the course of their existence have been treated under the provisions of Article 4 of Directive 81/602/EEC. This article authorizes the administration to farm animals of substances with oestrogenic, androgenic or gestagenic action approved in accordance with the Directives on veterinary medical products (other than substances referred to in Article 3 of Directive 81/602/EEC) for therapeutic use, synchronization of oestrus, termination of unwanted gestation, the improvement of fertility and the preparation of donors and recipients for the implantation of embryos. The administering of these substances shall be effected by a veterinarian, however, EC member States may allow the synchronization of oestrus and the preparation of donors and recipients for the implantation of embryos to be done not by a veterinarian but under his direct responsibility.
WT/DS26/R/USA
Page 4
derogation in Directive 88/299/EEC allows imports from third countries of treated animals and meat of such animals under guarantees equivalent to those for domestic animals and meat.
Directive 96/22/EC will replace Directives 81/602/EEC, 88/146/EEC and 88/299/EEC as from 1 July 1997. It will maintain the prohibition on the use of these hormones for growth promotion purposes; extend the prohibition on the use of beta-agonists; restrict the use of the hormones at issue for therapeutic or zootechnical purposes, reinforcing in particular the role of the veterinarian; and reinforce the provisions on control and testing. Penalties and sanctions in case of violations are to be increased where checks detect the presence of prohibited substances or products or residues of substances administered illegally. Such substances or products will be confiscated and any treated animals or meat placed under official supervision until penalties have been applied.
Hormones (chemicals) produced by the bodies of humans and animals are called endogenous or natural hormones. (Phyto-hormones are produced by some plants.) Compounds chemically synthesizedto mimic the effect of natural hormones are called syntheticor xenobiotic hormones. Natural hormones are secreted into the blood stream by specialized cells and travel throughout the body. Hormones act by binding protein receptors present in hormone-responsive tissues. The receptor undergoes a conformational change, binds to specific DNA sequences and regulates specificgenes within a cell. Synthetic hormones may differ from endogenous (natural) hormones in their rate of metabolism and excretion.
Hormones function in four broad areas: reproduction; growth and development; maintenance of the internal environment; and production, utilization and storage of energy. One hormone can have multiple actions. For example, the male hormone testosterone controls many processes from the development of the fetus to libido in the adult. One function may be controlled by multiple hormones: the menstrual cycle involves oestradiol, progesterone, follicle-stimulating hormone and luteinizing hormone.
Of the six hormones involved in this dispute, three are naturally occurring hormones produced by humans and animals: , progesterone and testosterone (hereafter also referred to as natural hormones).
is a sex steroidal hormone with oestrogenic action (i.e., responsible for female characteristics); testosterone is a sex steroidal hormone with androgenic action (i.e., responsible for male characteristics); progesterone is a sex steroidal hormone with gestagenic action (i.e., responsible for maintaining pregnancy). These three hormones are produced throughout the lifetime of each individual and are required for normal physiological functioning and maturation. Hormone levels vary with the tissue, with the species of animal and with the sex and individual. Hormone levels vary most dramatically with puberty, pregnancy and castration.
The other three hormones involved in this dispute are artificially produced hormones: trenbolone, zeranol and melengestrol acetate (MGA) (hereafter also referred to as synthetic hormones). These hormones mimic the biological activity of the natural hormones. Trenbolone mimics the action of testosterone; zeranol mimics the action of ; and MGA mimics progesterone.
In the United States, the three natural hormones may be used for medical treatment (therapeutic). Oestradiol-17is also permitted for zootechnical purposes. In the United States the six hormones are also approved for growth promotion purposes. Three of the hormones used for growth promotion purposes, trenbolone, zeranol, and MGA, have no zootechnical or therapeutic uses. For growth promotion purposes, five of these hormones (except MGA) are formulated as pellets (with approved and fixed amounts of compound) designed to be implanted in the ear of the animal. The ear is discarded at slaughter. MGA is administered as a feed additive.
WT/DS26/R/USA
Page 5
The SPS Agreement makes reference, in a number of provisions, to "the relevant international standards, guidelines and recommendations". Annex A:3(a) of the SPS Agreement states that the international standards, guidelines and recommendations relevant for food safety are those established by the Codex Alimentarius Commission relating to food additives, veterinary drug and pesticide residues, contaminants, methods of analysis and sampling, and codes and guidelines of hygienic practice.
The Codex Alimentarius Commission (hereafter the "Codex Commission") is a joint FAO/WHO advisory body established to implement the Joint FAO/WHO Food Standards Programme. The purpose of this programme is to protect the health of consumers and to ensure fair practices in food trade through the elaboration of food standards. These standards, together with notifications received from governments with respect to their acceptance or otherwise of the standards, constitute the Codex Alimentarius. The Codex Alimentarius (hereafter "the Codex") is thus a collection of internationally adopted food standards presented in a uniform manner.
Membership of the Codex Commission is open to all member Nations and Associate members of FAO and/or WHO and is composed of government representatives of these members. Most of its members, including the United States and the EC member States, are WTO Members. The European Communities has an observer status in the Codex Commission. The Codex Commission has established a number of subsidiary bodies, including the Codex Committee on Residues of Veterinary Drugs in Food ("CCRVDF").
The technical and scientific analysis of veterinary drugs, food additives and some other substances in foods and beverages is not undertaken by the Codex Commission itself but independently by the Joint FAO/WHO Expert Committee on Food Additives ("JECFA"). The JECFA is composed of independent scientists who serve in their individual capacities as experts, not as representatives of their governments or organizations. The goal of the JECFA evaluation of veterinary drugs is:
"to establish safe levels of intake by setting Acceptable Daily Intakes (ADI) and to develop maximum residue limits when veterinary drugs are used in accordance with good veterinary practice".5
The elaboration of Codex standards involves an 8-step process:
Step 1: The Codex Commission decides to elaborate a standard and identifies which subsidiary body or other body should undertake the work, taking into account the "Criteria for the Establishment of Work Priorities and for the Establishment of Subsidiary Bodies". Decisions to elaborate standards may also be taken by subsidiary bodies of the Codex Commission subject to subsequent approval by the Codex Commission or its Executive Committee.
Step 2: The Codex Commission secretariat arranges for the preparation of a "proposed draft standard". In the case of veterinary drugs, JECFA is in charge of preparing recommendations for maximum residue levels.
Step 3: The secretariat distributes the "proposed draft standard" to the members of the Commission for comments.
5Codex Alimentarius, Vol.3, Residues of Veterinary Drugs in Foods, p.vi.
WT/DS26/R/USA
Page 6
Step 4: The comments received are sent by the secretariat to the CCRVDF which considers the comments and prepares, if appropriate, a proposed draft standard.
Step 5: The proposed draft standard is submitted through the secretariat to the Codex Commission or to the Executive Committee with a view to its adoption as a "draft standard".
Step 6: The "draft standard" is sent by the secretariat to all members and interested international organizations for comments on all aspects, including possible implications of the "draft standard" for their economic interests.
Step 7: The comments received are sent by the secretariat to the CCVDRF, which considers such comments and may amend the "draft standard".
Step 8: The "draft standard" is submitted through the secretariat to the Codex Commission together with any written proposals received from members and interested international organizations for amendments at Step 8 with a view to its adoption as a "Codex standard". Adoption of standards is normally done on the basis of a consensus decision, however, if requested, a vote may be taken. In this case, a decision by the majority of Codex members is required. An accelerated elaboration procedure may be used when there is an urgent need for a standard.
The Codex Commission keeps under review and may revise Codex standards, generally following procedures similar to those used for the elaboration of standards. Codex standards are published and sent to governments for acceptance and to international organizations to which competence in the matter has been transferred by their EC member States. Acceptance of the standards is voluntary and Codex members are not required to indicate formal acceptance of Codex standards, guidelines or recommendations. The implementation of Codex standards at the national level is the responsibility of members.
Codex standards for veterinary drugs are normally stated in terms of an Acceptable Daily Intake ("ADI") and a Maximum Residue Limit ("MRL"). An ADI is "an estimate by JECFA of the amount of a veterinary drug, expressed on a body weight basis, that can be ingested daily over a lifetime without appreciable health risk (standard man = 60 kg)".6 An ADI is derived from the experimental No Observable Effect Level ("NOEL") in the most appropriate animal species, by applying an appropriate safety factor. To account for sensitivity variabilities between humans and animals, and dietary variabilities among humans, a safety factor is typically applied. When data from long-term animal toxicity studies are available, a safety factor of 100 is generally applied. Larger safety factors, up to 1000, may be used in certain cases.
A Codex MRL is one of the tools for ensuring that intake does not exceed the ADI and that "Good Practice in the use of Veterinary Drugs" ("GPVD") is observed. It is the maximum concentration of residue resulting from the use of a veterinary drug (expressed in µg/kg or µg/kg on a fresh weight basis) that is recommended by the Codex Commission to be legallypermitted or recognized as acceptable in or on a food. Test animals are first treated with the drug in accordance with proposed GPVD and, on the basis of this usage, tentative MRLs are set for various tissues. These MRLs are then compared with the ADI, considering dietary food intake. If the MRL established on the basis of proposed GPVD would cause the ADI to be exceeded, the MRL will be lowered to a level which ensures that the ADI is not exceeded, and the proposed GPVD will also be made stricter. If, on the other hand, the proposed
6Ibid., p.65.
WT/DS26/R/USA
Page 7
MRL would not cause the ADI to be exceeded (as is most frequently the case), the MRL will be proposed for adoption. Thus, MRLs are frequently set at levels below (even far below) the theoretical safe levels determined from an ADI. An MRL may also be reduced to be consistent with the GPVD as approved by national authorities or increased (to a level still below the safe level) to be detectable using practical methods.
"Good Practice in the Use of Veterinary Drugs" (GPVD), is defined as:
"the official recommended or authorized usage including withdrawal periods, approved by national authorities, of veterinary drugs under practical conditions." 7
According to the Codex expert advising the Panel, the terms "good veterinary practice" and "good veterinary and husbandry practice", when used in JECFA reports, are synonyms for GPVD.
For the hormones at issue, JECFA considered five of the six substances (all except MGA) and made recommendations on four of them (excluding trenbolone) during its 32nd Session in 1987. For trenbolone, further data was sought and a JECFA recommendation made in 1989. The CCRVDF considered the JECFA recommendations at its meetings in 1987 and recommended draft standards for the three endogenous hormones and zeranol. These draft standards were approved by the Codex Commission at Step 5 in 1989. Standards for these four hormones were considered at Step 8 by the Codex Commission in June 1991, but, following a vote on the matter, were not adopted. A draft standard for trenbolone at Step 5 was adopted on 1991. In June 1995, the Codex Commission adopted standards, at Step 8, for the five hormones, on the basis of a vote. These standards apply exclusively with respect to cattle, and meat and meat products of bovine origin, when these hormones are used for growth promotion purposes.
With respect to the three natural hormones in dispute, , progesterone and testosterone, similar Codex standards apply. For these three hormones it was considered "unnecessary" to establish an ADI or MRL.8 Specifically, the Codex states that:
"Establishing an ADI and an [MRL] for a hormone that is produced endogenously at variable levels in human beings was considered unnecessary by the Committee. Residues resulting from the use of this substance as a growth promoter in accordance with good animal husbandry practice are unlikely to pose a hazard to human health."9
The 32nd JECFA Report of 1988 ("1988 JECFA Report"), on which the Codex standards are based, concluded that residues arising from the use of testosterone and as a growth promoter in accordance with good animal husbandry practice are unlikely to pose a hazard to human health and that the amount of exogenous progesterone ingested in meat from treated animals would not be capable of exerting an hormonal effect, and therefore, any toxic effect, in human beings. Since, according to JECFA, the potential toxic effect of residues of these hormones is directly related to their hormonal effect, the report concluded that the additional residue levels in treated animals are not capable of exerting any toxic effect. On the basis of this safety assessment and in view of the difficulty of determining the levels of residues attributable to the use of these hormones for growth promoting purposes in cattle (residues of endogenous natural hormones in meat cannot be practically distinguished from those exogenously administered), JECFA concluded that it was "unnecessary" to establish an ADI or MRL for these hormones.
7Ibid.
8Codex Alimentarius, Vol. 3 - 1995, Section 1, pp.7, 12 and 14.
9Ibid., Section 1, footnote, pp.7, 12 and 14.
WT/DS26/R/USA
Page 8
With respect to two of the synthetic hormones at issue, zeranol and trenbolone, the Codex standards are the following: an ADI of 0-0.5 and 0-0.02 µg/kg body weight, respectively, and for both hormones an MRL of 2 µg/kg in bovine muscle and 10 µg/kg in bovine liver.
The 1988 JECFA Report, on which the Codex standard for zeranol is based, noted that zeranol was a weak oestrogen which mimicked the action of . The Report concluded that the toxic (in casu tumorigenic) effect of zeranol is associated with its hormonal (i.e. oestrogenic) properties and that an ADI could thus be established on the basis of a no-hormonal-effect level. Adopting what it considered to be a conservative approach by using as a basis studies on ovariectomized female cynomolgus monkeys (highly sensitive to oestrogenic substances) and using a safety factor of 100, JECFA set an ADI for human beings of 0-0.5 µg/kg of body weight. For a 70 kg person consuming 500 g of meat daily over an entire lifetime, the maximum permissible or safe level of zeranol residues in meat would then, according to JECFA, be 70 µg/kg of edible tissue. However, the Report noted that when zeranol is administered to cattle according to good animal husbandry practice, the maximum mean residue levels did not exceed 0.2 µg/kg in muscle, 10 µg/kg in liver, 2 µg/kg in kidney, and
0.3 µg/kg in fat at any time after implantation. These residue levels obtained on the basis of good animal husbandry practice are thus below the maximum permissible level of 70 µg/kg. However, in order to set a level which is detectable by routine residue analysis methods, the Codex MRL was increased to 2 µg/kg in muscle and set at 10 µg/kg in liver.
Trenbolone acetate is the chemical form or ester used for the administration of trenbolone. Trenbolone, or trenbolone acetate ("TBA"), an androgen which mimics the action of testosterone, is rapidly hydrolysed after administration to cattle. The major metabolite (i.e. compound into which TBA breaks down by chemical activity after entering the body) is occurring inter alia in liver, and
-trenbolone present in muscle. With respect to TBA, the 1988 JECFA Report concluded that its potential toxic effects only arise as a consequence of its hormonal activity. The report further concluded that, therefore, an ADI could be established on the basis of a no-hormonal-effect level. Adopting what it considered to be a conservative approach by using as a basis studies on castrated male rhesus macaquemonkeys (which are highly sensitive to compounds with antigonadotropic activity) and pigs (which are a sensitive model for assessing hormonal effects of TBA) and using a safety factor of 100, JECFA later set an ADI for human beings of 0-0.02 µg/kg of body weight (34th JECFA Report of 1989). The maximum ADI for a 60 kg person would thus be 1.2 µg of TBA residues. JECFA then set MRL's for
-trenbolone in muscle and
in liver of 2 µg/kg and 10 µg/kg, respectively, based on average residue levels in heifers at 15-30 days after implantation of 300 mg TBA, noting that concentrations would be even lower at proposed GPVD. According to JECFA, the MRL's thus obtained on the basis of conservative estimates should not exceed the Codex ADI or safe level at any time after implantation of the drug, that is, irrespective of the withdrawal period used.
European consumers' concern over the use of hormones for growth promotion purposes in livestock grew steadily throughout the 1970s as the result of the illegal use of dethylstilboestrol, commonly known as DES (see paragraph 4.123), in vealproduction in France and incidents, particularly in Italy, where adolescents had been reported to be suffering from hormonal irregularities and veal had come under suspicion as a possible cause. European consumer organizations called for a boycott of veal, and the market for veal was severely affected. On 20 September 1980, the EC Council of (Agriculture) Ministers adopted a declaration in favour of a ban on the use of oestrogen and endorsed the principle of greater harmonisation of legislation on veterinary medicines and of greater control on animal rearing, both at the production and slaughtering stages.
On 31 October 1980, the EC Commission proposed legislation aimed at banning the use of all hormone products (COM (80) 614), except for therapeutic purposes. This proposal was expanded later by documents COM(80)920 and COM(80)922, presented on 6 January 1981. These allowed for
WT/DS26/R/USA
Page 9
the controlled use for therapeutic and zootechnical purposes of three natural hormone products, and introduced a number of control measures on the production and handling of such products, together with proposals on the testing of animals. On 13 February 1981, the European Parliament adopted the "Nielsen Report" approving the Commission proposals. The EC Economic and Social Committee endorsed the proposals in February 1981. However, three EC member States (Belgium, Ireland and the United Kingdom) sought to have the three natural hormones remain available both as therapeutic drugs and as growth promotion agents, and Ireland and the United Kingdom also argued for the retention of the synthetic hormones, trenbolone and zeranol. Moreover, third countries, including Argentina, Australia, Canada, New Zealand, South Africa and the United States, also raised questions concerning the impact of any ban on their exports to the European Communities.
The EC Council of Ministers adopted its first Directive on the issue (81/602/EEC) on 31 July 1981. In that Directive, and in regard to five of the hormones at issue (all but MGA), the Council directed the Commission to provide, not later than 1 July 1984, a report on the experience acquired and scientific developments, accompanied, if necessary, by proposals taking into account these developments. Accordingly, the Commission set up a Scientific Group on Anabolic Agents in Animal Production, chaired by Professor G.E. Lamming (the "Lamming Group"). The question addressed to the Lamming Group was:
"Does the use for fattening purposes in animals of the following substances: , testosterone, progesterone, trenbolone and zeranol present any harmful effect to health".10
The Lamming Group issued an interim report on 22 September 1982 (the "Lamming Report"). The Lamming Report concluded as follows:
"The Scientific Working Group is of the opinion that the use of , testosterone and progesterone and those derivatives which readily yield the parent compound on hydrolysis after absorption from the site of application would not present any harmful effects to the health of the consumer when used under the appropriate conditions as growth promoters in farm animals.
"Evaluation of data on "trenbolone" and "zeranol" revealed that some data on the hormonal non-effect-level and the toxicology of these compounds and their metabolites are still missing.
"The Scientific Working Group considers it necessary that additional information be provided before a final conclusion can be given on trenbolone and zeranol.
"Proper programmes to control and monitor the use of anabolic agents are essential.
"It is necessary to continue scientific investigations on the relevance of the present use of the "no-hormone-effect" level related to the harmful effects of anabolic agents".
The EC Scientific Veterinary Committee gave its reaction to the Lamming Report on 9 November 1982, followed by the EC Scientific Committee for Animal Nutrition on 17 November 1982 and by the EC Scientific Committee for Food on 4 February 1983. These Committees supported the conclusions and recommendations of the Lamming Report, but stressed the need to lay down provisions regarding the establishment of proper programmes to control and monitor the use of anabolic agents with regard, in particular, to instructions for use, surveillance programmes and analysis methods. In January 1984, the Commission asked a group of experts within the EC Scientific Committee on Anabolic Agents to review the information on trenbolone and zeranol. On 12 June 1984, the Commission
10Report of the (EC) Scientific Veterinary Committee, Scientific Committee for Animal Nutrition and the Scientific Committee for Food on the Basis of the Report of the Scientific Group on Anabolic Agents in Animal Production, pp.1 and 12.
WT/DS26/R/USA
Page 10
published a proposal (COM(84)295 final) for a Council Directive amending Directive 81/602/EEC, which envisaged the controlled use of the three natural hormones for growth promotion purposes and proposed re-examining the ban on the two synthetic hormones after their scientific evaluation had been completed. However, the European Parliament, the EC Economic and Social Committee and the EC Council of Ministers rejected the Commission's proposal.
The EC Commission amended its proposal accordingly and on 31 December 1985 the EC Council adopted Directive 85/649/EEC. This Directive banned the use of all the substances concerned for growth promotion purposes and established more detailed provisions concerning authorized therapeutic uses. The Directive was challenged in the European Court of Justice, which annulled it on procedural grounds. The proposals were re-introduced by the EC Commission and re-adopted by the EC Council as Council Directive 88/146/EEC on 16 March 1988.
Following reports of significant use of illegal growth-promoting hormonal substances in a number of EC member States, on 26 September 1988 the European Parliament established a "Committee of Enquiry into the Problem of Quality in the Meat Sector". The Report of this Committee (the "Pimenta Report") endorsed the ban on the use of hormones and was adopted by the European Parliament on 29 March 1989 (see paragraphs 4.36-4.39). The essential findings of the Pimenta Report were that the prohibition of hormonal substances for non-therapeutic (i.e. growth-promoting) purposes must be maintained and expanded because:
this was the only way to restore consumer confidence in the meat sector;
10 out of 12 national veterinary experts indicated that a total ban would facilitate implementation and control;
The scientific conclusions regarding the use of natural hormones rested upon strict conditions of use which it believed could not in reality be attained. The Committee was of the opinion that use of the natural/nature-identical hormones carries the risk of inexperienced application, incorrect dosage and unsupervised injection which could pose a risk to the animal and the consumer, and also noted doubts with regard to long- term cumulative and interactive potential carcinogenicity. In addition, the Committee believed that proven necessity and socio-economic desirability should be criteria of acceptability for the use of (bio)chemical growth promoters in animal-rearing;
The Committee did not accept the argument that prohibiting the use for growth promotion of some natural or nature-identical hormones would result in an increase in the use of other, more dangerous growth-promoting substances to the detriment of the consumer;
The Committee believed that the Commission should promote the concept of animal welfare in agricultural production.
The European Parliament adopted another report on the issue of use of hormones for animal growth promotion, the "Collins Report" of 7 February 1989.11 This report argued that:
"Current licensing systems for the regulation of veterinary medicines (including at present, growth promoting products) require that a new product satisfy three criteria: safety, quality and efficacy. These criteria may well be satisfactory for therapeutic drugs. They are by no
11European Parliament, Committee on the Environment, Public Health and Consumer Protection, Report on "The USA's Refusal to comply with Community legislation on slaughterhouses and hormones and the consequences of this refusal", EP 128 381/B, 7 February 1989, named after its reporter Mr. Collins, MEP.
WT/DS26/R/USA
Page 11
means sufficient for growth promoting products. For the latter it is proposed here that the Community's veterinary medicine licensing system be adapted to include a "fourth hurdle", entailing an objective socio-economic and environmental impact assessment". In the Commission's July 1988 draft proposals for the reform of veterinary medicine licensing in the Community this idea was accepted in principle. The final version of the proposals (December 1988) does not include this concept. It is clear, however, that the social, agricultural and environmental implications of the use of growth and yield promoting pharmaceuticals require a licensing system somewhat different from that which exists for these products when used for therapeutic purposes".
The EC Commission organized a scientific conference on this subject in Brussels from 29 November to 1 December 1996. With regard to the natural hormones, the 1995 EC Scientific Conference on Growth Promotion in Meat Production (the "1995 EC Scientific Conference") concluded that:
"At present, there is no evidence for possible health risks to the consumer due to the use of natural sex hormones for growth promotion, since:
Residue levels of these substances measured in meat of treated animals fall within the physiological range observed in meat of comparable untreated animals.
The daily production of sex hormones by humans is much higher than the amounts possibly consumed from meat, even in the most sensitive humans (prepubertal children and menopausal women).
Due to an extensive first-pass metabolism, the bioavailability of ingested hormones is low, thus providing a further safety margin."12
With regard to the synthetic hormones, zeranol and trenbolone, the 1995 EC Scientific Conference concluded that:
"At the doses needed for growth promotion, residue levels [of trenbolone and zeranol] are well below the levels regarded as safe (the MRLs). There are, at present, no indications of a possible human health risk from the low levels of covalently-bound residues of trenbolone."13
In March 1987, the United States raised the issue of the EC ban under the Tokyo Round Agreement on Technical Barriers to Trade ("TBT Agreement"). Bilateral consultations between the United States and the European Communities failed to resolve the dispute. Arguing that the EC Directive was not supported by scientific information, the United States requested the establishment of a technical experts group ("TEG") under Article 14.5 of the TBT Agreement to examine the question. This request was denied following the EC response that the use of growth promotants was a process and production method (PPM), and that parties to the TBT Agreement only had an obligation not to use PPMs to circumvent the Agreement. The European Communities favoured the establishment of a panel "to
12"Assessment of Health Risk - Working Group II", in 1995 EC Scientific Conference Proceedings, pp. 20-21.
13Ibid.
WT/DS26/R/USA
Page 12
evaluate the rights and obligations of Parties deriving from Article 14.25 (of the TBT Agreement)". 14
The dispute went unresolved.
On 1 January 1989, the United States introduced retaliatory measures in the form of 100 per cent ad valorem duties on a list of products imported from the European Communities. The European Communities consequently asked for the establishment of a panel. This request was denied by the United States. In 1989, a joint US/EC Task Force agreed on certain measures which allowed imports into the European Communities of US meat certified to have not been produced with hormones. This resulted in the United States withdrawing some products from the retaliation list. The other EC products figuring in the list remained subject to the retaliatory action. On 19 June 1996, the European Communities requested the establishment of a panel to examine this matter. On 15 July 1996, after this Panel was composed, the United States terminated its retaliatory action in its entirety.
CLAIMS OF THE PARTIES
The United States claimed that the EC ban on the importation and sale of animals, and meat derived from animals, that had been administered any of the six hormones at issue for growth promotion purposes (oestradiol-17 , progesterone, testosterone, trenbolone, zeranol and melengestrol acetate (MGA)) was inconsistent with the SPS Agreement and the GATT.
Arguing that the EC measures were sanitary measures, the United States claimed, with regard to the SPS Agreement, that these measures: directly and indirectly affected international trade; were not based on an assessment of risk and were consequently inconsistent with Article 5.1 of the Agreement; were maintained without sufficient scientific evidence in contravention of Article 2.2; were not justified as a "provisional" measure under Article 5.7; breached Articles 2.2 and 5.6 in that they were not based on scientific principles; were not applied only to the extent necessary to protect human life or health and were more trade-restrictive than required to achieve the appropriate level of sanitary protection; arbitrarily or unjustifiably discriminated between Members where identical or similar conditions prevailed, in contravention of Article 2.3; constituted a disguised restriction on international trade, in breach of Article 2.3; contravened Article 3.1 because they were not based on the relevant international standards, guidelines or recommendations and that this departure from international standards was not justified by Article 3.3; and were based on arbitrary or unjustifiable distinctions in the levels of protection in different situations, resulting in discrimination or a disguised restriction on international trade in contravention of Article 5.
The United States claimed that the EC measures discriminated against imports and were inconsistent with Article III of GATT. Arguing that US meat and animals were "like" EC meat and animals, the United States claimed that the EC measures were inconsistent with Article III:4 of GATT because they prohibited the importation and sale of certain imported meat and animals, while permitting the sale of like domestic products. The EC measures therefore treated US imports less favourably than domestic production. The United States further claimed that the European Communities had no legitimate policy purpose for discriminating against US meat and animals. The United States also claimed that the EC measures were inconsistent with Article I:1 of GATT becausethey failed to accord to imports from the United States the advantages, favours, privileges or immunities granted to like animals and meat originating in the territories of other countries. Finally, the United States argued that the EC measures could not be justified by resort to Article XX of GATT, in particular Article XX(b), because the European Communities had put forward no evidence to support its measures on health grounds, and the measures were "applied in a manner which would constitute a means of arbitrary or unjustifiable discrimination between countries where the same conditions prevail, or a disguised restriction on international trade". The United States claimed that if the EC measures were not sanitary measures,
14GATT document TBT/M/Spec/7, p.9, para. 34.
WT/DS26/R/USA
Page 13
they would be inconsistent with the TBT Agreement; in particular they were technical regulations within the meaning of the TBT Agreement, and would be inconsistent with Articles 2.1, 2.2, and 5.1.1 and 5.1.2.
The European Communities submitted that the analysis of the SPS and/or TBT Agreements should take place only if alleged violations of GATT Articles were found. Therefore, in their defense, the European Communities first invoked GATT. With regard to the alleged violation of Article III:4 of GATT, the European Communities argued that the animals to which the hormones at issue had been administered for growth promotion, and meat from those animals, were not "like" other animals and meat from those animals, respectively. Furthermore, the European Communities argued that even if they were found to be "like", imported products were not given "less favourable treatment" than domestic products. Therefore, the European Communities claimed that its measures did not infringe Article III:4 of GATT. The European Communities claimed that in case its measures were found to be contrary to Article III:4, they were justified by Article XX(b), which did not affect the power of a Member to adopt a policy in order to protect human and animal health.
With regard to the alleged violation of Article I of GATT, the European Communities claimed that such violation was not mentioned in the Panel's terms of reference, was not mentioned in the documents by which the United States requested consultations and was not discussed during the consultations that were held subsequently. The European Communities claimed that even if animals and meat from animals to which the hormones at issue had been administered for growth promotion were found to be "like" (quod non), the measures at issue applied without distinction to all imports of meat irrespective of their country of origin and not only to imports originating in the United States. Accordingly, the European Communities argued that its measures did not violate Article I:1 of GATT.
The European Communities claimed that the measures at issue, in any event, did not violate any provision of the SPS Agreement because they satisfied all the conditions imposed by it. The measures were based on scientific principles as required by Article 2.2 of the SPS Agreement, and a risk assessment had been performed which established the scientific basis for regulatory action. The European Communities observed that the SPS Agreement recognized a Member's right to establish the level of protection which the Member determined to be appropriate. The European Communities claimed that the measures at issue aimed at achieving a level of protection which was higher than could be achieved if the recommendations of Codex Alimentarius for these hormones were followed. It also claimed that WTO dispute settlement panels were not competent to judge its level of sanitary protection nor the scientific evidence upon which it was based, but only whether its measures were in conformity with the provisions of the SPS Agreement. It further claimed that the US arguments in fact attacked the EC chosen level of protection, not its measures, because they suggested that residues of these hormones, above naturally present levels, did not pose any risk to health and, therefore, did not warrant the application of any measures to control them. The European Communities also claimed that its measures were based on the precautionary principle. Moreover, it claimed that the United States had failed to discharge its burden of proof because it had failed to show that the measures at issue were no more trade restrictive than required to achieve the EC appropriate level of sanitary protection. The European Communities claimed that its measures were applied in exactly the same way to all animals treated with these hormones and meat from such animals intended for consumption in the EC market, whatever its origin; there was consequently no discrimination nor disguised restriction on international trade. The European Communities did not submit arguments on TBT because it considered that the measures at issue fell with the SPS Agreement. The European Communities claimed that because the measures at issue did not violate any of the provisions of the SPS Agreement, they should be found also to be in conformity with the rules of GATT, in particular Article XX:b, in case a violation of one of its provisions were to be found.
WT/DS26/R/USA
Page 14
ARGUMENTS OF THE PARTIES
The European Communities considered that the SPS Agreement would apply only if a violation of the Articles of GATT were to be established. The European Communities argued that the General Interpretative Note to Annex 1A of the WTO Agreement provided that:
"In the event of conflict between a provision of the General Agreement on Tariffs and Trade 1994 and a provision of another agreement in Annex 1A to the Agreement Establishing the World Trade Organisation ...., the provision of the other agreement shall prevail to the extent of the conflict" (emphasis added).
The European Communities argued that the SPS Agreement codified, as the last recital of its preamble stated, the desire of the Members "to elaborate rules for the application of the provisions of GATT which relate to the use of sanitary or phytosanitary measures, in particular the provisions of Article XX(b)". Furthermore, it was well established GATT law and practice that GATT 1947 did not affect the power of its Members to set up a regulatory policy which they deemed necessary in order to protect human, animal or plant health. But the conformity of the measures it applied for that purpose could be reviewed under GATT.15 It was, therefore, admitted that a violation of GATT by a measure should be established first, before recourse to possible justifications (e.g. under Article XX(b)) could be considered. In such an event, the Panel could review the measure but not the underlying policy objective on which it was alleged to be based. The European Communities noted that there were several new provisions in the SPS Agreement that established a series of rights and obligations for Members. However, the SPS Agreement reaffirmed the right of Members to adopt or enforce measures the Member deemed to be necessary to protect human, animal or plant health (recitals 1 and 6 of preamble). It could therefore be argued that most of the obligations created by the SPS Agreement were already applied under GATT 1947 through interpretations of Article XX(b) by panel reports and the CONTRACTING PARTIES.
The European Communities claimed that Article 2.4 of the SPS Agreement established that SPS measures which conformed to the Agreement were "presumed" to be in accordance with the obligations of the Members "under the provisions of GATT which relate to the use of sanitary or phytosanitary measures, in particular the provisions of Article XX(b)" (emphasis added). Despite the fact that the words "in particular" appeared in Article 2:4, it was hard to imagine other provisions of GATT "which relate to the use of sanitary or phytosanitary measures".
The European Communities noted that, in their "Statement of Administrative Action", the USTR had explained that:
"The S&P negotiations initially began as an attempt to elaborate on the provisions of Article XX(b). The Agreement extends beyond an interpretation of existing GATT provisions, however, and includes new obligations - in particular, transparency requirements such as providing notice of, and an opportunity to comment on, proposed S&P measures."16
Therefore, if the SPS Agreement were to be defined as a self-standing agreement this was not because it interpreted provisions of GATT other that Article XX(b), but only because it laid down additional procedural requirements. The substantive role of the SPS Agreement, was to interpret Article XX(b)
15See, e.g. the panel report on "Thailand - Restrictions on Importation of and Internal Taxes on Cigarettes", adopted on 5 October 1990, DS10/R.
16US Statement of Administrative Action, p.87.
WT/DS26/R/USA
Page 15
of GATT. In other words, recourse to the substantive provisions of the SPS Agreement could be made only under the same conditions under which recourse could be made to Article XX of GATT, that was only after a violation of another provision of GATT was first established. As far as the additional, procedural obligations laid down by the SPS Agreement were concerned, these could be examined by the Panel directly and independently of any need to establish first a violation of the provisions of GATT.
The United States claimed that the EC ban was subject to the SPS Agreement because the Agreement applied to "all sanitary and phytosanitary measures which may, directly or indirectly, affect international trade"17 and the EC measures were sanitary measures, as defined by the SPS Agreement, which directly and indirectly affected international trade. The SPS Agreement was specifically tailored to SPS measures and was the lex specialis in this instance. It made sense to look first to the agreement most specifically designed to address the types of measures and issues involved. The SPS Agreement applied to all SPS measures, and applied whether or not a Member was invoking a GATT exception for such measures. It imposed requirements additional to those in Article III and to the conditions for justifying the application of the exception in Article XX(b). For instance, Article 3.2 of the SPS Agreement provided that an SPS measure conforming to international standards, guidelines or recommendations was presumed to be consistent with the "relevant provisions" of not only GATT, but also the SPS Agreement itself. The Agreement stood independently from GATT. Thus, a complaining party did not need to first show that an SPS measure was inconsistent with GATT before it might invoke the provisions of the SPS Agreement.
The United States submitted that the negotiating record confirmed that the SPS Agreement was intended to go beyond GATT and to impose requirements additional to those in GATT. In the earliest phases of negotiations on SPS issues, the discussions had focused on further definition of Article XX(b). Up to the Dunkel Draft text (MTN.TNC/W/FA of 20 December 1991), the SPS text was phrased as a Decision of the CONTRACTING PARTIES interpreting the GATT, simply because participants had wished the proposed disciplines to apply to all GATT contracting parties and only a Decision (as opposed to an optional-membership Code) would accomplish that result.18 However, in April 1992, when the possibility emerged of using the MTO Agreement (later renamed the WTO Agreement) to tie the Uruguay Round results together into a "single undertaking", negotiators had agreed to change the form of the text to a free-standing Agreement in Annex 1A of the MTO Agreement.19 In contrast, the various Understandings on GATT Articles began as decisions interpreting the GATT and were finally incorporated into GATT. Had the negotiators intended that the SPS Agreement be nothing but a clarification of the GATT, in whole or in part, the United States argued, they would have given its provisions the same form as, for instance, the Understanding on the Interpretation of Article XVII of the GATT. The European Communities' suggestion that the SPS Agreement could be divided between provisions that applied to all SPS measures (and applied under
17 Article 1.1 of the SPS Agreement.
18"Given the desire of participants for the application of the proposed disciplines to all contracting parties, this draft has been presented in the form of a Decision of the CONTRACTING PARTIES on the Application of Sanitary and Phytosanitary Measures. However, this is without prejudice to the final form the agreement might take." MTN.GNG/NG5/WGSP/7, 20 November 1990, p.1.
19P.21 of document 707 dated 15 April 1992, "Review of Individual Texts in the Draft Final Act (Texts on Agriculture), Informal Note by the Secretariat" (Secretariat proposal to rectify the form to change it from a decision to an agreement), and p.19 of document 963 dated 10 June 1992, "Review of Individual Texts in the Draft Final Act, Informal Note by the Secretariat" (record of rectifications agreed in the Legal Drafting Group in April 1992, including this change in form). The paragraphing of the text was rearranged into article format during the legal drafting process for the Marrakesh Final Act in February 1994. Secretariat proposals for rectifications at p.71 of MTN/FA/Corr.2 dated 18 February 1994, and agreed rectifications in MTN/FA/Corr.5 dated 11 March 1994.
WT/DS26/R/USA
Page 16
all circumstances), and other provisions which only applied if an SPS measure violated the GATT, had no basis in the text of the SPS Agreement.
The United States claimed that according to the definition of sanitary and phytosanitary measures in the SPS Agreement, whether a measure was a sanitary measure depended on its purpose20. The United States noted that the purported sanitary purpose of the EC measures was stated in the texts of the EC Council Directives themselves, and was apparent from the history of the EC measures and subsequent statements of the European Communities, including statements to Contracting Parties of the GATT 1947. In reviewing the proposal for Directive 81/602/EEC, the European Economic and Social Committee had recognized that "there is an urgent need to draw up EC provisions in order to protect human health".21 Similarly, the debate of the European Parliament on the proposal for Directive 88/146/EEC underscored the claim that the EC measures were designed to protect human health from risks associated with the use of the hormones at issue.22 The preamble to Directive 81/602/EEC stated in part: "Whereas, due to the residues that they leave in meat, certain substances with a thyrostatic, oestrogenic, androgenic or gestagenic action may be dangerous for consumers" and "Whereas, moreover, the harmless or harmful effects of the use of , Progesterone, Testosterone, Trenbolone and Zeranol still have to be examined in detail." This concern for human health was carried forward with Directive 88/146/EEC.23 In addition, during the discussion in the Tokyo Round Committee on Technical Barriers to Trade concerning the EC measures, the EC representative had stated that the European Communities "adopted the Directive as the best way of avoiding any health risks connected with the use of hormonal substances".24 The United States argued that although the European Communities had stated that the purpose of its measures was a sanitary purpose, ostensibly to protect human life or health within the European Communities from risks arising from additives, contaminants, or toxins in food, the measures were not legitimate sanitary measures.
The European Communities agreed with the United States that the measures challenged were better defined as measures falling within the scope of the SPS Agreement rather than the TBT Agreement.
20Article 2.1 of the SPS Agreement. The United States argued that the European Communities was mistaken in its perception of the coverage of the SPS Agreement. The EC claim that sanitary and phytosanitary measures were defined based in part on the type of product to which the measure applied was incorrect. Nothing in the SPS Agreement limited its coverage by the type of product. The SPS Agreement applied to any sanitary or phytosanitary measure. For example, measures restricting the importation of construction equipment to ensure that it was not carrying soil with plant pests or diseases, or that it did not harbour plant pests, were phytosanitary measures, even though they applied to construction equipment. The same would be true for measures applicable to lawn furniture to ensure that it did not harbour cocoons of a plant pest, or measures applicable to tires to make sure they did not host mosquitoes carrying infectious diseases for humans.
21Paragraph 1.2 of the opinion, EC Official Journal No C 138/29, 9 June 1981. This opinion is cited in the preamble to Directive 81/602/EEC.
22Debate published in the Annex to the EC Official Journal No 2-330 of 10 October 1985, beginning, p.232.
23Directive 88/146/EEC states that:
"Whereas the administration to farm animals of certain substances having a hormonal action is at present regulated in different ways in the member States; whereas while their immediate effect on animals from the farmer's point of view is clear, assessments of their effect on human health vary and this is reflected in the regulations governing their use; whereas this divergence distorts the conditions of competition in products that are the subject of common market organizations and is a serious barrier to intra-EC trade."
24Paragraph 12, TBT/M/Spec/7 (Committee on Technical Barriers to Trade, Minutes of Meeting held on 23 July and 28 July 1987). See also paragraph 6 of TBT/M/Spec/7 ("The complaint of the United States should not deter his authorities from applying the measures that they had adopted to protect the health and safety of their population."), as well as paragraphs 9 and 24. See also paragraphs 11 through 13 of TBT/M/Spec/5.
WT/DS26/R/USA
Page 17
The European Communities noted that an essential link of the SPS disciplines with the regulatory freedom of the parties was that the measures applied were of concern to the SPS Agreement (and the WTO system) only if they affected international trade. Moreover, the SPS Agreement defined the measures to which it applied on the basis of the objective of the measure and the type of product to which it was applied.
The United States claimed that the EC measures on their face applied to imports of animals and meat, prohibiting imports of animals to which any of the hormones at issue had been administered for growth promotion purposes, and imports of meat from such animals. They clearly affected international trade and, when they were applied, they closed market access for significant quantities of imports from the United States 25 and other Members and continued to block imports of animals and meat.
The United States observed that US exports of beef and veal to the European Communities in the three years prior to the ban (1986-88) averaged in the hundreds of millions of dollars, and trade to the European Communities during the 1985-87 period was growing at about 30 per cent per year.26 In 1989, when the ban went into effect, US exports plummeted to nearly zero. Moreover, as a result of various rounds of multilateral trade negotiations, the European Communities had bound its tariffs on all products affected by the measures at issue. In the Tokyo Round of Multilateral Trade Negotiations, concluded in 1979 prior to the adoption of the EC measures at issue here, the United States negotiated a quota for 10,000 tonnes of high-quality beef subject to a tariff of 20 per cent ad valorem and free of any variable levy. In the Uruguay Round, this 10,000-tonne quota was converted to a tariff-rate quota and the variable levy on beef imported in excess of 10,000 tonnes was "tariffied". The European Communities also had bound duties on offal products (e.g., hearts, livers, and kidneys). As a result of the EC beef import ban, US exports of these products to the European Communities had been substantially impaired.
The United States argued that the SPS Agreement was designed to ensure that no Member maintained protectionist barriers to trade in the guise of SPS measures and that Article 1.1 required that sanitary measures "be developed and applied in accordance with the provisions of this Agreement". The requirement that measures be "developed" in accordance with the SPS Agreement did not apply here since the EC measures were developed prior to the entry into force of the WTO. However, the European Communities were still required to "apply" these measures in accordance with the provisions of the SPS Agreement.
Moreover, in order to differentiate between legitimate SPS measures and other types of measures, the United States noted that the SPS Agreement provided a number of disciplines for sanitary measures. These included an obligation that each sanitary measure:
be based on an assessment, as appropriate to the circumstances, of the risks to human, animal or plant life or health27;
not be maintained without sufficient scientific evidence28;
25According to a table provided by the United States, an average 70 per cent of all US cattle were treated with one or more of these hormones.
26The United States estimated in 1988 that the EC measures cut off approximately $100 million in US exports.
27Article 5.1 of the SPS Agreement.
28Ibid., Article 2.2.
WT/DS26/R/USA
Page 18
be based on scientific principles29;
not be more trade-restrictive than required to achieve the appropriate level of sanitary protection, taking into account technical and economic feasibility30;
not arbitrarily or unjustifiably discriminate between Members where identicalor similar conditions prevailed31;
not be applied in a manner that would constitute a disguised restriction on international trade32; and
be based on international standards, guidelines or recommendations, where they existed33.
The United States claimed that the EC measures failed each of these requirements. The EC ban was not a legitimate sanitary measure because it severely restricted international trade, was not applied only to the extent necessary to protect human life or health, was not based on scientific principles, was maintained without sufficient scientific evidence, was not based on an assessment of the risks, and was more trade-restrictive than required to achieve the appropriate level of protection. It also constituted a disguised restriction on international trade.
The European Communities responded that, as envisaged in the EC legislation, imports might take place from third countries provided that guarantees were offered that no animal and no meat coming from an animal to which hormonal substanceshad been administered would be exported to the European Communities. The EC legislation in question did not impose a ban on US meat or on meat from any other origin. The US exported fresh meat not destined for human consumption and exported to the European Communities hormone-free meat products for human consumption on the basis of measures agreed by the joint US/EC Task Force. The US exports of meat to the EC market for the years 1989 to 1994 were as follows: 1988: 72,557 tonnes, 1989: 8,380 tonnes, 1990: 3,617 tonnes, 1991: 1,156 tonnes, 1992: 6,320 tonnes, 1993: 6,833 tonnes, 1994: 6,594 tonnes, 1995: 8,499 tonnes.34
The European Communities agreed that the SPS Agreement imposed the disciplines enumerated by the United States and considered that its measures satisfied all those disciplines. When the EC Commission had first proposed to the EC Council in 1980 to take measures in this area as well as in 1988, it had found itself facing certain factual, legal and scientific situations. From the factual point of view, consumer concerns over the use of hormones for growth promotion in livestock were very high. Despite the traditionally cautious approach to the regulation of dangerous substances followed by the EC member States, consumer confidence in science and regulators (especially among well- informed consumer organisations and groups) was very low. This factual situation had not changed during the entire period that had preceded the adoption of Directive 96/22/EC. Consumer concerns were even higher today.
29Ibid., Article 2.2.
30Ibid., Article 5.6.
31Ibid., Article 2.3.
32Ibid., Article 2.3.
33Ibid., Article 3.1.
34The European Communities explained that the figures were based on Eurostat statistics.
WT/DS26/R/USA
Page 19
The European Communities noted that from the legal point of view, the use of hormones in animal growth promotion had long been prohibited in most EC member States; only some EC member States allowed the use of some of these substances under certain conditions. 35 With the progressive establishment of the common market, the divergence in the legislation of EC member States was inhibiting free trade within the European Communities in animals and meat from animals treated with some of these hormones, and distorting the conditions of competition among EC meat producers.
Furthermore, the European Communities observed that from the scientific point of view, the situation was very unclear in the early 1980s. The relevant international organizations - FAO, WHO, the Office internationale des épizootics (OIE) and the Codex Commission - had started to seriously examine the safety of these hormones in meat production only during the 1980s. The first substantive and comprehensive scientific report had been published by OIE in 1983. JECFA had discussed and issued a substantive and comprehensive scientific report on these hormones only in 1988. Furthermore, none of these scientific reports established beyond doubt that the use of these hormones in animal growth was safe for human health. For example, in the 1983 OIE Scientific Report it was stated:
"Even though many unknown influences still exist with regard to the details of the mechanism of action, particularly with reference to how the message brought by the receptor is translated into action in the cell nucleus, we are nevertheless beginning to have some understanding of how anabolics work" (page 62).
With regard to the three so-called "natural" hormones, the European Communities noted that all of the above scientific reports suggested that they were "unlikely to pose a hazard to human health" if used "in accordance with good veterinary or animal husbandry practice".36 As regards the other so-called "synthetic" hormones, reliable toxicological data were still missing in the 1988 JECFA Report (page 26). Their use for animal growth promotion had been provisionally recommended in accordance with good animal husbandry practice and residue checking in order to ensure that the so-called Acceptable Residue Levels (ARLs) provisionally recommended were not exceeded. The European Communities argued that even today the state of scientific knowledge was not much different from what it was in 1981 and 1988 when the challenged measures had been adopted (as shown by the 1995 EC Scientific Conference).
The United States argued that although the purported major objective of the EC ban was to protect human health, the real purpose was to protect EC meat production from third-country competition. Noting that the European Communities cited as the purposes of these measures (i) protection of human health and (ii) a desire to alter the terms of competition for animals and meat in order to render these conditions "largely identical" within the EC market, the United States claimed that neither reason offered a valid justification under the WTO agreements. Furthermore, in developing the hormone ban, the European Communities were also motivated by the fact that the ban would help reduce surplus supplies of meat in the European Communities and lower the cost of the EC Common Agricultural Policy. The United States argued that this attempt to protect domestic production from more competitive imports was protectionism and was inconsistent with EC WTO obligations.
The United States noted in this regard that in April 1984, the European Communities had introduced milk quotas to reduce theoversupply of milk, which resulted in an increase in cattle slaughter.
35The European Communities indicated that 4 or 5 EC member States allowed the use of some of these hormones for growth promotion until their prohibition at EC level. The information on level of use, if available at all, would only be known to those individual EC member States. One EC member State, the United Kingdom, had indicated that anecdotal evidence suggested that growth promoting hormones might have been used up to 40 per cent of the United Kingdom cattle prior to the ban.
361988 JECFA Report, pp.19 et seq.
WT/DS26/R/USA
Page 20
EC intervention stocks of beef, which were slightly less than 400,000 tonnes in November 1983, soared to over 800,000 tonnes by the end of 1985.
The European Communities rejected the US argument that the EC measures had been adopted in order to protect domestic production from foreign competition. The reference to the divergence in the legislation of the EC member States which distorted the conditions of competition and affected free circulation of products in intra-EC trade (1st and 2nd paragraphs of Directive 88/146/EEC), was made for the sole purpose of justifying the need to take action at the EC level. Because of its internal constitutional structure, the European Communities was obliged to take action in order to eliminate distortions in the conditions of competition and to increase free intra-EC trade.37 The preamble of the Directive thus clearly explained that the underlying objective was to respond in an harmonized way to the effect of these hormones on human health. This objective, however, was not in itself independent or more important than the objective of safeguarding against risks to human and animal health resulting from the substances at issue.
The European Communities further argued that action at the EC level was explicitly permitted by Article XXIV:8(a) of GATT. It submitted that when the European Court of Justice examined the object and purpose of Directive 85/649/EEC (invalidated on procedural grounds and subsequently re- enacted in Directive 88/146/EEC) it had noted that "the aim of the Directive, according to the recitals in its preamble, is to protect human health and consumer interests with a view to eliminating the distortion of conditions of competition and bringing about an increase in consumption of the product in question".
Furthermore, the European Communities asserted that the historical record clearly demonstrated that the purpose of the EC measures was to protect human and animal health from risks arising from the use of the hormones at issue. The European Communities noted that the European Parliament had proposed a ban on imports of meat from animals treated with growth promoting hormones primarily because "scientific information about these substances is far from complete and that considerable doubt therefore exists about the desirability of their use and of their effect on human health".38 Serious health concerns had been clearly stated in different MEP reports, including the "Nielsen Report", the "Pimenta Report" and the "Collins Report", and the European Parliament had remained constantly opposed to the authorization of all these hormones for animal growth promoting purposes for reasons of possible hazards to human and animal health. The Economic and Social Committee had also rejected the use of these hormones for growth promotion purposes on health grounds. The European Communities claimed that, from the above sequence of events which had eventually led to the adoption of Directive 88/146/EC, it was clear that the European Communities had examined carefully the potential risks to human and animal health from the use of these hormones for animal growth promotion purposes; considered that scientific evidence, whilst suggesting that these substances, if used in accordance with good veterinary practice, did not seem to pose risks to human and animal health, did not exclude beyond doubt that they posed no potential risks to human health; examined carefully all of the available options for harmonizing the conditions of use of these substances before concluding that, at present, the ban on their use was the only scientifically, technically and economically feasible option.
The European Communities rejected the US references to milk quotas and observed that the increase in the EC intervention stocks of beef was unrelated to the objective and purpose of the challenged measure. Such high intervention stocks of beef did not exist in the EC member States that introduced the prohibition on the use of these hormones as early as 1961; such high intervention stocks of beef did not exist in 1981 when this prohibition was for the first time imposed at EC level; and such high
37The European Communities noted that this obligation resulted from Articles 100 and 100A, in conjunction, in this case, with Articles 30 and 36 of the EC Treaty.
38Point E of Parliament's Resolution, EC Official Journal No 288, 11 November 1985, p.158.
WT/DS26/R/USA
Page 21
intervention stocks of beef did not exist in April 1996 when the European Communities adopted Directive 96/22/EC which re-enforced the prohibition on the use of these hormones. The European Communities further submitted that the fact that the Directive offered equal access to the EC market for third-country meat was additional proof that there was no protectionist purpose in the EC measures. Statistics showed that the European Communities had continued to import about the same quantities of meat as before the application of the ban, it was, however, meat from animals to which hormones for growth promotion purposes had not been administered.
The United States recalled that Article 2.2 required Members to ensure that any sanitary measure was applied only to the extent necessary to protect human or animal health, based on scientific principles and not maintained without sufficient scientific evidence. The United States argued that the European Communities had failed to follow scientific principles, and lacked any scientific evidence to give credence to its asserted health concerns. The European Communities presented no evidence to contradict the long established scientific determination that the six hormones were safe for use as growth-promotion substances and that consumption of meat containing the minute quantities of residues from treated animals was not harmful to consumers.
The United States claimed that while the negotiators had not defined precisely "scientific principles", at a minimum this term incorporated the scientific method, which represented those principles and processes universally regarded as necessary for scientific investigation, in particular procedures for:
the observation of phenomena in nature or under controlled conditions;
the systematic classification of empirical data;
the measurement of empirical quantities and for calculating probable errors and significant deviations;
forming a hypothesis;
analysing experimental results using logic and mathematics; and
many other related techniques and processes.
The European Communities responded that the US explanation of the concept of "scientific principles" was a caricature of "the scientific method" which could have been taken straight from a school textbook circa 1960. There was absolutely no reason to suppose that the Members of the SPS Agreement had the list presented by the United States in mind when signing it. There were many theories of science and the "scientific method"; the European Communities relied on biological principles when assessing the risks of using hormones for growth promotion. Measures must be based on scientific principles, as opposed to non-scientific ones, such as superstition. If a measure was aimed at reducing or eliminating a risk to health, then it must actually address that risk in a manner which could be scientifically justified. If, for example, the measure was aimed at eliminating a pathogenic organism from a food, there were several methods, e.g. heating, salting, pickling, etc. which could be scientifically proven to be effective. If, however, a Member required prayers to be said over the food, or a ritual dance to be performed around it, that would not be compatible with the SPS Agreement because such methods could not be scientifically proven to be effective.
The European Communities further argued that during the meetings of the Panel with scientists on 17 and 18 February 1997, a number of experts advising the European Communities explained their
WT/DS26/R/USA
Page 22
views on the potential dangers to human and animal health from the use of these hormones for growth promotion. None of the experts advising the Panel argued that the experts referred to by the European Communities did not employ scientificprinciples in their research. The US Statementof Administrative Action stated that Article 2.2 of the SPS Agreement did not require the best science nor the weight of scientific evidence to be taken into account; it only stipulated that there should be "scientific principles" and "sufficient" (not absolute) scientific evidence. The European Communities further argued that for example, Dr. Lucier had agreed with the conclusions of the EC scientists that both the natural and synthetic hormones were carcinogenic at low levels (the natural hormones are carcinogenic even at existing physiological levels). Some of the scientists who attended the meetings of the Panel (and those of JECFA 1988) might not agree with the conclusions of the scientists advising the European Communities or with Dr. Lucier, but this was not relevant for the purposes of Articles 5.2,
2.2 and 3.3 of the SPS Agreement. What was important was whether, in the scientific research employed by the EC scientists (or the scientific reports to which they made reference in their reports), the minimal attributes of scientific inquiry were respected. The European Communities had not heard the opposite from any of the scientists advising the Panel nor from the United States and Canada. Therefore, the European Communities was allowed to take into account their scientific views (or the views of the scientists to which they referred in their reports) in its assessment of the risks of these hormones for growth promotion.
The European Communities indicated that a logical consequence of the requirement for measures to be based on scientific principles was that they must not be maintained without scientific evidence. All Members had measures in place before the SPS Agreement was drawn up, and in the absence of this requirement it could have been argued that the requirement for basing measures on scientific principles could not be applied retrospectively. The SPS Agreement understandably did not define the term "scientific evidence". It was easier to argue that there were scientific principles, scientific methods, scientific experiments and scientific data. But it was difficult to define what was "scientific evidence", since its content was relative in terms of time and was dependent on the principles, methods, experiments and data mentioned. For example, what might be an acceptable scientific method for one scientist might not satisfy another, who might be more interested in certain other scientific principles or aspects totally neglected or only partially examined by the first scientist. For that reason the SPS Agreement only required "sufficient", not clear or certain, scientific evidence (Article 2.2).39 The SPS also required Members, in their risk assessment, to taken into account "available scientific evidence". But from the "available" scientific evidence, a Member was entitled to rely on that which its own scientists said was appropriate and sufficient, and disregard other available evidence. It followed that neither the Panel nor any other member might judge the adequacy of the scientific evidence upon which a Member based its measure in order to achieve its level of sanitary and phytosanitary protection. For the same reasons, a Codex group of scientific experts could not judge the adequacy of the scientific evidence used by a Member of the SPS Agreement.40 In other words, if the "weight" of available scientific evidence indicated that a substance was not dangerous to human health, but another small or minority part of available scientific evidence (of the same Member) argued that there may be potential risks to human health, that Member was entitled under the SPS Agreement to take a precautionary approach and base its measure on the latter part of the available scientific evidence. As the United States has said in the Statement of Administrative Action, it is sufficient if the "government maintaining the measures has a scientific basis for it" (emphasis added).
The scientific basis of the EC Directives 81/602 and 88/146 were the following:
39On the other hand, the European Communities noted that the term "scientific evidence" had a different meaning from the requirement to provide "evidence" in a legal trial.
40See, for example, the article by D.A. Wirth (1994), "The role of Science in the Uruguay Round and the NAFTA Trade Disciplines", 27 Cornell International Law Journal, 817-859, at 856-57.
WT/DS26/R/USA
Page 23
the 1982 Report of the Scientific Veterinary Committee, Scientific Committee for Animal Nutrition and the Scientific Committee for Food on the basis of the Report of the Scientific Group on Anabolic Agents in Animal Production (the Lamming Report);
the 1983 OIE Scientific Conference Report;
the 1988 JECFA Report;
the various works of relevant international institutions, such as the International Agency for Research on Cancer (IARC);
the scientific works by individual scientists relevant to the issue of use of hormones in general and for animal growth in particular; and
information on the use of these hormones for growth promotion available from other countries, when relevant.
For the adoption of Directives 81/602/EEC, 88/146/EEC and 92/22/EC, apart from the above scientific evidence, additional technical information had been taken into account. This information consisted mainly of the internal studies of the EC Commission, the reports of the European Parliament, the reports of the Economic and Social Committee and the deliberations of the Council of Ministers. For their deliberations, the Ministers were assisted by scientificgroups and individual scientific experts, including experts from the relevant administrations of the EC member States. For the adoption of Directive 96/22 of 29 April 1996, the 1995 EC Scientific Conference Proceedings had also been taken into account.
The European Communities noted that in its first proposal to the Council on these substances, the EC Commission had explained that controls were needed "... in view of the effects of these substances in relation to carcinogenicity, physiological effects and meat quality. The latter aspect is also important as regards use of natural substances since physiological effects may also occur when exposure to exceptional levels incidentally takes place".41 In its second submission to the Council of Ministers the EC Commission had explained the (zero) tolerance levels which were being proposed and the need to take into account the natural levels of the endogenous hormones which might occur in animals.42 In its third expanded submission to the Council of Ministers, the EC Commission had explained the proposed therapeutic and zootechnical uses of the natural hormones, including the system of control of use and marketing.43
The European Communities indicated that, with respect to Directive 88/146/EEC in 1982, the Commission had created an ad-hoc Scientific Committee of Experts on Anabolic Agents in Animal Production (chaired by Professor Lamming). This Committee had issued an interim report on 22 September 1982 (the "Lamming Report"). The EC Scientific Veterinary Committee gaveits reaction on 9 November 1982. This was followed by the EC Scientific Committee on Animal Nutrition on 17 November 1982, and by the EC Scientific Committee for Food on 4 February 1983. In essence, these Committees supported controlled use of natural hormones, but were opposed to the use of trenbolone and zeranol. On 12 June 1984, the EC Commission published its proposal (COM(84)295) which envisaged the controlled use of the three natural hormones for growth promotion and proposed re-visiting the prohibitions on the two synthetic hormones after completion of their scientific evaluation.
41COM(80)614 of 31 October 1980.
42COM(80)920 of 6 January 1981.
43COM(80)922 of 6 January 1981.
WT/DS26/R/USA
Page 24
The European Commission's proposal was rejected by the European Parliament based on the report of the Committee on the Environment, Public Health and Consumer Protection and on the opinion of the Committee on Agriculture, Fisheries and Food.44 The European Parliament, which did not approve the proposed authorization of the three natural hormones except for therapeutic purposes, had explained that:
"... Considering that scientific information about these substances is far from complete and that considerable doubt therefore exists about the desirability of their use and of their effect on human health ...
"... whereas, in particular, there seems to be doubt as to the effects on the immunity against various diseases of animals treated with hormone cocktails and that this in turn may lead to an increased use of antibiotics ...
"... Recognizes that there is considerable difficulty involved in checking whether such substances have been used because, where they have been properly administered, the measurable residue concentrations are well within normal physiological limits relatively soon after application;
"Believes, therefore, thatthepermitted use of hormones for therapeuticpurposes must be strictly controlled and documented. ...
"Rejects the authorization of artificial and natural hormones as growth promoters".
The European Communities concluded that, therefore, the European Parliament and the scientific experts advising their members had carefully reviewed the scientific evidence presented to it by the European Commission and by its own expert committees, and had decided to reject the Commission's proposal. The Economic and Social Committee had also reviewed the European Commission' s proposal, taking into account a very wide range of technical and scientific expertise, and in its opinion (29 November 1984) stressed that:
"... even the lifting of the bans on the three endogenous anabolics (oestradiol-17 , testosterone and progesterone) in respect of foodstuffs of animal origin cannot be regarded as unobjectionable from the point of view of the safety, health and well-being of the population, until they have beenproved harmless on the basis of sound experience covering, inter alia, conditions of use ..."
The European Commission's proposal was also rejected by the Council of Ministers. As a result, the Commission amended its proposal and on 31 December 1985 the Council adopted Directive 85/649/EEC. This Directive confirmed a ban on the use of all the substances concerned for growth promotion purposes and established more detailed provisions concerning authorized therapeutic use. This Directive was challenged in the European Court of Justice, which annulled it on purely procedural grounds. Re-introduced, the proposal was re-adopted by the Council on 16 March 1988 (Council Directive 88/146/EEC), prohibiting the use in livestock farming of substances having a hormonal action.
The European Communities noted that in 1987, Professor Lamming had published an account of further work undertaken by members of his group on zeranol and trenbolone which concluded that these substances would probably be safe when used in accordance with "accepted husbandry practice".45 The European Communities argued that the report had not defined this concept, and that its conclusion was basedon theunsupported assumption that their carcinogenicity was related to their hormonal effects.
44European Parliament Doc. A2-100/85 of 11 October 1985.
45The Veterinary Record, 24 October 1987, p.389.
WT/DS26/R/USA
Page 25
The European Communities indicated that subsequently, in 1988, following discoveries of significant use of illegal growth-promoting hormonal substances in certain EC member States, the European Parliament had established a "Committee of Enquiry into the Problem of Quality in the Meat Sector". This Committee had received submissions from a wide range of scientific experts, organisations and institutions, including meat trade and farmers' organisations, third country producers, the pharmaceutical industry and consumers' organizations. The outcome of the work of the Committee of Enquiry into the Problem of Quality in the Meat Sector (the "Pimenta Report"), was adopted by the European Parliament on 29 March 1989. The report observed that there was no discernable health risk in the use of the three endogenous hormones at issue where these were, in particular, administered to castrated steers and where they were administered in the form of slow-release implants in a part of the animal which was not consumed (e.g. the ear) and where a specific withdrawal period was observed before slaughter. Of the two other hormonal substances, zeranol was regarded as being as safe as the three endogenous hormones referred to, while with regard to trenbolone acetate, a small residual doubt regarding potential long-term carcinogenicity persisted. The Pimenta Report also indicated that the Committee had been impressed with the thoroughness and objectivity with which the scientists concerned had produced and presented their findings. Nonetheless, the Committee endorsed the ban on the use of hormones, noting examples where scientific knowledge on the safety of substances had subsequently been reversed and that "[t]he scientific evidence does not address the question of potential interaction of these substances with other substances or the multiplier effect of these substances in a worst-case scenario (e.g. the administering of hormones to an animal with an already high level of endogenous hormones and the ingestion of the meat (or milk) of that animal by a female taking the oestrogen-based contraceptive pill" (page 6).
The European Communities noted that the Pimenta Report had concluded that successful regulations could not be based solely on scientific information, partly because of the "current state of knowledge" argument, and partly because the regulatory process had to resolve social and political conflicts that extended beyond scientific considerations. Although the decision-making process in controversial socio-economic and environmental issues had to take into account the scientific evidence, it was essential that the final decision rest with society itself through its elected representatives. The report noted that what might appear to be a public rejection of science was more plausibly related to a question of confidence in the adequacy of regulatory control. The Committee was unconvinced that the legalization of the five natural/nature-identical hormones for growth-promotion would prevent the use of harmful black-market alternatives. It observed that producers were still using easily-detectable synthetic substances because of the continued attractiveness of these products and the economic gain to be obtained from their use. These were also incentives not to follow the conditions for administration specified by the Lamming Committee.
In this respect, the Pimenta Report noted that "upstream" controls were essential, beginning with testing blood and urine samples of live animals on the farm. The processes of metabolism, dilution and excretion in the living animal meant that the longer the interval between administration of the substance and testing of the live animal or carcass, the more difficult it was to detect the metabolites characteristic of the use of illicit substances. In addition, oncethe carcass was slaughtered and butchered, detection of metabolites and tracing back to the animal became very difficult. However, testing for synthetic substances posed no significant problems, although the degree of capability and sophistication of detection varied enormously between EC member States. In addition, methods of chemical masking and other sophisticated techniques were being increasingly used. While testing for the three natural hormones posed a problem in that the range of levels of these hormones produced naturally in the body of the animal was so large and variable, ranges of levels could be drawn up, taking into account the sex, age, oestral cycle and condition of the animal. If tests showed that an animal, for no apparent reason, was at or above the upper limit, then testing on other animals in the herd could provide valid evidence since the statistical possibility of an entire herd possessing above-average hormone levels was minute.
WT/DS26/R/USA
Page 26
The Committee reported that ten out of twelve veterinary authorities from the EC member States agreed that a total ban on administered hormones would be easier to police than a partial ban, becauseveterinarians could generally tell by visual signs when an animal had beentreated with hormones (the shape of the animal and the sexual organs were altered) but not whether this had been a treatment with legal versus illegal substances. Furthermore, psychologically, a total ban was more effective than a partial one in that regulators, producers and consumers were left in no doubt that the administration of any hormonal substance for growth promotion purposes was not allowed. The control of the whole meat production chain in Europe would have to be equally rigorous whether all hormonal substances or only some hormonal substances were banned for growth promotion purposes. Inspection would need to be just as comprehensive and rigorous. The Pimenta Report concluded that severe sanctions would have to be imposed on those flouting the ban to overcome the strong economic incentives to use growth promotion substances, and that this problem had to be tackled in order for a ban to be effective and for confidence to be restored in the meat sector (page 13).
The European Communities added that the European Parliament had adopted another important report on the issue of use of hormones for animal growth promotion, the "Collins Report" of 7 February 1989.46 This report proposed that the EC veterinary medicine licensing system be adapted to includea"fourth hurdle", entailing an objectivesocio-economic and environmental impact assessment in addition to safety, quality and efficacy criteria. It argued that the social, agricultural and environmental implications of the use of growth and yield promoting pharmaceuticals required a licensing system different from that which existed for these products when used for therapeutic purposes.
The European Communities indicated that the European Commission, in close cooperation with EC member States, had continued to review the application of Directive 88/162 in the appropriate scientific and management committees. After the entry into force of the SPS Agreement, the European Communities had decided to review once again the situation from the scientific point of view. For that reason, it had organized the 1995 EC Scientific Conference in Brussels. In light of the findings of that conference and other evidence, the European Commission had proposed to the Council of Ministers to maintain the prohibition on the use of these hormones for growth promotion, to further restrict their use for therapeutic or zootechnical purposes, to reinforce the provisions on control and testing, and to increase substantially the penalties and sanctions in case of violations.47 Both the European Parliament48 and the Economic and Social Committee49 approved the Commission's proposal. The Council of Ministers adopted the new Directive 96/22 of 29 April 1996, which will replace Directive 88/146/EEC and the other relevant Directives as from 1 July 1997.
The United States argued that the European Communities had failed to comply with Article 2.2 which prohibited a Member from maintaining sanitary and phytosanitary measures "without sufficient scientific evidence". The proposal for a Council Directive submitted by the EC Commission on 13 June 1984 explicitly admitted that "on scientific grounds, it appears that the use of oestradiol-17ß, testosterone and progesterone, and those derivatives which readily yield the parent compound on hydrolysis after absorption from the site of application, would not present any harmful effects to the health of the consumer nor harm the consumer by altering the characteristics of meat when used under
46European Parliament, Committee on the Environment, Public Health and Consumer Protection, Report on "The USA's Refusal to comply with EC Legislation on Slaughterhouses and Hormones and the Consequences of this Refusal", EP 128 381/B, 7 February 1989, named after its reporter Mr. Collins, MEP. See para. 2.32.
47EC Official Journal C 302, 9 November 1993, p.8 and C 222, 10.8.1994, p.16.
48EC Official Journal C 128, 9 May 1994, p.105.
49EC Official Journal C 52, 19 February 1994, p.30.
WT/DS26/R/USA
Page 27
the appropriate conditions as growth promoters in farm animals". 50 The United States added that there was absolutely no new scientific evidence considered by the Commission prior to the change in the proposed Directive to ban the use of these hormones. Indeed, in response to the 1995 EC Scientific Conference Proceedings, EC Agriculture and Rural Affairs Commissioner Fischler had confirmed that these hormones did not pose a danger to health when used in beef production. However, rather than deciding to modify the ban to reflect the scientific evidence, Mr. Fischler had indicated that this was a political matter and had noted that a 10 million ECU advertising campaign extolling the virtues of eating hormone-free meat had been launched and that the EC Council would be pressed to introduce tougher monitoring measures. The United States concluded that, in the absence of any identifiable risk from these hormones and without any supporting scientific evidence for the ban, the EC measures were not "applied only to the extent necessary to protect human, animal or plant life or health" and were not "based on scientificprinciples", contrary to Article 2.2. Indeed, the EC approach was contrary to scientific principles.
The European Communities responded that Members were not obliged to demonstrate a scientifically confirmed adverse effect from a particular hazard before they might take measures. The SPS Agreement could not have been intended to operate in such a way that Members must wait until people were actually sick or dying before being allowed to take measures. For example, it had been recently reported that scientists only now had discovered the exact mechanism by which smoking could cause cancer. But governments all over the world had been taking measures to prevent or reduce smoking, even in the absence of such clear scientific evidence.51 The European Communities added that the closest the SPS Agreement came to defining sufficient "scientific evidence" was in the footnote to Article 3.3. It followed from that definition in that scientific justification required an examination and evaluation of available scientific information, based on scientific principles. But at the end it was still the prerogative of the Member in question to decide whether the international standard, guideline or recommendation was sufficient to achieve its appropriate level of sanitary protection. The level of protection was decided by the Member alone and it was not a judgment that must be based on scientific principles or scientific evidence. Moreover, it was incorrect to state, as the United States did, that the choice of a measure was a purely scientific judgement. Measures must be based on scientific principles and not maintained without scientific evidence (conditions which the EC measures clearly meet) but these conditions could be met by a number of measures which could all achieve the same level of protection. If a Member was faced with choosing between several equally effective and scientifically justified measures, the SPS Agreement laid down other criteria, i.e. degree of trade- restrictiveness and technical and economic feasibility. This was the choice faced by the European Communities in the present case; either a ban on use for growth promotion or the imposition of strict veterinary controls. An assessment of trade-restrictiveness and technical and economic feasibility clearly showed that a ban on the use of hormones for growth promotion was the only measure available to achieve its chosen level of protection.
The European Communities indicated that Mr. Fischler had been much more cautious on the possible risks to human and animal health arising from the nature and improper use of these hormones for animal growth promotion than suggested by the United States.
The European Communities claimed that the question was whether the scientific evidence on which the contested measures were based was "sufficient" in the sense of Article 2.2. The European Communities observed that the term "sufficient" was nowhere defined in the SPS Agreement and noted that it was generally agreed that "sufficient" could not mean other than the "minimal" level of scientific
50COM(84)295 final, clause 13 of the preamble, 29 June 1984, Official Journal C 170, p.4 seq.
51The European Communities noted that, for example, the GATT panel on Thailand - Restrictions on Importation of and Internal Taxes on Cigarettes (1991) had accepted that "smoking constitutes a serious risk to human health" and, therefore, fell within the scope of Article XX:B of GATT 47.
WT/DS26/R/USA
Page 28
evidence required.52 None of the publicly available scientific reports established beyond doubt that the use of these hormones in animal production was safe for human health. The European Communities also claimed that as regarded the three so-called "natural" hormones, all of the scientific reports suggested that they were unlikely to pose a hazard to human health if used in accordance with good veterinary or animal husbandry practice. The scientific evidence for the necessity to maintain its measure was the evidence from the 1995 EC Scientific Conference that there was still a lack of information on, for example, the mode of action, the effect of administering combinations and the effect of ingesting residues over a long period. For example, the United States, in the Statement of Administrative Action, had chosen the so-called Delaney Clause to explain the differences between the basic concepts of the level of protection that a government chooses and the measure that the government used to achieve that level of protection:
"[T]he Delaney Clauses, in the first instance, establish a level of protection. They reflect a decision by the Congress that there should be no risk of cancer to humans from the substances those clauses cover .... A determination that a particular food additive poses a health risk is made on scientific grounds... Importantly, "risk assessment" as used in the SPS Agreement is not limited to quantitative risk assessment, which is a particular type of risk assessment used to evaluate the potential for carcinogenesis. The Delaney Clauses are entirely consistent with the Agreement's requirements in this regard. The determination that a particular substance poses a risk of cancer is a scientific determination, based on an evaluation of the potential for a substance to induce cancer. Based on scientific principles, the United States has determined that if a substance induces cancer in animals, it poses some risk of human carcinogenesis . And since the level of protection under Delaney requires that there be zero risk of carcinogenesis, the United States prohibit the substance" (pages 94-95).
The United States replied that conclusions of any scientific review could not be absolute because science was not absolute. Experimental results might either disprove or lend support to a particular hypothesis, but never prove it, and certainly would never prove it "beyond doubt". Yet the European Communities appeared to claim for itself the right to maintain their ban because scientific reports had not "established beyond doubt" that the use of these hormones for growth promotion was safe. Science could be used to determine whether there was a risk associated with the use of a particular substance; it could not eliminate the possibility that a potential risk might be found in the future. The SPS Agreement required a Member to base its measures on a risk assessment and prohibited a Member from maintaining SPS measures without sufficient scientific evidence. The Agreement did not provide that measures might be maintained without scientific evidence of a risk until such time as science proved "beyond doubt" that there was no risk. Moreover the United States noted that the lack of knowledge could not itself be the basis for taking a sanitary measure. The SPS Agreement required the European Communities to demonstrate scientific evidence of a particular risk. Scientists did not claim to know everything about everything. Scientific knowledge was always progressing and evolving. Accordingly, to claim that a Member was justified in banning an activity wherever there were still areas for science to explore would be to render the SPS Agreement meaningless - there were always areas for science to explore. Furthermore, in the US view, scientists knew more about the nature and mode of action of hormones than they did about most, if not all, other classes of compounds.
The European Communities claimed that there were three possible methods by which it could achieve its chosen level of protection: by the imposition of controls on hormone use to prevent misuse, by a ban on hormone use, and by a combination of controls and a ban. After a careful and extensive evaluation of the risks to human health and the technical and economic feasibility of each, the European Communities had chosen the third option, for the reasons explained in paragraphs 4.127 to 4.201. The European Communities also argued that the US claim that "it is important to distinguish
52D.A. Wirth (1994) "The role of Science in the Uruguay Round and the NAFTA Trade Disciplines", 27 Cornell International Law Journal, 817-859, p.856.
WT/DS26/R/USA
Page 29
between having a risk assessment and basing a measure on that risk assessment" revealed nothing. The European Communities based its measures on the risk assessment it conducted for that purpose. Neither the 1982 EC Expert Report nor the 1988 Lamming Committee Report, nor the 1988 JECFA Report constituted in themselves "risk assessment" in the sense of Articles 5.1 to 5.6 of the SPS Agreement; they only formed part of "available scientific evidence." The other factors mentioned in Article 5.2 and in paragraphs 3 to 6 thereof, were not dealt with by the scientific reports on which the United States was basing its claim. Indeed, the assessment of these factors was not a scientific question in the strict sense, and thus they fell within the responsibility of the appropriate political authorities of each Member.
The European Communities further argued that the United States had failed to appreciate the difference between the risk assessment required by the SPS Agreement when adopting measures, and the hazards against which members set a level of protection. Paragraphs 4.127 to 4.201 examined the hazards from hormones which were a source of risk to human health, irrespective of the way in which they were used. The European Communities was not obliged by the SPS Agreement to set different levels of protection against hormones used for growth promotion and hormones used for other purposes; on the contrary, one of the few requirements in the SPS Agreement in respect of the level of protection was that it be consistent (Article 5.5). The SPS Agreement imposed more, and stricter, disciplines on measures than it did on the level of protection. This was because measures were capable of being judged against objective universal criteria, while the level of protection was not. The choice of the level of protection might legitimately be influenced by factors such as political and financial priorities. As the United States had said in its Statement of Administrative Action, "it is a societal value judgment". However, once a level of protection had been decided, the measures used to achieve it could be analyzed scientifically. The European Communities claimed that the point was not that science did not know everything about hormones. The point was that it knew a lot, including the fact that they were carcinogenic. The problem was that science did know exactly how, and under what circumstances, this carcinogenic effect occurred. This was why the European Communities took a precautionary approach.
The United States indicated that in the United States, as in other countries, hormones were regulated as animal drugs when they were intended for use in animals for either therapeuticor production purposes. The use of hormones for growth promotion purposes had been approved by more than 20 countries, including the United States, Canada, Australia, New Zealand, Japan, Korea, the Philippines, South Africa, Mexico and most Latin American countries, and had been under intense scrutiny by scientists world-wide for over 15 years. The United States affirmed that no scientific review had ever concluded that there was a basis for banning the sale of animals to which one of these hormones had been administered in accordance with good animal husbandry practice or for banning the sale of meat from such animals. The reports which the European Communities had relied on, including the 1983 OIE Scientific Report and the 1988 JECFA report, as well as its internal reports, had all concluded that these hormones were safe when used for growth promotion purposes in accordance with good animal husbandry practice. The fact that the scientists, in their study of these hormones, had also reviewed the data on the effects of hormones at high doses or under other methods of administration did not constitute evidence that the approved uses for animal growth promotion purposes were unsafe. The United States contended that the issue was not whether any hormone was safe for all conceivable purposes at any conceivable level of concentration. The issue was whether the EC ban on the importation of animals, and meat derived from such animals, that had been administered the hormones in accordance with good animal husbandry practice, was supported by sufficient scientific evidence. The European Communities had offered absolutely no evidence that such use of these hormones presented any risk to human health. The United States noted that the experts advising the Panel had confirmed that there was no scientific evidence to support the EC ban.
The United States pointed out the Lamming Committee report for the three natural hormones had found that: "Thus no questions of safety arise in relation to the proper use of oestradiol-17 ,
WT/DS26/R/USA
Page 30
progesterone and testosterone in an appropriate form of preparation" (emphasis added, page 7). In this regard, Part B of the Pimenta Report stated on page 6 that:
"The present scientific evidence indicates that there is no discernable health risk in the use of the three endogenous hormones where these are, in particular, administered to castrated steers (whose hormone levels are therefore considerably lower than normal) and where they are administered in the form of slow-release implants in a part of the animal which is not consumed (e.g. the ear) and where a specific period with no further administration of hormones is observed before slaughter."
As indicated in paragraph 2.33, the scientific experts and other interested parties which the European Communities had called together for the 1995 EC Scientific Conference had concluded that there was no evidence of possible health risks from the use of the natural hormones for growth promoting purposes.
The 1995 EC Scientific Conference had furthermore concluded that:
"At the doses needed for growth promotion, residue levels [of trenbolone and zeranol] are well below the levels regarded as safe (the MRLs). There are, at present, no indications of a possible human health risk from the low levels of covalently-bound residues of trenbolone."
The United States claimed that these conclusions were consistent with other scientific reviews of the hormones by scientists around the world, including the reviews conducted for the Codex Commission.
The United States observed that the concept of the "appropriate level of sanitary or phytosanitary protection" or "acceptable level of risk" was one of the key concepts of the SPS Agreement. The level of protection or risk related to the chance or probability of an adverse health effect resulting from a particular activity. For example, a level of protection or risk might be expressed as a risk of 1 in 1 million, meaning that out of every 1 million humans in the relevant population, one person could be expected to suffer the adverse health effect from engaging in a particular activity. The appropriate level of protection or acceptable level of risk was the maximum level of risk that was considered tolerable. The United States repeated that the choice of what level of protection was appropriate was a social value judgement, not a scientific determination. There was no scientific basis for choosing one level of protection over another. For example, science could not say whether a risk of one in one million of kidney failure from a particular activity was acceptable, or whether a risk of one in 1,000 or one in 10 million should be acceptable. The level of protection was the goal the Member sought to achieve in terms of health protection. Sanitary and phytosanitary measures were the means to achieve that goal.
The European Communities claimed that the terms "appropriate level of protection" and "acceptablelevel of risk" were alternative ways of expressing the concept but they were not synonymous. In the case in question, the European Communities had a level of protection but no acceptable level of risk, because it did not accept any level of residues of added hormones in meat. It rejected the US suggestion that level of protection equalled level of risk; in the EC view, it did not. The level of protection was not the "chance or probability" of an adverse effect. It was particularly incorrect to say that the appropriate level of protection was the maximum levelof risk that was considered acceptable. The European Communities did not "accept" any risk from adding carcinogens to food for the sake of a little extra profit for farmers and drug companies. The European Communities was aware, of course, that despite its measures there might be occasions when such contaminated meat slipped through its detection system, but that did not alter the fact that its objective, in setting its level of protection, was to prevent any such residues. The "risk" (i.e. the "potential for adverse effects on human and animal health") from the use of hormones had been explained in detail by the European Communities, especially in paragraphs 4.127 to 4.201. The 1988 JECFA, on which the United States was arguably basing its contention, had also found that there was a potential risk to human health: if there were
WT/DS26/R/USA
Page 31
no such potential risk, the JECFA would have recommended no ADI and MRL for the two synthetic hormones. For the three natural hormones, JECFA had not recommend ADIs and MRLs because of problems in detecting the level of residues in meat. The European Communities further claimed that an ADI or an MRL recommended by Codex could never be considered to be a measure for the purposes of the SPS Agreement. Thus, science might help in setting a standard, guideline or recommendation (i.e. ADI or MRL) designed to exclude theprobability that an individual would develop cancer after a lifetime of exposure to a particular chemical substance, or to limit that probability to no more than one chance in a million. The choice of where to set the probability, for example the one-in-a-million as opposed to one-in-a-thousand or zero chance, was one of public policy and belonged to the competent democratic authorities of the Members. The level of protection was subject only to Articles 5.4 and 5.5, not to the whole of Article 5. Contrary to the US view, the European Communities argued that a sanitary measure was not a measure to "limit the risk"; it was a measure to protect human or animal life or health, etc. In the case in question, the EC measure was aimed at eliminating the potential for adverse effects by preventing the occurrence of residues in meat.
The United States claimed that the EC reliance on its supposed appropriate level of protection as a justification for their ban was misplaced. First, the European Communities had never articulated what their appropriate level of protection was in respect of the six hormones. The requirement of "no residues", which the European Communities claimed to be its level of protection, was a measure (essentially a measure that set a maximum residue limit, or "MRL", of zero) not a level of protection. Moreover, this was a measure which in fact was not maintained by the European Communities. The European Communities permitted residues of these hormones in meat. First, all meat had residues of the natural hormones ("residues" included amounts in the meat resulting from the animal’s natural production of hormones). Second, the EC derogation for herd management and other purposes meant that meat would also have residues of administered hormones. The United States observed that the experts advising the Panel had confirmed that there was no scientific justification, where the concern was with residues of a substance, for only banning one use of the substance and not regulating residues. The United States argued that the EC statement that the aim of the EC Directives was to protect human and animal health by "seeking a level of protection which requires the presence of no residues in meat" suggested that the EC purported "level of protection" was really a way to justify their total ban. In the US view, aside from the fact that it was physically impossible to require the presence of no residues of hormones in meat, this was a case of the measure driving the level of protection, rather than, as it should be, the other way around.
The United States submitted that a level of protection with respect to animal drug residues in meat might be "no risk of cancer in humans". If that were the EC appropriate level of protection, then the European Communities could assess whether there were any risks presented by hormone residues in light of this level of protection. But that was not the EC appropriate level of protection. The European Communities studiously avoided naming a level of protection (naming instead the measure it had established), perhaps because naming a level of protection made clear that the EC ban was unjustified. Indeed, there was no scientific evidence that there was a risk of cancer to humans consuming meat from animals to which any of the six hormones had been administered for growth promotion purposes in accordance with good animal husbandry practice. Quite the opposite, there was scientific evidence that the levels of residue under discussion did not pose a risk of cancer in humans. In other words, there was no appropriate level of protection that would justify the EC ban. A "zero risk" level of protection was the most stringent level of protection possible. Since the banned animals and meat already achieved this level of protection, then the ban could not be justified on the basis of the EC appropriate level of protection.
WT/DS26/R/USA
Page 32
The United States further argued that the EC own guidelines for the control of residues of oestradiol and testosterone were inconsistent with the EC position in this dispute.53 For oestrogen, the EC guideline was 40 picograms ("pg")/ml (40 parts per trillion); for testosterone, it was 10 nanograms ("ng")/ml (10 parts per billion) in male cattle younger than 6 months and 30 ng/ml in males 6-18 months; in non-pregnant females it was 0.5 ng/ml (no guideline level had been set for progesterone.) Many animals that had not been administered any of the six hormones for growth promotion purposes violated the EC guidelines. For example, recent data were available to suggest that the very means of slaughter might cause increases in these hormones above these identified limits. At the same time, levels of hormones in animals to which any of the six hormones had been administered for growth promotion purposes fell below the guideline levels set by the European Communities. This was because, as had been confirmed by the experts advising the Panel, the residues in meat of the three natural hormones, when administered for growth promotion purposes in accordance with good animal husbandry practices, were within the normal physiological levels.
The United States noted that this was in part related to the fact that producers in EC member States raised bulls (non-castrated adult male cattle) to maximize the growth of animals, whereas in the United States steers (castrated adult male cattle) were predominantly raised and hormones were administered to obtain more efficient growth. Bulls naturally had far higher levels of hormones than steers, including steers to which hormones had been administered for growth promotion purposes. Similarly, in the European Communities there were quite a number of pregnant animals entering the normal slaughter process.54 Like bulls, pregnant animals also had far higher than average levels of hormones. In other words, the agricultural practices in the European Communities, in light of the ban, resulted in consumers being routinely exposed to far higher levels of natural hormones in their meat of EC origin than meat of other origins. The selective EC ban was not justified even under their own professed approach. If one were to accept the EC claim that any residue of these hormones would pose a risk of cancer, then the European Communities would be required to prohibit the use of the hormones for herd management and other purposes. However, the European Communities permitted these uses. The United States further noted that meat could have residues identical to those of zeranol, even though zeranol had not been administered to the animal, if the animal had eaten feed contaminated with the common mould Fusarium.
The European Communities claimed that there was no obligation to follow scientific principles or to use risk assessment in setting the level of protection. As the United States had said in the Statement of Administrative Action: "The SPS Agreement thus explicitly affirms the right of each government to choose its levels of protection, including a "zero risk" level if it so chooses. A government may establish its levels of protection by any means available under its law, including by referendum." It was, of course, implicit that in order to need a level of protection there must be some hazard against which a Member needed to protect. However, this only implied the identification of a hazard, not an assessment of the probability that it would cause damage. Many different kinds of biological, chemical or physical hazards might occur in foods, and governments had to decide to what degree they aimed to protect their populations from these hazards. These decisions were taken, as were any political decisions, in the light of a number of factors including the potential danger to health and the cost and feasibility of achieving effective protection. For example, in the case of a substance which was fatally poisonous, or which caused a very serious or fatal condition such as cancer, governments would generally
53"Residues in food producing animals and their products; Reference materials and methods", Final Report, Directorate- General for Agriculture, Commission of the European Communities, 1992. Discussed at p.552 of the 1995 EC Scientific Conference Proceedings.
54B. Hoffmann "Problems of Residues and Health Risks of Anabolic Agents with Sex Hormone-like Activities", 1995 EC Scientific Conference Proceedings, p.284. Note also the conclusion on that same page that: "Thus, due to the lack of formation of extra residues, in principle no risk for public health can be seen following the presently recommend use of endogenous sex hormones as anabolic agents, at least in cattle".
WT/DS26/R/USA
Page 33
set a very high level of protection, i.e. they aimed to avoid the presence of such a substance in food altogether. However, in the case of a lesser hazard, such as an organism which caused an uncomfortable, but transient and non life-threatening effect, governments might set a level of protection which aimed to minimize, but not necessarily eliminate, the presence of the organism in food. There would obviously be differences between governments in their approach to setting a level of protection, as a function of their economic priorities and cultural habits. For example, a developed country might consider it desirable to set a high level of protection against contamination of foods by waste material from the chemical industry, whereas a developing country might consider it a higher priority to encourage the establishment of a chemical industry rather than, for the present, worry about some chemical residues in food. This was particularly the case for countries where food supply was more of a problem than food quality. Similarly, cultural habits such as the eating of certain foods in a raw state might cause a government to set a higher level of protection against some pathogens than would be the case for a government whose population cooked the same food thoroughly before eating it. For economic and cultural reasons, the decision on where to set the level of protection was not a purely scientific one, and it was not one which could be subject to international rules, given the great diversity among the Members of the WTO.
The European Communities pointed out that, contrary to the US allegations, it had stated clearly, on several occasions, that it aimed to avoid the presence of residues of added hormones in meat. The reference to "zero risk" was misleading. The EC interpretation of "zero risk" in this context was that it was not prepared to accept any residues of added hormones for growth promotion.
The European Communities further clarified that it had argued that the Codex recommendation was a level of protection as opposed to a measure. However, it was by no means clear that the MRLs were "standards", as they were adopted under a heading of "Standards and related texts"; they could better be regarded as guidelines or recommendations. It was interesting that now the United States did not refer to Codex MRLs as "measures" in the sense of the SPS Agreement.
The European Communities noted that the divergencies in the regulatory approach of Members regarding theuse of thehormones at issuedemonstrated that countries viewed and appreciated differently the potential risks to human and animal health resulting from the use of these hormones for growth promotion purposes. While the United States allowed the use of all six hormones, Canada allowed the use of all of these hormones except trenbolone55; Australia, New Zealand and South Africa allowed the use of all of these hormones except MGA56; Argentina allowed the use of zeranol and trenbolone, but not the other four hormones; and at least 25 meat-producing countries (other than the 15 EC member States) prohibited the use of these hormones for growth promotion.
The United States claimed that the issue in this dispute was the EC ban, not the measures of other Members. Differences among countries in the regulation of these hormones were not scientific evidence supporting the EC ban. As Australia and New Zealand had indicated, a hormone might not be registered due, for example, to a lack of interest associated with differences in production practices. Moreover, it could not be ruled out that the EC ban had influenced other countries to forgo approving the use of the hormones for growth promotion purposes in order to ensure access to the EC market for meat.
The European Communities replied that what was significant was that by virtue of the fact that they regulated hormones, all of these countries must consider that a risk existed. The differences in regulation reflected different opinions about the degree of risk. It was untrue that the EC rules had influenced other countries due to fears about a trade ban; there was not, and never had been, a ban
55Canada corrected the EC claim that Canada does not permit the use of trenbolone (see Section V, para. 5.12).
561995 EC Scientific Conference Proceedings, pp.597-598.
WT/DS26/R/USA
Page 34
on trade. It was entirely possible that other countries had carefully considered the scientific evidence and come to the same conclusion as the European Communities.
The United States recalled that Article 2.3 required Members to ensure that their sanitary and phytosanitary measures (a) did not arbitrarily or unjustifiably discriminate between Members where identical or similar conditions prevailed, including between their own territory and that of other Members; and (b) were not be applied in a manner which would constitute a disguised restriction on international trade. The United States argued that the EC ban arbitrarily or unjustifiably discriminated between Members where identical or similar conditions prevailed and discriminated against imported animals and meat, both compared to domestic products and to products imported from other Members. The EC ban discriminated between Members that permitted the use of the hormones for growth promotion purposes and those that prohibited their use or permitted the use of the natural hormones for therapeutic purposes (including the European Communities itself); the question for purposes of Article 2.3 was whether the discrimination was justifiable. The United States argued that there were no relevant differences in the conditions prevailing in the United States compared to the European Communities or other supplying Members with respect to these animals or meat and that there was no scientific or other basis for the European Communities' discrimination against US meat and animals. The EC ban was designed, in part, to remove any competitive advantage arising from the use of the hormones at issue for growth promotion purposes and to thereby protect domestic production. As the European Court of Justice had judged, "the possibility of a reduction in surpluses was indeed taken into consideration during the process leading to the adoption of the Directive".57 Furthermore, the EC Parliament, in calling for the European Communities to adopt its ban, had explicitly cited the fact that "there is overproduction of meat and meat products in the European Communities which adds considerably to the cost of the CAP".58 In March 1989, the draft of the Pimenta Report had also made it clear that control of beef supply was one of the true motives for this Directive: "The Inquiry Committee believes that only a total ban on the use of growth-promoters is concordant with the strategic aims now adopted for the Common Agricultural Policy, in particular the reduction of surpluses and thesafeguardingof aviable regionally-diversified farming community ”. Finally, Directive 88/146/EEC itself cited the desire to alter the conditions of competition in the meat sector.
The European Communities responded that European Court of Justice had pronounced itself twice on the aim of Directive 88/146/EEC. In the first case, the Court of Justice found that the aim of the Directive was to protect human health and consumer interests.59 This objective was pursued with a view to improving the quality of meat through regulating the conditions for the production and marketing of meat. The United States had totally disregarded this part of the Court's judgment and cited selectively from the other judgment of the Court.60 Furthermore, the Court ruling cited by the United States had concluded that "it does not follow that such a reduction, which is not cited in the preamble to the Directive as one of the objectives pursued, was in fact the exclusive or main purpose of the rules adopted". This made evident that the reduction in meat surpluses was no more than a possibility, and any Member was entitled to evaluate the side effects of an act it adopted for other
57"The Queen v. The Minister for Agriculture, Fisheries and Food and the Secretary of State for Health" (Case C-331/88), Judgment of the Court at p.I-4065.
58Resolution "closing the procedure for consultation of the European Parliament on the proposal from the Commission of the European Communities to the Council for a Directive amending Directive 81/602/EEC concerning the prohibition of certain substances having an hormonal action and of any substances having a thyrostatic action," para. I. (Doc. A2-100/85, 11 November 1985, EC Official Journal No. C 288, p.158 et seq.).
59Case 68/86, United Kingdom v. Council (1988) EC Reports p.855, p.897, para. 20.
60Case 331/88, Fedesa and others (1990) EC Reports p.I-4023, p.4065.
WT/DS26/R/USA
Page 35
purposes. Moreover, surpluses of meat had not existed in April 1996 when the European Communities had decided to maintain the prohibition by adopting Directive 96/22/EC. Furthermore, the evidence clearly showed that even the first EC Directive 85/649/EEC, which had been replaced by Directive 88/146/EEC, had not in fact contributed to reducing the small beef surpluses. Statistics showed that the European Communities had continued to produce domestically about the same quantities and had continued to import from third countries about the same quantities of meat as before the prohibition. The only difference, after the adoption of these Directives, was that the nature and composition of meat destined for human consumption in the European Communities changed (i.e. came from animals to which no hormones for growth promotion had been administered).
The European Communities also noted that there were several countries which did not allow the use of any or most of these hormones for animal growth promotion, but some of them did not impose any restrictions on imports of hormone-treated meat because they hardly imported any meat as their domestic production sufficed to cover demand. Argentina, for instance, which was a major beef producer, did not allow the use of the three natural hormones for growth promotion (because residues of these hormones could not easily be detected), but did not apply a prohibition on imports of animals or meat treated with these three hormones as there were virtually no imports taking place. In contrast, the European Communities had always imported large quantities of meat and had to adopt the measures in question in order to ensure that the objectives of its domestic sanitary policy were not circumvented through imports from third countries.
The European Communities argued that its measures offered equal opportunities of access to the EC market for all third-country animals and meat from animals to which no hormones had been administered for growth promotion purposes. Of the 31 countries which were authorized to export meat to the European Communities, only six apparently allowed the use of some or all of these hormones for growth promotion purposes. All of these 31 countries (including the United States) had continued to export to the European Communities animals and meat from animals to which no hormones had been administered for growth promotion purposes. Thus, overall the same competitive pressure as before was maintained from third-country imports on domestic meat production and US allegations to the contrary were unfounded. The intention of the EC measures at issue, therefore, was not to shield domestic meat production from foreign competition. The EC legislation did distinguish between countries which permitted the use of hormones for growth promotion and those which did not, but this was a justified distinction in view of the chosen EC level of protection. The argument that its measures were a disguised restriction on trade was unsubstantiated. The EC measure did not, therefore, contravene Article 2.3.
The United States noted that the European Communities admitted to distinguish between Members that permitted the use of the hormones for growth promotion and those that did not. The United States considered that the EC claim that this distinction could be justified by its chosen level of protection was unfounded. The European Communities had not demonstrated any scientific basis or other basis for the discrimination, and it was not justified by any health risks. None of the experts advising the Panel gave any response which would justify a determination that animals to which any of the six hormones had been administered for growth promotion purposes, and meat from such animals, should be considered "unlike" other animals or meat. The European Communities had admitted that residues of hormones resulting from implants could be impossible to detect, and scientists had concluded that such residues were meaningless given the variation in different meat sources and the levels of hormone production in humans. Moreover, the European Communities had not demonstrated that its chosen level of protection was any different from that of any of the Members who had approved the use of these hormones for growth promotion. More importantly, the right of a Member to select its own appropriate level of protection was not a justification for a breach of its obligations under Article 2.3. The United States concluded that the European Communities had failed to show how its asserted defense was even relevant. This discrimination was consequently not justified by any health risks, it was arbitrary and unjustifiable. Accordingly, the United States concluded that the EC ban
WT/DS26/R/USA
Page 36
unjustifiably discriminated between Members where identical or similar conditions prevailed and was inconsistent with the obligations of the European Communities under the first sentence of Article 2.3.
Furthermore, the United States contended, the ban constituted a disguised restriction on international trade, in breach of Article 2.3, as demonstrated by the record which showed that economic reasons related to reduction of meat surpluses and to altering the conditions of competition in the EC market were an important purpose of the ban. Moreover, the EC Directives required the EC member States to permit widespread use of hormones for purely economic reasons connected with efficient herd management. The EC measures depended on drawing a distinction between, for instance, domestic beef from cattle injected with hormones for herd-management purposes and imported beef from cattle implanted with hormones for growth promotion purposes. In this connection the United States noted that the doses of hormones used for therapeutic and zootechnical purposes were equivalent to, or in some cases higher, than the average daily dose used for growth promotion purposes. The United States stated that if low level residues of the hormones caused adverse health affects, as claimed by the European Communities, then therapeutic and zootechnical uses should be banned as well. These facts confirmed that the hormone ban was not a valid health-based measure, but an economic measure, and a means of protecting EC producers against imports. The EC measures were designed to protect domestic production. In sum, since the EC measures were not applied only to the extent necessary to protect human life or health, were not based on scientific principles, were maintained without sufficient scientific evidence, were not based on an assessment of the risks, and were more trade-restrictive than required to achieve the appropriate level of protection, it was apparent that the measures were not legitimate sanitary measures. Instead, they were measures designed to protect domestic production in the guise of sanitary measures. That was the essence of a disguised restriction on international trade.
The European Communities responded that, although the EC legislation permitted the use of the three natural hormones for therapeutic or zootechnical reasons, this was allowed under strict conditions. Thus, the three hormones might be administered, for therapeutic or zootechnical reasons,
(i) only by a veterinarian, (ii) only by injection or vaginal spiral (to the exclusion of implantation), and (iii) only to farm animals which had been clearly identified. Furthermore, such treatments must be registered by the veterinarian and these animals could not be slaughtered for meat production before a waiting period long enough to ensure that no residues were present in their meat. The doses of these hormones required for therapeutic or zootechnical purposes were several times lower than the doses required for effective growth promotion. In the case of animals which were at the end of their reproductive career, such treatments were prohibited during the fattening period following the end of their breeding life. All these conditions ensured that the hormones were administered properly and that no residues of hormones, other than those naturally produced by the animals themselves, were present in the meat destined for human consumption.
The European Communities noted that the 1988 JECFA Report had also recognized that a clear distinction should be drawn between the use of hormones for therapeutic or zootechnical reasons and animal growth promotion, indicating that "... residues left after the use of a drug for growth promotion should be considered separately from residues left after the use of that drug for other purposes".61
The European Communities claimed that it was misleading to state that there was widespread use of hormones for therapeutic or zootechnical purposes in the European Communities. Although the European Communities did not possess the exact figure of the number of animals treated for these purposes, of those EC member States which kept records62, it was estimated that only between 1 per cent and 2 per cent of breeding cattle in the European Communities were treated each year for such purposes. This percentage corresponded to about the same proportion of total bovine meat of EC origin
611988 JECFA Report, p.16.
62E.g. Denmark, the Netherlands and Finland.
WT/DS26/R/USA
Page 37
consumed in the European Communities, to which once in their breeding life such treatments may have been administered. Regardless of the insignificant quantities involved, what needed to be underlined was that allowing the administration of the three natural hormones for therapeutic or zootechnical reasons was in full compliance with the EC policy of ensuring no residues of hormones in meat for human consumption since the strict conditions imposed by EC law effectively ensured its policy objective (no residues at all). Furthermore, the European Communities pointed out that meat from animals which had, at some time in their life, been treated with hormones for therapeutic or zootechnical reasons, under the strict conditions laid down in the EC law, might be marketed for human consumption in the European Communities. Equally, animals treated for therapeutic or zootechnical reasons and meat of such animals were allowed to be imported from third countries under guarantees which were equivalent to those applied for domestic animals and meat from such animals. Therefore, this provision of EC law applied regardless of the country of origin of the animals or meat of such animals. The United States had not brought forward any evidence to show that animals and meat from animals treated in the United States for therapeutic or zootechnical purposes, if they complied with the conditions laid down in EC law, were not allowed to be imported into the European Communities. The European Communities further contended that "identical or similar conditions" did not prevail between the United States and the European Communities in respect of the use of hormones for growth promotion. Neither were the products similar or identical. A measure might be applied only to the extent necessary to protect life or health. If, for example, it could be demonstrated that pasteurizing milk at 72 C for 15 seconds was sufficient to kill tuberculosis, a Member would not be justified in requiring milk to be heated to a higher temperature for a longer time for the purpose of protecting against the same hazard. But in this case, the European Communities had demonstrated that if it had followed the US example in applying the Codex recommended MRLs in the way they were applied in the United States, it would have certainly failed to achieve its level of protection, which was no residues of these hormones in meat for human consumption and also to protect animal health, in view of the very limited number of tests carried out by the United States and the number of violations found even in this limited number of tested samples.
The United States noted that the EC claims that it had no solid information about the percentage of meat produced in the European Communities from animals to which hormones had been administered for herd management and other purposes was at odds with the European Communities' constant reference to the strict controls on such use and the need to identify every single animal to which hormones had been administered. Industry sources in the European Communities had indicated to the United States that 3.75 to 4 per cent of all cattle were treated each year based on an average of numbers in Belgium, France, Germany, the Netherlands and the United Kingdom. In addition, about 6 per cent of sheep in the European Communities were treated each year. In 1995, from one company alone, 2 million doses of hormones had been sold (for sheep) and ½ million doses were sold for cattle. Furthermore, the one guess that the European Communities offered was misleading. By stating that hormones were administered to 1 to 2 per cent of the EC herd each year, the European Communities failed to note that this figure should be multiplied by the number of years that the animals in the herd remained in production in order to determine the quantity of the meat supply from animals that, during their lifetime, had been administered these hormones. For example, if the hormones were administered to different animals each year, then over the course of 5 years, 5 to 10 per cent of the herd would have been administered these hormones. Alternatively, if the hormones were administered to the same animals, then an animal might have received five separate sets of treatment. Furthermore, there was no requirement that the hormones administered always be the same substances. Nothing prevented an animal in the European Communities from having been exposed to multiple hormones (or in the EC’s words, “combinations ” or “cocktails” of hormones).
The United States claimed that, contrary to the requirements of Article 3.1 to base its sanitary measures on international standards, guidelines or recommendations where they exist, the EC ban was
WT/DS26/R/USA
Page 38
not based on the relevant international standards. The relevant international standards in this respect were those of the Codex Commission. The Codex standards for , progesterone and testosterone in foods of bovine origin stated that there was no need to set any Acceptable Daily Intake level (ADI) or any Maximum Residue Limit (MRL).
Codex had based its decision not to set ADIs and MRLs for these three hormones on the basis of the 1988 JECFA Report63, which had found that residue levels in treated animals fell well within the normal range of levels found in untreated bovine animals of different types and ages. JECFA thus deemed it unnecessary to set an ADI for hormones that were produced endogenously in human beings and showed marked physiological variations in levels according to age and sex, and had concluded that residues arising from the use of , testosterone and progesterone for growth promotion purposes in accordance with good animal husbandry practice were unlikely to pose a hazard to human health. JECFA, in its assessment of the hormones, had defined what was meant by an MRL not specified. It meant that: "Available data on the identity and concentration of residues of the veterinary drug in animal tissues indicate a large margin of safety for consumption of residues in food when the drug is used according to good practice in the use of veterinary drugs. For that reason, and for the reasons stated in the individual evaluation, the Committee has concluded that the presence of drug residues in the named animal product does not present a health concern and that there is no need to specify a numerical MRL".64
With respect to trenbolone and zeranol, the United States noted that the Codex standards specified both an ADI and an MRL. The Codex definition of an MRL was as follows:
"... the maximum concentration of residue resulting from the use of a veterinary drug (expressed in mg/kg or µg/kg on a fresh weight basis) that is recommended by the Codex Alimentarius Commission to be legally permitted or recognized as acceptable in or on a food.
"It is based on the type and amount of residue considered to be without toxicological hazard for human health as expressed by the Acceptable Daily Intake (ADI), or on the basis of a temporary ADI that utilizes an additional safety factor. It also takes into account other relevant public health risks as well as food technological aspects.
"When establishing an MRL, consideration is also given to residues that occur in food of plant origin and/or the environment. Furthermore, the MRL may be reduced to be consistent with good practices in the use of veterinary drugs and to the extent that practical analytical methods are available."65
The Codex standard for trenbolone and zeranol was an MRL of 10 µg/kg of bovine liver and 2 µg/kg of bovine muscle.
The United States noted that as early as December 1987, the CCRVDF agreed on MRLs for trenbolone and zeranol, and agreed that no MRLs were necessary for the three endogenous hormones. At the June 1991 biennial meeting of the Codex Commission, four of the hormones (oestradiol, progesterone, testosterone, and zeranol) were at the end of the lengthy 8-stage approval process of which the decision by the CCRVDF was a key point. At that stage in the process, the science on a particular drug had been thoroughly reviewed, and Codex Commission members had been provided numerous chances to voice concerns about the scientific assessment of these hormones - either in writing
631988 JECFA Report, p.19.
64Fortieth report of the Joint FAO/WHO Expert Committee on Food Additives, pp.5-6.
65Codex Alimentarius, Vol. 3, Residue of Veterinary Drugs in Foods, Rome 1993, pp.65-66.
WT/DS26/R/USA
Page 39
to the Codex Commission secretariat or at CCRVDF plenary meetings. So, the United States contended, the scientific consensus on the four hormones was clear by the time they came before the Codex Commission. However, the issue continued to become politicized. The United States stressed that in 1994, the Codex Alimentarius Executive Committee (which includes a member from the European Communities) developed four principles clarifying, inter alia, that "[t]he food standards, guidelines and other recommendations of the Codex Alimentarius Commission shall be based on the principle of sound scientific analysis and evidence, involving a thorough review of all relevant information, in order that the standards assure the quality and safety of the food supply." Despite the EC opposition, these principles were adopted by the Codex Commission in June 1995. This paved the way for the Codex Commission to finally adopt standards for the five hormonal substances (by this time, trenbolone was also at the Step 8 of the Codex Commission procedures) recommended by the CCRVDF. Notwithstanding the adoption of the standards, the United States noted that the European Communities had maintained its ban and, in fact was expanding its ban for example by expanding it to new species.
The European Communities argued that at the time it had adopted and applied Directive 88/146/EC (1 January 1989), there existed no Codex standards on the hormones. Moreover, the Codex had never studied scientifically the hormone MGA and had never adopted any standard on its use. Furthermore, the European Communities argued that the Codex standards on the five hormones at issue had been adopted by a majority of only 33 votes in favour, 29 votes against and 7 abstentions (i.e. a minority of those participating), that was a very close vote in favour of the adoption of the standards. There were, therefore, at least 14 countries other than the 15 EC member States which had voted against the standard. This close vote clearly indicated that the issue of hormones had been, and continued to be, very controversial both from the scientific and the regulatory policy point of view. The European Communities argued that, despite the fact that the United States knew of the strong and wide opposition to the adoption of these recommendations on hormones for growth promotion purposes throughout the preparation of the Codex Commission's decision of July 1995, it had still insisted on and pressed, the issue for economic reasons, and despite the long established practice of the Codex Commission to adopt decisions on MRLs by consensus. In fact, the unusual procedure of voting by secret ballot had been requested in this case by the United States.66 The European Communities argued that even today, no Codex member had notified Codex of its acceptance of the standards adopted in July 1995 on the hormones.67 The United States itself seemed to support Codex standards, guidelines and recommendations on a very selective basis, only when they favoured their strict economic and trade interests.68
The European Communities added that Article 3.1 made it plainly clear that there was no absolute obligation on Members to always follow standards on SPS measures adopted by Codex. There was no doubt that the two systems of the Codex and the SPS Agreement did not interact properly, because a member of Codex, which had different views about other considerations (e.g. health concerns of consumers) and in good faith abstained from blocking the adoption of a Codex standard knowing in advance that in doing so it would not be required to follow the standard whose adoption it did not block, would later find itself to have an obligation to follow under the SPS Agreement. This was an inherent contradiction in the functioning of the two systems, of which both Codex and WTO Members were well aware and efforts were now being made to resolve it in an appropriate way.
66Report of the 21st Session of the Codex Alimentarius Commission, Rome, 3-8 July 1995, Alinorm 95/37, paras 45-46.
67In the meantime, South Africa notified acceptance of the MRLs for veterinary drugs (Codex response to written questions raised by the Panel.).
68The European Communities noted that a Codex press release (Codex Facts, 05-95) had explained that "[t]he countries which voted in favour of the hormone MRL are primarily meat-producing/ exporting countries where the cost of production is either not subsidised or is subsidised to a lesser extent than in the EU. Production efficiencies, therefore, exercised a major influence on the Codex Commission. The animal health industry (commercial producers of the hormones) was also pleased with the Commission's decisions for obvious reasons".
WT/DS26/R/USA
Page 40
The European Communities further observed that JECFA could not propose an ADI for natural hormones because when it had considered the matter, in the early and mid-1980s, the technology for measuring increases in levels of the natural hormones was not sufficiently advanced to be appropriate for routine use. Noting that the United States itself set limits for levels of the natural hormones in meat above those naturally present, but did not check annually for the presence of residues from these hormones, the European Communities further argued that the only "international standard, guideline or recommendation" relevant to the EC measure was the "Codex Code of Practice for Control of the Use of Veterinary Drugs". The MRL laid down by Codex was a level of protection, not a measure, and there was no obligation in the SPS Agreement to adopt Codex recommended levels of protection. The European Communities recalled its earlier arguments that neither the Panel nor any other Member, nor a Codex Alimentarius group of scientific experts, could judge the adequacy of the scientific evidence used by a Member. In view of the potential hazards to human health presented by the misuse of hormones in food animals, the European Communities considered that they should only be administered for the purposes and under the conditions defined in Directive 96/22/EC, which were in accordance with the "Codex Code of Practice for Control of the Use of Veterinary Drugs".
With respect to the synthetic hormones, the European Communities observed that JECFA (on whose scientific advice Codex had based its recommended MRLs) had established an ADI for zeranol of 0 - 0.5 µg/kg body weight, and for trenbolone of 0 - 0.02 µg/kg body weight. The European Communities was therefore justified, in respect of zeranol and trenbolone, in accepting the lower limit,
i.e. an ADI of zero. Codex had made no recommendation for MGA. The European Communities argued that it was contrary to common logic to claim (as it was argued by the United States) that an MRL of 2 mg/kg of bovine meat for zeranol was a "measure" in the sense of the SPS Agreement. Scientists had defined the MRL as the quantity of a particular residue in meat that would not exceed the ADI for the particular substance concerned, if its administration was done in accordance with the required practice. The decision where to set the level, i.e. at 2 mg/kg or at 10 mg/kg of bovine meat, was a decision on the level of protection, not a "measure" as defined in Annex A of the SPS Agreement.69 Moreover, the European Communities claimed that as there was no international standard relevant to its measure, there was no contravention of Article 3:1. Codex had not established standards for five of the hormones at issue; all it had done was to recommend maximum residue levels for the two synthetic hormones and decided that it was too difficult to do so for the three natural ones. Maximum residue levels were only of relevance to Members which accepted residues of these hormones in meat. The European Communities did not. Furthermore, as six of the scientists advising the European Communities had pointed out in their written opinions to the Panel70, the JECFA Report on the basis of selected studies in laboratory animals had "concluded" that the carcinogenic effect of progesterone (and oestradiol and testosterone) was linked in some way to its hormonal effect, so that if it was present in a concentration insufficient to produce a hormonal effect in animals it would not be carcinogenic; they had then extrapolated this assumption to humans. But the JECFA Report had no scientific basis for making this assumption; contrary to their own stated intent (section 2.1.2 of the 1988 JECFA Report) they had not known (and a decade later it was still not well known) how these hormones produced
69The definition of SPS measures in Annex A of the SPS Agreement specifies that "(...) [s]anitary or phytosanitary measures include all relevant laws, decrees, regulations, requirements and procedures including inter alia end product criteria; processes and production methods; inspection, certification and approval procedures; quarantine treatments including relevant requirements associated with the transport of animals or plants, or with the materials necessary for their survival during transport; provisions on relevant statistical methods, sampling procedures and methods of risk assessment; and packaging and labelling requirements directly related to food safety."
70Dr. Liehr, Adlercreutz, Cavalieri, Letzler, Pinter and Epstein. The European Communities indicated that Dr. Liehr, for example, in his scientific advice on this issue had stated the following:
"In the 1988 JECFA Report, the authors considered only the hormonal receptor-mediated activities of the natural hormones. In view of the considerable amount of scientific evidence which has accumulated since the release of that Report, particularly in respect of the genotoxicity of oestradiol, the Report can no longer be considered applicable to a risk assessment of the use of hormone growth promoters."
WT/DS26/R/USA
Page 41
their effects. Nevertheless, on the basis of these unjustified extrapolations and assumptions, and because detection was too difficult, they had "deemed" it unnecessary to set an ADI.
The United States claimed thatthe European Communities appeared to misunderstand Article 3 and confused a level of protection with a measure when it claimed that the MRLs established by Codex were a "level of protection" rather than a measure. The United States submitted that an MRL was a measure, not a level of protection. If a government wished to ensure, for example, that its public was not exposed to a risk of more than 1 in 10 million of nerve damage from lead in food, the government could establish an MRL to limit total exposure. However, the MRL was not the level of "protection", it was the maximum limit of residue. What level of protection it achieved would depend in large part on where the MRL was set. In other words, the level of protection determined where to set the MRL.
The United States rejected the EC argument that Article 3.1 did not apply to its ban because Codex did not have a standard for the use of these hormones, only for their residues. The United States argued that a requirement to base a measure on the relevant international standard includeda requirement to base the form of the measure on the international standard. Otherwise, it would be easy to evade the requirements of Article 3.1 by deliberately adopting a measure different in form from the international standard. Under that approach, as soon as a Member had chosen to adopt a measure different from the international standard, there would no longer be any requirement to base its measure on the international standard. In other words, Article 3.1 would cease to apply as soon as a Member acted inconsistently with it. The United States added that to claim, as the European Communities did, that since the Codex ADI was a range from zero to 0.5 µg/kg body weight, setting an ADI of zero was “based on” the Codex standard, was a distortion of the international standard. The ADI was expressed as a range because it referred to the "acceptable" daily intake. In that regard, an intake of zero of these residues was acceptable, but it was not the Codex standard.
The European Communities responded that international standards, guidelines or recommendations were based on a certain concept of level of sanitary protection. The SPS Agreement explicitly allowed Members not to follow the international standards, guidelines or recommendations in order to achieve their appropriate level of protection. The objective of Article 3.1 was to harmonize, as far as possible, the sanitary or phytosanitary measures of Members, given that a different level of protection might require different types of measures. An ADI or MRL recommended by Codex could therefore, never be considered to be a measure, because measures were adopted by Members. Science might help in setting a standard or recommendation (i.e. ADI or MRL) designed to exclude the probability that an individual would develop cancer after a lifetime of exposure to a particular chemical substance, or to limit that probability to no more than one chance in a million. But the choice where to set the probability, for example one-in-a-million as opposed to one-in-a-thousand or zero chance, was a choice of public policy, and belonged to the competent democratic authorities of the Members. It was a goal of public policy, not of scientific nature. The EC policy in case of potentially carcinogenic substances like these hormones aimed to avoid exposing consumers to them. Therefore, the European Communities would not consider a level of protection of 1 in one million to be acceptable. That was why its level of protection in this case aimed at reducing that probability to zero.
The European Communities argued that the "appropriate level" a Member decided to apply in its territory did not have to be expressed in the same technical fashion, i.e. as an MRL. There were usually several ways of dealing with any given hazard, and the SPS Agreement did not require Members to change the type of measures they chose. For example, one of the many ways of tackling pathogenic organisms in food was irradiation, and some Members allowed it to be used for certain foods. Codex had adopted standards for irradiation of foods but this did not mean that every Member was now obliged to allow irradiation of foods. Similarly, if a Member chose a level of protection against a contaminant on the basis of an MRL, this did not mean that it should set its MRL at the level recommended by Codex. Also, if a Member chose not to allow any residues of dangerous substances in food, this did
WT/DS26/R/USA
Page 42
not mean that it was obliged to base its protection on the concept of an MRL recommended by Codex, if there existed one for that substance. One of the weaknesses of the SPS Agreement was that a measure in conformity with a Codex standard or recommendation was deemed to be consistent with Article 3.2. Unfortunately, many Codex standards were quite out of date, having been adopted decades before the development of sophisticated analytical methods.
The European Communities argued that Article 3.3 permitted Members to "... introduce or maintain sanitary or phytosanitary measures which result in a higher level of sanitary or phytosanitary protection than would be achieved by measures based on the relevant international standards, guidelines or recommendations, if there is a scientific justification, or as a consequence of the level of sanitary or phytosanitary protection a Member determines to be appropriate ..." The Codex recommendations were designed to achieve a level of protection which was lower than that applied in the European Communities for the hormones at issue. Moreover, Codex and the other relevant international organisations mentioned in the SPS Agreement might issue standards, guidelines or recommendations, but they were relevant only for those Members which chose to follow them and base their measures on them. If a Member elected not to follow them because it had a different level of sanitary protection, the Member was entitled to take another type of measure necessary to achieve its chosen level of protection. In such a case the SPS Agreement only required the Member to respect the provisions of Articles 5.1 to 5.8 thereof.
The European Communities claimed that the "Codex Statements of Principle Concerning the Role of Science in the Codex Decision - Making Process and the Extent to which other Factors are taken into Account", approved on July 1995, explicitly recognized that other legitimate factors, relevant for the health protection of consumers, could be taken into account when elaborating and deciding food standards.71 In addition, Members which had different views about other considerations (e.g. health concerns of consumers) could abstain from accepting the relevant standards. The SPS Agreement permitted departure from international standards for the same reasons for which Codex standards, guidelines or recommendations were voluntary and could not be made legally binding, i.e., because each Member was free to decide its appropriate level of sanitary or phytosanitary protection. This was not a scientific judgment and scientific committees or expert groups could not replace the democratically elected authorities of Members.
The United States argued that the EC ban was not covered by the exceptions in Article 3.3 to the requirements of Article 3.1. The European Communities had not shown that its level of protection was different from that achieved by applying the Codex standards. In fact, the European Communities had not even identified its own level of sanitary protection. Noting that the SPS Agreement defined
71Alinorm 95/37, Appendix 2. The "Statements of Principle" is an annex to the report of the 21st Session of the Codex Alimentarius Commission. The "Statements of Principle" provide:
"1. The food standards, guidelines and other recommendations of Codex Alimentarius shall be based on the principle of sound scientific analysis and evidence, involving a thorough review of all relevant information, in order that the standards assure the quality and safety of the food supply.
"2. When elaborating and deciding upon food standards Codex Alimentarius will have regard, where appropriate, to other legitimate factors relevant for the health protection of consumers and for the promotion of fair practices in food trade.
"3. In this regard it is noted that food labelling plays an important role in furthering both of these objectives.
"4. When the situation arises that members of Codex agree on the necessary level of protection of public health but hold differing views about other considerations, members may abstain from acceptance of the relevant standard without necessarily preventing the decision by Codex."
WT/DS26/R/USA
Page 43
"appropriate level of sanitary or phytosanitary protection" as "[t]he level of protection deemed appropriate by the Member establishing a sanitary or phytosanitary measure to protect human, animal or plant life or health within its territory", the United States submitted that "level of protection" referred to protection from a particular sanitary or phytosanitary risk. For the level of protection to form the basis for taking action, there must first be identified some activity which gave rise to the identified risk, and the activity must present a level of such risk higher than the level deemed acceptable by the Member.
The United States submitted that the level of protection did not refer to protection from foreign competition, nor did it refer to protection from a particular substance or activity per se, only from the risk associated with that activity. Thus, for example, a Member might deem a chance of 1 in 1 million to be the appropriate level of protection from the risk of the establishment of a particular plant pest, or 1 in 10 million to be the acceptable level of risk of harm to nerves from the ingestion of lead in food. In contrast, the measures taken to achieve that level could for example, in the case of the plant pest risk, involve quarantine to ensure the pest was not harboured on imported plants or plant material, or fumigation. In the case of the nerve damage risk, the measures could consist of a maximum residue limit for lead in foods, or a prohibition on the use of lead in ceramic glazes for food utensils such as pitchers or glasses. Finally, the United States agreed that the choice of the appropriate level of protection was a societal value judgment. The SPS Agreement imposed no requirement to establish a scientific basis for the chosen level of protection because the choice was not a scientific judgment.
The European Communities responded that the United States' arguments were contradictory. It followed from these arguments that for the United States, the concept of "measures" for the purpose of the SPS Agreement included quarantine, fumigation or prohibition on the use of a substance. In criticizing the EC formulation of "no residue at all", the United States had also admitted that this was also a measure, not a level of protection. However, the United States had also argued that an MRL was a measure. If an MRL was a measure, then what was a prohibition laid down in law on the use of hormones for growth promotion?
The European Communities noted that there were two exceptions provided in Article 3.3. The first exception was fulfilled when the international standard was inadequate, faulty or obsolete from a scientific point of view. According to the second exception, a Member was in any case entitled to introduce or maintain measures which aimed at achieving its appropriate level of protection, to be determined in accordance with Article 5 of the SPS Agreement. The European Communities agreed that to invoke either of the options in Article 3.3, which were available to a Member when taking a sanitary or phytosanitary measure, there must exist "a potential risk for adverse effects". But the SPS Agreement left Members free to define the level of probability they wanted to assume: this might range "from zero to infinite" and it also left them free to decide the type of measure they might choose to ensure that the level of protection they considered to be appropriate was achieved. The European Communities further agreed that the choice of the appropriate level of protection was a social value judgement and that the SPS Agreement did not impose a requirement to establish a scientific basis for the chosen level.
The United States argued that both exceptions contained in Article 3.3 had the same effect, since both referred to a situation where the basis for departing from the relevant international standard was that the international standard was not sufficient to achieve the Member's appropriate level of protection. The concept of the appropriate level of protection interlaid a number of the provisions of the SPS Agreement. For example, Article 5.6 used the appropriate level of protection to determine if sanitary or phytosanitary measures were more trade-restrictive than required. Article 3.3 also used this concept in permitting a Member to introduce or maintain a sanitary or phytosanitary measure that was more stringent than the relevant international standard where the international standard would not be sufficient to achieve the Member’s appropriate level of protection. In the US view, none of the different EC formulations for the purported level of protection was appropriate for purposes of the SPS Agreement. For example, on the one hand, the European Communities had stated that its
WT/DS26/R/USA
Page 44
appropriate level of protection was "the assurance that meat will not contain residues of administered hormones". Elsewhere the European Communities had stated that its level of protection was the "guarantee that EC consumers did not ingest residues of added hormones in meat". In yet another place, the European Communities had stated that its level of protection was "no residue at all".
The United States argued that the first formulation, "the assurance that meat will not contain residues of administered hormones" was not a level of protection from sanitary risk. Prohibiting certain residues was a measure, not a level of protection. This formulation said nothing about whether those residues posed a risk. In fact, the United States could only assume that it was a misstatement since the European Communities permitted hormones to be administered for herd management and other purposes. Nor did this formulation say anything about how any residues from administered hormones compared to other residues, such as residues of hormones that were naturally occurring in meat. The second formulation, "guarantee that EC consumers did not ingest residues of added hormones in meat", suffered from the same deficiencies and misstatement as the first formulation. The third formulation, "no residue at all" was also a measure, not a level of protection. The United States also assumed that this so-called "level of protection" applied only to the synthetic hormones since meat would always have residues of the naturally occurring hormones. In fact, the EC measure was not a requirement that there be no residue of these substances. The EC measure was a ban on the use of the substances for growth promotion irrespective of any residue.
The European Communities contended that it was obvious that the EC level of protection in this case was no residue of added hormones in meat. This level resulted from the provisions of the EC Directives which prohibited the administration to farm animals, by any means whatsoever, of "substances having a thyrostatic action or substanceshaving an oestrogenic, androgenic or gestagenic action"; or the slaughter or sale of any animal to which the above-mentioned substances had been administered, or of the meat or meat products from such animals. The aim of these provisions was to protect human and animal health by seeking a level of protection which required the presence of no residues in meat. In contrast, the level of protection of Codex was the recommended MRLs. The European Communities noted furthermore that its measure applied not only to the six hormones in dispute, but to all hormones, substances and combinations thereof which exerted the above-mentioned action on all farm animals (i.e. not only on beef cattle). Conversely, the MRLs recommended by Codex did not exclude, when followed by a Member, that thyrostatic, oestrogenic, androgenic and gestagenic actions could be exerted on farm animals. Moreover, the EC level of protection was concerned only with added hormones, over and above those which occurred naturally; the reference to natural levels was irrelevant to this case. Its rule were strict and avoided the presence of residues in animals treated for therapeutic or zootechnical purposes. So the fact was that the EC rules did not permit residues of added hormones at all.
The United States claimed that the EC ban was not designed to, nor did it, achieve any particular level of protection. Meat naturally had widely varying residues of the endogenous hormones, and many foods had levels of residues which were orders of magnitude greater than that found in the banned meat (as shown in the following tables).
WT/DS26/R/USA
Page 45
Comparative androgen intakes
Food | Weight of portion (g) | Androgen intake (ng) |
Unimplanted bull meat | 500 | 1,560 |
Steer or female implanted w/ trenbolone1 | 500 | 135-150 |
Heifer implanted w/ testosterone | 500 | 35 |
1Assuming 25 per cent fat, 75 per cent muscle ng = nanograms
Comparative oestrogen intakes from food sources
Food | Weight of portion (g) | Oestrogen intake (ng) | |
Unimplanted steer meat1 | 500 | 61.1 | |
Oestradiol-Implanted steer meat1 | 500 | 11.4 | |
Zeranol-Implanted steer meat3 | 500 | 7* | |
Cow meat1,2 | 500 | 75 | (7.2-540)* |
Hen's egg | 50-60 | 1,750* | |
Cabbage | 100 | 2,400* | |
Peas | 100 | 400* | |
Wheat germ | 10 | 200* | |
Soybean oil | 10 ml | 20,000* | |
Milk | 500 ml | 75* |
E.g. 1 egg equivalent in oestrogen content to 76.5 kg of implanted steer beef
1Assuming 25 per cent fat, 75 per cent muscle
2Oestrone only
3Muscle tissue only
*Oestradiol equivalents ng = nanograms
The United States noted that the European Communities permitted the sale and consumption of meat without regard to the levels of endogenous hormones; did not regulate exposure to hormones found in any other foods, and finally, permitted the sale and consumption of meat from animals that had received the natural hormones for purposes other than growth promotion. If the supposed health problem was exposure to residues of any of these hormones, then the European Communities should regulate dietary exposure to residues of the natural hormones. For the same reasons, the European Communities could not claim that its appropriate level of protection was "zero", since if "zero risk" was intended to refer to the five categories of risk identified by the European Communities (discussed in paragraph 4.126 below), then achieving zero risk was impossible. For example, two of the categories referred to by the European Communities were risks arising from detection and control (or more accurately arising from difficulty in detecting the presence, or controlling the use, of these hormones) and risks arising from the administration and use of hormones (or more accurately from incorrect administration or misuse of these hormones). In both cases the activity of concern to the European Communities was already illegal in the European Communities and with respect to exports to the European Communities. As a result, there was no reason to believe that lifting the ban would mean that there would be a risk that was not now already present. The United States stated that it would also be interested in the EC explanation of how it could guarantee that there was a zero probability that illegal activity would occur. The United States argued that history (including the history of the European Communities’ own failure to prevent illegal use of hormones by EC producers) showed that it was impossible to guarantee that there was a zero probability that illegal activity would occur.
WT/DS26/R/USA
Page 46
The European Communities claimed that it did not insist on "zero risk"; it aimed to prevent any residues. The European Communities have never claimed to "guarantee" a "zero probability that illegal activity will not occur". There was always a possibility that any controls would be evaded but that did not mean that the controls should be abandoned. For example, there were measures in the United States which banned murder, and the sale of heroin, but apparently these measures were sometimes evaded. This did not mean that the measures should be dropped. The European Communities contended that the sixth preambular paragraph of the SPS Agreement made clear that it did not require Members to change their appropriate level of protection or to "downward" harmonize them to less stringent Codex sanitary and phytosanitary measures. Also Article 3.3 clearly indicated that a Member was not required to accept international standards that would result in a lower level of protection than the Member had determined to be appropriate. Under Article 3.2, if a sanitary or phytosanitary measure conformed to a relevant international standard, it was then deemed to be necessary to protect human health, and presumed to be consistent with the relevant provisions of the SPS Agreement and of GATT. However, as the USTR itself had stated: "... the fact that a sanitary or phytosanitary measure differs from a relevant international standard, guideline, or recommendation does not, in itself, create any adverse presumption concerning that measure".72
The European Communities noted that the approach adopted by the SPS Agreement was in conformity with previous GATT law and practice, and was a sensible approach for the purpose of establishing multilateral rules and disciplines to guide the development and progressive harmonisation in order to minimize the negative effects on trade from national sanitary and phytosanitary measures. This approach was also in conformity with democratic regulatory procedures, where frequently a dichotomy was operated in the decision making process between risk assessment, which established strictly the scientific basis for regulatory action and risk management which was the process by which the competent authority of a Member decided what action to take in the face of the assessment submitted to it by the scientists. Such action was based on factors such as public health and environmental protection, relevant legislation and legal precedent, application of social, economic and political values and consumer concerns. The European Communities argued that the risk managementphase, therefore, in a democratic legislative system, expressly recognized the importance of social value choices.
The European Communities argued that its interpretation of the disciplines the SPS Agreement imposed on Members was apparently also supported by the United States in the US Statement of Administrative Action, presenting to the US Congress the results of the Uruguay Round for approval, which stated as follows:
"The S&P Agreement thus explicitly affirms the right of each government to choose its levels of protection, including a "zero risk" level if it so chooses. A government may establish its levels of protection by any means available under its law, including by referendum. In the end, the choice of the appropriate level of protection is a societal value judgement. The Agreement imposes no requirement to establish a scientific basis for the chosen level of protection because the choice is not a scientific judgement"73 (emphasis added).
72US Statement of Administrative Action, p.92.
73US Statement of Administrative Action, point 3(b). The European Communities noted that this Statement used the example of the so-called "Delaney clause" to illustrate the point :
"The Delaney clauses are entirely consistent with the Agreement's requirement in this regard. The determination that a particular substance poses a risk of cancer is a scientific determination, based on an evaluation of the potential for a substance to induce cancer. Based on scientific principles, the United States has determined that if a substance induces cancer in animals, it poses some risk of human carcinogenesis. And since the level of protection under Delaney requires that there be zero risk of carcinogeneses, the United States prohibits the substance." (point 9).
WT/DS26/R/USA
Page 47 In another public document also submitted by the USTR to Congress it was stated that:
"... the requirement for a scientific basis applies to SPS measures; it does not apply to the level of food safety that those measures are designed to achieve. The SPS Agreement explicitly recognizes the right of governments to choose the level of food safety that they consider appropriate.
"... Moreover, by requiring that a measure be based on scientific principles (rather than, for instance, requiring that a measure be based on the "best" science or on the "weight of evidence") the Agreement recognizes the fact that scientific certainty is rare and many scientific determinations require a judgement among differing scientific views. The SPS Agreement preserves the ability of governments to make such judgments."
The European Communities argued that the table of foods such as potatoes and Lima beans was irrelevant. In the first place, consumers had always been exposed to a certain amount of hormones, or hormone-like substances in their diet; humans evolved alongside them and could cope with them. This was not the case for added hormones. Secondly, the figures in the table were fictitious, in the sense that they were examples and might bear little or no relation to the situation where growth hormones were used in practice. Thirdly, plant substances were not identical to the animal hormones (see in this regard the statement of one of the EC scientific experts, Dr. Liehr, in the transcripts of the meeting of 18 February 1997, paragraphs 757-758. The European Communities also claimed that the comparison which the United States cited in its table was incorrect because it compared meat from a treated heifer with meat from an untreated bull. The correct comparison (which had been omitted in the table) must be with an untreated heifer (not bull). This comparison could be found in the 1995 EC Scientific Conference Proceedings (page 278, table 3):
meat from untreated bull - 1 560 ng;
meat from untreated heifer - 8 ng;
meat from treated heifer - 35 ng.
The human diet normally contained a variety of meat; evidently if all the meat came from treated animals the amount of hormone residue ingested would be significantly increased (more than fourfold on the above comparison). If a proper comparison were to be made involving male animals, it would have to be between treated and untreated males to see how much the levels of female hormones were increased in male animals. According to the US argument, heifer meat would be "within the normal range" if it contained as much testosterone as bull meat.
The United States insisted that the US Statement of Administrative Action (SAA) did not support the EC ban. The US position in this dispute was fully consistent with the SAA. The United States explained that the SAA was a communication between the President and the Congress of the United States describing the President's interpretation, and intentions with respect to the implementation, of the Uruguay Round Agreements. It was not a document that was applicable for purposes of the generally accepted rules of treaty interpretation embodied in Article 31 of the Vienna Convention on the Law of Treaties. For example, it was not part of the text of the Marrakesh Agreement Establishing the World Trade Organization ("WTO Agreement") nor was it an international agreement made in connection with the WTO Agreement. However, the SAA confirmed the understanding of the US negotiators that an SPS measure must be based on scientific principles, have a basis in science, be based on an assessment of the risks, and not be maintained without sufficient scientific evidence. But these obligations were clearly set forth in the SPS Agreement itself, which should be construed in accordance with the ordinary meaning to be given to the terms of that Agreement in their context and in the light of its object and purpose. The EC ban failed to meet any of these obligations, and nothing in the SAA relieved the European Communities of those obligations. In contrast, the SAA clearly reflected the US understanding that the SPS Agreement fundamentally required an examination of whether a measure
WT/DS26/R/USA
Page 48
was adopted with a basis in science and based on a risk assessment, rather than as a protectionist measure. The United States noted that this dispute would not have arisen if the European Communities had followed the SPS rules referred to in the SAA. The European Communities had not articulated its appropriate level of sanitary protection, nor had it based its ban on scientific grounds. In contrast, the SAA noted that the determination that a particular substance "poses a health risk is made on scientific grounds". The United States consequently claimed that, for all these reasons, the EC ban was inconsistent with Article 3 (and Articles 2 and 5) of the SPS Agreement.
The European Communities responded that the Statement of Administrative Action itself explained that it "represents an authoritative expression by the Administration concerning its views regarding the interpretation and application of the Uruguay Round Agreements, both for purposes of US international obligations and domestic law", and that the "Administration will observe and apply the interpretations and commitments set out in it". When approved by the US House of Representatives, the SAA was said "to carry particular authority". The European Communities concluded that the SAA was a highly informative piece of evidence which indicated the way the SPS Agreement had be interpreted officially by the United States in tempore non suspecto. The European Communities did not contest that Article 3.3 required a scientific justification for the measures taken by a Member as a consequence of the level of protection it deemed appropriate. However, the footnote to Article 3.3 confirmed the prerogative which the SPS Agreement left to Members to make risk managementdecisions that reflected social value choices distinct from the strict scientific process of risk assessment. 74 The purpose of the footnote to Article 3.3 was to clarify what was meant by "scientific justification" in Article 3.3, not to define the justification for choosing a level of protection. The level of protection was decided by the Member alone and it was not a judgement that must be based on scientific principles or scientific evidence. The European Communities noted that the SPS Agreement specified that sanitary and phytosanitary measures must be "necessary for the protection of human, animal, or plant life or health", must be applied only to the extent necessary to protect human, animal or plant life or health", must not "arbitrarily or unjustifiably discriminate between Members where identical or similar conditions prevail", must not "be applied in a manner which would constitute a disguised restriction on international trade", and must not be "more trade restrictive than required to achieve [a Member's] appropriate level of protection, taking into account technical and economic feasibility". The application of these requirements did not address the scientificunderpinnings for a national regulatory requirement. Instead, these tests targeted the choice by national regulatory authorities among a variety of potential measures, as determined by such factors as impacts on international trade, discriminatory effect, economic efficiency and technical feasibility, and the relationship between the regulatory goal and the measure chosen.
The European Communities further argued that under Article 13, national governments were fully responsible for the observance of the Agreement by sub-national governments. However, nothing in the SPS Agreement required that state or local governments adopt, or comply with, federal sanitary or phytosanitary measures. The reasons for this were twofold. First, because it was quite possible that within a single Member with federal constitutional structure there could exist different levels of protection and different types of sanitary and phytosanitary measures. Second, the inability of the federal government of that Member and of the WTO dispute settlement panels to substitute their scientific judgment for that of the local government on the level of protection and the measures applied to achieve its chosen level of protection. The European Communities claimed that if such differences might exist within onesingle Member, it was obvious that such differences were also likely to exist between different Members with different perceptions of what constituted risk to public health.
74Footnote 2 to Article 3.3 reads as follows: "For the purposes of paragraph 3 of Article 3, there is a scientific justification if, on the basis of an examination and evaluation of available scientific information in conformity with the relevant provisions of this Agreement, a Member determines that the relevant international standards, guidelines or recommendations are not sufficient to achieve its appropriate level of sanitary or phytosanitary protection".
WT/DS26/R/USA
Page 49
The European Communities argued that the United States attacked not only the level of protection chosen by the European Communities, but also its measures, by insisting that residues of hormones above naturally present levels did not pose any risk to health and, therefore, did not warrant the application of any measures to control them other than the MRLs recommended by Codex. If this argument were to be accepted, then it would in the future be open to any country to oppose any health measure which was based on the precautionary principle. There were very few examples of health hazards where all scientists agreed on the degree of risk, and countries must be allowed to make judgments as to what degree of risk they were willing to accept. The European Communities argued, by way of example, that there was no scientific proof that bovine spongiform encephalopathy (BSE) was transmissible to humans; there was only a "likelihood" based on the appearance of a very small cluster of human cases in the only country with a high incidence of the disease in cattle. Nevertheless, the European Communities (and most other countries) had taken severely trade-restrictive measures to protect public health as a precaution, because if BSE were transmissible to humans the consequences would be devastating.
With regard to MGA, the European Communities noted that it was authorized for growth promotion only in the United States and Canada, but apparently nowhere else. Codex had never scientifically examined this hormone and had never recommended any standard. Still, the United States in essence argued that the European Communities should accept meat treated with MGA and check only for residue limits as they were set by the United States. This argument flew in the face of the SPS Agreement, which explicitly allowed the European Communities to adopt a level of protection which it deemed appropriate (in this case, no residue at all).
The European Communities insisted that the US position in this case contrasted sharply with the legal position of the United States when it had presented the SPS Agreement to Congress for approval, and was dictated by short-sighted economic interests (of maximum US$80 per animal), regardless of the danger to human and animal health posed by the use of these hormones for growth promotion. All countries, including the United States, regulated the use of hormones in farm animals, and the difference between the European Communities and the United States in this respect was the extent to which their use was regulated. The European Communities concluded that this difference in degree of regulation was a reflection of the different levels of consumer protection adopted by the European Communities and the United States. The European Communities had taken a precautionary approach, placing the attainment of a high level of consumer protection before the commercial interests of farmers and pharmaceutical companies. Where there existed a doubt over the safety of a product, the European Communities gave the benefit of doubt to the consumer, especially in cases where the potential risks might affect very large parts of the population. The European Communities claimed that, in the case of growth hormones, the United States had taken an opposite approach, giving the benefit of doubt to the producer. It noted that a 1986 report on Human/Food Safety and Regulation of Animal Drugs, approved by the US House of Representatives, had concluded that "the Food and Drug Administration (FDA) has consistently disregarded its responsibility, because it has repeatedly put what it perceives are interests of veterinarians and the livestock industry ahead of its legal obligation to protect consumers thus jeopardizing the health and safety of customers of meat, milk and poultry".75 These criticisms appeared equally valid in the present case.
The United States rejected the EC argument that because there was no guarantee that science would not one day identify a risk associated with residues resulting from the use of the six hormones for growth promotion purposes, the European Communities were justified to use a "precautionary approach" to ban such use. Speculation that some day there might be a risk identified, even if there was none now, was not a basis for banning something. If it were, then Members could ban anything. To invoke the precautionary principle, there still must be some scientific information indicating a risk,
75US Committee on Government Operations, 27th Report, 99th Congress, 31 December 1985.
WT/DS26/R/USA
Page 50
not just mere speculation, but the European Communities had produced no such scientific information with respect to their ban. On the other hand, to invoke the exception of Article 3.3, the European Communities was required to compare its level of protection against that afforded by a measure based on the Codex standard and show that its ban was a consequence of the EC level of protection being more stringent than the level achieved by use of the Codex standard.
The United States also stated that the final sentence of Article 3.3 provided, however, that even if Article 3.3 applied, the measure at issue might not be inconsistent with any other provision of the SPS Agreement. Thus, Article 3.3 was limited only to justifying a departure from international standards and did not justify a breach of any other obligation arising under the SPS Agreement. As a result, the fact that the EC ban was inconsistent with Articles 2 and 5 of the SPS Agreement meant that the EC ban was not justified under Article 3.3 even if the European Communities could establish that the Codex standards would not achieve the EC appropriate level of sanitary protection.
The European Communities responded that a hazard (cancer) had been identified and a lack of scientific knowledge on the exact mechanisms by which it arose was not a sufficient excuse for failing to take strict measures to prevent it. Article 5.7 was not applicable to this case because EC measures were not based on insufficient knowledge and were not temporary. The European Communities knew about the hazard and had taken measures to prevent it. In fact, there was clear evidence, from IARC, JECFA and others, that these hormones were carcinogens. They posed a risk and the European Communities was therefore justified in acting. There was not enough scientific evidence to conclude that these hormones, particularly when used by farmers without any official control, did not pose a risk to human health. The European Communities added that international standards, guidelines or recommendations were based on a certain concept of level of sanitary protection. But a different level might require different type of measures. This was what the European Communities applied in the case of these hormones: to achieve its level of protection, the only reasonably available and less restrictive measure was to apply a prohibition on the use of these hormones for growth promotion. The Codex recommendations were designed to achieve a level of protection which was lower than that applied in the European Communities.
The United States recalled that Article 5.1 reads as follows:
"Members shall ensure that their sanitary or phytosanitary measures are based on an assessment, as appropriate to the circumstances, of the risks to human, animal or plant life or health, taking into account risk assessment techniques developed by the relevant international organizations".
The United States claimed that the central concept underlying the SPS Agreement was "risk". With limited exceptions not applicable to this dispute, a government must scientifically demonstrate the existence of an identified risk to the life or health of humans, animals or plants resulting from a specific activity before that government could impose or maintaina sanitary measure. While the SPS Agreement did not define "risk" as such, it defined "risk assessment" with respect to food safety as "the evaluation of the potential for adverse effects on human or animal health arising from the presence of additives, contaminants, toxins or disease-causing organisms in food, beverages or feedstuffs" (Annex A of the SPS Agreement). The United States concluded that, as a result, in the context of the present dispute "risk" consisted of the "potential for adverse effects on human health" arising from the presence in meat of residues resulting from the administration to animals of any of the six hormones for growth promotion purposes.
The United States claimed that the European Communities had never performed any risk assessment, or relied on any risk assessment, that could serve as a basis for its ban with respect to the six hormones. If there was no risk from a substance, then it could not be said that a sanitary or
WT/DS26/R/USA
Page 51
phytosanitary measure was necessary to protect against a risk posed by the substance. By definition, in order to be a legitimate "sanitary" measure, a measure must protect against one or more specified risks. The United States submitted that the remarkable characteristic of the public debate in the European Communities on these hormones was that the “risk” was usually described in terms of consumer anxieties rather than any observable adverse effect on human health. During the consultations, the European Communities had failed to identify any specific risk to human or animal life or health against which the ban was designed to protect. On two occasions, the European Communities had formed groups of scientists to examine the safety of the hormones (excluding MGA), but both the Lamming Group convened in 1982 and the 1995 EC Scientific Conference had concluded that these hormones were safe when used for growth promotion purposes in accordance with good animal husbandry practice. The United States submitted that the European Communities had failed to identify a risk assessment on which its ban was based and, while suggesting that a risk assessment (or risk assessments) had been performed by "scientific groups and/or conferences" or by "groups of experts established by the EC member States and the European Communities", they had failed to provide the time or place of any such risk assessment, the names of the scientists who had conducted the risk assessment, or the documentary evidence of the results. Furthermore, although the European Communities had argued before the Panel that there was new scientific evidence now available, the European Communities had not relied on any of the studies of the scientific experts whom it claimed had new or differing evidence when it introduced the ban in 1989, nor when it had promulgated its new Directives in 1996. Indeed, the European Communities had apparently not consulted any of those scientists until it was well into a WTO dispute settlement proceeding to review its ban. A risk assessment by the European Communities' own regulatory scientists had never been produced.
The United States claimed that the European Communities could not just assert that there was a risk associated with these six hormones when used for growth promotion and then prohibit their use. Article 5.1 required that a Member base a sanitary measure on an assessment of the risks to human or animal life or health. The European Communities had failed to produce anything resembling an assessment of the risks that would form the basis for its ban. The European Communities had in fact rejected every scientific review of the six hormones and had been unable to produce any scientific risk assessment of these hormones that it did accept. It was important to distinguish between having a risk assessment, and basing a measure on that risk assessment. This was not a situation in which there was no risk assessment. To the contrary, the European Communities certainly had risk assessments available for its use.
The United States noted that during the Article XXIII consultations, the European Communities had stated that the risk assessment on which it had based its ban was the report of the 1995 EC Scientific Conference. Subsequently, the European Communities had indicated that the 1995 EC Scientific Conference "was not to perform for it a risk assessment, but to provide a public forum for discussion of the scientific aspects of the use of growth promoters", a statement which concurred with the US view of the Conference. In any case, the United States argued that the 1995 report failed to satisfy the requirements of Article 5.1 in at least two key ways: first, the 1995 EC Scientific Conference had found that the five hormones it reviewed were safe for use as growth-promoting hormones and that there was no basis for banning animals, or meat from animals, to which these hormones had been administered for growth promotion; and second, the 1995 EC Scientific Conference had not looked at all six hormones - it had not discussed MGA. Accordingly, it could not constitute a risk assessment for MGA. In fact, the 1995 EC Conference did not appear to have actually performed a risk assessment on any of the hormones; it simply "discussed the safety assessment" of the five hormones. The United States contended that in the case of the six hormones, since the European Communities had failed to establish through an assessment of the risks that there was in fact any identifiable risk posed by these hormones, there was no basis for the EC ban on the importation of animals to which had been administered these hormones or meat from those animals. The United States consequently claimed that the EC ban was not supported by an assessment of risk and was consequently inconsistent with Article 5.1.
WT/DS26/R/USA
Page 52
The European Communities responded that Codex had not developed risk assessment techniques. In the absence of such international guidelines, each Member must apply its own methodology. That followed by the European Communities was, and had been throughout the consideration of the hormone issue, to obtain the best scientific information available, to have that information evaluated by technical experts and for the EC Commission to make a proposal based on an assessment of that scientific and technical advice. In assessing the advice and drawing up its proposals, the European Communities also took account of the factors set out in Article 5.2, including"relevant processes and production methods". Livestock production in the European Communities and in most of the countries which exported to the European Communities was characterized by large-scale intensive production in a free market. If producers were at liberty to administer hormone growth promoters at their own discretion there was a considerable risk that mistakes and misuse would occur. There was no guarantee that the correct products or doses would always be used or that animals would be injected or implanted in the correct site. There was also a risk that animals would be slaughtered for human consumption soon after treatment if, for example, the market price suddenly became favourable. Article 5.2 also required Members to take into account "relevant inspection, sampling and testing methods". Due to the scale of livestock production, it was not possible to carry out regular inspections of livestock holdings to check that universally available growth promoters were being used properly. Sampling and testing for these substances was extremely expensive and time-consuming and, in the case of substances analogous to endogenous natural hormones, it might be difficult to distinguish marginal cases of misuse. Moreover, there was nothing in the text of the contested measures, the legislative history or in any other document to suggest that "consumer anxieties" was the purpose for which the measures were adopted, although it was likely that consumer concerns had been taken into consideration during the "risk management" phase, since consumer concerns on potential risks to human health resulting from the use of hormones were very high at that time (and they were even higher today). There was little question that consumer concerns about safety and other issues also influenced the US agencies' decisions, even where there was arguably little, if any, basis for those concerns.
The European Communities argued that it had considered a number of scientific reports and judgements regarding the potential risks from the hormones at issue. As previously indicated76, this included:
the 1982 Report of the Scientific Veterinary Committee, Scientific Committee for Animal Nutrition and the Scientific Committee for Food on the basis of the Report of the Scientific Group on Anabolic Agents in Animal Production (the Lamming Report);
the 1983 OIE Scientific Conference Report;
the 1988 JECFA Report;
the various works of relevant international institutions, such as the International Agency for Research on Cancer (IARC);
the scientific works by individual scientists relevant to the issue of use of hormones in general and for animal growth in particular;
information on the use of these hormones for growth promotion available from other countries, when relevant.
For the adoption of Directives 81/602/EEC, 88/146/EEC and 92/22/EC, apart from the above scientific evidence, additional technical information had been taken into account. This information consisted
76See paras. 4.28-4.29.
WT/DS26/R/USA
Page 53
mainly of the internal studies of the EC Commission, the reports of the European Parliament, the reports of the Economic and Social Committee and the deliberations of the Council of Ministers. For their deliberations, the Ministers were assisted by scientificgroups and individualscientific experts, including experts from the relevant administrations of the EC member States. For the adoption of Directive 96/22 of 29 April 1996, the 1995 EC Scientific Conference Proceedings had also been taken into account.
The European Communities submitted that it had based its measures on the risk assessment it had conducted for that purpose. Neither the 1982 Lamming Report, nor the 1988 JECFA Report constituted in themselves "risk assessment" in the sense of Articles 5.1 to 5.6 of the SPS Agreement. They only formed part of "available scientific evidence." The other factors mentioned in Article 5.2, and in paragraphs 3 to 6 thereof, were not dealt with by those scientific reports on which the United States had based its claim. The assessment of these factors was not a scientific question in the strict sense, and thus they fell within the responsibility of the appropriate political authorities of each Member.
The European Communities further argued that the 1988 JECFA Report had found that there was a potential risk to human health because if there were no such potential risk, the JECFA would not have recommended any ADI and MRL for the two synthetic hormones. For the three natural hormones, JECFA had not recommended an ADI and MRL because of problems in detecting the level of residues in the meat on a routine basis. Moreover, the 1988 JECFA Report, apart from being deficient in several respects from the scientific point of view, had not performed a "risk assessment" in the sense of Article 5. All it had done was to "evaluate the safety of hormones used for growth promotion". This evaluation was not a proper risk assessment for the purposes of Article 5, since it dealt only with one of the elements required by Article 5.2, i.e. it formed part of the "available scientific evidence" but did not constitute the only relevant scientific evidence available. In addition, it did not take into account relevant process and production methods, relevant inspection, sampling and testing methods and relevant ecological and environmental conditions, as required by Article 5.2. Even if the European Communities was to assume that they did (quod non), still the European Communities considered that Article 3.3 explicitly permitted it to adopt a level of sanitary protection it determined to be appropriate in its territory. The European Communities added that it had explained in the scientific opinions which it had submitted to the Panel and in the meetings of 17 and 18 February several defects of the 1988 JECFA Report from the scientific point of view, in particular the so-called hormonal effect level at which these hormones considered to be carcinogenic, its failure to set ADI values for the three natural hormones and its incorrect figures cited for the daily production of for prepubertal boys. (See transcript of the meetings with scientific experts, paragraphs 315-318. See also paragraph 740 with regard to prepubertal boys).
The European Communities claimed that another point where the JECFA Report was deficient related to the question whether ADI (and consequently MRL) values could eve be fixed for the three natural hormones. The European Communities referred to an official letter which the European Communities received from the US Government in 1988, where it was stated that tolerance levels were set in the United States even for the three natural hormones, but meat would not be monitored for residues. The European Communities further claimed that as Dr. Lucier had argued on 18 February, ADIs could have been established in the 1987 JECFA Report. But they had not been established precisely because the scientific approach of JECFA was to examine only the hormonal effect levels of these hormones, because only at such levels were they supposed by JECFA scientists to be carcinogenic. But this line of argument had been shown to be incorrect or at least there was a substantial part of the scientific community which did not agree with this view (see also paragraphs 469-479 of the transcript of the 17 February meetings with scientific experts) The European Communities added that in this case the quantification of the risk was said by Dr. Lucier, one of the scientific experts advising the Panel, to range between 0 and 1 per one million. This was the minimum risk from the use of these hormones in accordance with good veterinary practice. The European Communities argued that this risk would certainly be higher if good veterinary practice was not respected, and would be even much
WT/DS26/R/USA
Page 54
higher if synergistic and long-term exposure risks were taken into account. (See the statement of Dr. Lucier in this regard in paragraph 742 of the transcript of the 18 February meetings with scientific experts. See also the opinion of Dr. Epstein, a scientist advising the European Communities, in the case of prepubertal boys in paragraph 214. As regards the quantification of the risk of cancer, the European Communities also drew the attention of the Panel to the opinion of the Panel expert, Dr. Lucier, reflected in paragraphs 6.110 and 6.111, as well as to the opinion of Dr. Liehr, an EC scientist, in paragraph 330 of the 17 February meeting with the scientific experts.
The European Communities noted in this regard that scientists had said that "the ADI-model is not a risk model and does not predict the risk of the occurrence of adverse effects, when ADI-values are exceeded".77 The same was basically true for the MRL-model, "which specifies the quantity of a particular residue in meat that, even in dietary extremes, would not exceed the ADI for the particular substance concerned".78 The European Communities concluded that the MRL was based on the notion that there was a risk, but if the proposed threshold was observed the probability of the adverse effect arising from the use of hormones would not materialize. In this case, the JECFA recommended standards which would achieve "a level of protection" against the risk. But the SPS Agreement allowed Members to determine another level, i.e. the level they deemed appropriate. This level might be higher or lower. The European Communities argued that all the above scientific data, reports, papers, conferences and other relevant information constituted a body of scientific evidence which was carefully considered by the competent EC institutions and the EC member States.
The European Communities claimed that there were substantial differences in the arguments not only between the European Communities and the United States and Canada, but also in the arguments between Canada and the United States. Canada had argued on 19 February that the European Communities did not have any of the scientific evidence presented to the panel by the scientists advising the European Communities in this case, when it decided to perform a risk assessment. Canada said that the European Communities had discovered this scientific evidence some weeks earlier, searching for an ex post facto justification of the measures it had taken before. The European Communities claimed that this argument of Canada was not correct and did not corroborate with what the United States had said, because as the United States had explained that the views of Dr. Liehr were already mentioned twice in the 1988 JECFA report (footnotes 34 and 35). Dr. Liehr was also mentioned in the 1995 EC Scientific Conference Proceedings (e.g., at page 386). The views of Dr. Liehr were therefore known to the European Communities before, and were not discovered some weeks earlier as Canada had argued. The work of almost all of the scientists that provided advice to the European Communities in these disputes were aware to the responsible EC authorities, such as the work of Dr. Epstein and Dr. Hertz who were mentioned in the book of Orville Schell of 1985, Dr. Liehr was already mentioned in the 1988 JECFA Report and in the 1995 EC Conference, Dr. Metzler who has been doing research in this area since a long time and Dr. Adlercreutz.
The European Communities said that the reason why the United States had never requested Codex to set a "standard" for MGA was that MGA was possibly even more dangerous than the other five hormones. The European Communities noted that the United States applied a so-called "safety factor" of 200, instead of the usual 100, and argued that, apart from its carcinogenicity, MGA exerted a more powerful progestagenic effect than progesterone. Its principal use was in fattening heifers which were kept under abnormally intensive conditions and could injure themselves or their inexperienced handlers when they came into oestrus and start mounting each other. MGA, although not a very effective growth promoter, was very effective at suppressing oestrus and was added to the feed mainly for this purpose. Although it had a withdrawal period, there was a strong incentive on the handlers not to observe it since the heifers would start coming into oestrus before going to slaughter. It was, therefore,
77H.A. Kuiper, "Risk Assessment Strategies for Xenobiotics", 1995 EC Scientific Conference Proceedings, p.375.
78"Report and Conclusions", 1995 EC Scientific Conference Proceedings, p.4.
WT/DS26/R/USA
Page 55
probable that meat from heifers regularly contained significant residues of this powerful sex hormone. The European Communities also claimed that as regarded oestrogens, progestins and combinations thereof (i.e. mainly , progesterone and MGA), the 1987 IARC Report explained that "Steroid hormones are essential for the growth, differentiation and function of many tissues in both animals and humans. It has been established by animal experimentation that modification of the hormonal environment by surgical removal of endocrine glands, by pregnancy or by exogenous administration of steroids can increase or decrease the spontaneous occurrence of tumours or the induction of tumours by applied carcinogenic agents .... The incidence of tumours in humans could be altered by exposure to various exogenous hormones, singly or in combination" (emphasis added). The European Communities further claimed that the United States was offering a solution to safeguard against the potential risk to human health from the use of these hormones for growth promotion which the European Communities had considered carefully and rejected, because it did not meet its appropriate level of protection. It stressed that the hormone MGA was authorized for growth promotion only in United States and Canada, but apparently nowhere else. Codex had never scientifically examined this hormone and had never recommended any standard. Still, the United States in essence argued that the European Communities should accept meat treated with MGA and check only for residue limits as they were set by the United States. The European Communities submitted that this argument contradicted the SPS Agreement, which explicitly allowed the European Communities to adopt a level of protection which it deemed appropriate (no residue at all).
The European Communities argued that the Lamming Committee had not considered itself competent to make recommendations to the European Commission on the conditions under which these hormones could be allowed to be used for growth promotion purposes. This showed that the Lamming Committee could not be considered competent to perform a risk assessment for the European Communities in the sense of Article 5, because risk assessment could only be performed by the governments of Members. The Lamming Committee had not provided anything other than scientific advice, which the responsible EC institutions had used as part of the available scientific evidence in their risk assessment. The European Communities further argued that the US claim that the European Communities stated during the consultations that it based its "ban" on the 1995 EC Scientific Conference was clearly wrong. As the purpose of the 1995 EC Scientific Conference was not to perform for it a risk assessment, but to provide a public forum for discussion of the scientific aspects of the use of growth promoters, the reference to Article 5.1 and related comments were irrelevant as well as inaccurate.
The United States claimed that the Codex risk assessment and other scientific reviews stood in sharp contrast to the EC purported risk assessment. Yet the European Communities took the curious position that the Codex and other scientific reviews were not "risk assessments" for purposes of the SPS Agreement because they did not "constitute the only relevant scientific evidence available". The United States argued that, first, under the SPS Agreement, a risk assessment was to "take into account" available scientific evidence (Article 5.2); it was not supposed to "constitute" scientific evidence. Second, the European Communities suggested that Codex had not taken into account all the elements listed in Article 5.2, without explaining why elements which were omitted would be relevant to the risk assessment. Yet the elements listed by the European Communities were taken into account by Codex, as reflected in the Codex definition of an MRL. The United States observed that the experts advising the Panel confirmed that there was no scientific basis for the EC refusal to use the Codex standards as a basis for the EC measures. Third, under Article 3.2, measures that conformed to Codex standards were presumed to be consistent with the relevant provisions of the SPS Agreement. This would include Articles 5.1 and 5.2, so that the Codex risk assessment might be presumed under the SPS Agreement to be consistent with those Articles. Fourth, the EC approach was to selectively misquote parts of various risk assessments while ignoring the conclusions of those risk assessments as well as those portions and the scientific evidence that did not support its ban. The European Communities did not explain how this approach of selectively ignoring the scientific evidence could be considered to take into account the relevant scientific evidence. Furthermore, the European Communities could
WT/DS26/R/USA
Page 56
not establish that the Codex standards would not achieve the EC appropriate level of sanitary protection. The European Communities had suggested that it had adopted "zero risk" as the appropriate level of protection, and this "zero risk" level would justify its measures under Article 3.3. The United States submitted that an "appropriate level of sanitary protection" was a level of protection from a risk and that the European Communities had not identified any particular risk from these hormones, which were all safe.
The European Communities claimed that it was not required to accept only the two scientific reports which the United States contended had carried out a proper risk assessment, i.e. the Lamming Report and the Codex assessment. These were just two reports of a certain scientific orientation, but there were other scientific reports by individual scientists, scientific conferences, (e.g. the 1995 EC Scientific Conference) and responsible international institutions (eg. IARC) which did not entirely agree with them. Some of the scientific reports on which the European Communities had based its measures questioned the very basic scientific assumptions on which the reports supported by the United States were founded, including the carcinogenic effects of these hormones when used for animal growth promotion. The EC scientific evidence was therefore constituted by the scientific studies which concluded that these hormones could be carcinogenic irrespective of the dose administered because they may alone or in combination be "genotoxic" (i.e. act directly on the genome to cause cancer). The European Communities added that this was exactly what some scientists had feared for the hormone DES. It had warned the US authorities at an early stage, but those authorities had failed to pay attention in time.79 Moreover, the sanitary policy of the European Communities on these hormones for animal growth promotion purposes was exactly the policy pursued by the United States under the so-called Delaney Clause for dangerous food additives.
The European Communities noted that the "Delaney Clause" was the most notable example of the adoption by the United States of a zero-risk policy. Enacted in 1958 in the context of public fear over the growing incidenceof cancer, the Delaney Clause provided that"no additive shall be deemed to be safe if it found to induce cancer when ingested by man or animal."80 The Delaney Clause required that all carcinogenic food additives be banned, regardless of their degree of risk or offsetting benefits. The Delaney Clause had become an increasing burden to regulatory agencies since its passage in 1958. The scientific methods available then did not have the sensitivity of today's technology, which could detect minimal carcinogenic effects or minimal "migration" of substances from packaging materials to food. The Delaney Clause therefore banned chemicals with insignificant carcinogenic properties, without permitting agencies to conduct the usual risk/benefit analysis employed when considering approval of new substances. The United States had - with only modest exceptions - an inflexible, zero tolerance policy for carcinogenic substances added to the food supply. No matter what its concomitant benefits, a substance having a possible carcinogenic effect could not be approved as a food additive.
79The European Communities explained that Diethylstilboestrol (DES) was a synthetic hormone which improved both rate of gain and feed efficiency in poultry, cattle and sheep. DES was perhaps the most widely known of oestrogenic growth promoting agents. It was first used in the US therapeutically in pregnant women in the 1940's to prevent abortion. The US Government approved its use as a feed additive for poultry in 1947, for cattle in 1954 and later for sheep. In 1971, its carcinogenic effects in humans and in laboratory animals was clearly established. Additional studies showed that DES had other effects on daughters and sons of mothers treated with it, the most common being vaginal adenosis and other gross abnormalities of the reproductive tract. At the time (1976) these studies were published, an estimated 25 million cattle and 7 million sheep were being treated with DES. Still, its complete withdrawal from the US market was achieved only in 1978, while in the meantime the US FDA was trying to discover and fix acceptable "no residue" limits.
8021 U.S.C. § 348(c)(3)(A). The European Communities noted that the Delaney Clause also applied to colour additives (U.S.C. § 379e(b)(5)(B)) and compounds administered to food-producing animals (21 U.S.C. § 360b(d)(1)(I)). On 3 August 1996, new legislation had been enacted which changed US regulation of pesticides and encouraged international harmonization of pesticide tolerances (Food Quality Protection Act of 1996, Pub. L. No 104-170). This new legislation did not affect food additives other than pesticides residues. Thus, the Delaney Clause still governed the regulation by FDA of food additives and compounds administered to food-producing animals.
WT/DS26/R/USA
Page 57
The European Communities argued that none of the scientific reports the United States had cited had concluded in favour of an unqualified use of these hormones for animal growth promotion. All these reports had concluded that all thesehormones individually were potentially dangerous to human and animal health in general as well as when used for growth promotion purposes and that all these hormones individually were not likely to pose risks to human or animal health only if used in accordance with good agricultural/animal husbandry/veterinary practice. Moreover, the 1982 Lamming Report and the 1988 JECFA Report had not examined the potential risks arising from these hormones when used in combination with other hormones, and had not explained what constituted good agricultural/animal husbandry practice.
The European Communities contended that available scientific evidence showed the use of the natural and synthetic hormones subject to this dispute could be dangerous to human and animal health. The potential danger resulted from (i) their nature and mode of action; (ii) the action of their metabolites;
combinations of hormones and multiple exposure; (iv) problems linked to their detection and control; and (v) problems linked to their administration. In submitting its supporting arguments, the European Communities indicated that it focused on the potential risks to human beings, but that this did not imply that there were no risks for animal health.
The European Communities claimed that even though the precise mode of action of the hormones at issue had only been partly understood, all scientists agreed that these hormones were dangerous. The principle sex hormone in men, testosterone, contributed to acne, baldness, prostatic disease, coronary heart disease and peptic ulceration. Oestrogens, and to a lesser extent progesterone, exposed women to premenstrual tension, dysmenorrhoea, some disorders of the reproductive system and an increased tendency to develop gallstones. More importantly, these hormones also played a major role in the development of tumours in sex-hormone dependent tissues (for example prostatic cancer, breast cancer, cancer of the uterus and neoplastic liver disease).
The European Communities noted that the carcinogenic risks from these hormones was well known. For instance, in the 1983 OIE Scientific Report a scientist had concluded that "although it is unlikely that the small amounts of anabolic steroids which might be ingested in agricultural produce could (cause cancer), this possibility should not be ignored entirely". Moreover, the same 1983 OIE Scientific Report stated that: "... However, as pointed out by Williams, the initiation-promotion concept does not adequately explain all carcinogenic mechanisms. Therefore, it is not possible to state with any certainty that anabolic steroids act only as promoters in induction of cancers in other hormone dependent tissues such as prostate, uterus and breast, although present evidence suggests that levels of endogenous steroids are probably not sufficient to initiate carcinogenesis in those tissues" (emphasis added).
The European Communities further noted that the carcinogenic effects of hormones on humans (and animals) had been documented by several reports of the International Agency for Research on Cancer (IARC). IARC had classified steroidal oestrogens (including oestradiol) in Group 1, meaning the agent was carcinogenic to humans; androgenic (anabolic) steroids (including testosterone) in Group 2.A, agents probably carcinogenic to humans and progestins in Group 2.B or agents possibly carcinogenic to humans.81 A 1987 IARC Report indicated that, in the case of androgens:
"The fact that castration palliates prostatic cancers suggests that testosterone may be involved in the genesis of these tumours, and a number of epidemiological observations suggest that increased testosterone levels may increase the risk for prostatic cancer" (page 96). ... The
81IARC Monograph s, Supplement 7, pp.31, 96 and 280 (1987).
WT/DS26/R/USA
Page 58
evidence that anabolic steroids can cause both benign and malignant liver tumours is quite strong. However, because no analytical epidemiological study has been done, the Working Group felt constrained to classify the evidence for carcinogenicity in humans as no more than "limited" (page 97).
The European Communities noted that, according to the same IARC Report, in the case of oestrogens, however, the evidence was clear:
"A number of studies, utilizing a variety of designs, have shown a consistent, strongly positive association between exposure to a number of oestrogenic substances and risk of endometrial cancer, with evidence of positive dose-response relationships both for strength of medication and duration of use" (page 280).
The European Communities added that in a recent article published in the review Science82, the author argued that oestrogen's (natural hormone) mode of action may be different from what it was thought to be until now:
"... while it is still too early to tell just how many alternate paths there are, other recent results are further undermining the idea that there is a single main pathway for oestrogen action."
The European Communities asserted that the scientific assumption that, despite their carcinogenic potential, these hormones were not likely to cause problems to human health was based on the assumption that "the carcinogenic effect of these hormones is related to the hormonal activity of these compounds,
i.e. an increase in tumour incidencein tissues of animals with high levels of specific hormone receptors, which would not occur at normal physiological levels".83 But scientists could not define what those "normal physiological levels" were since they varied widely from animal to animal, and for that reason these hormones were considered by some scientists to be "promoters rather that primary inducers of cancer in hormonally sensitive tissues".84 According to the European Communities, it was on this assumption that was also based the 1988 JECFA Report, on which Codex had based its recommendations. With regard to , the 1988 JECFA Report stated that "[t]he results of studies ... [showed that] .... [o]ral and parenteral administration of oestradiol -17ß can increase the incidence of tumours in experimental animals. These tumours largely occur in tissues with high levels of specific hormone receptors that are normally responsive to stimulation by the particular hormone concerned. The Committee concluded that the carcinogenic response was related to the hormonal activity of oestradiol-17ß at levels considerably higher than those required for a physiological response" (page 18).85
The European Communities argued that the general scientific assumption was that these substances were not "genotoxic"86, but exerted only "epigenetic" action. But this was an assumption which had
82Science Vol. 273, 30 August 1996.
83H.A. Kuiper, "Risk Assessment Strategies for Xenobiotics", 1995 EC Scientific Conference Proceedings, pp.370-371.
84H.A. Miller, J.K. Leighton, "Risk Assessment Strategies for Hormones and Hormone-like Substances", 1995 EC Scientific Conference Proceedings, p.386, and 1983 OIE Scientific Report, p.339 sqq.
85The European Communities noted that the same assumptions were made in the 1988 JECFA Report for progesterone (p.20), testosterone (p.22), trenbolone acetate (p.24), and zeranol (p.27).
86"Genotoxic carcinogens" were defined by the European Communities as "chemicals whose mode of action is mediated by a series of molecular events, initiated by direct interaction of the compound or activated reactive intermediate species with cellular DNA eventually leading to uncontrolled cell replication and formation of tumours". The European Communities added that the concepts of NOEL (no observed effect level) and ADI (acceptable daily intake), on which for instance was
(continued...)
WT/DS26/R/USA
Page 59
not yet been clearly proved. Scientists had agreed that "further development of metabolic studies is required, particularly with regard to exogenous anabolic agents (synthetic hormones). Such advanced studies would lead to a better understanding of their mechanism of action and any toxic effects".87 The European Communities also claimed that the distinction between agents which acted as tumour initiators and those considered as tumour promoters was over-simplistic. This distinction permitted some scientists to make some predictions by extrapolating tumour incidence observed in test animals at high exposure levels to potential human exposure to low levels of such hormones. However, other scientists had agreed that "these models do not indicate a true risk but provide a calculated risk for the occurrence of the (carcinogenic) effect. The relevance of these models is questionable since the biological basis for high to low dose extrapolation is lacking".88 The European Communities submitted that it was on this disputed extrapolation, based on the no hormonal effect level, that JECFA had based its 1988 Report and Codex its July 1995 recommendations. The example of DES illustrated the danger of such an approach (see paragraph 4.123, footnote 79).
The United States rejected the EC claim that "all scientists agree that the six hormones are dangerous" and submitted that instead, hormones used for growth promotion purposes were not inherently safe or dangerous, but protection of public safety depended upon the dose of the compound, its toxicity and the manner of human exposure. Under the conditions of use in the United States, these hormones were safe when used for growth promotion purposes. The safety of the hormones as growth promotants was confirmed by the fact that they had been registered for use in more than 20 countries (including the largest meat-producing countries). These countries had strong food safety laws. The 1983 OIE Scientific Report had reviewed trenbolone and zeranol and had also come to the conclusion that these hormones were safe when used for growth promotion. The 1983 OIE Scientific Report had concluded that:
"1. Hormones generally pose no cancer risk where exposure is to levels below those required for detectable hormonal activity.
"2. Mutagenicity and carcinogenicity test data for trenbolone and zeranol suggest that these agents and their metabolites are neither mutagenic nor clastogenic and that they would only influence cancer risk - either increase it or decrease it - if there was exposure to hormonally effective levels.
"3. Therefore, in judging whether it is safe to use trenbolone or zeranol as anabolic agents in meat production the emphasis needs to be on making sure that any residue of these agents in meat are below the levels that could have any hormonal effect on the meat-eater. ..."89
86(...continued)
based the 1988 JECFA report, were not applicable to such genotoxic agents Recently (1990), oestrogen induction of DNA adductformation hadbeen described (1995 EC Scientific Conference Proceedings, Kuiper, p.373, Miller and Leighton, p.386, with a reference to a study by J.G. Liehr, "Genotoxic Effects of Oestrogens", 1990). "Epigenetic" agents were believed not to induce gene mutations but to operate via other mechanisms like cytotoxicity, cell proliferation, peroxisome proliferation or hormone disbalance, and interference with hormone production. These effects were considered as dose-dependent for which a threshold level (NOEL) might be established (ibid, Kuiper, p.374). Once a NOEL had been set up, then an ADI level or MRL (maximum residue limit) was usually established.
871983 OIE Scientific Report, pp.263-269.
881995 EC Scientific Conference Proceedings, p.373.
89F.J.C. Roe, “Anabolics in Animal Production”, Symposium held at OIE, February 1983, p.339.
WT/DS26/R/USA
Page 60
Furthermore, the United States noted that the IARC work had been reviewed by JECFA.90 IARC's work did not support the EC argument because IARC had not looked to see if there would be a risk at the low levels of residues involved in this dispute. The IARC had specifically stated that "[t]he operating principle is to determine the ability of the chemical to produce cancer or other genetic and related effects without the strictures of mode of human use or the magnitude of the doses"91 (emphasis added). The IARC work involved looking at much higher doses, and different modes of administration, than those involved in the use of hormones in livestock for growth promotion purposes. The United States explained that for these six hormones, one could not extrapolate from high doses to low doses. In scientific terms, this was referred to as high-to-low dose or linear extrapolation, and it was not a general principle of toxicology or pharmacology but was instead one model used to explain experimental observations. This model was also used in some approaches to risk assessment. An alternative approach to explain scientific data would be to use a threshold model. Under a threshold model some minimal amount of a compound was necessary before any response occurred. These were not contradictory models; experimental data for some compounds were better explained by linear extrapolation whereas data for other compounds, including the hormones involved in this dispute, could be better explained using a threshold approach. The threshold approach implied that a no-effect level could be determined through scientific studies, an approach that had been determined to be valid for the implant hormones used for growth promotion by the Lamming's Committee92 and the 1995 EC Scientific Conference.93 MGA had not been not examined in those reports, but the US Food and Drug Administration had conducted scientific reviews of MGA and haddetermined thatthe threshold approach applied to MGA as well.
The European Communities noted that the IARC tests were not necessarily only at high doses. IARC did not differentiate between the effects of the substance at high doses versus low doses. With regard to oestrogens, progestins and combinations thereof (i.e. mainly , progesterone and MGA), the 1987 IARC Report explained that:
"It has been established by animal experimentation that modification of the hormonal environment by surgical removal of endocrine glands, by pregnancy or by exogenous administration of steroids can increase or decrease the spontaneous occurrence of tumours or the induction of tumours by applied carcinogenic agents... The incidence of tumours in humans could be altered by exposure to various exogenous hormones, singly or in combination .
"These statementsmake explicit the facts that oestrogens and progestins occur naturally, and that the hormonal milieu and dose-effect relationships are generally inextricably involved in the carcinogenic effects of oestrogens and progestins.
...
"Thus, there is a basic incongruity between the human data and the animal carcinogenicity data. As noted earlier, however, the effects of these chemicals in humans appear, at least in most cases, to be linked to the hormonal milieu" (emphasis added).
901988 JECFA Report, pp.17, 20 and 21.
91IARC Monograph s, Supplement 7, p.272.
92Veterinary Record, 24 October 1987, pp.389-392
93Assessment of Health Risk Working Group II, pp.19-21.
WT/DS26/R/USA
Page 61
It followed that the findings of the IARC reports aimed at establishing the carcinogenic or other effects of these substancesirrespective of the modes of human use and the magnitudeof the doses. Furthermore, the phrase "at least in most cases" did not mean in "all cases" and under all conditions.
The European Communities argued that the US statement that the threshold approach implied a "no-effect" level deliberately concealed the fact that in respect of the hormones under discussion the scientific groups to which the United States referred looked only for a "no hormonal effect" level. They did not look for a "no toxic effect" level; the JECFA report stated that no toxicologicalmonograph was prepared for any of the hormones which they considered. The "rationale" for this approach was that any toxic effects in humans, including cancer, would be linked to the hormonal effects in laboratory animals; without the latter the former would not occur. However, this was a mere guess. There was no scientific justification for such an assumption. Noting that the United States had banned DES without requiring such a link, the European Communities concluded that if the hormones exerted their carcinogenic effects by acting directly on the genome, there was no "safe" level. The evidence for this was increasingly strong, particularly in the case of oestradiol.
The United States contended that an international scientific consensus had evolved on the safety and efficacy of the natural hormones (oestradiol, progesterone and testosterone) and their mimics (zeranol, MGA and trenbolone) for use as growth promoting agents in bovine animals when used according to good veterinary practice. The United States recalled that after imposing the ban, the European Communities had twice convened groups of scientific experts to review the use of these hormones (except MGA) and that on both occasions the European Communities’ own chosen scientific experts had concluded that it was safe to use these hormones for growth promotion purposes (as reflected in the Lamming Report and in the 1995 EC Scientific Conference proceedings). Moreover, Codex had reviewed these hormones, with the exception of MGA for which JECFA had never been requested to perform a safety assessment, and had concluded that they were safe to promote growth in cattle. Codex standards had been adopted94 after an extensive review of the relevant science including studies on biological activity, carcinogenicity, embryo toxicity, and mutagenicity; use patterns; residues in animals; analytical methodology; toxicological data from laboratory animal experiments; and observations in humans. The Codex process had afforded the opportunity for input by scientists around the world.
The United States added that after thorough review, the US Food and Drug Administration (FDA) had also approved oestradiol, progesterone, testosterone, trenbolone and zeranol for use as ear implants to promote growth in cattle and, in addition, FDA had approved MGA for use as a feed additive for increased weight gain, improved feed efficiency and suppression of oestrus. Oestradiol, progesterone, and testosterone had been approved in the 1950's, zeranol had been approved in 1969, trenbolone in 1987, and MGA in 1968. In the case of the three synthetic hormones, the FDA review had included a standard battery of toxicology tests including studies designed to assess each compound’s toxicity, carcinogenicity, mutagenicity and any reproductive effects. FDA had also required special studies designed to examine the hormonal action of the compound.95
The United States claimed that oestradiol, progesterone and testosterone occurred naturally in all mammals and that the amount produced daily varied widely among species and among animals of the same species, and was influenced by factors such as age and breed. Furthermore, endogenous hormonal substances were also found in a variety of non-meat foods that formed a part of the normal human diet. Over 300 species of plants that were used for food were recognized as having oestrogenic substances, including soybeans, cherries, apples, green beans, alfalfa, palms and licorice. Moreover, these hormones occurred at much higher levels in some of these other foods. For example, a hen’s
94ALINORM 91/31 APPENDIX IV, as adopted by the 21st Session of the Codex Alimentarius Commission.
95The United States submitted an explanation of the FDA review of these hormones as well as a table providing details on the approved uses of these hormones in the United States.
WT/DS26/R/USA
Page 62
egg had a higher concentration of oestradiol equivalents than meat from an implant-treated steer. The relative concentration in an egg was more than one thousand times higher than that in the steer, and the total amount ingested by consuming a hen’s egg, even considering the difference in quantity consumed, was slightly greater than 600 times that from consuming meat from an implant-treatedsteer.96 Many fruits, vegetables, grains and nuts contained substantial quantities of phytoestrogens.97
The European Communities argued that consumers had always been exposed to a certain amount of hormones, or hormone-like substances in their diet; they had evolved alongside them and could cope with them. This was not the case for added hormones. Furthermore, the plant substances referred to by the United States were not identical to the animal hormones.
The United States further observed that wide variations occurred in the production of oestradiol, progesterone and testosterone in humans and at amounts considerably greater than the amounts of these hormones in meat from treated animals. While human production of oestradiol, progesterone and testosterone was measured in micrograms per day, ingestion of these hormones was measured in nanograms or picograms (one thousand to one million times less, respectively, than a microgram). Thus, ingestion of residues of oestradiol, progesterone and testosterone in meat was very small when compared to those amounts produced by humans, whether or not the meat was from an animal to which hormones had been administered. These amounts were so small that they did not have any effects on humans. In addition, the residues of the natural hormones had very low biological activity when ingested because more than 90 per cent of the residues (including metabolites of the three natural hormones) passed through an individual and were rapidly excreted. Thus over 90 per cent of the residues were not absorbed, so they could not have any effect on the consumer. Finally, the United States submitted that any residues of the natural hormones in meat from animals to which one of these hormones had been administered could not be distinguished from the residues resulting from the endogenous production of these hormones.
The European Communities argued that the various assessments or reviews undertaken by groups of scientists, conferences, etc. were not a risk assessment in the terms of Article 5. Article 5 required Members to ensure that their measures were based on a risk assessment. It did not provide for them to delegate this responsibility to scientific groups. Article 5.2 specifically required Members to "take into account available scientific evidence" as one of a number of factors. It did not say that Members should adopt without question the conclusions of whatever scientific group had happened to pass an opinion on the subject. No democratic system of government could abdicate its responsibilities to technicians in this manner. Moreover, the conclusion of all scientific experts did not support an unqualified and free use of these hormones for growth promotion, as was the practice in the United States. They all concluded cautiously that, under the present state of scientific knowledge, the use of the five hormones for growth promotion was not likely to pose risks to human or animal health, if used in accordance with good agricultural husbandry (veterinary) practice.
The European Communities contended that the United States apparently sided (as had done the 1988 JECFA report and the Codex) with those scientists who believed that these hormones caused cancer due to their hormonal activity and that, therefore, they were promoters rather than inducers of cancer in hormonally sensitive tissues. But this was the belief of just one part of the scientific
96J. Riboleau (1983), "Teneur en substances oestrogènes de l’oeuf vierge et l’oeuf faconde des pietaux". Compt. Redn. Sco. Biol, 129-914; V.C. Craknell, and F.L. Mauld "Anabolic Agents in Animal Production" (unpublished Lilly Research Ltd.); J.B. et al. Tarr (1984), "Pharmaceutical metabolic and tissue residue studies of 3H-Zeranol in cattle" (unpublished).
97K. and R. Vertual, D.S. (1990), "Naturally Occurring Oestrogens in Plant Foodstuffs--A Review". Journal of Food Production, Vol. 43, pp.577.581; A.N. Booth, S.M. Tithoff, and C.M. Hehier, (1960), "Oestrogen-like Activity in Vegetable Oils and Milk By-products",. Science, Vol. 131, p.1807; E.L. Monk, R.E. Erf, and T.A. Mellest (1975), "Relationships between Immunoreactive Oestrone and Oestradiol in Milk, Blood and Urine of Dairy Cows", Journal of Dairy Science, Vol. 44, pp.34-40.
WT/DS26/R/USA
Page 63
community. In the absence of clear evidence to the contrary, the European Communities considered that it was entitled to follow a cautious approach and sided with the other body of scientists who placed more attention on the carcinogenic risks arising from the possible "genotoxic"98 action of these hormones, irrespective of the dose in which they were administered to animals for growth promotion. This choice was in compliance with the proper definition of the concepts of "risk" and "risk assessment" which the SPS Agreement had explicitly established in Articles 5.1 to 5.6 and in Annex A(4). In this respect, the European Communities referred to a number of scientific opinions.
The European Communities observed, for example, that Dr. Liehr argued that "oestrogens have been implicated for some time in the induction of human cancers, and there is increasing evidence of a similar role for progesterone and testosterone".99 In particular, Dr. Liehr had stated:
"In the case of oestradiol, experiments have shown that its hormonal potency is not linked to its carcinogenic activity. Investigations of the metabolic pathways of oestradiol have revealed that free radicals are continuously produced, and DNA damage by these free radicals, as well as DNA adduct formation, has been demonstrated in a range of tissues. As damage to DNA and DNA adduct formation are known to be associated with tumour formation, these experiments provide strong evidence that oestradiol is a genotoxic carcinogen. In the case of progesterone and testosterone, it is not clear whether their carcinogenic effects are mediated by their action on specific receptors in target tissues or due to mutagenic DNA damage.
"In the present state of knowledge, it is very difficult and in any case not advisable to determine a safe threshold dose of hormone below which tumour formation will not occur. Hormone-associated tumours in the breast, uterus or prostate may be induced by a combination of exogenous and endogenous hormones and/or their metabolites. As the total amount of hormone or metabolite necessary for tumour induction is not known, the amount of exogenous hormone or metabolite necessary for tumour induction in addition to unknown amounts of endogenous hormone or metabolite has not yet been determined. In the 1988 JECFA report, the authors considered only the hormonal receptor-mediated activities of the natural hormones. In view of the considerable amount of scientific evidence which has accumulated since the release of that Report, particularly in respect of the genotoxicity of oestradiol, the Report can no longer be considered applicable to a risk assessment of the use of hormone growth promoters."100
In addition the European Communities argued that Dr. Adlercreutz indicated that:
"... the 1987 JECFA report must now be regarded as too old. The main problem with this report is that it relates any negative effects of hormonal drugs to their hormonal effects, i.e. if a compound occurs in amounts not giving clear "hormonal" responses it is regarded as safe. The anabolic steroid trenbolone has both anabolic and oestrogenic effects and it is not clear which effects are considered. It is frequently used in combination with oestradiol and also with zeranol, which are oestrogenic. Furthermore, testosterone and trenbolone seem also to affect the glucocorticoid receptor and testosterone increases growth hormone and growth factor concentrations and both affect steroid biosynthesis. Many other hormonal effects of the five compounds dealt with are known. When studying hormonal effects only gross effects are considered, but nowadays it is obvious that effects may occur at the cellular level at low concentrations of hormones and without being observable by the relatively coarse methods used in toxicology in vivo studies. Such effects may be observable in human subjects first after
98J.G. Liehr (1990), "Genotoxic effects of oestrogens"; and the 1995 EC Scientific Conference Proceedings, p.386.
99J. Liehr, "Potential genotoxicity of Hormones", 23 December 1996.
100Ibid.
WT/DS26/R/USA
Page 64
many years or even longer time. Another problem is that the JECFA report does not deal with or discusses very little the carcinogenicity of metabolites of hormones, particularly those of oestrogens. At the time when the 1987 JECFA Report was written very little was known about the genotoxic effects particularly of estrogen metabolites. Since the report was written it has also been more and more obvious that hormones have nongenomic effects not related to their hormonal effect via the receptors and that steroid hormone metabolites may be as important or even more important with regard to various effects on biochemical or genotoxic events than the parent compound.
"All these hormones have been shown to cause cancer in various experimental systems if given in high amounts but it is obvious that the effect is not always due to their hormonal activity. For oestrogens and trenbolone it has been shown that also other mechanisms occur and the mechanisms by which the other hormones cause cancer is far from clear.
(...)
"Anabolic steroids in meat has been shown to result in positive doping tests. A compound giving positive doping tests cannot be administered for meat production because these may result in lifetime problems for any sportsman or sportswomen who is caught with a positive doping test without having used any doping agents.
"The 1987 JECFA report is definitely too old and does not take into account recent scientific evidence changing our view on the carcinogenicity of oestrogens and does not take into account non-hormonal and other negative effects of the administration of steroids to animals intended for human consumption". 101
The European Communities also presented the argument of Dr. Cavalieri that:
"... the presently available knowledge about tumour induction in hormone-responsive tissues suggests a specific pathway of oestrogen metabolism, namely 4-hydroxylation followed by oxidation to E-3, 4-Q and reaction DNA. This DNA damage is responsible for generating the critical mutationsthat can lead to tumour formation, if the appropriate mechanismsof promotion occur. Therefore, oestrogens, in particular oestradiol, have genotoxic effects and no threshold dose can be established. In addition, the effects of the other natural and synthetic hormones used for meat production on the metabolism of oestrogens are not known. Thus, Acceptable Daily Intakes and Maximum Acceptable Levels cannot be established for residues of these hormones in meat for human consumption.
"The 1988 and 1990 FAO/Codex Alimentarius expert group reports recommended ADIs and MRLs for these hormones based on their hormonal effects. Genotoxic effects were unknown at that time and, thus, were not considered. In light of the new knowledge about the genotoxic effects of oestrogens, the entire subject of these residues in meat for human consumption must be reassessed."102
Furthermore, the European Communities argued that Dr. Metzler reported that the available data show that:
101H. Adlercreutz, "Evaluation of the Thirty-Second Report of the Joint FAO/WHO Expert Committee on Food Additives, and Discussion of Dr. J. Liehr's Report on this Topic", 7 January 1997.
102E. Cavalieri, "Genotoxicity and Potential Carcinogenicity of Hormones Administered to Animals for Promotion of Growth in Meat Production", 7 February 1997.
WT/DS26/R/USA
Page 65
"17-oestradiol,
-trenbolone and zeranol can be metabolically activated to products capable of covalent DNA binding. The level of DNA binding is low as compared to typical chemical carcinogens, but may have significant toxicological implications, as discussed below.
"17-oestradiol and some of their metabolites have the potential to cause aneuploid and are therefore chromosomal mutagens. For
-trenbolone, the reports about aneuploidogenic potential are controversial. For zeranol and its metabolites, aneuploidogenic potential has not been studied adequately.
"The metabolism of -trenbolone and zeranol in humans and the genotoxicity of the metabolites have not been fully elucidated.
"The mechanisms of hormonal carcinogenesis are not yet fully understood. The hormonal and genotoxic effects of an agent may well act in concert. Therefore, it is conceivable that an hormonally active compound is retained in target cells by the hormone receptor and stimulated cell division, which in turn makes the cell more vulnerable for the induction of gene mutations and aneuploid. Xenobiotic hormones have longer half-life and may accumulate to higher levels after repeated exposure.
"The standard systems used in toxicology for assaying genotoxicity may not be suitable to detect the rather subtle genotoxic effects of carcinogenic hormones. As the hormones under consideration and/or some of their metabolites exhibit both DNA-damaging and aneuploidogenic potential, it is not permissible to establish values for No-Adverse-Effect-Levels and for Acceptable Daily Intake. The conclusions and recommendations of the 1988 and 1990 FAO/Codex Alimentarius Reports are solely based on the hormonal activities of the agents under consideration. As the evidence for genotoxic effects of these agents has increased considerably over the past years, an update is required which takes these recent findings into account."103
The European Communities also referred to the report submitted by Dr. Epstein in which he argued that:
"... there is substantial evidence challenging thevalidity of classifying carcinogens as epigenetic, for which thresholds or Acceptable Daily Intake (ADI) levels are claimed, or genotoxic. Apart from this, hormonal anabolics are mutagenicin mammaliantest systems and are thus genotoxic. There is also substantial scientific evidence challenging the existence of thresholds for any carcinogen. This evidence is even more persuasive for exposures involving infants and young children, in view of their enhanced sensitivity to carcinogens and for exposures involving unpredictable synergistic interactions. There is no scientific basis for claims that ADI levels can be set for natural and synthetic anabolic carcinogens, or for claims that ADI levels can be based on "no-hormonal-effectlevels" of synthetic anabolic carcinogens (FAO/WHO, 1990).
"These conclusions on the hormonal effects of anabolics are consistent with, and an amplification of, those detailed in an earlier review of endocrine factors in human carcinogenesis. In almost all cases two postulated mechanisms: "alteration of the susceptibility of tissues to the initiation of cancer" and "promotion of the development of cancer from initiated cells", could not be separated. The carcinogenic effects of DES in relation to breast cancer were considered to be "more in keeping with an effect on initiation rather than on promotion". It may further be noted that no mechanism of action, whether promotion, initiation or other, could be determined for the carcinogenic effects of oestrogens on salivary, ovarian, renal and thyroid
103M. Metzler, "Genotoxic Potential of the Natural Sex Hormones -oestradiol and Testosterone, and of the Synthetic Compounds 17 -trenbolone and Zeranol", 6 February 1997.
WT/DS26/R/USA
Page 66
cancers, and malignant melonoma in humans. A most recent report has further summarized evidence that parent oestrogens and their catecholamine metabolites induce several types of DNA damage, adduct formation, and gene mutations. Additionally, the author concluded, on the basis of experimental evidence, that "hormonal activity of oestrogens was considered to be necessary but not sufficient for tumour induction to occur".
"There is no scientific basis for making distinctions between genotoxic and epigenetic carcinogens on the basis of available bioassay data. Thus, there is no basis for attempts to derive threshold levels for hormonal anabolics, or other carcinogens, from such data. ...
"It should further be emphasized that extrapolation from high dose bioassay data is likely to underestimate, rather than overestimate, the carcinogenic effects of low dose chronic human exposure. A recent publication endorsed this conclusion with particular reference to metabolic considerations. "Limited evidence would indicate that proportionately less active metabolite is formed at high concentrations where Phase 2 enzymes predominate, while at lower concentrations pathways leading to active metabolites are favoured. The overall effect would lead to an underestimate of risk from high dose animal experiments when extrapolating to low level, chronic human exposure".
"Apart from the invalidity of distinctions between genotoxic and epigenetic carcinogens, it is generally accepted that threshold levels cannot be determined for genotoxic carcinogens. While hormonal anabolics are inactivein bacterial gene tests, and hence dismissed as epigenetic, they are nevertheless clastogenic in mammalianevidence of genotoxicity is clearly more relevant to human cancer than are data based on bacterial gene tests.
"There are other cogent reasons for rejecting the threshold hypothesis. These include:
The enhanced sensitivity of neonatal rodents to the carcinogenic effects of anabolic hormones, supported by substantial evidence on the increased susceptibility of infant rodents and humans to a wide range of carcinogens, including natural anabolics and diethylstilbestrol. (IARC, 1989, NRDC, 1989)
Synergistic interactions between different anabolics administered in combination. Illustrative, are the synergistic effects of oestrogen and progesterone in the induction of mammary tumours in mice. (IARC, 1987)
The possibility of additive and/or synergistic interactions between natural and synthetic anabolic carcinogens and endogenous hormones, particularly in infants.
Synergistic interactions, not as yet investigated, between anabolic carcinogens and carcinogenic and/or xenoestrogenic contaminants in meat products, such as chlorinated hydrocarbon pesticides.
The absence of routine monitoring and residue analysis for parent anabolics following their legal administration, and of sensitive and practical analytic techniques further preclude attempts to estimate threshold or Acceptable Daily Intake levels for hormonal anabolics. Still further complicating problems of residue analysis is the impracticality of assaying for biologically active oestrogen metabolites."104
The European Communities also presented the arguments of Dr. Pinter that:
104S.S. Epstein, "Report to the EC on Cancer Risks from Hormonal Meat Products", 5 February 1997.
WT/DS26/R/USA
Page 67
"... there is evidence that endogenous and exogenous hormones represent potential carcinogenic risk to humans. The risk is associated with the level of hormones, the time of exerting the hormonal effect and the general status of the hormone-responsive organs. Exogenously administered hormones have been proved to be carcinogenic to experimental animals and there is also evidence for hormones to be casually associated with human tumours. In contrast to many animal carcinogens, in the case of hormones, one has to bear in mind that we deal with "human" carcinogens. Therefore, any consideration should be dealt with more seriously than with "animal" carcinogens. Although the level of endogenous hormones varies greatly during the life time, there is no evidence that low level of additional hormones do not represent additional risk.
"Oestrogens and/or their metabolites can react with DNA causing DNA damages, can alter proteins including tubulin, resulting in aneuploid. There is also evidence that this effect is different from the hormonal one, therefore relying on no-hormonal effect might be inappropriate. Administration of oestrogens and progestins in humans is proved to be tumorigenic. In case of hormone mixtures and their residues, it is possible that similar risks exists. Although the level of administered or ingested mixtures may be different, the tumorigenic hazard should not be excluded.
"All effort should be made to avoid additional hormonal effect unless it is absolutely necessary (mediation, etc.). The risk associated with consumption of hormone-containing meat products can be regarded as unnecessary risk which can be avoided."105
The United States argued that the work referenced by the European Communities did not provide evidence of any risk involving residues of the six hormones when used for growth promotion purposes. Contrary to the EC suggestion in submitting the reports it had requested from individual scientists, oestradiol-17ß was not genotoxic when administered by the oral route. Most of the studies cited in these reports involved compounds that were genotoxic but were not related to oestradiol-17ß (or to any of the other five hormones involved in the dispute) either structurally or biologically. The only studies presented as evidence that oestradiol-17ß was genotoxic discussed a Syrian hamster model system. The hamsters were given oestradiol-17ß at doses millions of times higher than the levels present in food, as shown below. The route of administration was directly into the animal (systemic) not oral. The purported DNA damage (adduct formation) was attributed to a minor metabolite of oestradiol-17ß, not oestradiol-17ß per se. The studies failed to demonstrate that the adducts were permanent and irreparable- a necessity for compounds that were genotoxic. Rather the studies suggested that the DNA changes were secondary to the cell’s inability to inactivate these extremely high doses of oestradiol-17ß, a common finding in toxicology.
The United States observed that the dosage used in Dr. Liehr’s study was 61µg/day in male Syrian hamsters weighing approximately 100 g. This was 610 µg/kg/day. For a 60 kg adult human male, this was the equivalent to 36.6 mg/day. On average an adult male produced 48 µg of oestradiol/day. This meant that Dr. Liehr’s study involved delivery to male hamsters of 762.5 times more than the comparable average daily production rate in adult men. The average daily production rate in men was, in turn, 15,000 times more than the residue of oestradiol in meat from treated animals. The difference between the dose used in treatment of Syrian hamsters and the residue in 500 g of meat from treated animals was approximately 11.5 million. In other words, if one extracted all the oestradiol out of 11.5 million 500 g portions of meat, and injected all this quantity of oestradiol every day for a number of months into human males, one would replicate the dose used by Dr. Liehr.
The United States maintained that the final step in the scientific process was to determine that the data generated in a model system accurately predicted the adverse effects in humans. The scientific
105A. Pinter, "Some Aspects of Hormonal Carcinogenesis", 5 February 1997.
WT/DS26/R/USA
Page 68
data on the oral activity of oestradiol-17ß demonstrated that the results obtained in the Syrian hamster model system were not relevant for assessing the toxicity of oestradiol-17ß administered orally to humans.
The United States argued that a more definitive test, a two-year rodent bioassay, had already been conducted for oestradiol-17ß and the results of the bioassay together with the results of other short-term tests had been reviewed by various scientific experts and used in determining the safety of oestradiol-17ß. Based on the results of these tests and other relevant scientific data, it had been determined that oestradiol-17ß was not a genotoxic carcinogen. Since long-term in vivo carcinogenicity tests had already been conducted for these hormones, the results of additional short-term assays were of little value. The full scientific process necessary to make the regulatory decision that oestradiol-17ß was not a genotoxic carcinogen had been completed. The very low levels of oestradiol-17ß present in food were not adding to the cancer risk of women, children or men.
Furthermore, the United States observed that oestradiol was the same whether produced endogenously in women, naturally present in food of animal origin, or present as a result of treatment of animals for therapy or growth promotion. Oestradiol-17ß was present in milk, including human breast milk. Oestradiol content was very high in human breast milk early in lactation. Although infants and young children were very susceptible to genotoxic agents, there was no evidence that oestradiol-17ß was genotoxic in this very sensitive population. Genotoxic agents caused irreparable damage to cells. Newborn human infants were orally exposed to relatively large doses of oestradiol-17ß in human breast milk with absolutely no evidence of genotoxic or other harmful effects.
The United States also argued that there was no evidence that use of the hormones in question resulted in adverse health effects to treated animals.
The United States acknowledged that scientists did not claim to know everything about everything and noted that scientific knowledge was always progressing and evolving. However, lack of knowledge could not itself be the basis for taking a sanitary measure. The SPS Agreement required the European Communities to demonstrate scientific evidence of a particular risk. Accordingly, to claim that a Member was justified banning an activity wherever there were still areas for science to explore would be to render the SPS Agreement meaningless: there would always be areas for science to explore.
The European Communities responded that it had not argued in this case thatlack of knowledge by scientists of all aspects of hormones' mode of action could alone justify the prohibition on their use for growth promotion. What the European Communities had argued was that it was concerned with substances which scientists knew were potentially carcinogenic. This scientific evidence, coupled with the other possible sources of risks, enabled a Member to adopt a precautionary approach by prohibiting their use for growth promotion. This was all the more so when under a proper risk assessment, performed in accordance with what was now required by Articles 5.1 to 5.6, the European Communities had concluded that no other measure was reasonably available to it, taking into account technical and economic feasibility, that would have achieved its appropriate level of sanitary protection and which would have been significantly less restrictive to trade.
The European Communities emphasized that the real policy question on how regulatory authorities should treat hormones depended on the definition of the terms "harm" and "risk". If risk was defined as "the probability (either measured or estimated) that harm might occur from exposure to these hormones"106, Members were free to decide whether such measured or estimated risks were desirable from their regulatory policy point of view. No scientific report to date, including the 1988 JECFA Report on which Codex had based its recommendations, had ever suggested that there was no potential
106J. Bridges and O. Bridges, "Hazards of Growth Promoting Agents and Strategies of Risk Assessment", 1995 EC Scientific Conference Proceedings, p.247 seq.. and "Risk Assessment Strategies for Xenobiotics", 1995 EC Scientific Conference Proceedings, p.365 seq.
WT/DS26/R/USA
Page 69
risk from the use of these hormones. The simple fact that Codex had considered it necessary to recommend MRLs for some of these hormones was sufficient proof in itself that there was such a risk. Such a risk might also result from their administration, if this were not done in accordance with good veterinary practice.
The United States argued that the European Communities attempted to imply that there were large gaps in the knowledge about hormones, and that this gap compared unfavourably to scientific knowledge about other types of compounds. However, in the US view, scientists knew more about the nature and mode of action of hormones than they did about most, if not all, other classes of compounds. The United States asserted that the European Communities was wrong in claiming that the fact that there were ADI levels or MRLs for a substance was evidence that there was a risk from the substance at levels below that ADI or MRL. The concept of ADIs and MRLs was only applicable to residues of substances that did not occur naturally, such as trenbolone and zeranol. The United States explained that using laboratory animals, scientists determined at what level the hormones had any effect on animals (not necessarily an adverse health effect, but simply where the hormone started to have a hormonal effect). Then, using a large safety margin (100 times or even 200 times), the scientists calculated what level of these hormones the most sensitive member of the human population could be exposed to every day of their lives and still have no effect. This was the ADI level (acceptable daily intake level). The scientists then determined what was the maximum residue level (MRL) that food could contain and still ensure that the ADI was met. This MRL was calculated using conservative consumption estimates, for example assuming that a person ate 500 grams of beef per day. There were no evidence that there was any risk of harm to human health at levels below the ADI and MRL. Moreover, the European Communities' claim assumed that if there was any risk of harm from hormones at high levels of consumption, then there must be a risk, albeit smaller, from low levels. The United States argued that this was unsupported by evidence and was also untrue in the case of a number of common foods. These common foods were safe at normal levels of consumption, but posed a risk of harm to human health at very high levels. Some examples of this were potatoes, which contained alkaloids which were toxic in high quantities; lima beans, which contained cyanide which was toxic in high quantities; and rye and other small grains, which in Europe often contained ochratoxin-A that at high levels could cause kidney disease and cancer.
The European Communities responded that the "safety factor" applied when setting the values of ADI and MRLs did not provide the necessary protection, especially in the case of potentially carcinogenic substances whose mode of action was not clear, as the example of the hormone DES had shown. The "safety factor" was in reality no more than a useful tool and applied to those scientific reports which depended on extrapolations from imperfect data on laboratory animals to setting ADIs and MRLs for human beings. The European Communities argued also that the use of simplistic examples of a number of common foods (eg. potatoes, lima beans, rye and other small grains) bore no relevance whatsoever to the scientific issues in discussion. The United States had not explained how many kilogrammes of potatoes one would have to consume first before absorbing alkaloids in toxic quantities. Conversely, a very minuscule portion of meat from an animal to which these hormones had been improperly administered might contain a very high dose of residues which were dangerous to human health. The European Communities argued that the example of the premature sexual development and ovarian cysts involving about 3,000 Puerto Rican infants and children demonstrated the above point.107 In addition, potatoes, lima beans and the other products mentioned by the United States had always formed part of the normal diet of human beings and naturally-occurring substances in these products had entered the metabolism of the human body throughout the course of human evolution and could not be compared with exogenously administered carcinogenic substances given to animals for growth promotion purposes. The European Communities argued that, faced with the identified cancer risks, it was entitled to be as cautious as possible and noted that the United States actually applied
107A. Pérez-Coma, C.A. Saenz, "Anomalous Sexual Development in Puerto Rico - 28 Years of Experience".
WT/DS26/R/USA
Page 70
the same type of sanitary and phytosanitary policy with regard to potentially carcinogenic food additives through the "Delaney clause".
The United States clarified that it did not apply a zero risk policy to all potential risks presented by the use of a food additive. To receive approval for use in the United States, the sponsor of a food additive must demonstrate that the additive was "safe" for its intended uses. Under US law, "safe" did not mean the absence of all risk but rather a "reasonable certainty" that no harm would result from the intended use of the additive. The same standard also applied to the setting of tolerances for pesticide residues in raw agricultural commodities and processed foods. This concept of safety was articulated in both the US Senate and House reports to the 1958 bill enacting the Food Additives Amendment, which stated that "[t]he concept of safety used in this legislation involves the question of whether a substance is hazardous to the health of man or animal. Safety requires proof of a reasonable certainty that no harm will result from the proposed use of an additive. It does not - and cannot - require proof beyond any possible doubt that no harm will result under any conceivable circumstance". 108
The United States noted that the European Communities in essence had argued that there was not a threshold level below which residues of these hormones would be safe.109 In other words, the European Communities claimed that any residue, no matter how infinitesimal, posed a risk of carcinogenicity. Yet the European Communities had not provided scientific evidence to support this position. Moreover, if the European Communities truly believed that there was no threshold, and were unwilling to accept any risk from residues of these hormones, then the European Communities would have to ban the use of hormones for herd management and other purposes. But the European Communities did permit the use of hormones for herd management and other purposes. In the US view, the European Communities’ own arguments belied its position. For example, the European Communities had claimed that the levels of hormones that occurred naturally in meat and other foods were not a risk to human health becausehumans evolved alongside them. This conclusion, while lacking any scientific foundation, nonetheless demonstrated that the European Communities believed that there was a threshold of exposure below which there was no risk.
The European Communities contended that it did not permit the administration of the three natural hormones "to entire herds of cattle". Although the European Communities did not possess the exact figure of the number of animals treated for therapeuticor zootechnical purposes, partly because some of the EC member States did not keep such a detailed statistical record, it could be estimated from those EC member States which did keep such records (e.g. Denmark, The Netherlands and Finland) that only between 1 per cent and 2 per cent of breeding cattle in the European Communities were treated each year for such purposes. This percentage corresponded to about the same proportion of total bovine meat of EC origin consumed in the European Communities. This might be partly explained by the fact that there were now better treatments available which were not based on these hormones. Regardless of the insignificant quantities involved, what needed to be underlined was that allowing the administration of the three natural hormones for therapeutic or zootechnical reasons was in full compliance with the European Communities' policy of ensuring no residues of hormones in meat for human consumption since the strict conditions imposed by EC law effectively ensured its policy objective (no residues at all).
The United States responded that it was wishful thinking, not fact, to think there were no residues present at the end of a withdrawal period. As had been confirmed by the experts advising the Panel, residues might be below the level of detection, but there would always be some amount left. This was because hormones (as almost all compounds, with very few exceptions) depleted by first-order kinetics. This meantthat after one half-life, only 50 per cent of the original amount remained;
108H.R. Rep. No. 2284, 85th Cong., 2d Sess. (1958); S. Rep. No. 2422, 85th Cong., 2d Sess. (1958).
109See para. 4.135 with respect to EC arguments regarding the threshold approach.
WT/DS26/R/USA
Page 71
after 2 half-lives, 25 per cent of the original amount remained, and so on until the amount remaining became vanishingly small but still present. The EC claim only demonstrated the overbroad nature of the EC ban. If the European Communities had been correct in thinking that there were no residues after a particular waiting period, then the European Communities at a minimum would not have been entitled to prohibit all imports of meat from animals to which any of the six hormones had been administered for growth promotion purposes, but only meat from animals that had not been subject to a sufficient waiting period. Similarly, instead of prohibiting imports of all animals, the European Communities could have prescribed a waiting period for those animals prior to slaughter. The EC view would have meant there was no basis for an absolute ban with respect to the three synthetic hormones either, since under that view there would be no residue of these hormones after a sufficient waiting period.
The United States also responded that the EC claim that it had no solid information about the percentage of meat produced in the European Communities from animals to which hormones had been administered for herd management and other purposes was at odds with the constant EC reference to the strict controls on such use and the need to identify every single animal to which they had been administered. Industry sources in the European Communities informed the United States that 3.75 to 4 per cent of all cattle were treated each year, based on an average of information, known in Belgium, France, Germany, the Netherlands and the United Kingdom. In addition, about 6 per cent of sheep in the European Communities were treated each year. In 1995, from one company alone, 2 million doses of hormones were sold (for sheep) and ½ million doses were sold for cattle. Furthermore, by stating that hormones were administered to 1 to 2 per cent of the EC herd each year, the European Communities failed to note that this figure should be multiplied by the number of years that the animals in the herd remained in production in order to determine the quantity of the meat supply from animals that, during their lifetime, had been administered these hormones. For example, if the hormones were administered to different animals each year, then over the course of 5 years, 5 to 10 per cent of the herd would have been administered these hormones. Alternatively, if the hormones were administered to the same animals, then an animal may have received five separate sets of treatment. Furthermore, there was no requirement that the hormones administered always be the same substances. Nothing prevented an animal in the European Communities from having been exposed to multiple hormones (or in the EC view, "combinations" or "cocktails" of hormones).
The European Communities argued that other risks arose from the hormones at issue because of their metabolites. When drugs were introduced into the body, their pharmacological and toxic effects were directly correlated to their concentration in tissues and fluids. The concentration of a drug was a function of its resorption, excretion and metabolism. Drugs were metabolized (broken down) and their metabolites might also have pharmacological effects, which might differ from the effect of the parent drug. Metabolites might have different side or toxic effects. Anabolic steroids had a large variety of metabolites, some of which were only recently identified. The pharmacological and toxic behaviour of many of these metabolites were still unknown. In addition, the same parent substances might produce different metabolites in different species, making the extrapolation from studies on laboratory animals to humans unreliable.
The European Communities asserted that there were scientists who argued that knowledge about the toxicity of the metabolites of these hormones was as yet very limited. Moreover, metabolites of such substances which, even in low concentrations, might have highly toxic effects were generated in the human body.110 These scientists argued that "the use of hormones in growth promotion of animals should not be allowed, as it cannot be excluded that unchanged agents, their metabolites and, above all, unknown highly effective and toxic metabolites are distributed with the meat which is purchased
110W. Schänzer, "The Illegal use of Anabolic Agents in Sport", 1995 EC Scientific Conference Proceedings, p.352.
WT/DS26/R/USA
Page 72
by consumers". 111 Moreover, in some of the experimental studies on which the 1988 JECFA Report was based, no extensive research and risk assessment of the potential risks arising from the metabolites of the hormones under consideration had been carried out.112 Noting that despite all the uncertainties, JECFA had decided to establish a temporary ADI for trenbolone acetate while requesting additional information to be submitted to it by 1990, the European Communities argued that in light of the arguments advanced by some scientists, it was questionable whether the decision of the JECFA could be regarded as a reasonable one. The European Communities remarked that the United States had already approved the use of trenbolone for livestock growth promotion in 1987, whereas Codex had established its final recommendation only in 1995. In 1987, the Standing Commission on Hormone Toxicity of the German Society for Endocrinology and the Commission for Toxicology of the German Society for Pharmacology and Toxicology had stated that "the present information in respect to a likely genotoxic potential did not allow a final evaluation for trenbolone".113
The United States argued that EC claim that there were health risks arising from metabolites of the hormones were unfounded. The European Communities appeared to simply rely on the fact that "there are scientists who argue" that "knowledge is very limited" and metabolites"may have adverse effects", with no evidence of any such adverse effects, much less in relation to specific hormones or to the levels of exposure from the use of hormones for growth promotion purposes. Furthermore, the European Communities cited studies on the illegal use of anabolic hormones in athletes which had nothing to do with residues of hormones in meat. The United States disagreed that Codex had failed to examine potential risks from metabolites. The additional studies on metabolites requested by the 1988 JECFA report had been performed and accepted in 1989. The Lamming Committee had also examined the safety of the metabolites for the hormones it had reviewed and had found them to be safe. Moreover, in the United States, metabolism of a compound was an integral part of the safety assessment of all animal drug products prior to approval of the drug or the setting of a final MRL for the hormone products. All scientific evidence showed that there were no risks of harm to human or animal health due to residues of any metabolite of the six hormones when used for growth promotion.
The European Communities responded that data derived from the directly-observed effects of these hormones on human beings (whether for therapeutic purposes or by athletes) was equally or even more important than data derived from laboratory animals. The fact that these anabolic steroids were used "illegally by athletes" did not diminish the relevance of the EC argument, since themetabolites of these substances, whether used legally or illegally, were the same. The 1988 JECFA Report on which the United Statesbased its contentions in this case statedthat information from human case reports on drugs also used therapeuticallyin human beings "may provide important evidence of possible adverse effects in human beings that are not detectable in animal models". The Codex had also noted that "[h]uman data of insufficient extent may be considered in the setting of [a Codex Acceptable Daily Intake] on the basis that extrapolation of data from animal studies contain greater uncertainty".
111Ibid., p.353.
112As an example, the European Communities noted that, as regards trenbolone acetate (TBA), the JECFA had stated that "[i]n the absence of satisfactory toxicological data the Committee was unable to establish a separate no-effect level for the a-TBOH metabolite. It also noted that this metabolite was not produced in significant amount in the rat, which made it inadvisable to extrapolate from data generated from ß-epimer, experiments in that species".
113B. Hoffmann, "Problems of Residues and Health Risks of Anabolic Agents with Sex Hormone-Like Activities", 1995 EC Scientific Conference Proceedings, p.291.
WT/DS26/R/USA
Page 73
Referring to scientists' reports at the 1995 EC Scientific Conference of evidence of the illegal use of growth-promoting substances in many European countries and elsewhere114, the European Communities submitted that the illegal use of mixtures of veterinary drugs and growth promoters might result in unpredictable residue levels in edible foods, which might constitute a risk for the consumer. This might be due to variations in physiological and pathological parameters, altering drug disposition and elimination. Biotransformation patterns of synthetic drugs and hormones might differ markedly depending on the animal species and gender. Interactions might occur between compounds as had been reported for sulphadimidine and oestrogenic and androgenic hormones administered to dwarf goats. This might require longer withdrawal times in order to arrive at safe residue levels. For those reasons, the 1995 EC Scientific Conference had concluded that "the use of combinations (or cocktails) of hormones poses serious risks for the health of consumers because the use of mixtures, which usually contain illegal substances, is made in the form of injections, rather than implants, and this is an impediment to their detection at slaughter. This potentially allows the sale of meat containing very high concentrations of hormones".115
In the European Communities, the terms "combinations" and "cocktails" were used interchangeablyto denote the higher potency and multipliednegative effects that such substances exerted when used in combination, compared to when used individually. Whether the combinations were legally marketed was irrelevant, as there was nothing to prevent an unscrupulous farmer from administering one, two, three or even more of these legally available substances, in an attempt to yield potentially higher growth rates. The European Communities stressed that the use of cocktails in the United States did not appear to be subject to strict control, since a number of products were freely available on sale in the market. On the other hand, several of the licensed products in the US market were commercially manufactured combinations of hormones, with adjuvants (i.e. additives used to speed up or slow down the active element of the combination).
The European Communities argued that when JECFA had established its recommendations on the ADIs and MRLs of the hormones at issue, it had considered only information on the single substances, although scientists had agreed that the concept of ADI was not applicable to assess the risk from exposure to mixtures (or cocktails) of compounds. The 1988 JECFA Report had, however, pointed clearly to the need to examine the effects of combinations:
"The Committee noted that several of the hormonally active substances on the agenda were used in combination one with another, and recommended that, where substances having similar physiological activities were combined, evidence that their hormonal effects were additive, rather than synergistic, should be provided. The Committee agreed that data on the residues of each of the substancesthat are used in combination should be available for evaluation, whether or not their physiological activities were similar."
The European Communities concluded that there was considerable potential risk arising from the use of combinations, containing authorized or unauthorized hormones, and that existing international rules did not deal adequately with this potential source of risk.
The United States submitted that the EC claims regarding health risks arising from the use of hormones in combination were unfounded. There was a difference between combinations of hormones and theso-called"cocktails." Combination products contained only authorized hormones and adjuvants.
114"Report and Conclusions", Steering Committee, 1995 EC Scientific Conference Proceedings, p.9.
115Ibid., page 9, and H.A. Kuiper, "Risk Assessment Strategies for Xenobiotics", 1995 EC Scientific Conference Proceedings, p.376.
WT/DS26/R/USA
Page 74
The product intended for commercial sale was the product tested and reviewed in the United States and subjected to both animal and food safety analysis prior to release for commercial purposes. Furthermore, they were manufactured according to required good manufacturing practices so that they were consistent in identity, strength and purity. Since hormones might be legally purchased in formulations that had been demonstrated to be safe and efficient, there was no reason for a farmer or a rancher to buy other combinations and use them illegally. Furthermore, because approved formulations were tested for optimal growth promotion effects, the addition of other hormones or higher dose levels, under conditions different from the label, did not result in additional growth effects. Accordingly, such use would result in increased economic costs with no accompanying economic gain.
On the other hand, the United States observed that the term "cocktails" had no legal definition and was generally used to refer to the mixture of illegal drugs prepared in unlicensed facilities that were sold on the black market in EC member States. The United States agreed that the use of such cocktails was dangerous and stressed that they were illegal for use in the United States. Finally, it was not clear how concern over the use of illegal "cocktails" could justify the EC ban on the use of strictly regulated single or combination products, or even how the concern over "cocktails" could justify the EC ban on the use of all hormones for growth promotion purposes, including the administration of single hormones.
The European Communities claimed that illegal marketing of hormones and veterinary drugs in general in the United States might be as great as, or even greater than, in the European Communities. The European Communities referred to a 1985 report produced by the US Congressional Committee on Government Operations, entitled "Human Food Safety and the Regulation of Animal Drugs". This report noted that "... the problem of illegal veterinary prescription drug sales is of such magnitude and pervasiveness that it threatens the credibility of the veterinary drug approval and the regulatory process". These conclusions were based on a focused investigation by the US Food and Drug Administration which found that "indiscriminate sale of prescription veterinary drugs" was "pervasive" throughout eastern Iowa, and that other, ongoing investigations seemed to confirm that "... the illegal sales problem was probably nationwide in scope".
The European Communities contended that the risks that might potentially arise from the multiple exposure of humans to hormones and other chemical substances were similar to those resulting from the illegal use of combinations (or cocktails) of hormones. All foodstuffs, whether of animal or plant origin, were likely to contain trace amounts of several substances derived from various sources, a fact currently not addressed in any risk assessment strategy for food. Because the ADI and MRL assessed the safety of a single compound and not the exposure to mixtures of compounds, a sufficient protection of humans was only guaranteed if the MRL was not exceeded. Thus, there was a substance-related risk to consumers if the acute ingestion of a residue amount exceeded the established MRL and possibly the ADI. But the question also arose on how to protect consumers from a potential adverse effect resulting from long-term treatment with growth- promoting substances. Such adverse effect was the possibility that the biotransformation of other compounds might be altered, as had been shown in the case of steroids. This might result in altered residue kinetics for those compounds. 116 The European Communities claimed that the idea that there might be a risk associated with the long-term exposure to a mixture of substances, resulting in a possible biotransformation of other compounds, was a precursor to the so-called "precautionary" principle and it was at the heart of the policy followed by the European Communities on such issues.
The United States claimed that with respect to multiple exposure, the European Communities had once again failed to identify any particular risk or provide any evidence or risk assessment demonstrating a risk. Instead, the European Communities had simply mused about the "idea that there
116Working Group II, "Assessment of Health Risk", 1995 EC Scientific Conference Proceedings, p.20.
WT/DS26/R/USA
Page 75
may be a risk". In this musing, the European Communities had also chosen to ignore the fact that one of the reasons for applying a safety factor of 100 (200 in the case of MGA) was to take into account, among other things, the possibility of exposure to more than one compound over a life time.
The European Communities responded that the Codex itself had noted that "[u]ncertainty in the safety evaluation process is primarily addressed through the use of safety factors. Their respective values are arbitrary and have no measured biological significance, however, their appropriateness is somewhat borne out by experience".117 The safety factor, therefore, might not provide the necessary protection, especially in the case of potentially carcinogenic substances whose mode of action was not clear, as the example of the hormone DES had shown. The "safety factor" was in reality no more than a usefultool and applied to those scientific reports which depended on extrapolations from imperfect data on laboratory animals to set ADIs and MRLs for human beings. Although these concepts had served well in the past and would continue to do so in the future, a Member was entitled to deviate from these concepts as used by Codex in establishing recommendations. Codex had proven to be useful in setting guidelines and recommendations for food additives precisely because it had left its members free to decide whether they would accept its recommendations. The SPS Agreement was cast largely on the same basis, because it allowed Members to depart from international standards, guidelines or recommendations under the conditions laid down therein (e.g. Article 3.3).
The European Communities noted that none of the scientific conferences and expert review groups (including the Lamming Report and the 1988 JECFA Report), which had examined the potential risks from the use of these hormones for animal growth, had concluded in favour of an unqualified use of these hormones for growth promotion purposes. This was because the administration of these hormones to laboratory animals, at high doses but also at low doses in some experiments, had caused serious health effects, including cancer. Furthermore, all the scientists which had expressed themselves in favour of allowing the use of these hormones at low doses for growth promotion purposes did not exclude risks to human and animal health from improper administration or from combinations and metabolites of hormones and multiple exposure. This critical caveat on conditions of use had been recognized in the EC Commission proposal (COM(84)295 final) which finally led to the adoption of Directive 88/146/EEC. This Directive proposed to allow the use of the three "natural" hormones only under the following conditions in order to safeguard public health:
only in the form of an implant to be administered in a part of the animal discarded at slaughter (usually the ear);
to an identified animal only, to allow control of the withdrawal period;
only by a veterinarian;
substances allowed to be administered must be on an EC list, setting out clearly conditions of use; and
these substances must be shown to be effective and safe.
The European Communities observed that the Lamming Report had also urged the attention of the European Commission to the need to lay down certain essential provisions before the three natural hormones could be used for growth promotion, in particular as regarded the following:
117Codex Alimentarius Commission (1993), "Risk Assessment Procedures used by the Codex Alimentarius Commission and its Subsidiary Advisory Bodies", p.11.
WT/DS26/R/USA
Page 76
"(a) Instructions for use:
Specification of the doses, the type of pharmaceutical preparation, the number and frequency of administrations.
Association of anabolic agents.
Localization of implant and ablation of zone treated.
Withdrawal period before slaughter.
Identification of animals treated, with indication of the period of treatment.
(b) Surveillance programme and analysis methods
Control of production and trade in anabolic agents.
Veterinary control of authorized uses.
Means and methods of control."
The European Communities noted that JECFA had recognized that the use of these hormones was unlikely to pose a hazard to human health if used in accordance with good animal husbandry practice, but that there were no internationally agreed rules on what constituted "good animal husbandry practice" for the administration of these hormones. As far as "good veterinary practice" was concerned, the 1988 JECFA Report had noted that the Code of Practice for the Use of Veterinary Drugs was under preparation by the Codex Committee. This Codex Code of Practice had finally been adopted in 1993.118 However, it was not clear how strictly this Code was adhered to by countries. The European Communities wondered what concrete measures the United States had taken in the case of hormones to comply with the Codex Code.
The European Communities submitted that difficulties arose because the identification of animals or meat from animals which had been treated with the three natural hormones was difficult due to the wide variation in natural levels of these hormones arising because of factors such as sex, age, diet, physiological condition, etc. While residue analysis for natural hormones in treated animals could experimentally show differences between them and untreated animals, it was currently not possible through routine tests to identify treated animals on the basis of assays of edible tissue samples, and it was also difficult using blood, urine or faeces. A country allowing imports of animals and meat from animals treated with hormones would have to rely largely on checks and controls on use of hormones carried out by the exporting country if trade was not to be impeded, but the importing country would also have to carry out checks and controls in order to ensure that its consumers ate meat which conformed to the requisite standards. Detection and surveillance of hormones was based on screening methods, principally immunoassays (radio or enzyme immunoassays) because they were rapid, relatively low cost and could be applied simultaneously to a large series of samples.119 However, the European Communities claimed that all the available methods used for detection and control had their limits and drawbacks. Furthermore, although Codex had established an ad hoc working group on the matter, there were still no internationally agreed rules for the methods to be used for residue control programmes.120 The European Communities argued that in light of the potential risks to human and animal health resulting from the uncertainties in residue control programmes and the absence of
118International Code of Practice for Control of the Use of Veterinary Drugs, adopted by Codex in 1993.
119M-L Scippo and G Maghuin-Rogi ster, "Methods of Detection and Surveillance of Natural Sex Steroid Hormones", 1995 EC Scientific Conference Proceedings, p.541 seq.
120The European Communities stressed that in 1991 the Codex Commission had only established provisional methods to serve as a potential source of use. See Joint FAO/WHO Food Standards Programme, Residues of Veterinary Drugs in Food, Vol. 3, 2nd Ed., 1993.
WT/DS26/R/USA
Page 77
internationally agreed rules, any Member (including the European Communities) was entitled to adopt a regulatory system to achieve the level of protection for its public that it considered to be appropriate.121
The United States claimed that the European Communities had moved to a discussion of "control, inspection and approval procedures" which were covered in Article 8 and Annex C of the SPS Agreement, because it had been unable to produce any risk assessment or evidence with respect to the substances themselves to support its ban. The experts advising the Panel had confirmed that even where there was a failure to follow good animal husbandry or veterinary practice, resides of the three natural hormones administered for growth promotion purposes might well still be within the normal physiological range. The European Communities was unable to identify any particular risk or provide any evidence or risk assessment demonstrating a risk. The European Communities could not explain why any risks from its failure to adequately detect residues of hormones would be higher in the case that it permitted the importation of animals to which any of the six hormones had been administered for growth promotion purposes compared to its current ban. The United States stressed thatpresumably, under its current ban, the European Communities still ran the risk that it would not be able to detect all banned animals and meat. Furthermore, the European Communities had not explained why it was unable to implement MRLs for the six hormones when it was able to implement MRLs for other foods and substances.
The European Communities agreed that a potential for meat to slip past its detection systems existed, and that it could not expect to detect every instance of illegal use. But the European Communities stressed that it operated comprehensive and costly residue control programmes designed, and effectively applied, to detect, as far as technicallypossible and economically feasible, such residues from all sources.122 The fact that there was a potential for prohibited residues to evade its control and detection was not in itself a reason for the European Communities to adopt the US approach on this issue. To the contrary, this risk of residues of potentially carcinogenic substances evading detection was an important factor in carrying out a proper risk assessment and seriously influenced, right from the beginning, the decision of cautious EC member States in setting the level of sanitary protection they deemed appropriate in their territory. The European Communities also did not dispute the fact that MRLs were useful tools in regulating drugs and food additives in appropriate cases. But it was obvious that in the case of mass medicationof the food animal population, the amount of testing required to apply the EC level of protection would be very much higher than under the present EC legislation. It was equally obvious that the chance of detecting violations would be much lower. The European Communities argued that the difficulties arose because exporting countries (like the United States) were not willing and capable of ensuring, on a reasonable basis, that the ADI and MRLs recommended by Codex would not be exceeded in their exported meat. The technical and financial burden had to be shared mainly by the importing Members which wished to see the Codex's recommendations strictly observed.
The European Communities argued that US authorities practically did not control the sale of these hormones either individually or as compounds, and noted that they were freely available on sale, even by mail. There was no testing and checking of residues for the natural hormones and for the
121The European Communities noted that, for example, Argentina prohibited the use of the three natural hormones (while allowing the use of the two synthetic ones), citing as an essential motive for the restriction the lack of suitable methods for measuring hormone residue levels in meat. 1983 OIE Scientific Report, p.411.
122Directive 88/146/EEC explicitly required thatundertakings in the EC member Statesproducing theprohibited hormones, those companies authorized to market these hormones for whatever purposes and undertakings producing pharmaceutical and veterinary products based on those substances, must keep a detailed register recording (in chronological order) the quantities produced or acquired and those sold or used for the production of pharmaceutical and veterinary products. These requirements had now been tightened by requiring the names of the persons to whom such quantities had been sold or from whom they had been purchased to be made available to the competent authorities (Article 9 of Directive 96/22/EC). In addition, heavy fines and penal sanctions, including imprisonment, were imposed when violations were detected.
WT/DS26/R/USA
Page 78
synthetic hormones. The United States had even dropped the three natural hormones from their control programme, even though the United States applied MRLs on them, indicating that no testing had been performed because implant-associated residues could not be distinguished from normal background concentrations. The European Communities added that MGA had been tested for the last time in 1993, when only 22 samples had been checked of which one had tested positive. Zeranol had been last included in 1989 and trenbolone had never been included in the control programme. Moreover, the European Communities noted that the US FSIS had reported a number of cases of improper use of hormones in 1986. More recently, in 1993, the FSIS had stated that "one fourth of one per cent of animals tested showed positive for residue violation. In other words, almost 99.74 per cent of animals tested showed no residue violation". The European Communities argued that the FSIS 1993 data indicated that none or very few of approximately 130 million head of livestock slaughtered in the United States had been monitored for residues of natural or synthetic hormones.
The European Communities claimed that the situation on residue controls and checks in the United States was not clear and concluded that if it had followed the example of the United States in applying the Codex recommended MRLs in the way they were applied in the United States, it would have certainly failed to achieve its level of protection which was no residues of these hormones in meat for human consumption, and it would also have failed to protect animal health. The European Communities was correct in its analysis when performing its risk assessment in the sense of Article 5, that the use of these hormones for growth promotion purposes under the conditions identified by all scientific reports (i.e. in accordance with good veterinary practice) would not be technically or economically feasible. Therefore, the responsible EC institutions had rightly concluded that prohibition of use of these hormones was the only reasonably available least trade-restrictive measure.
The United States first noted that the EC claim that an MRL would not be feasible because of the associated amount of testing for residues was directly at odds with the EC claim that "at least in the European Communities there are comprehensive and costly residue control programmes designed and effectively applied to detect, as far as technically possible and economically feasible, such residues from all sources". Second, the United States noted that for purposes of the EC ban it did not matter how stringent an exporting country's requirements were or what type of programme was in place with respect to the use of these hormones. Animals and meat were banned, period. The issue in this dispute was not whether the United States subjected its exports of animals and meat to any particular controls or whether the US sanitary measures were equivalent to the EC measures123, but whether the EC ban was in accordance with the WTO Agreements.
The United States then noted that contrary to the EC allegations, it had an extensive system in place to ensure that growth promoting hormones were used domestically according to good animal husbandry practice. Sponsors were required to determine conditions for safe and efficacious use of their products and to supply that data to the government for review. Labels for products had to accurately describe these conditions for use and be written in a way likely to be followed in practice. All sponsors were required to manufacture their drug products according to specifications and Good Manufacturing Practices to ensure consistency of strength, purity and identity, and the United States inspected these manufacturing facilities periodically. The United States also required a manufacturer to keep records on the amount of product sold and to whom (the specific retailer); overselling in a particular area relative to need could trigger investigative action. The United States also had authority to investigate live animals on the farm or feedlot to ensure correct use of the products. The United States operated an adverse drug reporting system to ensure that products were being used in a safe manner. At the slaughterhouse, the United States inspected animals to ensure the proper use of the hormones.
123The United States noted that issues of equivalency were examined under Article 4 of the SPS Agreement.
WT/DS26/R/USA
Page 79
These practices, along with directed monitoring (“for cause” inspections) and random sampling through the Residue Monitoring Programme, ensured that public safety was safeguarded. The United States operated a national residue programme (NRP) as part of its regulatory oversight of meat production. The NRP did not apply to the naturally occurring hormones because implant-associated residues could not be distinguished from normal background concentrations. Zeranol, trenbolone and MGA were among the compounds covered by the NRP. The compounds actually tested under the NRP varied from year to year, with emphasis given to those most susceptible to improper use. The most recent year in which hormone growth promoters had been tested was 1991, when 185 steers and 129 heifers were randomly selected nationwide at slaughter for DES/zeranol testing. All findings met FDA tolerances. The last year in which testing had been done for MGA was 1988. All of the 373 randomly selected heifers were in compliance with FDA requirements. Testing could be reinstituted if a problem were perceived. In addition, US slaughter establishments producing meat and offal products for export to EC member States were required to participate in an expanded residue testing programme conducted by private laboratories. Under this programme, the number of samples collected during the last five years, for the cattle, sheep, swine and equine species, had been approximately as follows: zeranol = 3500; MGA = 1300; trenbolone = 3500; and DES = 3500. To the best of the knowledge of FSIS employees working closely with that programme, all of the findings met FDA tolerances.
The United States clarified the conditions of use for these hormones. In the case of cattle in the United States, hormones were not generally administered by individual farmers. Instead, the vast majority of the hormones were administered when the animal first entered the feedlot. At this time, a team under the supervision of a veterinarian assessed the overall health of the animal and provided veterinary treatment as required. An implant was usually inserted by team members at this time. As large commercial operations, the feedlots had great incentive to properly care for these animals, including proper placement of implants, in order to provide the consumer the highest quality product possible. Implants were developed for use with specialdelivery devices designed for ear implantation. The release rate of active ingredient from the implant was maximally effective for growth promotion purposes when implanted in the ear. Implantation at sites other than the ear was difficult in practice as the hide was very tough and not easy to penetrate. Other implantation was also not as effective as ear implantation, might damage the carcass with scar tissue thus decreasing its value, and would entail legal liability and a condemned product. Therefore it was in the feedlot's interests to use the implants under the conditions of use specified on the label.
The United States claimed that with respect to US imports from third countries, any exporting country was required to have a programme in place to ensure that food produced from animals treated with any drug was safe for consumers. This programme had to achieve a level of protection at least equal to the US appropriate level of sanitary protection. The exporting country’s programme must include a systematic national residue testing programme. Exporting countries were required to report the results of their NRP to retain eligibility to export to the US market. To export meat to the United States, all foreign countries had to obtain an annual US residue certification stating that the country maintained, among other requirements, a control programme to ensure compliance with US standards for residues in meat destined for the United States. One element of the US audit of residue control programmes was the evaluation of annual testing results supplied by each country. The United States recommended that each country determine which compounds should be included in its annual residue sampling plan, and provide information with the plan that explained the decisions to include (or exclude) compounds. The selectionof compounds to be included in the testing plan should be based on conditions present in the country, e.g., availability of specific compounds, conditions that warranted use, and the residue hazard when used. In determining whether a country maintained an acceptable residue control programme, the United States considered the following, in addition to the residue sampling plan and test results: (i) US port of entry testing results; (ii) the findings of on-site reviews of plants exporting product to the United States, and laboratories performing the analytical testing; (iii) compound selection, statistical design, and implementation of sampling plans; (iv) trace back and enforcement procedures; and (v) timely receipt of investigative reports following a US port of entry violation.
WT/DS26/R/USA
Page 80
The European Communities submitted that it did not argue that it was "unable" to implement MRLs for growth-promoting hormones, but that it was unwilling to do so. In this it was acting consistently, as it did not operate MRLs for other drugs which had powerful adverse physiological effects, including carcinogenicity, at low concentrations. Therefore, if after conducting a risk assessment a Member concluded that its level of sanitary protection could not be ensured by applying the MRLs recommended by Codex, this Member was allowed to depart from those MRLs. This was the principle on which both Codex and the SPS Agreement were based.
The European Communities referred to the reports of Dr. Pérez-Comas in relation to detection, control and administration of hormones. Dr. Pérez-Comas argued that:
"... anomalous sexual development in Puerto Rico is a major and serious public health problem. (...) We have been detecting an increase of such cases since early in 1969 (...) Up to this date over 1,500 patients have been documented.
"From our patients, we have deducted that environmental contaminants are a strong factor in etiology of the condition that affects children of both sexes, different races and nationalities residing in PR, and all ages in the form of Anomalous Sexual Development. (...) We have documented clinical evidence of increased serum total oestrogens in a significant number of patients, with over 60 per cent of the females with ovarian cysts, variable levels of FSH, LG & Prolactin, that recede or diminish in a significant number of patients after a meat and poultry depletion diet.
"We have been informed that anomalous sexual development cases are increasing in Chile, Bolivia and Colombia. As to our knowledge no definite cause have been found, but poultry and meat contamination is being considered, for in Latinamerica multiple use of growth promoting hormones have been documented. We consider that in PR the Anomalous Sexual Development in children is complex with different etiological factors such as zeranol, other unknown oestrogenic substances, and pesticides. Serious emphasis must be done on the control of these substances and its avoidance in animal husbandry. (...)"124
Finally, the European Communities argued that the US unwillingness to comply with the EC measures was the only obstruction to its exports. But the experience of the "Interim Measure" showed that individual farmers in the United States had no difficulty in complying with the EC level of protection. The number of companies in the United States which promoted their meat to the consumer as having been produced without the use of hormones was significant and growing.
The European Communities claimed that there were additional risks to human and animal health arising from the administration and potential misuse of hormones. At the 1995 EC Scientific Conference, it had been observed that a misuse could be hazardous to consumers, if instead of the authorized implant cheaper black-market products were administered to the animals via injections at undefined parts of the animals' bodies. Moreover, injection could be given in addition to an authorized implant. Such misuse would probably not be analytically detectable, and not only in the case of endogenous substances. If the use of hormonally active substances were authorized, this would probably also lead to a dramatic reduction of the possibilities to reveal such a misuse with the help of detected injection sites, because there would no longer be indications to select animals during the post-and ante- mortem examination that were considered to be suspicious. On the contrary, each animal would have to be controlled for injection sites. Since this was completely unrealistic, a randomised selection of
124A. Pérez-Comas and C.A. Saénz, "Anomalous Sexual Development in Puerto Rico: 28 Years of Experience", 28 February 1997.
WT/DS26/R/USA
Page 81
the carcasses was the only control for injection sites. Consumers could thus be subject to the risk of eating undetected, maybe highly contaminated, injection sites. Irrespective of the assessment of a number of hormonally active substances by the Codex, the principle that the health hazard of a substance depended only on its dosage continued to be valid also for the risk of injection sites. Particularly in the case of illegally administered black-market products, the dosage was not predictable. In the context of preventive protection of consumers' health, it was necessary to take the misuse potential of a substance into consideration when assessing health risks. If health risks to consumers could not be excluded when a substance was misused, single substances with an extreme misuse potential should also be banned even if the substances were not hazardous when legally used. Risks to consumers' health could not be excluded if clenbuterol and hormonally active growth promoters were abusively administered. It was questionable if the authorization of these substances as growth promoters would end their extreme misuse potential; the authorization of a substance distinctly restricted its controllability. Thus the misuse of authorized substances was hardly controllable. Because of the effect on sampling and inspection, the authorization of a substance always went hand-in-hand with a facilitated use of the corresponding black-market product.125
The United States rejected the EC attempt to justify its ban on the grounds that it could not be sure that the hormones would always be administered correctly. It was not clear to the United States how this differed from the purported "detection and control risks", since in both cases the European Communities claimed to be concerned that unauthorized meat would slip past its detection. In any event, the same problems applied to this category of purported risk as applied to the EC detection and control concerns. The European Communities did not identify any particular risk, nor did it produce any risk assessment or evidence of risk. Furthermore, the European Communities did not explain why there was not always a potential for meat containing prohibited residues to evade its detection, even under its current ban. The European Communities had permitted the sale and consumption of domestic meat even in the face of admitted and well-publicized illegal use of hormones in its territory. The European Communities also permitted the importation of meat, including meat from animals to which had been administered hormones for purposes other than growth promotion. There could never be an absolute guarantee that the European Communities could detect every instance of an incorrect administration of hormones. With regard to the potential misuse of these products, the United States claimed that experts had acknowledged that even if misused, the residues of the six hormones might well be below the Codex (or FDA) MRLs (in the case of the three synthetic hormones). Indeed, direct injection (rather than use of an implant) of doses 100 times higher than approved doses still resulted in residues well below the established ADI.
The European Communities responded that it had no difficulty in admitting that a potential for meat to slip past its detection always existed and did not expect to detect "every instance" of illegal use. But at least in the European Communities there were comprehensive and costly residue control programmes designed and effectively applied to detect, as far as technically possible and economically feasible, such residues from all sources, unlike the practice in the United States. The fact that there was a potential for prohibited residues to evade its control and detection was not in itself a reason for the European Communities to adopt the US approach on this issue. To the contrary, this risk of residues of potentially carcinogenic substances evading detection was an important factor in carrying out a proper risk assessment and seriously influenced, right from the beginning, the decision of cautious Members in setting the level of sanitary protection they deemed appropriate in their territory.
The European Communities argued that a potential source of risk frequently mentioned by scientists stemmed from uncertainty about the nature and reliability of the methods used to establish
125B. Jülicher, "Sampling Strategies", 1995 EC Scientific Conference Proceedings, pp.537-539.
WT/DS26/R/USA
Page 82
the relevant data. The procedure for characterising the risk from the use of hormones had remained largely unchanged for 30 to 40 years and scientists had argued that the process of the scientific assessment of the risk should draw on the science base to assess the risk, then address uncertainties, rather than fuse these two elements. However, a much sounder scientific approach was to encourage strongly the investigation of the mechanism of toxicity of each new promoter. Unless more effort were directed to understanding why a particular adverse effect of concern occurred (i.e. assessment of the mechanism of toxicity), risk assessment would always be vulnerable to the criticism that its scientific basis was flawed. Science continued to make rapid advances and no risk assessment could be regarded as the definitive statement for all time. A regular review procedure was needed, the frequency of which should be determined by such factors as the availability of new toxicological data; the level of uncertainty identified in the products' risk assessment; and findings from exposure monitoring investigations. All too often, further investigations of the risk ceased at the point at which a particular promoter was either accepted for use or rejected. Means of progressing the knowledge base, in the areas where substantial uncertainties existed, had to be found.126
The European Communities also submitted that the JECFA 1988 report stated that "insufficient data were available on normal levels in some classes of animals likely to be treated with oestradiol-17ß would fall within the normal ranges for untreated animals in each class". The European Communities argued that this lack of data, however, was thought not to be important and JECFA had based its conclusions on extrapolations from other data. In the end, JECFA did not think it necessary to establish MRLs, because of the impossibility of detecting residues of natural hormones in treated meat.
Furthermore, the European Communities argued that scientists had raised concerns with regard to the confidentiality of data. It had been suggested that a separate risk assessment must be conducted on each growth promoting substance drawing on all the available data, not simply that provided by a company wishing to market a product. This required improved accessibility to information classified as confidential. For this purpose, it had been suggested that a time-limit should be set on the confidentiality of safetydata.127 The European Communities submitted that the rationale for maintaining confidentiality was to maintain commercial secrets. The normal procedure to obtain an authorization to market an animal drug (for instance as followed in the United States) was for the drug sponsor to demonstrate to the relevant authority that an animal drug was safe and effective in the target animals, safe for humans consuming edible products from treated animals, and safe for the environment.
Moreover, the European Communities observed that frequently the relevant authorities of some countries were said to operate under a "zero risk" concept, although zero risk could not be determined and zero residues were probably not achievable.128 But for the general public, the idea that zero risk residues were not achievable was difficult to accept. Rightly or wrongly, public perception frequently was that relying solely on safety data provided by a company was unsound.
The United States claimed that the EC arguments on uncertainty about the nature and reliability of the methods used to establish the relevant data and their confidentiality was misplaced. The European Communities had once more identified no particular risk, offered no risk assessment showing any risk, and produced no evidence of a risk. Its reliance on general uncertainty was again a generalized concern that was no more than unfounded speculation rather than an actual risk. The fact that there were limits
126Dr. J. Bridges and Dr. O. Bridges, "Hazards of Growth Promoting Agents and Strategies of Risk Assessment", 1995 EC Scientific Conference Proceedings, pp.258-259.
127Ibid., p.262.
128The European Communities noted that the United States was supposed to apply a "zero" risk standard for carcinogenic substances under the so-called "Delaney Clause". Furthermore, when presenting the WTO Agreement to Congress for approval, the US Federal Drug Agency had argued that its "zero-risk" policy was not affected by the WTO Agreements.
WT/DS26/R/USA
Page 83
to science was not a basis under the SPS Agreement to maintain a measure. The United States and other Members agreed that there were limits to science, but the Uruguay Round participants had negotiated and agreed to the SPS Agreement with its reliance on science in full recognition of those limits. Moreover, to state that the "evidence" on which the European Communities relied to maintain its measure was "a lack of information on, for example, the mode of action, the effect of administering combinations and the effect of ingesting residues over a long period" and claim as its evidence only that hormones in general exerted powerful physiological effects and were associated, under certain ill-defined circumstances, with adverse effects on health amounted to no more than speculation and scare tactics. In the US view, the European Communities had put forth no scientific evidence to support the existence of these risks.
The European Communities argued that all countries, including the United States, regulated the use of hormones in farm animals. The difference between the European Communities and the United States in this respect was the extent to which their use was regulated. This difference in degree of regulation was a reflection of the different levels of consumer protection adopted by the European Communities and the United States. The European Communities took aprecautionary approach, placing the attainment of a high level of consumer protection before the commercial interests of farmers and pharmaceutical companies. The United States took a different approach. Where there existed a doubt over the safety of a product, the European Communities gave the benefit of doubt to the consumer, especially in cases where the potential risks might affect very large parts of the population. The United States had, in the case of growth hormones, given it to the producer.
The European Communities stressed that the difference in degree of regulation between the United States and the European Communities was due to the greater attachment of the European Communities to the precautionary principle.129 Such an approach was required to avoid situations as those portrayed by many cases of health hazards which only became apparent long after substances or products had been assumed to be safe, such as Thalidomide and DES. Two cases of recent interest in the European Communities illustrated the desirability of taking a precautionary approach to consumer protection: E. Coli and BSE.130
129 D. Freestone & E. Hey (1995), "The Precautionary Principle and International Law".
130The European Communities explained that E. Coli was a bacterium which inhabited the intestines of all animals. In the early days of food bacteriology it had been assumed to be ubiquitous and harmless. In the first FAO monograph on meat hygiene, published in 1957, it was stated that: "There are not convincing reports which directly implicate E. Coli in outbreaks of food poisoning". The second report of the Joint FAO/WHO Expert Committee on Meat Hygiene (1962) had considered that knowledge of the importance of E. Coli in causing food-borne disease was lacking and should be investigated. Even in 1974, Thornton's Textbook of Meat Hygiene stated that animal strains of E.Coli were "unlikely to become established in the human bowel". When the European Communities, in 1964, had adopted its first rules on meat hygiene, it had taken a precautionary approach, requiring hygienic structures, equipment and handling procedures with a view to preventing all bacterial contamination of meat. This approach had been strongly criticised by the United States on the grounds that it was not necessary to lay down detailed production and processing methods, as consumer protection could be assured by end-product testing for dangerous pathogens, which did notinclude E. Coli. Subsequently, outbreaks of bacterial food poisoning, including several large E. Coli outbreaks in which children had died, had persuaded the United States to review and propose radical changes to its meat hygiene rules. Yet, only twenty years ago, E. Coli was not considered to be a significant danger to human health.
With regard to BSE, the European Communities noted that this was a new disease of cattle. It had been first recognised in the United Kingdom only 10 years ago. When it had been found to be a member of the group of transmissible spongiform encephalopathies, it had been widely assumed to be the bovine form of scrapie, a disease of sheep recognised for over 250 years. As there was no evidence of a danger to human health from eating meat from scrapie-affected sheep, it had been assumed in some quarters that there would similarly be no danger to humans from BSE. That assumption had recently been challenged by the appearance of a cluster of cases of a new form of Creutzfeldt-Jakob disease, a human spongiform
(continued...)
WT/DS26/R/USA
Page 84
The European Communities noted that Dr. S.S. Epstein of the School of Public Health, University of Illinois at Chicago, had explained the risks from the use of these hormones for animal growth promotion as follows:
"Although the carcinogenicity of DES in test animals was known as early as 1938, its use as a feed additive was approved by the USDA and the US FDA, in 1947 .... However, misleading assurances of safety and stonewalling by the FDA and USDA, including the deliberate suppression of residue data, managed to delay a US ban on DES until 1979.
"The meat industry then promptly switched to other carcinogenic additives, particularly natural sex hormones, which are implanted in the ears of commercially raised feedlot cattle. Unlike the synthetic DES, whose residues can be monitored and whose use was conditional on a seven- day preslaughter withdrawal period, residues of natural hormones are not routinely detectable because they cannot be differentiated from the same hormones produced by the body. Since 1983, the FDA has allowed virtually unregulated use for these natural additives right up to the time of slaughter, subject only to the non-enforceable requirement that residue levels in meat must be less than 1 per cent of the daily hormonal production in children.
"A dramatic warning of the dangers of growth-promoting additives was triggered by an epidemic of premature sexual development and ovarian cysts involving about 3,000 Puerto Rican infants and children from 1979 to 1981. These toxic effects were traced to hormonal contamination of fresh meat products and were usually reversed by simple dietary changes. Using highly specialized research techniques, independent testing found that samples of the meat products were contaminated with oestrogen residues more than tenfold in excess of normal ranges. Additionally, elevated levels of oestrogen and the synthetic zeranol were found in the blood of affected children. Increased rates of uterine and ovarian cancers in adult women were also associated with the epidemic.
"More than a decade ago, Roy Hertz, then director of endocrinology of the National Cancer Institute and a world authority on hormonal cancer, warned of the carcinogenic risks of oestrogenic feed additives, particularly for hormonally sensitive tissues such as breast tissue, because they could increase normal body hormonal levels and disturb delicately poised hormonal balances. Hertz pointed to evidence from innumerable animal tests and human clinical experience that such imbalance can be carcinogenic. Hertz also warned of the essentially uncontrolled and unregulated use for these extremely potent biological agents, no dietary levels of which can be regarded as safe. Even a dime-sized piece of meat contains billions of trillions of molecules of these carcinogens.
"Virtually the entire US population of consumers, without any warning, labelling, or information, ingests unknown and unpredictable amounts of hormonal residues in meat products over a lifetime. In 1986, as many as half of all cattle sampled in feedlots as large as 600 animals were found to have hormones illegally implanted in muscle rather than the ear skin, to induce further increased growth. This practice results in very high residues in meat, which even the FDA has admitted could produce "adverse effects". Left unanswered is whether such chronic and uncontrolled oestrogen dosages are involved in increasing cancer rates (now striking one in three Americans), particularly the alarming 50 per cent increase in the incidenceof breast cancer
130(...continued)
encephalopathy, which experts considered was likely linked to BSE. In this case, it was fortunate that a precautionary approach was taken at an early stage, despite the generally held view that any risk to humans was uncertain or extremely small. The tissues known to contain the BSE agent, as well as some in which infectivity had not been demonstrated but which could possibly be affected, were removed from the human food chain. At the time, this action was criticised in some quarters as an over-reaction but it now appeared to have been far-sighted.
WT/DS26/R/USA
Page 85
since 1965. These questions are of further concern in the light of recent evidence confirming the association between breast cancer and oral contraceptives, whose oestrogen dosage over a fraction of a lifetime is known and controlled, in contrast with that from residues of growth hormones in meat products."131
The United States responded that the article by Dr. Epstein cited reports over a decade old, inaccurate for several reasons, and did not constitute scientific evidence in support of the EC ban. First, Dr. Epstein had not cited any evidence with respect to the safety of residues of the six hormones for growth promotion purposes. Second, Dr. Epstein’s facts were wrong. In the case of Puerto Rico, in contrast to Dr. Epstein's assertion, the cause of the premature development had never been identified and the initial reports of zeranol in samples of patients’ blood could not be confirmed.132 Third, the reference to the article by Dr. Hertz was not relevant. Dr. Epstein attributed conclusions to Dr. Hertz that he had never made. Also, the reference to feed additives was a reference to DES, not to the six hormones. The United States submitted that with almost 30 years experience with the use of zeranol, there had not been any adverse effects reported on human health resulting from the legal use of this product in cattle or sheep.
The United States concluded that the "evidence" put forward by the European Communities did not support the risks it alleged. It noted that the report of the 1995 EC Conference set forth on page 4 a list of the "Principles of risk-assessment" which illustrated that the European Communities had stopped its analysis at what was supposed to be the starting point for a risk assessment. Those Principles stated that "the starting point in any risk-assessment is to identify all the potential (or even hypothetical) hazards (or untoward outcomes) that may be caused by eating meat containing residues of growth-promoting substances". In the US view, the European Communities had failed to move to the next step. As the report notes: "Risk is a different concept; it is the chance, or the likelihood, that a particular hazard will occur as a result of eating meat containing residues of growth-promoting substances". The United States argued that by failing to consider whether there was a chance or likelihood of any of the hypothesized hazards, the EC ban was not based on an "assessment of the relevant risks", as required by Articles 5.1 and 5.2.
Furthermore, the SPS Agreement required for any sanitary measure that there be "sufficient scientific evidence" and an "assessment of the risks" or, under Article 5.7, "available pertinent information". The United States claimed that the "precautionary principle", which the European Communities attempted to use to justify its ban, was not an approach unique to the European Communities. A wide number of the world’s governments hadsubscribed to the precautionary approach in Principle 15 of Agenda 21 of the Rio Declaration on Environment and Development (“Rio Declaration”). Nonetheless, the United States argued that the European Communities had misunderstood the precautionary approach. The precautionary approach called for taking action based on preliminary or inconclusive scientific information where it might be important to act in the absence of conclusive scientific information. One example of the application of the precautionary approach was global climate change. In that instance, preliminary scientific information indicated the need to act as a precaution even though a full scientific understanding had not yet been achieved. The United States claimed that Article 5.7 reflected the precautionary approach, but the European Communities had admitted that its ban was not pursuant to Article 5.7.
131S.S. Epstein (1990), "The Chemical Jungle: Today's Beef Industry", International Journal of Health Services, Vol. 20, pp.278-279.
132Hannon et al. (1987), Archives of Environmental Contaminants Toxicology, Vol. 16, pp.255-262.
WT/DS26/R/USA
Page 86
The European Communities argued that one of the aims of the measures at issue were to protect animals from the use of these hormones for growth promotion. In the preparatory discussions, the European Parliament had considered "the possible adverse effects of these substances on the immunity against various diseases of animals and that this in turn may lead to an increased use of antibiotics". 133 Furthermore, the preamble and text of Directive 88/146/EEC clearly established that concerns of potential risks to animal health were also taken into account. The strict conditions imposed by the Directive on the use of these hormones for therapeutic and zootechnical purposes had as one of its objectives the protection of animal health from improper administration. Article 5 provided that the EC member States must ensure that "no animals are dispatched from their territory to that of another member State which have had administered to them in any way whatsoever substances with a thyrostatic, oestrogenic, androgenic or gestagenic action...", except for therapeutic or zootechnical purposes. Article 6(1) provided that "member States shall prohibit importation from third countries of animals and of meat from animals to which have been administered in any way whatsoever substances with a thyrostatic, oestrogenic, androgenic or gestagenic action", except for therapeutic or zootechnical purposes and under conditions equivalent to those applied in the European Communities. One objective of this prohibition was to protect animal health, since the prohibition on trade in such animals clearly discouraged the administration of these hormones in the first place.
The European Communities claimed that scientists agreed that doses of these hormones at levels which produced a hormonal activity were carcinogenic to laboratory animals. The 1988 JECFA Report had documented these carcinogenic effects. It had reported that oral and parenteral administration of could increase the incidence of tumours in experimental animals. These tumours largely occurred in tissues with high levels of specific hormone receptors which were normally responsive to stimulation by
. JECFA had also noted that the incidence of tumours of the mammary gland, ovary, uterus and vagina were higher in animals treated with progesterone alone than in control animals. The incidence of uterine tumours was "surprisingly high" and the incidence of prostatic tumours was higher in rodents treated with high doses of testosterone, than in control animals. JECFA concluded that the carcinogenic responses were related to the hormonal activities of each of these three hormones.
The European Communities noted that JECFA had also found liver hyperplasia and tumours in mice fed high doses of trenbolone acetate and slight increases in the incidence of islet-cell tumours of the pancreas of rats, which JECFA concluded arose as a consequence of the hormonal activity of the trenbolone metabolites. The JECFA review of zeranol noted oestrogenic but not carcinogenic effects in rats. In mice, however, higher incidence of anterior lobe tumours of the pituitary gland occurred than in mice in the negative control group. JECFA reported that such tumours rarely occurred spontaneously in mice but were known to result from administration of oestrogenic hormones. JECFA again concluded that the tumorigenic effect of zeranol was associated with its oestrogenic properties.
The various carcinogenic effects mentioned by the 1988 JECFA Report on laboratory animals were explained more fully in the reports of the International Agency for Research on Cancer (IARC). IARC Monograph, Supplement 7, 1987, reported sufficient evidence of carcinogenicity to animals from and progesterone. IARC noted that administration to mice of
and its esters increased the incidenceof mammary, pituitary, uterine, cervical, vaginal, testicular, lymphoid and bone tumours. In rats, there was an increased incidence of mammary and/or pituitary tumours. In hamsters, a high incidence of malignant kidney tumours occurred in intact and castrated males and in ovariectomized females, but not in intact females. In guinea-pigs, diffuse fibromyomatous uterine and abdominal lesions were observed. IARC reported that progesterone increased the incidence of ovarian, uterine and mammary tumours in mice. Dogs treated with progesterone for four years developed
133European Parliament's Resolution, EC Official Journal, No. C288/158, 11 November 1985.
WT/DS26/R/USA
Page 87
a dose-related incidence of mammary-gland nodules. Importantly, IARC indicated that there was evidence that low doses of progesterone administered over long periods acted in combination with carcinogenic agents, such as some viruses or chemicals, to enhance tumour developments. Therefore, long-term administration of synthetic progestins could produce a comparable hazard by increasing the incidence of tumours due to other agents.134
The IARC had also found that zeralenone, a metabolite of zeranol, in the diet of laboratory animals had carcinogenic effects135, and had caused other effects, including:
"... atrophy of the seminal vesicles and testes, squamous metaplasia of the prostate gland, osteopetrosis, myelofibrosis of the bone marrow, cytoplasmic vacuolization of the adrenal glands, hyperkeratosis of the vagina and endometrial hyperplasia ... pseudopregnancy [and] infertility."136
In addition, consumption of zeralenone by pregnant animals caused a reduction in fetal weight and an increased prevalence of skeletal problems and incidence of stillbirths.137 As regards the carcinogenic effects of the hormone MGA, published data documented the induction of a statistically significant incidence of mammary tumours in female mice.138
The European Communities added that the above carcinogenic effects were not the only negative effects which the administration of these hormones had on animals. The 1995 EC Scientific Conference had documented a number of other effects which resulted from these hormones.139 Such negative effects included increases in liver abscesses, fertility problems, effects on male reproduction, alteration of gonadal function, decrease in pregnancy rates, increase of embryonic or fetal mortality rates, etc. The risks to animals of these negative effects were as high as (or even higher than) the risks these hormones posed to human beings, since they were administered directly to the animals and improper use could easily multiply the potential negative effects.
The United States noted that the protection of animal health was irrelevant to this case. The European Communities did not claim that its ban was designed to protect animal health in the territory of other Members, and imports of meat in this case were not relevant to the protection of the health of animals within the European Communities. In any case, data suggested that animals to which had been administered these hormones for growth promotion purposes were in fact healthier.
The European Communities stressed that it did not argue that the measures at issue were designed to protect animal health in the territory of other Members (regardless of how desirable that would have been). But the European Communities believed that it had the right under the WTO rules (and in particular the SPS Agreement) to prohibit imports from third countries of animals to which these hormones had been administered for the reasons explained above. This prohibition was not discriminatory, as it applied in exactly the same way to trade in such animals within the European Communities.
13421 IARC Monograph s, p.132.
13556 IARC Monograph s, p.431; and 31 IARC Monographs, p.287.
136Ibid., p.417.
137Ibid., pp.421-422.
138J.W. Lauderdale et al. (1977), "Studies of a progesterone (MGA) as related to residues and human consumption", Vol. 3, pp.5-33.
139R. Renaville, S. Massart, A. Prandi, U. Fazzini, M. Sindic, L. Nicolay, M. Falaki, A. Burny & D. Portetelle, "Aspects on the Use of Anablic Steroids in Animal Production", 1995 EC Scientific Conference Proceedings, pp.75-77.
WT/DS26/R/USA
Page 88
The European Communities claimed that Article 5.4 provided that Members should (not shall) take into account the objective of minimizing negative trade effects. This was complied with both in respect of intra-Community trade and trade with third countries. Only a high level of protection chosen by the European Communities would have ensured the flow of trade in the internal market. The imposition of veterinary controls on the use of hormones as growth promoters would have caused insuperable barriers to trade with third countries. Moreover, a government was not obliged to lower its level of protection below what it considered to be necessary. It should only keep in mind the objective of not going beyond what was necessary.
The United States further observed that the European Communities also claimed that its measures were consistent with Article 5.5. This was not accurate. The United States had shown that there was no scientific evidence that there was a risk to human health from the residues of the hormones when used for growth promotion. However, even if one were to disregard all of the scientific evidence and risk assessments that had been conducted and were to assume instead that these residues did pose a risk, the EC ban would still be inconsistent with Article 5.5. Article 5.5 provided in its relevant part that "each Member shall avoid arbitrary or unjustifiable distinctions in the levels it considers to be appropriate in different situations, if such distinctions result in discrimination or a disguised restriction on international trade." Thus, Article 5.5 established a two-step test with respect to distinctions between a Member’s appropriate level of protection in different situations: was the distinction in the levels of protection (i) either arbitrary or unjustifiable and (ii) did this distinction result in either discrimination or a disguised restriction on international trade.
The European Communities argued that Article 5.5 required consistency only in terms of avoiding arbitrary or unjustifiable distinctions in applying the chosen level of protection, and then only if such distinctions resulted in discrimination or a disguised restriction on trade. It could not be interpreted as requiring consistency in setting or deciding the level of protection, since the text would have specified it. This interpretation was consistent with the preamble, which stated that the harmonization of SPS measures did not require Members to change their appropriate level of protection. It followed from this that Article 5.5 did not require Members to achieve consistency between the level of protection which they chose against different hazards to human, animal and plant life and health. This would in any case be an illogical requirement; no Member protected individual plants to the same extent that it protected individual humans. Consistency of application meant that Members must in all circumstances apply measures which were capable of achieving the same level of protection against a given hazard unless there was justifiable reason in a particular situation to apply difference measures to achieve a different level of protection.
The European Communities had followed exactly the same principles for other drugs and food additives as it had for hormones. All such substances were evaluated on a case-by-case basis and where substances were deemed too dangerous, or potentially dangerous, to allow their residues in food, a zero tolerance was fixed. As an example, the use of antibiotics in livestock was regulated in the European Communities by forbidding the use as growth promoters of those antibiotics which were of therapeutic value in humans, in order to avoid problems of antibiotic resistance.
The United States argued that the EC claim that its measures were applied in exactly the same way to achieve the same level of protection in respect of all meat intended to be placed on the market in the European Communities, whatever its origin, was not accurate. The European Communities in fact applied different levels of protection to different meat. The European Communities had suggested that its appropriate level of protection with respect to the six hormones when used for growth promotion purposes was "zero risk". However, it had chosen a different, less stringent, level of protection with
WT/DS26/R/USA
Page 89
respect to carbadox. Carbadox was a substance that the European Communities permitted to be used for growth promotion purposes in the production of swine. It was a feed additive that was a known genotoxic carcinogen140, unlike the six hormones. The experts advising the Panel all confirmed that carbadox was genotoxic. Genotoxic meant that scientists considered carbadox to induce cancer. By permitting its use and the sale and consumption of meat from animals to which carbadox had been administered, the European Communities had chosen an appropriate level of protection that was less stringent than for the six hormones involved in this dispute. There was no principle or criterion that accounted for the different treatment by the European Communities of harmless residues, in the case of the hormone ban, as compared to residues of carbadox. If the different treatment resulted from the EC selection of differing "appropriate" levels of sanitary protection, and in the absence of any principle or criterion that accounted for the distinction, then the distinction in the levels of protection was arbitrary and unjustifiable.
The United States noted that Article 5.5 required that if the challenged distinction in levels of protection were arbitrary or unjustifiable, consideration must be given to the second test, whether the distinction in levels of protection resulted in a disguised restriction on international trade. Carbadox was used in swine production. The six hormones were used mainly in cattle production. It was no coincidence that the European Communities had banned the use of the six hormones for growth promotion purposes while permitting the use of swine growth promotants. The EC swine industry was relatively more efficient and market-oriented than the beef sector. Efficiency in beef production was not as important in the European Communities, since that sector relied more heavily on EC domestic price support measures, import protection and export subsidies. Unlike beef, there was virtually no acquisition of pork under the EC intervention scheme. Also, in contrast to beef, EC pork could be exported to the United States without the use of export subsidies. As the United States had demonstrated, the European Communities clearly wanted to reduce beef supplies when it extended the hormone ban. The European Communities had no comparable concerns with respect to its pork sector. On the contrary, it needed to preserve competitiveness in order to maintain export markets. Therefore, it continued to permit the use of growth promotants for swine. As a result, the European Communities had set its appropriate level of protection for the six hormones at a more stringent level than that for carbadox because it needed to protect its beef industry from competition. The distinction in the EC levels of protection resulted in a disguised restriction on international trade.
The United States claimed that despite its arguments to the contrary, the European Communities must believe that there was a threshold of safety for synthetic hormones, since it permitted the use of synthetic hormones, for example allyl-trenbolone 141 (altrenogest), a feed additive for oestrus control in swine and horses. Allyl-trenbolone was a synthetic sex steroid with progestational activity142, not approved for use in the United States. In addition, the European Communities allowed the use of medroxyprogesterone acetate, not a natural progesterone, as a vaginal implant for oestrus control in sheep. In all these instances the European Communities allowed meat from animals so treated to be sold and consumed. Furthermore, the European Communities accepted that there was a threshold for harm in the case of other compounds administered to farm animals. One such example was sulfamethazine, which caused thyroid cancer in laboratory animals, yet the effect was considered by JECFA to occur only above a threshold level. In that case the European Communities had no difficulty accepting the JECFA assessment and had set an MRL for this compound. Sulfamethazine was used in the European Communities for the treatment of respiratory ailments in chickens, cattle and swine. Apparently the European Communities only believed that there was no threshold for harm when such a position assisted it in discriminating against the banned meat and animals.
14036th report of JECFA (1990), pp.45-50.
141EC Directive 96/22/EC, Art 4, 2(I).
142A.R. Peters (1992) "Endocrine Manipulation-Toxicological Frontiers", J. Reprod Feter Suppl 45, pp.193-201.
WT/DS26/R/USA
Page 90
The United States recalled that Article 5.6 required Members, when establishing or maintaining a sanitary measure to achieve the appropriate level of sanitary or phytosanitary protection, to ensure that such a measure was not more trade-restrictive than required to achieve their appropriate level of sanitary protection, taking into account technical and economic feasibility. 143 The EC measures were "more trade-restrictive than required to achieve their appropriate level of sanitary or phytosanitary protection"144, contrary to Article 5.6. The EC ban was not intended to achieve any particular level of sanitary protection from any identified risk, and regulated meat based on the production method used rather than on any level of residues of hormones in the meat. For the United States, the level of any risk, if one could have been identified, would have presumably been related to the level of residues. Moreover, an "appropriate level of sanitary or phytosanitary protection" required that there be a particular risk to be protected against and the European Communities had failed to identify any particular risk from any of these hormones when used in accordance with good animal husbandry practice. The United States submitted that the EC ban stopped meat with one level of hormone residues from being imported and sold while permitting meat with higher levels of residues of the same hormone to be sold, because it continued to allow the human consumption of domestically-produced meat from animals administered hormones for therapeutic purposes, as well as meat with levels of endogenous hormones higher than those found in meat from animals administered the same hormones for growth promotion. Similarly, the European Communities permitted meat from pregnant dairy cows or from bulls and other food products (such as milk, butter and eggs) to be produced, marketed, consumed and used in food processing, although these products contained far higher levels of endogenous hormone residues than meat derived from steers and heifers treated with the six hormones. Consequently, the European Communities permitted the human consumption of endogenous hormone residues without any limitation while banning imported meat from animals administered safe hormones for growth promotion purposes. Furthermore, the EC ban was more trade restrictive than required because it severely restricted trade, although it was not designed to protect against any particular risk, nor to achieve a particular level of protection from that risk. The EC import ban applied regardless of whether the hormones were administered properly, regardless of any withdrawal period, and regardless of the levels of any residues of the hormones in the meat. Thus, the EC measures were of the most extreme form possible short of a total ban on imports of meat.
The European Communities submitted that from the sequence of events which had led to the adoption of Directive 88/146/EC, it was clear that the European Communities had examined carefully the potential risks to human and animal health from the use of these hormones for animal growth promotion purposes. It had considered the scientific evidence which suggested that these substances, if used in accordance with good veterinary practice, did not seem to pose risks to human and animal health, but which did not exclude beyond doubt that they posed potential risks to human health. The European Communities had examined carefully all of the available options for ensuring the proper conditions of use and control of these substances before concluding that, at present, the ban on their use was the only scientifically, technically and economically feasible option.
The United States claimed that, even if (putting aside the lack of a scientific basis for the EC ban) one were to accept the EC "scientific evidence", the ban would be overbroad and thus was more trade-restrictive than required to achieve the EC appropriate level of sanitary protection. The European
143The footnote to Article 5.6 reads: "For purposes of paragraph 6 of Article 5, a measure is not more trade-restrictive than required unless there is another measure, reasonably available taking into account technical and economic feasibility, that achieves the appropriate level of sanitary or phytosanitary protection and is significantly less restrictive to trade."
144"Appropriate level of sanitary or phytosanitary protection" is defined in paragraph 5 of Annex A of the SPS Agreement as "[t]he level of protection deemed appropriate by the Member establishing a sanitary or phytosanitary measure to protect human, animal or plant life or health within its territory".
WT/DS26/R/USA
Page 91
Communities did not contest that any human health risk from hormones in meat would be due to the residues of the hormones in the meat. Yet the European Communities had not regulated residue levels of these hormones. If the European Communities had serious concerns about health risks from hormones, then it would have regulated the level of residues of hormones. Instead, the European Communities banned selected uses of these hormones while ignoring the actual levels of residues to which consumers were naturally exposed from meat or other dietary sources. This was also true with respect to the three synthetic hormones. The European Communities did not regulate residue levels. According to the European Communities’ own inaccurate claim, after a sufficient waiting time prior to slaughter, there would be no residues of the three synthetic hormones in meat from animals to which these hormones had been administered for growth promotion purposes. Yet the European Communities had still banned the meat. In other words, if it believed that hormones posed a health risk at exposures above a set level, the European Communities would have regulated the levels of hormones allowed in food in order to ensure that the set level was not exceeded, rather than imposing an absolute ban on the use of these hormones for growth promotion purposes with respect to animals, and meat thereof, exported to the European Communities. The European Communities had failed to do so, thus casting doubts on the credibility of its arguments.
The United States claimed that the obvious way to regulate the levels of residues in meat was through MRLs (bearing in mind that Codex had concluded that it was unnecessary to establish an MRL for residues of the three natural hormones administered for growth promotion purposes in accordance with good animal husbandry practice). To pretend that MRLs were not technically or economically feasible was directly at odds with the EC claim that "at least in the European Communities there were comprehensive and costly residue control programmes designed and effectively applied to detect, as far as technically possible and economically feasible, such residues from all sources".
The European Communities argued that its appropriate level of protection was the assurance that meat would not contain residues of administered hormones and that this was achieved by requiring stringent veterinary controls on the use of hormones in food animals, including a restriction on such use to certain specifiedtherapeuticor zootechnical applications. The European Communities considered that an examination of the possibility of extending this permission to the routine use as growth promoters revealed that the scale of meat production would render controls technically impossible, and that in any case the costs of control would be so high as to render such use uneconomic.
The European Communities indicated that the required control and monitoring system had been examined by the competent democratic institutions of the European Communities and rejected for a number of reasons. One concern was costs: it would be so expensive to restrict the supply and administration of hormone growth promoters by veterinarians thatthe costs would outweigh thebenefits. Furthermore, control was a problem due to the scale of livestock production and the fact that implantation of growth promoters must be performed at certain restricted times of the growth period; it would be technically impossible to ensure at all times, and at reasonable administrative and economic costs, a proper administration of hormones. Another concern related to the bureaucracy; the recording of the treatment of individual animals and their checking at slaughter for the presence of implants (to ensure they had been correctly administered) would be a heavy and expensive bureaucratic burden and would further add to the cost of the process.
The European Communities argued that inspection raised other problems. The inevitable failure to find remnants of implants at slaughter in some animals (due to migration from the implant site, or mal-administration) would entail further expensive examination and residue testing, adding further to the cost. Residue testing would, of course, have to include the same testing and corresponding costs mentioned above in respect of third-country meat imported into the European Communities, as the European Communities would not be able, in the absence of internationally agreed rules on controls and testing, the widely diverging practices of Members, and the obvious and powerful economic
WT/DS26/R/USA
Page 92
incentives for meat producers to disregard the rules, to rely entirely on testing and controls carried out by third countries.
Furthermore, the European Communities argued that the installation and control of a labelling system to ensure integrity of identification from the live animal through slaughter and processing to a meat product would also be extremely expensive and laborious. The cost would have to be borne by the product, which again would render the whole process uneconomic. Retail premises, including restaurants, which supplied unwrapped meat directly to the consumer would have to carry a prominent warning that they used meat from animals treated with hormones. This would be a strong disincentive to using such meat and would thus amount to a restriction on trade. In addition, consumers would not be able to make reliable authentic choices (e.g. of meat served in restaurants).
All of these considerations had led the European Communities to consider that the only technically and economically feasible measure to ensure that its appropriate level of protection was met was the ban. Any other requirements would have had the effect of stopping trade. The only alternative, a ban on the use of hormones for growth promotion purposes, however, had permitted trade in meat to continue.
The United States considered the EC arguments about technical feasibility and costs to be unfounded. With respect to administration costs, there was no justification for limiting the supply and administration of the hormones to veterinarians in all circumstances. Who administered the hormone had no connection to the residue level. The European Communities did not require administration by a veterinarian in all circumstances as could be seen for example from Directive 96/22/EC, Article 1.2(c)(I) and the second paragraph under Article 5. Furthermore, a concern that costs might outweigh the benefits was a consideration for the marketplace. It was not for the European Communities to prejudge the question by banning the importation of animals and meat. The WTO agreements provided for competitive opportunities, and the question here was why the European Communities would not provide the opportunity to supply animals and meat.
The United States argued that the European Communities had provided no reason to conclude that the costs of inspection under an MRL regime would be different from the costs under their current ban. One assumed that the European Communities were exercising oversight and enforcement with respect to imports of animals and meat now to ensure that no banned animals or meat were imported. It was difficult to understand why there would be less monitoring in the case of a ban than there would be in the case of MRLs. The European Communities was saying that it did not even have to try to identify programmes that would permit importation. This was contrary to the approach taken in the report of the Appellate Body on United States - Standards for Reformulated and Conventional Gasoline (AB-1996-1). The EC claims were also directly at odds with its statement that "at least in the European Communities there were comprehensive and costly residue control programmes designed and effectively applied to detect, as far as technically possible and economically feasible, such residues from all sources".
With regard to control, the United States noted that the European Communities had failed to make clear why the cost of control programmes undertaken in producing countries was relevant to the import ban. Countries had already undertaken expensive and extensive control programmes to comply with the current ban (for example, Australia had reported a cost in the order of $A10 million per year for its control programme to comply with the EC ban). These were not costs that would be borne by the European Communities. Why would these costs make an MRL regime not a technically or economically feasible alternative to the absolute ban currently in place? Presumably a country exporting now to the European Communities could choose not to adopt any new measures to take advantage of any trade opportunities presented by an MRL regime, and that country would be no worse off than at present.
WT/DS26/R/USA
Page 93
The United States dismissed the EC arguments on bureaucracy. Any bureaucratic burden in third countries that might result from an MRL regime did not constitute a reason for concluding that the MRL regime was not a technically or economically feasible alternative to the EC current ban. The same criticisms that applied to inspection and control also applied to this reasoning. The United States suggested that the European Communities could choose to maintain its ban with respect to domestically produced animals and meat if that was less costly for the European Communities, while applying an MRL regime to imports of meat.
The United States concluded that the European Communities had not explained, and could not explain, how it was unable to control the use of these hormones without a complete ban while other countries managed to effectively regulate the use of these hormones for growth promotion purposes without enormous costs. It was clear that the ban restricted trade. It had severely disrupted exports of meat and animals from the United States, and was not required for health protection. The risk assessments that had been performed had all concluded that these hormones could be safely used for growth promotion. The European Communities had presented no new scientific evidence to contradict those scientific conclusions or to demonstrate that the ban was required for protection from any identifiable risk. If there had been evidence of risks associated with certain levels of hormone residues in meat, then the European Communities should control the level of the residues rather than banning the use of the hormones altogether. Moreover, the European Communities had been unable to articulate any particular level of protection that the ban was required to achieve. As a result, in the absence of any clearly articulated level of protection, it could be presumed that the EC ban was more trade- restrictive than required, in contravention to Article 5.6.
The European Communities claimed that the United States had failed to discharge its burden of proof in this case. In particular, it had failed to establish that, on the face of the overwhelming scientific evidence that theuse of hormones for animal growth promotion was potentially very dangerous to public and animal health, there were other type of measures which, whilst significantly less restrictive on trade, were capable of ensuring that the level of protection which the European Communities had chosen in this case (zero residue of hormones in animals and meat) was effectively achieved. Conversely, the European Communities had examined all of the alternatives in 1984 and again in April 1996, and had concluded that the prohibition on use of these hormones in animal growth promotion was the only reasonably available and less trade restrictive measure.
The United States argued that if the European Communities could not justify its ban as falling within the exception provided in Article 5.7, which dealt with "cases where relevant scientific evidence is insufficient" and permitted a Member to "provisionally adopt sanitary or phytosanitary measures on the basis of available pertinent information". These hormones had all been thoroughly tested, examined and reviewed. The scientific community had been able to conduct thorough risk assessments on the use of these hormones. The assessments had all concluded that these hormones did not pose any identified risk to human life or health when used in accordance with good animal husbandry practices. Both groups of scientists assigned by the European Communities to review the hormones had been able to reach a conclusion that the hormones were safe, based on the available scientific evidence. The United States concluded that the European Communities had never presented any "available pertinent information" on which it had based its ban, as required by Article 5.7. It would also be difficult, over 10 years after adoption and over seven years after they had gone into effect, to describe the EC measures as "provisional" as called for in Article 5.7.
The European Communities responded that the EC measures were not "provisional". They were definitive. Moreover, no risk assessment of hormones had been completely "thorough", due to lack of information. The European Communities stressed that the SPS Agreement required Members to carry out a risk assessment, not the scientific community. More importantly, the scientific evidence
WT/DS26/R/USA
Page 94
concerning the need to regulate the use of hormones was in itself sufficient to justify its legislation and the European Communities did not need to rely on the exception provided for in Article 5.7 concerning cases where relevant scientific evidence was insufficient. Nevertheless, the European Communities stressed that all expert opinion, including that of the 1988 JECFA Report on which Codex had based its decision, had concluded that scientific evidence did not exist concerning the mode of action, the effect of using mixtures of hormones and the long-term effects of exposure on consumers. The European Communities concluded that its measures were not "provisional" but, on the contrary, they were definitive.
The United States submitted that the Agreement on Technical Barriers to Trade (TBT Agreement) was designed to ensure that a Member's technical regulations did not create unnecessary obstacles to trade. This was reflected both in Article 2.2 of the TBT Agreement and in clause 5 of the preamble.145 The TBT Agreement applied to "[a]ll products, including industrial and agricultural products"146, and applied to "technical regulations" which were defined in the Agreement. The Agreement established a number of disciplines designed to ensure that a Member's technical regulations did not create unnecessary obstacles to trade, including the obligations contained in Articles 2.1 and 2.2 of the TBT Agreement. Although Article 1.5 of the TBT Agreement recognized that some sanitary and phytosanitary measures met the terms of the definitions of "technical regulation", "standard" and "conformity assessment procedure", the drafters of the TBT Agreement had made explicit that where a technical regulation, standard, or conformity assessment procedure was a sanitary or phytosanitary measure, the SPS Agreement, not the TBT Agreement, applied.147 Whether a measure was a sanitary measure under the SPS Agreement depended on the purpose of the measure. The United States had shown that the EC measures were sanitary measures as defined in the SPS Agreement. Accordingly, they were not subject to the TBT Agreement by virtue of Article 1.5 of the TBT Agreement, but were subject to the SPS Agreement. However, the TBT Agreement’s provisions demonstrated the full degree to which the EC ban was fundamentally incompatible with the EC WTO obligations. Even if the EC ban did not consist of sanitary measures, the European Communities still could not maintain the ban since it would be inconsistent with the TBT Agreement.
The United States submitted that the EC measures were technical regulations. They laid down mandatory process and production methods (PPMs) related to product characteristics for animals and meat. They prohibited the sale of animals to which had been administered any of the six hormones for growth promotion purposes, and meat from such animals. The European Communities did not contest that its measures were mandatory PPMs and during the debate on its measures in the Tokyo Round Committee on Technical Barriers to Trade, the European Communities had insisted that its measures were "requirements which had been drafted in the EEC Directive in terms of a PPM".148
145Article 2.2 of the TBT Agreement provides: "Members shall ensure that technical regulations are not prepared, adopted or applied with a view to or with the effect of creating unnecessary obstacles to international trade." Clause 5 of the preamble expresses that the Members enter into the TBT Agreement "Desiring ... to ensure that technical regulations ... do not create unnecessary obstacles to international trade."
146Article 1.3 of the TBT Agreement.
147The United Statesnoted that a similar “carve-out” was provided for government purchasing specifications in Article 1.4 of the TBT Agreement .
148See, for example, TBT/M/Spec/5 (Minutes of Meeting held 22 May 1987), paragraph 9. See also the communication from the European Communities (TBT/Spec/21), of 23 July 1987, in which the European Communities states: "The EC Directive constitutes a PPM. The EC Directive establishes the principle of non-administration of hormonal substances for fattening purposes to animals of which the meat is exported to the EC. It thus constitutes a regulation in the form of a process or production method and not a standard expressed in terms of product characteristics."
WT/DS26/R/USA
Page 95
Similarly, the European Communities’ own statements established that the PPMs were related to the characteristics of the product. For example, Directive 81/602/EEC cited the EC concern over "the residues [the substances] leave in meat".149 The United States claimed that (as demonstrated by the discussion under Articles I and III of GATT) the EC measures treated meat imported from the United States less favourably than like products produced within the European Communities and originating in the territory of other Members. The term "like product" was not defined in the TBT Agreement, but given the direct parallel between the languageof Article 2.1 of the TBT Agreement and Article III:4 of GATT, the GATT approach to "like product" should apply. Similarly, given the direct parallels between Article 2.1 of the TBT Agreement and Articles I and III of GATT, a technical regulation that was inconsistent with Articles I or III of GATT should also be inconsistent with Article 2.1 of the TBT Agreement. Accordingly, for the same reasons that the EC ban was inconsistent with Articles I and III of GATT, it would be inconsistent with Article 2.1 of the TBT Agreement.
The United States claimed that inherent in Article 2.2 of the TBT Agreement was the requirement that a technical regulation fulfil a legitimate objective. Protection of domestic production was not a legitimate objective. The benefits of the WTO agreements would be nullified or impaired if Members were permitted to design their technical regulations in order to protect domestic production. This was reflected both in the prohibition in Article 2.2 on creating "unnecessary obstacles to trade" and in the statement in Article III:1 of GATT that internal measures should not be applied "so as to afford protection to domestic production". Yet, as shown in the discussion of Article III of GATT and the SPS Agreement, the EC measures were disguised protectionism, designed to restrict trade. Therefore, these measures were more trade-restrictive than necessary to fulfil a legitimate objective.
The United States stressed that the European Communities had required under paragraph 7 of Article 6 of Directive 88/146/EEC that conformity assessment procedures be established with respect to imports. However, this provision required that "[i]n accordance with the procedure laid down in Article 8, a control programme shall be drawn up regarding imports from third countries, to ensure that imports do not receive more favourable treatment than EC products."150 In other words, the EC approach had reversed the general principle that imports were to be accorded no less favourable treatment than domestic production. Instead, treatment of imports was to be subverted to the objective of ensuring that imports received no more favourable treatment than domestic production. Not only was this contrary to Article 5.1.1 of the TBT Agreement, but it was an approach that clearly required that the EC conformity assessment procedures with respect to imports be prepared, adopted or applied with a view to, or with the effect of, creating unnecessary obstacles to international trade. Requiring that the conformity assessment procedures for imports be made more onerous wherever necessary to ensure that imports did not receive more favourable treatment than domestic production was not consistent with the requirement in Article 5.1.2 of the TBT Agreement that "conformity assessment procedures shall not be more strict or be applied more strictly than is necessary".
The European Communities responded that in the present case, it agreed that the measures challenged were better defined as measures falling within the scope of the SPS Agreement rather than the TBT Agreement. The European Communities therefore did not respond to the arguments made by the United States as regards the TBT Agreement.
149Clause 5 of the Preamble.
150The United States noted that similar language was found in Article 6 of Directive 81/602/EEC.
WT/DS26/R/USA
Page 96
The United States claimed that despite the requirements of Article III:4 of GATT, the EC ban prohibited the importation and sale of certain imported meat, while permitting the sale of like domestic products. The EC measures distinguished between:
animals, and meat from animals, to which had been administered any of the hormones at issue for growth promotion purposes; and
animals, and meat from animals, to which any of the hormones had never been administered, or to which the three natural hormones had been administered for therapeutic purposes.
The United States maintained that there was no legitimate basis for the EC distinction. Not only was there no basis for distinguishing between these like products due to any risk to human health, because there was no basis for claiming a health risk from meat produced using the six hormones, but the European Communities' other basis for the measures stated a proscribed purpose, the protection of domestic production.151
Although the term "like product" had not been defined in GATT 1994, the United States observed that under GATT 1947 and GATT 1994 the approach had been to examine whether products were "like" on a case-by-case basis. In making a determination whether products were "like", panels had considered factors such as physical characteristics, end-uses, tariff classification and substitutability.152 The United States argued that in this instance, animals, and meat from animals, to which had been administered any of the hormones for growth promotion purposes were clearly "like" other animals and meat from other animals. In fact, in the case of the three natural hormones, meat from an animal to which the hormone had been administered for growth promotion purposes could be physically identical to meat from an animal to which no hormone had been administered, or to which a hormone had been administered for therapeutic purposes. There was no difference at all between such physically identical products and thus no basis to distinguish between them for health or other legitimate purposes.
The United States claimed that as it had demonstrated in the discussion of the EC measures under the SPS Agreement, these hormones were naturally occurring in food animals, and their levels varied widely even among animals of the same species. Thus, for example, a heifer would naturally have higher levels of oestrogen than a steer to which had been administered. Whether there was a residue did not depend on whether the hormone was administered for growth promotion or for therapeutic purposes. Similarly, with respect to the synthetic hormones, there was nothing about these hormones that altered the physical characteristics, end-uses, tariff classification or substitutability in a way that would render animals, and meat from animals, to which these had been administered
151Article III:1 of GATT provides that:
"1. The contracting parties recognize that internal taxes and other internal charges, and laws, regulations and requirements affecting the internal sale, offering for sale, purchase, transportation, distribution or use of products, and internal quantitative regulations requiring the mixture, processing or use of products in specified amounts or proportions, should not be applied to imported or domestic products so as to afford protection to domestic production" (emphasis added).
152The United States noted for example the panel reports on "United States - Measures Affecting Alcoholic and Malt Beverages", DS23/R, BISD 39S/209; "Japan - Customs Duties, Taxes and Labelling Practices on Imported Wines and Alcoholic Beverages", BISD 34S/83; and paragraph 6.9 of the panel report on "United States - Standards for Reformulated and Conventional Gasoline", adopted on 19 June 1996, WT/DS2/R.
WT/DS26/R/USA
Page 97
"unlike" animals, and meat from animals, to which they had not been administered. The synthetic hormones were rapidly metabolized by the animals and any remaining residues in the meat were safe.
The European Communities responded that because female animals are treated for growth promotion with male hormones, and vice-versa, the residues in an animal treated with a natural hormone are different from, as well as greater than, those which it would otherwise contain. The human diet normally contained a variety of meat; evidently if all the meat came from treated animals the amount of hormone residue ingested would be significantly increased. If a proper comparison were to be made involving male animals, it would have to be between treated and untreated males to see how far the levels of female hormones were increased in male animals. According to the US argument, heifer meat would be "within the normal range" if it contained as much testosterone as bull meat.
The United States argued that Article III of GATT had been interpreted to permit governments to distinguish between otherwise like products for legitimate regulatory purposes. As stated in the panel report on United States - Taxes on Automobiles: "Article III serves only to prohibit regulatory distinctions between products applied so as to afford protection to domestic production. Its purpose is not to prohibit fiscal and regulatory distinctions applied so as to achieve other policy goals. This view had been expressed in a recent panel report, which states:
"The purpose of Article III is ... not to prevent contracting parties from using their fiscal and regulatory powers for purposes other than to afford protection to domestic production. Specifically, the purpose of Article III is not to prevent contracting parties from differentiating between different product categories for policy purposes unrelated to the protection of domestic production. The Panel considered that the limited purpose of Article III has to be taken into account in interpreting the term "like products" in this Article. Consequently, in determining whether two products subject to different treatment are like products, it is necessary to consider whether such product differentiation is being made "so as to afford protection to domestic production".153
The United States did not contest the right of the European Communities to distinguish between products for legitimate policy purposes. Members’ ability to so distinguish was confirmed by the text and history of Article III of GATT and the TBT Agreement. However, the United States did contest the legitimacy of the policy goal advanced by the European Communities for its discrimination. The European Communities had claimed health concerns as the basis for its ban, but all six of the hormones involved in this dispute had been determined to be safe by all the scientific experts that had reviewed them. Moreover, the European Communities had never put forward any basis to contest the safety of these hormones. The European Communities had also stated a competition rationale for its ban. However, in the US view, whatever the merits might be of harmonizing regulatory treatment within a Member’s territory, such a policy was difficult to reconcile with GATT when the policy was applied to imports with no other health, environmental, or other legitimate policy purpose. It had been noted that "a primary purpose of the General Agreement was to lower barriers to trade between markets, and not to harmonize the regulatory treatment of products within them".154 The United States further claimed that the EC measures had been motivated by a desire to reduce the supply of beef in the European Communities and they had worked to ease the European Communities’ over-supply problems. By requiring that foreign producers adopt the same processes and production methods as EC producers, the EC measures had severely restricted imports, thus providing relief to domestic EC producers.
The United States claimed that it had been common, prior to the entry into force of the WTO Agreement, for governments to take action to restrict imports as part of an effort to control domestic
153Panel report on "US - Measures affecting Alcoholic and Malt Beverages", adopted 19 June 1992, DS23/R, BISD 39S/206.
154Paragraph 5.8 of the panel report on "US - Taxes on Automobiles", (DS31/R).
WT/DS26/R/USA
Page 98
supplies of agricultural products. Article XI:2(c) of GATT provided an exception for this purpose, subject to strict conditions. 155 However, a primary component of the Agreement on Agriculture was to eliminate this practice.156 The EC measures, as applied to imports, were yet another example of measures to restrict imports in order to control the total supply of an agricultural product. Finally, the United States argued that not only animals, and meat from animals, to which had been administered the hormones for growth promotion were "like" other animals and meat from other animals, but the EC measures resulted in less favourable treatment than that accorded to domestic sales and imports from other countries of other animals, and meat from other animals, in breach of Article III:4 of GATT.
The European Communities argued that the measures at issue had their domestic counterpart in the prohibition on the administration of the same types of hormones to animals reared in the European Communities and in the prohibition on the marketing of those animals and of meat from those animals. The measures applied to domestic products had identical scope and were subject to the same exceptions as the measures applied to the imported products. The only difference between the measures applied to domestic products and those applied to imported products was that while the former were applied at the stage of production and sale, the latter were enforced at the point of importation. Furthermore, the measures applied to both domestic and imported products as "products", even if their scope was defined in terms of processes and production methods ("PPM") rather than in terms of product characteristics. The present case was to be distinguished from the situation considered by the panel report on US - Restrictions on Imports of Tuna I, in which the panel had reasoned that the Note ad Article III covered only those measures that were "applied to products as such". Yet, according to the panel, the US internal regulations "could not possibly affect tuna as a product".157 In contrast, in the present case the reason for prohibiting the importation of animals which had been treated with certain hormones, and of meat from those animals (as well as their domestic production and sale), was precisely that this treatment, unlike the PPMs for fishing tuna prescribed by the US regulations, modified the physical characteristics and biological composition of the products.
The European Communities disagreed with the US interpretation of "like products". In the EC view, a determination of "likeness" must be exclusively based on objective criteria related to the characteristics of the products themselves. The purpose of a regulatory distinction between "like products" might only become relevant in deciding whether a measure which was inconsistent with Article III:4 could be justified under any of the general exceptions provided for under Article XX of GATT. The European Communities contended that the purpose of the measures at issue was not to afford protection to domestic production but to protect human and animal health, as well as the interests of consumers and, therefore, would not, in conformity with the United States' own test, infringe Article III:4. Although two products need not be identical in order to be "like" for the purposes of Article III, previous panel reports had taken a consistently narrow approach when interpreting this term. In European Communities - Measures on Animal Feed Proteins, the panel had rejected a claim by the United States to the effect that all products used for the purpose of adding proteins to animal feeds were "like". The panel noted, inter alia, that the products concerned had different protein contents and different origins (vegetal, animal or synthetic).158 More recently, the panel report on Japan - Taxes
155The United States noted that the panel report on "Japan - Restrictions on Imports of Certain Agricultural Products", adopted on 2 February 1988 (L/6253 - 35S/163), illustrated just how strict these conditions were.
156Article 4.2 of the Agreement on Agriculture.
157Panel report on "US - Restrictions on Imports of Tuna I", DS21/R (unadopted), dated 3 September 1991, 39S/155, pp.193-195, paras 5.8-5.14.
158Panel report on "EC - Animal Feed Proteins", at para. 4.2. The other panel report dealing with a regulatory measure discriminating between non-identical products is the panel report on US - Measures affecting Alcoholic and Malt Beverages.
In this case, the panel found that low alcohol and high alcohol beer were "similar" in terms of physical characteristics but
(continued...)
WT/DS26/R/USA
Page 99
on Alcoholic Beverages, a case concerning the application of Article III:2 of GATT, had concluded that:
"... only vodka could be considered as like product to shochu since, apart from commonality of end uses, it shared with shochu most physical characteristics. Definitionally, the only difference is in the media used for filtration. Substantial noticeable differences in physical characteristics exist between the rest of the alcoholic beverages at dispute and shochu that would disqualify them from being regarded as like products. More specifically, the use of additives would disqualify liqueurs, gin and genever; the use of ingredients would disqualify rum; lastly, appearance (arising from manufacturing processes) would disqualify whisky and brandy"159 (emphasis added).
The European Communities reaffirmed its view that meat from animals to which had been administered any of the six hormones at issue for growth promotion had substantially different properties, composition and appearance than meat from animals to which those hormones had not been administered or to which they had been administered only for therapeutic purposes. Moreover, meat from treated animals was perceived by European consumers as a distinct product. For these reasons, animals treated with hormones for growth promotion, and meat from those animals, could not, in light of the criteria used by previous panel reports, be considered as "like" to other animals, and meat from those animals, for the purposes of Article III:4. The European Communities claimed that animals, and meat from animals, treated with any of the hormones at issue for growth promotion were different from animals and meat from untreated animals or from animals treated for therapeutic purposes, and referred to its arguments with regard to the SPS Agreement. Taking into account the consumer's concerns as detailed with respect to the SPS Agreement, meat from animals treated with the hormones at issue for growth promotion and meat from untreated animals or from animals treated for therapeutic purposes were not substitutable for most European consumers and, accordingly, not "like" for the purposes of Article III:4.
The European Communities argued that even if animals treated with hormones for growth promotion, and meat thereof, and other animals, and meat thereof, were found to be "like products", its measures would still be consistent with Article III:4 because they afforded identical treatment to imported products and to domestic like products. The European Communities disagreed with the US interpretation of Article III:4. In the EC view, Article III:4 did not prevent Members from establishing regulatory distinctions within a given category of like products, provided that such distinctions did not, formally or in effect, give less favourable treatment to imported products than to domestic products. For this purpose, the comparison must be made between the treatment given to all imported goods as a whole and the treatment accorded to all like domestic products as a whole. Thus, in the present case the relevant comparison was not between imported products treated with hormones for growth promotion and domestic untreated products, but between imported meat and domestic meat.
The European Communities submitted that the wording of Article III:4 called for a comparison between the treatment given to "the products of the territory of any contracting party imported into the territory of any other contracting party ...." and the treatment accorded to "like products of national origin..." and not between the treatment given to "any imported product" and the treatment accorded to "any domestic like product". The ordinary meaning of Article III:4 did not require that a comparison be made on an import-by-import basis, but a comparison of the treatment given to all imported products
158(...continued)
were not "like" for the purposes of Article III:4 because the distinction was not applied so as to afford protection to domestic production, BISD 39S/206, pp.93-295, paras 5.70-5.77.
159Panel report on "Japan - Taxes on Alcoholic Beverages", para. 6.23.
WT/DS26/R/USA
Page 100
as a whole vis-a-vis all domestic like products as a whole. In the EC view, the object and purpose of Article III had been made explicit in the general principle set forth in the first paragraph of that Article, which stated that internal taxes and regulations "should not be applied to imported or domestic products so as to afford protection to domestic production". Article III:4 gave effect to this principle with respect to internal regulations, other than tax regulations and mixing regulations, and should be interpreted in conformity with it. Where an internal regulation laid down different requirements for imported and domestic products, it might be assumed that domestic production was "protected" if any imported product received "less favourable treatment".160 The situation was different where, as in the present case, the internal regulation did not distinguish between imported and domestic products but between different types of like products, irrespective of their origin. The mere fact that an internal regulation accorded more favourable treatment to certain types of like products vis-a-vis other types of like products did not necessarily mean that domestic production of the like product as a whole was protected. For that, it would be necessary that those types of like product which were accorded more favourable treatment were inherently or at least predominantly domestic and/or that the types which received less favourable treatment were inherently or at least predominantly imported. The EC views found support in the panel report on US - Measures affecting Alcoholic and Malt Beverages, which stated that:
"The panel recognized that on the basis of their physical characteristics, low alcohol beer and high alcohol beer were similar. It then proceeded to examine whether, in the context of Article III, this differentiation in treatment of low alcohol beer and high alcohol beer is such "as to afford protection to domestic production". The panel first noted that both Canadian and United States beer manufacturers produce both high and low alcohol beer. It then noted that the laws and regulations in question in various states do not differentiate between imported and domestic beer as such, so that where a state law limits the points of sale of high alcohol content beer or maintains different labelling requirements for such beer, that law applies to all high alcohol content, regardless of its origin. The burdens resulting from these regulations thus do not fall more heavily on Canadian than on United States producers"161 (emphasis added).
Although in the EC view the panel report on US - Measures affecting Alcoholic and Malt Beverages erred by ascribing the analysis of the effects of a measure to the determination whether two products were "like" and not exclusively on the characteristics of the products, this panel report confirmed that Article III:4 was not infringed unless a regulatory distinction between like products had the effect of protecting domestic production.162 The European Communities claimed that in case
160Hence the "no-balancing" principle established by previous panelreports which have dealt with formally discriminatory internal regulations. See panel report on US - Section 337 of the Tariff Act of 1930, adopted on 7 November 1989, BISD 36S/345, 387, para 5.14; and panel report on "US- Standards for Reformulated Gasoline", WT/DS2/R, paras 6.14-6.15.
161Panel report on "US - Measures affecting Alcoholic and Malt Beverages", BISD 39S/206, para. 5.73.
162The European Communities indicated that as noted above, the panel report on "EC - Measures on animal feed proteins" (the only other adopted panel report that has applied Article III:4 to an origin neutral internal regulation) found that the products concerned were not "like" by reason of their differences in physical characteristics and, therefore, did not have to rule on the relevance of the effects of the measure. Nevertheless, it is worth noting that the arguments made in that case by the US, acting as a complainant, were based on the premise that an internal regulation which is not applied to all like products does not infringe Article III:4, unless it is shown to have a protectionist effect:
"The representative of the United States took the view that the measures focused the impact more directly on imported vegetable proteins, particularly soybeans, because they did not apply to animal, marine and synthetic proteins, even though such proteins were, with vegetable proteins, substitutable for use in feeds. He maintained that animal, marine and synthetic proteins were excluded from the measures because there was substantial domestic production in the EEC and not because these products were not like products, taking account of their generally higher protein content and certain technical advantages".
(continued...)
WT/DS26/R/USA
Page 101
its measures were found to be contrary to Article III:4, they were justified by Article XX:(b) which did not affect the power of a Member to adopt a policy in order to protect human and animal health. Moreover, the European Communities submitted that in case the Panel were to find that its measures infringed Article III:4, for the reasons stated, they satisfied the requirements of the SPS Agreement and, consequently, must be presumed to be in accordance with the GATT, and in particular with the provisions of Article XX(b).
The United States noted that Article I:1 of GATT prohibited discrimination against products from one Member in favour of like products from other Members. However, the European Communities permitted the importation and sale from other Members of animals, and meat from animals, to which the three natural hormones had been administered for therapeutic purposes or to which hormones had not been administered. Noting that, as demonstrated, the banned beef was "like" beef from an animal to which hormones had been administered for therapeutic purposes and was like beef from an animal to which hormones had not been administered, the United States argued that the EC ban discriminated against the like US products. Consequently, similar to all the reasons described under Article III, the EC measures failed to accord immediately to imports from the United States the advantages, favours, privileges or immunities granted to like animals and meat originating in the territories of other countries.
The United States claimed that the EC discriminatory measures could not be justified by resort to Article XX of GATT, in particular Article XX(b). The European Communities had put forward no evidence to support its measures on health grounds, and as shown with respect to the SPS Agreement, these measures were "applied in a manner which would constitute a means of arbitrary or unjustifiable discrimination between countries where the same conditions prevail, or a disguised restriction on international trade" within the meaning of the chapeau to Article XX. Furthermore, the SPS Agreement elaborated "rules for the application of the provisions of GATT which relate to the use of sanitary or phytosanitary measures, in particular the provisions of Article XX(b)"163 and sanitary measures that complied with the SPS Agreement were presumed to be in accordance with Article XX(b) of GATT.164 It had been shown above that the EC measures were inconsistent with the EC obligations under the SPS Agreement. The European Communities had failed to establish that its ban was justified by invoking Article XX(b).
For all these reasons, the United States claimed that the EC ban was inconsistent with the European Communities' obligations under the GATT, in particular Articles I and III.
The European Communities argued that the United States alleged violation of Article I was a claim which was outside the Panel's terms of reference. Consequently, the EC arguments were made only in the alternative, in case the Panel were not to uphold the EC argument. The European Communities stressed again that animals to which the hormones had been administered for growth promotion, and meat from those animals, had substantially different properties, composition and appearance than other animals, and meat from those other animals. Moreover they were perceived by consumers as a different product. For these reasons, they could not be considered as being "like" for the purposes of Article I:1 of GATT. In the European Communities' view, even if they were found to be "like", the measures would still not infringe Article I:1. The most favoured nation principle
162(...continued)
(Panel report on "EC - Measures on animal feed proteins", para. 3.39)
163Clause 8 of the preamble to the SPS Agreement.
164Article 2.4 of the SPS Agreement.
WT/DS26/R/USA
Page 102
contained in that provision was concerned with discrimination between countries and not with discrimination between products. Article I:1 did not prevent Members from giving different treatment to goods falling within a single category of "like products", as long as the distinction did not discriminate, formally or in effect, according to the country of origin of the goods.
The European Communities maintained that this interpretation had been supported by a long series of previous panel reports, including the panel report on Belgian Family Allowances which had found that "the Belgian legislation would have to be amended in so far as it introduced a discrimination between countries having a given legal system of family allowances and those which had a different system or no system at all, and made the granting of the exemption dependent on certain conditions" 165 (emphasis added). According to Dr. Jackson, this panel report stood for the proposition that "permissible distinctions between the treatment of goods may be based only on characteristics of the goods and not on characteristics of their country of origin".166 In European Communities - Measures on Animal Feed Proteins, the US claim that all feed proteins were like products and that by excluding some types of feed proteins from the scope of the contested measures the European Communities had discriminated between imports originating in the United States and imports of like products originating in third countries, thereby violating Article I:1, had been rejected by the panel which had noted that "a significant portion of EC imports of like products, including soybeans, subject to the measures originated from contracting parties other than the United States."167
In addition, the European Communities noted that in Spain - Tariff Treatment of Unroasted Coffee, the panel had concluded that by applying different tariff rates to different varieties of coffee which had beenpreviously found to belike products, Spain hadinfringed its obligation under Article I:1. In reaching this conclusion the panel had noted that the complainant, Brazil, exported to Spain mainly those varieties which were subject to the higher duty rates. Thus, imports from Brazil were discriminated de facto vis-a-vis imports from other countries that grew mainly the varieties subject to the lower duty rates.168
Finally, the European Communities pointed out that in United States - Restrictions on Imports of Tuna Fish I, Mexico had claimed that the labelling provisions of the Dolphin Protection Consumer Information Act violated Article I:1 because, in the case of tuna harvested in the Eastern Tropical Pacific Ocean ("ETP"), the right to use the label "dolphin safe" was accorded only if such tuna was accompanied by documentary evidence showing that it was not harvested with certain fishing techniques. Even though there was no question that tuna caught in the ETP by Mexico was "like" any other tuna, the panel rejected this claim by reasoning that:
"According to the information presented to the panel, the harvesting of tuna by intentionally encircling dolphins with purse-seine nets was practised only in the ETP [....] By imposing the requirement to provide evidence that this fishing technique had not been used in respect of tuna caught in the ETP the United States did not discriminate against countries fishing in this area. The panel noted that, under United States customs law, the country of origin of fish was determined by the country of registry of the vessel that had caught the fish; the geographical area where the fish was caught was irrelevant for a determination of origin. The labelling regulations governing tuna caught in the ETP thus applied to all countries whose
165Panel report on "Belgian Family Allowances", adopted on 7 November 1952, BISD 1S/59, p.60, para. 3.
166J. Jackson, "World Trade and the Law of the GATT", pp.258-259.
167Panel report on "EC - Measures on Animal Feed Proteins", para. 4.20.
168Panel report on "Spain - Tariff Treatment of Unroasted Coffee", adopted on 11 June 1981, BISD 28S/102, pp.111-112, para. 4.10.
WT/DS26/R/USA
Page 103
vessels fished in this geographical area and thus did not distinguish between products originating in Mexico and products originating in other countries. "169 (emphasis added).
The European Communities argued that the measures at issue applied without distinction to all imports of meat irrespective of their country of origin and not only to imports originating in the United States. Furthermore, the measures did not discriminate de facto against US imports. Animals treated with hormones for growth promotion, and meat from those animals, were not an inherently US product or a product which was predominantly produced in the United States. In addition to the United States, there were other countries which authorized some or all of the prohibited hormones. On the other hand, untreated animals, and meat from those animals were produced in virtually all countries of the world, including the United States. In fact, imports from the United States and from other countries which authorized the use of all or some of the prohibited hormones accounted for a major proportion of the EC imports.
The European Communities thus maintained that the measures at issue were also consistent with Article I:1 of GATT. In case the Panel were to find the contrary, the European Communities submitted that, for the reasons stated, they satisfied the requirements of the SPS Agreement and, consequently, must be presumed to in accordance with the GATT, and in particular with the provisions of Article XX(b).
The United States noted that the issue of the inconsistency of the EC ban with Article I:1 of GATT was within the terms of reference of this Panel, despite the EC argument to the contrary. The Panel had standard terms of reference established in accordance with DSU Article 7.1 and the United States request for the establishment of the Panel cited GATT, which included Article I:1. Furthermore, the United States had made explicit that its reference to Article III or Article XI was a non-exhaustive listing of the relevant GATT articles.170 Moreover, the EC claim raised a general issue of interpretation: whether the phrase "cited by the parties to the dispute" in the standard terms of reference in the DSU modified the words "covered agreement(s)" or whether this phrase modified the words "relevant provisions". In the US view, "cited by the parties to the dispute" was clearly meant to modify "covered agreement(s)". This conclusion was clear from the text and the context of DSU Article 7.1.
The negotiating background of Article 7.1 confirmed this meaning. The provision on standard terms of reference in paragraph F(b)(1) of the Montreal Decision on "Improvements to the GATT Dispute Settlement Rules and Procedures" had provided that a panel would examine the matter referred to it by the complaining party "in the light of the relevant GATT provisions". This phrase had been modified to "in the light of the covered Agreements cited by the parties to the dispute" in the section on "Terms of Reference" at page T.3 of MTN.TNC/W/FA (the "Dunkel Draft" of 20 December 1991) in the text on "Elements of an Integrated Dispute Settlement System". The alteration was made in response to an initiative by the European Communities, inspired by its difficulties in the Airbus case. The objective was to make it possible for a defending party to "cite" agreements additional to those cited by the complaining party, and to have the panel apply all agreements cited by both sides.171 Moreover, the European Communities' own arguments in this dispute contradicted its terms of reference claim. Even though the panel request did not refer to Article XX of GATT, and the European Communities did not "cite" this provision either in a dispute-related document or in DSB proceedings when the panel
169Panel report on "US - Restrictions on Imports of Tuna I", para. 5.43.
170See paras. 1.4 and 1.5
171The United States noted texts dated 12 December and 17 December 1991, in which the same reference to “agreements cited” in terms of reference appeared under the descriptive subheading “Forum/Norm Shopping.” This term had been used in earlier discussions in 1990, in the wake of the European Communities' inability to obtain terms of reference in the Tokyo Round Subsidies Code Airbus dispute which would permit application of the Agreement on Trade in Civil Aircraft as well.
WT/DS26/R/USA
Page 104
was established, the European Communities had sought to have the Panel examine the application of Article XX in this case. The European Communities could not have it both ways.
THIRD PARTIES SUBMISSION
Australia argued that the SPS Agreement established, inter alia, the basic rights and obligations of Members in respect of sanitary measures necessary for the protection of human life and health which may directly or indirectly affect international trade. Notwithstanding any other WTO rights and obligations, Australia considered that if a Member were acting inconsistently with its obligations under the SPS Agreement, it stood in breach of its WTO obligations. In addition, as the SPS Agreement represented an elaboration of the GATT in respect of SPS measures and as conformity with SPS obligations gave rise to a presumption of conformity with Article XX(b) of the GATT, Australia considered that the EC measures should be examined in the first instance against the rights and obligations of the SPS Agreement as it was this agreement which established the basic rights and obligations of Members in respect of SPS measures.
Australia placed particular stress on the fact that in observing that an international standard as provided for in Article 3.1 existed in relation to the EC hormone ban, the European Communities had failed to demonstrate its compliance with Article 3.3. In particular, Australia argued that the European Communities had failed to demonstrate either that there was a scientific justification for its adoption of measures resulting in a higher level of protection or that it had carried out an examination and evaluation of available scientific evidence, including in accordance with the relevant provisions of Article 5 to show that the international standard was not sufficient to achieve its appropriate level of protection.
Referring to the conclusions of the 1995 EC Scientific Conference [paragraph ...], Australia argued that although these conclusions provided clear support to the view that the five substances considered at the Conference were safe to use within the conditions specified in Australia and within the Codex standards, the European Communities had confusingly claimed that "... [t]he scientific evidence for the necessity to maintain our measures is the evidence from the 1995 Scientific Conference ...". The European Communities had made no attempt to substantiate this assertion or to explain the contradiction between this statement and the Conference conclusions.
Although the European Communities had asserted that for the natural hormones it was known that they have adverse effects which, combined with a lack of knowledge of their action, lack of data
WT/DS26/R/USA
Page 105
on the effect of combinations and the lack of definition of "good veterinary practice", permit the European Communities to adopt a different level of protection, i.e. ensure the EC consumers that there are no residues left other than the naturally produced ones by the animals themselves, Australia argued that the European Communities had failed to establish that these statements reflected the findings of a risk assessment or an examination and evaluation of available scientific information in conformity with the provisions of the SPS Agreement. The European Communities had also not established the basis for its claim that there was a lack of definition of "good veterinary practice", which was generally taken to mean using veterinary chemicals in accordance with conditions of registration, including observing intended purpose, correct dose rates and withholding periods that were determined by the technical registration process.
With regard to the EC arguments that hormones had serious adverse effects on health, including cancer, Australia noted that such claims had also been made in respect of drugs containing hormonal substances which were directly administered to humans for medicinal purposes. The European Communities had not explained why it practised a zero risk policy in the issue under examination by the Panel but not in other comparable circumstances when it was known that the hormonal product would directly enter the human system. Good veterinary practice was the basis for use of the hormones at issue and to focus only on the worst case scenario as the European Communities had done could render the use of all veterinary chemicals illegal, just as the use of all chemicals used in human medicinal preparations might be banned. Furthermore, the statement that the intensive large-scale production of meat rendered effective control of hormone growth promoters technically and economically unfeasible, flew in the face of the fact that major meat exporting nations apart from the European Communities were able to effectively control inappropriate hormone use without the imposition of a total ban on the use of these substances. All agricultural and veterinary chemicals used within Australia must undergo assessment and registration prior to distribution and use. This required athorough assessment of detailed technical data packages by the National Registration Authority. The outcome of registration was to control the use, labelling, packaging and supply of the product. This was underpinned by legislation, with penalties where these conditions were not followed. Additionally, Australia backed up the system with individual property identification of cattle to the point of slaughter, a National Residue Survey to assess compliance with MRLs, and a National Vendor Declaration system for cattle through which vendors must declare the chemicals used on cattle offered for sale and the date of treatment.
Australia claimed that there was no evidence that the European Communities had undertaken a risk assessment on this matter in accordance with the provisions of Article 5. The European Communities had stated that "the European Communities has twice formed groups of scientists to examine the hormones at issue". However, they had stated themselves that "the purpose of the 1995 EC scientific Conference was not to perform for it a risk assessment, but to provide a public forum for discussion of the scientific aspects of the use of growth promoters". Therefore, the European Communities had explicitly accepted that this Conference was not part of a risk assessment process. At the Canada/EC consultations on 25 July 1996, at which Australia was present as a third party, the EC representative had said that there had been no risk assessment done since the Lamming Committee and that no work had been done in this area since 1988. However, nowhere within the EC first submission in this case, had it claimed that the Lamming Committee had undertaken a risk assessment process which could be considered to have met the requirements of Articles 5.1 and 5.2 of the SPS Agreement.
Emphasizing that Article 5.4 required Members, when determining the appropriate level of sanitary or phytosanitary protection, to take into account the objective of minimizing negative trade effects, Australia submitted that there was no evidence to suggest that the European Communities had taken into account negative trade effects when imposing the prohibition on imports of meat treated with hormones for growth promotion purposes. This prohibition had imposed significant compliance costs on the Australian meat industry.
WT/DS26/R/USA
Page 106
With regard to Article 5.5, Australia argued that the European Communities had failed to demonstrate its compliance with this provision. "Consistency" did not demand a quantified level of risk, or that Members must have the same level of risk in respect of every measure. However, the European Communities must justify why it had adopted a lower level of risk in this case than in comparable circumstances. Australia agreed that in a general sense it was true that, through risk management, each country's acceptable level of risk might reflect societal values, but, as specifically stated by Article 5.5, the acceptable level of risk determined by a country through this process must be applied in a consistent manner, and could not be applied on an arbitrary or unjustifiable case-by-case basis (for example, the European Communities choosing to accept zero risk in the case of meat from animals treated with growth promoting hormones).
Australia furthermore argued that the European Communities had failed to demonstrate its compliance with the obligation of Article 5.6. It observed that the European Communities had claimed that its "measures are not provisional: they are definitive"172, and therefore submitted that Article 5.7 was not relevant in this case. Noting that as the European Communities had failed to comply with the provisions of Article 3.3, failing to demonstrate that there was a scientific justification for its level of sanitary or phytosanitary protection or that its maintenance of a level of protection higher than the relevant international standard was a consequence of the level it had determined in accordance with the provisions of Article 5, Australia concluded that there must be a presumption that the EC measures were not in conformity with the basic obligations of Articles 2.2 and 2.3 of the SPS Agreement.
Canada indicated that all the six hormones at issue were authorized in Canada for cattle growth promotion. With respect to the 1995 EC Scientific Conference on growth promotion in meatproduction, Canadian scientists had expressed concern that the published report of the Conference proceedings did not reflect the most current scientific knowledge and that, therefore, caution should be exercised in making use of the proceedings in scientific evaluation.
Canada agreed with the US conclusion that the EC measures were inconsistent with the SPS Agreement. Canada disagreed with the EC presentation of the arguments in this case which, from a legal point of view, appeared to suggest that the SPS Agreement only became relevant after a contravention of GATT had been determined and questioned the relevance of the General Interpretative Note to Annex 1A of the WTO Agreement cited by the European Communities in support of its position. In Canada's view, GATT and the SPS Agreement were agreements of equal status contained in Annex 1A of the WTO Agreement (Article II of the WTO Agreement). Canada believed that one must first examine the SPS and TBT Agreements because these contained rules that were more detailed and more precise than those of GATT. Canada noted that the last preambular paragraph of the SPS Agreement indicated that the Agreement was, in part, intended to be an elaboration of "rules for the application of the provisions of GATT which related to the use of sanitary and phytosanitary measures in particular the provisions of Article XX(b)". However, Articles 3 and 4, of the SPS Agreement on "Harmonization" and "Equivalence", did not correspond to any provision in GATT. The fourth and six preambular paragraphs, reflected the broad scope of the SPS Agreement. This reinforced the point that the SPS Agreement was an independent agreement.
172See para. 4.239.
WT/DS26/R/USA
Page 107
Canada stressed that the "Basic Rights" of Members under the SPS Agreement were provided in Articles 2.1 and 2.4. Article 2.1 set out a limited right to take sanitary or phytosanitary measures and Article 2.4 confirmed one role of the SPS Agreement as an elaboration of the rules for the application of GATT which related to the use of sanitary or phytosanitary measures, and in particular the provisions of Article XX(b). The "Basic Obligations" of the SPS Agreement were found in Articles 2.2 and 2.3. Paragraph 2 expanded on the Article XX(b) requirement that the measure be "necessary", setting out three conditions: that any SPS measure be applied only to the extent necessary to protect human, animal or plant life or health, be based on scientific principles and not maintained without sufficient scientific evidence, except as provided for in Article 5.7. Article 2.3 was based on the requirements of the chapeau of Article XX, clarifying that the point of comparison for "where identical or similar conditions prevail" included the territory of the Member taking the measure.
Canada noted that Article 2 by its own terms was said to establish the "Basic Rights and Obligations" of Members. The other provisions of the Agreement set out more specific rights and obligations, illustrative of these fundamental rights and obligations. For example, the requirements set out in Article 5 that SPS measures must be based on a risk assessment and that the assessment must take into account available scientific evidence, were rational extensions of the basic obligations in Article 2.2 to ensure that SPS measures were based on scientific principles and not maintained without scientific evidence. Similarly, Article 5.6 gave precision to the obligation in Article 2.2 that a Member must ensure that any SPS measure was applied only to the extent necessary to protect human, animal or plant life or health. In effect, Article 5.6 set out how a Member was to meet this basic requirement of Article 2.2. Likewise, Article 3.2 provided a prescription for satisfying the presumption of consistency with GATT in Article 2:4.
Canada argued that in the panel report on United States - Standards for Reformulated and Conventional Gasoline ("Reformulated Gasoline") (WT/DS2/AB/R refers) the Appellate Body had reviewed the chapeau to Article XX of GATT. Given the close relationship between the text of the chapeau and the obligation in Article 2.3, Canada submitted that the Appellate Body's interpretation of the requirement that a measure shall "not be applied in a manner which would constitute a disguised restriction on international trade" was apposite to the present case. The Appellate Body found:
"Arbitrary discrimination", "unjustifiable discrimination" and "disguised restriction" on international trade may, accordingly, be read side-by-side; they impart meaning to one another. It is clear to us that "disguised restriction" includes disguised discrimination in international trade. It is equally clear that concealed or unannounced restriction or discrimination in international trade does not exhaust the meaning of "disguised restriction". We consider that "disguised restriction", whatever else it covers, may properly be read as embracing restrictions amounting to arbitrary or unjustifiable discrimination in international trade taken under the guise of a measure formally within the terms of an exception in Article XX. Put in a somewhat different manner, the kinds of considerations pertinent in deciding whether the application of a particular measure amounts to "arbitrary or unjustifiable discrimination", may also be taken into account in determining the presence of a "disguised restriction" on international trade. The fundamental theme is to be found in the purpose and object of avoiding abuse or illegitimate use of the exceptions to substantive rules available in Article XX."173
In Canada's view, this reinforced the assertion in paragraph 140 of the US submission that a protectionist measure in the guise of a sanitary measure was the essence of a disguised restriction on international trade.
Canada agreed with the US conclusion that the EC measures contravened GATT and that Article XX could not justify these measures. Canada noted that the interpretation of GATT Article III
173WT/DS2/AB/R, p.25.
WT/DS26/R/USA
Page 108
was being considered by the WTO Appellate Body in "Japan - Taxes on Alcoholic Beverages" and that the report of the Appellate Body in that case might be relevant to the present case.
Norway argued that this case concerned the right of every state to make its own sovereign decision on the level of protection it afforded its citizens with respect to health hazards. This principle had always been part of GATT law and several panels, when deciding disputes related to Article XX(b), had reaffirmed such right.174 The SPS Agreement, within its scope of application, might be considered an exemplification of Article XX(b). This was not a case of discrimination between like products, or discrimination between producers from different countries. The question was purely one of reaffirming the Member’s right to decide, when a risk to its population was present, the limits to the risk to which it would expose its citizens, and its freedom to choose the measure to achieve this protection as long as the measure itself was consistent with WTO obligations. In an era of increasing consumer concern over the safety of food and the presence of potentially harmful residues from the use of veterinary drugs, pesticides etc., and at a time when old “scientific truths” had often been overturned by new evidence, it was of utmost importance that the WTO uphold this right of the Member to protect its citizens against the risks connected with the use of such substances.
Noting that the SPS Agreement restated the right of a Member to set its health standards in its preamble and in Article 3.3, Norway argued that it was clear that a Member had the right to determine the level of protection it considered appropriate for its population, subject to general WTO requirements with respect to the existence of a risk, and requirements with respect to justification of the measure the Member applied. In determining the appropriate level of protection, Norway argued the obligation upon the Member was only to show that a risk to its population was present. As long as the existence of such a risk had been established, the WTO was only concerned with the justification of the measure the Member choose to apply to achieve the level of protection it had deemed appropriate. According to Article XX(b), as well as Article 5.2 of the SPS Agreement, a Member did not have to scientifically prove the extent of the risk. The Member only had to show that such a risk existed (risk identification), and thereafter it was for the Member to define the level of probability of harm it wanted to assume - be it zero or greater (risk management).
Norway submitted that a Member must consider present day scientific evidence when making a risk assessment. Such evidence might ascertain or disprove the risk, and might be more or less complete, but this was only one of the elements to be taken into account in the risk assessment. Other relevant elements might be, for instance, the existence of adequate testing methods and internationally approved production methods or control requirements. Norway argued that the SPS Agreement did not require a Member to base its levels of protection solely on suggested maximum tolerances presented in current scientific evidence. The cases referred by the European Communities in its submissions
174Norway noted that, e.g. the panel report on "US - Restrictions on the Imports of Tuna" stated that:
"The panel further noted that Article XX(b) allows each contracting party to set its human, animal or plant life or health standard. The conditions set out in Article XX(b) which limits the resort to this exception, namely that the measures taken must be "necessary" ... refer to the trade measure requiring justification ... not to the life or health standard chosen by the contracting party. (...)."
Furthermore, the Appellate Body in the case United States - Standards for Reformulated and Conventional Gasoline, in its concluding remarks stated:
"Members have a large measure of autonomy to determine their own policies on the environment (including its relationships with trade), their environmental objectives and the environment legislation they enact and implement. So far as concerns the WTO, the autonomy is circumscribed only by the need to respect the requirements of the General Agreement and the other covered agreements."
WT/DS26/R/USA
Page 109
clearly showed from history how often old scientific “truths” had been disproved by new scientific evidence. It had never been the intention of the GATT/WTO, nor should it be the task of this Panel, to infringe upon a Member's right to decide to be on the safe side where current evidence did not exclude beyond doubt a potential risk to human health.
Norway stressed that once the right of the Member to determine its appropriate level of protection was reaffirmed, it had to be ascertained that the chosen measure complied with the general obligations spelled out in GATT or the SPS Agreement. In Norway's view, the test to be applied according to Article XX of GATT and Article 5 of the SPS Agreement was the following:
the measure was not applied in a manner which would constitute a means of arbitrary or unjustifiable discrimination between countries where the same conditions prevailed;
the measure was not applied in a manner which was a disguised restriction on international trade; and
the measure was necessary, i.e. not more trade restrictive than required to achieve the appropriate level of protection, taking into account technical and economic feasibility.
Having applied this test to the measure, its consistency with the WTO obligations should be readily ascertained.
Norway submitted that the European Communities had the right to set its own level of protection in this case, and at the chosen level. It had clearly shown the uncertainties related to the possible harmful effect of long term exposure to the substances at issue and had clearly shown that its measure did not discriminate between countries, nor had there been any disguised restriction on international trade. The measure had been applied uniformly to all products containing the substances in question, be they domestic or foreign. In selecting the measure which was necessary to minimise the potential risk (risk management) so as to ensure the appropriate level of protection (zero or other acceptable level of risk), many factors had to be taken into consideration, of which maximum residue levels (MRL) was but one. The technical feasibility of ensuring adequate control of the use of the hormones, i.e. strict veterinary supervision of each implant, were of utmost importance, as well as the technical feasibility of control of the residue level in all the meat in question. Furthermore, the economic cost of the measure must be taken into account. An MRL system with extensive controls and verifications at the producer level and at point of importation might be so prohibitively expensive as to be ruled out. The European Communities had shown that other measures than the contested ban were not available to achieve the appropriate level of protection, which was consequently not more trade restrictive than necessary.
WT/DS26/R/USA
Page 110
Noting that the European Communities had justified the import ban on the basis that the importation of meat from animals treated with any of the substances might pose a risk to human health and safety175, New Zealand concluded that the ban appeared to be a measure that fell within the definition of a "sanitary or phytosanitary measure" as defined in the SPS Agreement.176 New Zealand noted that Article 1 of the SPS Agreement did not make provision for any Member to exclude specific sanitary or phytosanitary measures from the ambit of the Agreement. Moreover, the effect of Article XVI:4 of the WTO Agreement was to require Members to ensure the conformity of all their laws, regulations and administrative practices with the SPS Agreement. Accordingly, theimport ban and all other sanitary or phytosanitary measures affecting international trade in place within the jurisdiction of any Member, were subject to the requirements of the SPS Agreement.
With regard to Article 2.2, New Zealand claimed that no evidence had been produced of an appreciable risk of an adverse health effect arising from the use of any of the substances. Even if it were to be shown that there was an appreciable risk, the European Communities would be required to demonstrate that the import ban was necessary to address it. In this regard, New Zealand noted that the European Communities had expressed particular concern about the perceived risks associated with the inappropriate or illegal use of hormonal growth promotants, particularly when used in "combinations" or "cocktails". However, even if it was demonstrated that an appreciable risk existed from such inappropriate or illegal use, the European Communities would need to demonstrate that a ban on imports was in fact "necessary" to address this specific concern. Finally, from the history of the scientific investigations into the five substances, under intense scrutiny by scientists in the last two decades, New Zealand did not consider that the importation ban could be presented as being "based on scientific principles" nor sustained by "sufficient scientific evidence". Moreover, in the case of the sixth substance, MGA, New Zealand was not aware of the existence of any scientific information which would warrant the import ban in its regard.177
New Zealand submitted that scientific evidence demonstrated that the level of hormones present in meat varied considerably. Meat from animals treated with the substances for growth promotion had a similar range or even lower hormonal residues than meat from some animals that had never been treated. Additionally, the European Communities allowed meat from animals that had been treated with the three naturally occurring substances for therapeutic purposes to be produced, imported and
175In this context, New Zealand noted that in the preamble of EC Council Directive 81/602/EEC of 31 July 1981 the banning of new licences for anabolic agents for growth promotion was characterised as being in the "interests of the consumer" and it discussed the potential "harmful effects" of the use of the substances in question. EC Council Directives 85/649/EEC of 31 December 1985 and 88/146/EEC of 7 March 1988 cited concerns relating to the effect on human health from the use of hormonal growth promotants. The preamble of the Proposal for a Council Regulation (93/C 302/06) on 14 October 1993 stated that the presence of such substances "may be dangerous for consumers and may also affect the quality of foodstuffs of animal origin". And finally, EC Council Directives 96/22/EC and 96/23/EC stated in their preambles that such substances may be dangerous for consumers.
176New Zealand argued that the import ban would also be inconsistent with Article 2.2 of the TBT Agreement were it not that Agreement explicitly noted (in Article 1.5) that it does not cover sanitary or phytosanitary measures. In the event that the import ban were determined to be neither a sanitary measure, nor a technical barrier to trade, the provisions of GATT would then need to be considered.
177New Zealand indicated that because of differing use requirements and commercial factors, there were differences in the range of agricultural chemicals (such as MGA) registered around the world and hence differences in the standards contained in national laws. The absence of an international standard for a particular substance should not in itself be seen as justification for a particular import ban. In addition, it should be noted that Article 4 of the SPS Agreement encouraged Members to mutually recognise, for the purposes of trade, the equivalence of different sanitary or phytosanitary measures in jurisdictions other than their own.
WT/DS26/R/USA
Page 111
marketed within the European Communities.178 New Zealand considered that for these reasons, the EC measures contravened Article 2.3 of the SPS Agreement.
New Zealand stressed that Codex had adopted international standards, satisfyingitself that there was sufficient scientific basis. It was unclear what "available scientific information" the European Communities could rely on to provide "scientific justification" for the import ban and a selective examination and evaluation of available scientific information would not be sufficient to meet the requirements of Article 3.3. Moreover, there should be a scientifically established risk to human or animal life or health before any determination could be made as to a Member s appropriate level of protection. In this case, there did not appear to have been any scientifically defensible determination of an appreciable risk in the case of the import ban.
New Zealand claimed that the SPS Agreement did not allow for the adoption of a sanitary or phytosanitary measure that provided for a higher level of protection than an international standard in the absence of a scientific justification or a risk assessment process that had identified an appreciable risk. In the case of all six hormones, there would not appear to have been any EC assessment meeting the requirements of Article 5 which had identified any such risk. While it was evident that consumers in the European Communities might harbour concerns about the use of these substances, consumer perceptions as to risk did not provide a basis for the adoption of measures under Article 5. Furthermore, New Zealand submitted that the usual EC risk assessment and evaluation processes for veterinary medicinal products179, which might otherwise have ensured a consistent approach as required by Article 5.5, did not form the basis for the EC measures.180 Equally, given that many other products freely marketed within the European Communities contained much higher levels of the hormonal residues that the European Communities were seeking to avoid through the import ban, there would appear, in the absence of evidence to the contrary, to be arbitrary and anomalous levels of protection in place. Finally, Article 5.6 made it clear that even if it could be shown that there was an appreciable risk to health, given the significant trade implications of the import ban, the European Communities would have had to consider whether there were alternative, less trade-restrictive measures that could be applied to address such a risk. Noting that the European Communities did not seek to rely on Article 5.7, which was a manifestationof the "precautionary principle", as the basis for its import ban, New Zealand emphasized that the present situation appeared to be one in which the scientific evidence was neither insufficient nor inadequate.181 Accordingly, this was clearly not an instance where the "sufficient evidence" standard could be substituted with the less rigorous standard of "available pertinent information". The fact that Codex had adopted an international standard would always create a strong presumption that sufficient scientific evidence existed.
178 In New Zealand's indicated that it understood that the European Communities had not set any ADIs or MRLs for meat from animals treated for therapeutic purposes.
179New Zealand clarified that the guidelines for assessments of veterinary medicinal products were contained, at that time, in EC Council Directives 81/851/EEC and 81/852/EEC. There were also guidelines presented in a proposed draft directive entitled "Testing of medicinal products for their mutagenic potential" which was published as Annex II in the Official Journal of European Communities, C 293/11, dated 5 November 1984.
180New Zealand added that, furthermore, since that time, the European Communities has not utilised the applicable assessment and evaluation processes to assess the use of these substances for growth promotion although it had assessed and approved their use for other therapeutic purposes without stipulating any MRLs for example in Commission Regulation (EC) 3059/94 of 15 December 1994).
181New Zealand noted that even the 1995 Scientific Conference sponsored by the European Communities, which was procedurally defective as a risk assessment process in that it excluded certain relevant participants and information, nonetheless, concluded that there was sufficient scientific information to determine that the five growth promotant substances could be used safely.
WT/DS26/R/USA
Page 112
PANEL'S CONSULTATION WITH SCIENTIFIC EXPERTS
The United States noted that Article 11.2 of the SPS Agreement referred to establishment of an "advisory technical experts group". In the view of the United States, this phrase referred to an expert review group established under Article 13.2 of the Understanding on Rules and Procedures Governing the Settlement of Disputes (DSU). Thus, the rules relevant to the establishment and procedures for such a group were those provided in Appendix 4 of the DSU. The Panel's conduct of the expert review process in the dispute should assure transparency, avoid conflicts of interest, affirm the integrity of the dispute settlement process, and aid public confidence in the outcome of the dispute. In this case, the Panel was to review the EC ban to determine if it was based on scientific principles and on a risk assessment, and if there was sufficient scientific evidence to support the ban. The scientific experts' role was a narrow one, they might help the Panel determine if there was a scientific basis for the EC ban.
As regarded the selection procedures and criteria, the European Communities considered that the use of experts by the Panel for the purposes of scientific and technical advice should respect general principles of law. In particular, it should be transparent, avoid conflicts of interest, affirm the integrity of the dispute settlement process and aid public confidence in the outcome of the dispute182. In addition, Appendix 4 of the DSU included general rules of law relating to the use of expert advice on the part of Panels. Even though the experts would not constitute an Expert Review Group but would be consulted individually and separately, such rules should be applied by analogy whenever relevant and appropriate183. In order to ensure that the principles outlined above were respected, the European Communities believed that a number of general criteria should be followed in the selection of the scientific experts by the Panel.
The European Communities indicated that, first, the experts nominated by the Panel should not be nationals of the parties to the dispute nor nationals of third parties in the same dispute. This principle aimed at ensuring a fair and impartial selection of experts and appeared all the more necessary in light of the possibility given by the Panel to the parties to nominate one expert each, without any limitation of nationality. Second, the Panel, in selecting experts with different fields of expertise, should ensure that all the areas which it had identified were covered, so that all the questions of the Panel received replies to the fullest extent possible. Third, given the very small number of experts that the Panel would consult, only scientists with proven expertise in the use of hormones in general and for animal growth promotion should be selected. Fourth, the experts should not have served as members of the Codex/JECFA group which produced the 1988 JECFA report on the use of hormones for animal growth promotion. Fifth, past or present significant ties with the industry producing these hormones would not provide sufficient guarantees of lack of conflict of interest, of integrity and of aiding public confidence in the outcome of the dispute. The European Communities considered that the above comments constituted compelling reasons for the European Communities and underscored the limited pool proposed by the Codex secretariat from which the Panel would make its choice, despite the existence of a large number of scientific experts on these issues internationally. The European Communities indicated that Panel's favorable response to the EC request to add a fifth scientist with expertise in the area of carcinogenic effects of hormones in this case addressed one of the concerns of the European Communities.
As regarded the purpose of requesting information from, and the type and nature of the questions to be asked to, experts, the European Communities stated that this should be "to further the Panel's understanding of the scientific facts relevant to the dispute". In order to fulfil this objective and be
182The European Communities noted that in a letter addressed to the Panel on 21 October 1996, the United States had also agreed to these general principles.
183The European Communities noted that this had also been acknowledged in the Panel's letter of 30 October 1996.
WT/DS26/R/USA
Page 113
in conformity with the provisions of the SPS Agreement and the DSU, the questions the Panel intended to ask must relate directly to the scientific issues of the case. These questions must be addressed to persons who on the basis of their recognized fields of expertise, could provide direct, reliable and verifiable information to the Panel. The questions should not concern legal issues or issues of interpretation of any of the WTO Agreements under consideration, becauseit was the sole responsibility of the Panel to interpret the legal provisions which had been invoked by the Parties and to apply them to the facts of the present case. The European Communities observed that the questions the Panel intendedto ask should also not address nor request pure factual information, which was the responsibility of the parties to the dispute to provide, in accordance with the DSU rules and its provisions on burden of proof. They should not address issues of trade policy either, which were again the sole responsibility of the parties and on which scientists were not competent and could not be expected to pronounce themselves. The European Communities claimed that scientists must be asked only scientific questions directly relevant to this dispute and in the light of the scientific arguments made by the parties to this dispute. The European Communities recalled that it was the Panel's responsibility to interpret and apply the relevant legal provisions. Accordingly, the European Communities considered necessary to stress that it reserved all its rights with regard to the conformity of the Panel's approach on these issues with the relevant provisions of the SPS Agreement and the DSU.
Following consultation with the parties, the Panel decided to seek scientific and technical advice as foreseen in paragraph 1 and 2, first sentence, of Article 13 of the DSU, and pursuant to paragraph 2, first sentence, of Article 11 of the SPS Agreement.
Names of individuals expert in the subject matter before the Panel were provided by the Codex Commission secretariat as well as by the International Agency for Research on Cancer ("IARC"). Brief curricula vitae were solicited from all experts who were prepared to assist the Panel.
The parties were provided the opportunity to comment on these potential experts on the basis of the curricula vitae, and in particular to state any compelling objections they might have with regard to any individual. The parties were invited to nominate one expert each, not necessarily from the list provided by the Panel. The Panel than selected two additional individuals from the list taking into account the comments of the Parties. At the request of the European Communities, the Panel decided to select an additional expert in the area of carcinogenic effects of hormones, on the understanding that the European Communities agreed that the Panel established at the request of Canada would seek advice from the same five experts. These experts were requested to serve, in their personal capacities, as individual advisers to the Panel. The Panel also sought information from the Codex Commission secretariat.
The Panel, in consultation with the parties, prepared specific questions which it submitted to each expert individually. The experts were requested to provide their responses, in writing, to those questions they felt qualified to address. The parties agreed that their written submissions to the Panel, including the written versions of their oral statements, be provided to each of the selected experts. Written questions were also submitted to the Codex Commission secretariat. The written responses of the experts and of the Codex Commission secretariat were provided to the parties.
The experts were invited to meet with the Panel and the parties to discuss their written responses to the questions and to provide further information. A summary of the responses provided by the experts is presented below.
The experts selected to advise the Panel were:
Dr. Francois André, Laboratoire des dosages hormonaux, France
WT/DS26/R/USA
Page 114
Dr. Dieter Arnold, Deputy Director, Federal Institute for Health Protection of Consumers and Veterinary Medecine, Germany
Dr. George Lucier, Environmental Toxicology Programme, National Institute of Environmental Health Sciences, United States
Dr. Jock McLean, University of Swinburne, Pro Vice Chancellor, Division of Science, Engineering and Design, Swinburne University of Technology, Australia
Dr. Len Ritter, Executive Director, Canadian Network of Toxocilogy Centres, University of Guelph, Canada
Dr. Alan Randell of the Codex secretariat also advised the Panel.
"Good veterinary practice" described each country's national or professional guidelines or rules which provided guarantees for professional activities. The use of veterinary drugs was controlled by means of an international code of practice of the Codex Alimentarius which described the use of veterinary drugs, from the necessity of a pertinent diagnosis to residue management (respect of withdrawal periods) in meat-producing animals. Dr. André noted that special attention should be paid to the prescription and to using the correct dosage, site and route of administration for medicines. Moreover, specific conditions for the use of particular drugs was regulated; ie., in Europe, EC Directive 96/22 described precisely the conditions of therapeutical or zootechnical use of the hormones in dispute.
Dr. André noted that "good veterinary and animal husbandry practice" did not appear to be well defined but could be considered as a combination of the two above-mentioned practices. The strict observance of "good veterinary and animal husbandry practice" was the only way to guarantee that each drug would be used according to the conditions for which it had been registered. Any disrespect of such practices, in term of dosage, injection route, withdrawal period or other conditions specified for use, would change the expected level of residues. Concerning the hormones in dispute, it could easily be demonstrated that overdosage, shorter withdrawal periods (when defined), or liquid solution injections instead of implants might lead to higher residue levels. Potential hazards to human and animal health might then appear.
WT/DS26/R/USA
Page 115
Dr. André indicated that the real problem was one of control: how to ensure that these practices were really respected. There were official control services charged with this control. But now more and more veterinarians, farmers, and farmer organizations were taking the control and were ensuring that they only used registered drugs as they were officially meant to be used.
Dr. Arnold observed that "good animal husbandry practice" was a broad term relating to the proper management and handling of animals. Particularly intensive production methods required good management. The terms management and husbandry included, for example, efforts toward improving rate of gain and feed efficiency and management of reproduction (e.g., hormonal control of oestrus, embryo transfer, treatment of infertility, prevention of pregnancy, termination of pregnancy, induction of parturition, etc.). Veterinarians were involved in the management of several of these conditions and were inter alia responsible that certain drugs were only administered after thorough diagnosis and in compliance with the approved label instructions. It was also a fundamental requirement that drugs (hormones, growth promoters, parasiticides and other substances) should not be used to replace good veterinary and animal husbandry practices.
Dr. Arnold indicated that the hormones under discussions were not per se hazardous or innocent substances. If ingested with food, their possible biological effects depended inter alia on the amounts ingested. Therefore, it was important that during the approval process conditions of use were established which guaranteed that even preferential consumption of food from treated animals would not lead to the ingestion of residues causing biological effects. The essential issue was compliance with the established conditions of use which were normally laid down in the label instructions. If "good veterinary and animal husbandry practices" were not applied, in particular if "good practices in the use of veterinary drugs" were not observed when using these substances, residue levels higher than those resulting from the authorized conditions of use may occur/were likely to occur in edible portions of the treated animals. These elevated levels did not always have biological effects - because the established margins of safety were usually high.
Dr. McLean indicated that departure from "good practice" was undesirable and a breach of the registration conditions. However, in practice, it was unlikely that minor departures from good practice with registered products would cause significant increases in hormone levels in meat.
WT/DS26/R/USA
Page 116
good practice in the use of veterinary drugs". The maximum permitted residue was based on the type and amount of residue considered to be without toxicological hazard for human health while taking into account other relevant public health risks. When the hormones at issue were used in accordance with the prescribed use conditions, the levels that were present were in fact indistinguishable from those which would be present in an animal which had not been treated at all.
Dr. Ritter further noted that the principle or philosophy of good veterinary and good husbandry practice also implied that all veterinary products to be utilized in food producing animals should only be administered in strict compliance with relevant product information, approved by the appropriate regulatory authority, and in strict accordance with instructions issued by qualified veterinarians. Violations of any of these principles might result in higher than expected residues or residues of unexpected drugs. The implications of such violations were proportional to the frequency with which they occurred, the magnitude of the violation and the prevalence of such violations within the food production system. Notwithstanding, given the exaggerated estimates of safety inherent in the approach to the establishment of MRLs, it was unlikely that occasional violations could increase overall residue levels to those which, on an infrequent exposure basis, could pose a risk to consumers.
Furthermore, sophisticated detection methods allowed differentiation of meat produced with or without synthetic hormones, even if a long withdrawal period was respected. For natural hormones, this differentiation was possible when esters of natural hormones were still present in the meat. In some cases it is possible to distinguish the two by physical means of isotope relations, but it would be very difficult to do at very low levels, Theoretically, quantitative differences might be expected between the legal use of natural hormones as growth promoters (i.e. as an implant) and their use as veterinary drugs (i.e. usually as injection); from the qualitative point of view, chemical differences could be expected in terms of metabolites and/or conjugates, owing to the dosage, the route of administration and the length of treatment.
WT/DS26/R/USA
Page 117
Of the categories of residues of biological concern, only the residues of the parent drug administered to the animals (e.g. the hormones itself or its esters) could be analytically differentiated with currently available analytical methods. Theoretically, hormones could be labelled with stable isotopes to permit discrimination between added hormones and endogenous hormones, but this had no relevance in practice. It was possible to discriminate whether or not animals had been treated with exogenous hormones. Despite their short half-lives in circulation, it was even possible, but not normally routinely feasible, to measure the concentrations of the esters at sites distant from the injection site.
It appeared that the substances used for growth promotion in the United States were (with the possible exception of testosterone-propionate) also used in the European Union for therapeutic or zootechnical purposes. Working-Group III of the 1995 EC Scientific Conference had concluded that: "While residue analyses for natural hormones on animals treated experimentally can show differences between them and untreated animals, it is not possible in routine use to identify treated animals on the basis of assays of edible tissue samples and it is difficult in blood, urine or faeces. Successful identification of treated animals may be made by analysis of injection sites or detection of esters (e.g. oestradiol benzoate) in blood shortly after treatment. Identification of animals treated with natural hormones on the basis of a single time-point analysis is difficult. The situation is more complex for food-producing animals, where typically a mixture of androgen only is used. Identification of meat from animals treated with natural hormones and imported from third countries is not at present possible"184. (emphasis added)
Dr. Lucier noted that it was likely that, on the average, meat from growth-promoted animals would be leaner (i.e. less fat) than meat produced without those agents. However, it was unlikely that, for a given piece of meat, one could physically differentiate whether or not it was from a growth- promoted animal. Residues arising from administration of the three natural hormones would not be chemically different from those in the body of non-treated animals although the residues were likely to be slightly higher in the growth-promoted animals.
184Working Group III, Detection and Surveillance, in 1995 EC Scientific Conference Proceedings, p.31.
WT/DS26/R/USA
Page 118
hormones, and might be maximized by the use of certain combinations of hormones. In general, the lean meat content of the carcass was improved and the fat content was decreased. Except in the case of some veal calves, effects on meat quality were generally not significant. Thus it was most unlikely that consumers could differentiate physically between meat produced with or without these hormones. JEFCA concluded that it would be very difficult to determine residue levels attributable to the exogenous use of the naturally occurring gonadal hormones.
While it was most unlikely that consumers could physically differentiate meat produced with synthetic hormones, residues could be reliably detected well below those levels regarded as safe. However, qualitatively the residue would be indistinguishable. Quantitatively, the amount that could be detected would be a function of the amount that would have been administered either for growth promotion or therapeutic uses.
Other foodstuffs than meat from hormone-treated cattle contained hormones, especially oestrogens. This was the case of offals, of certain plants, of meat from specific animals (such as pregnant cows); all contained high levels of hormones. However, the consumption of these foodstuffs was always occasional and would not enhance risks for human health. This was not comparable to the systematic enhancement of the hormonal content of meat.
Whether animals were slaughtered without any withdrawal time or at the end of the legally established withdrawal time, there would always be a residue. The exogenous compound was not absent at the end of the withdrawal time; however, the sum of both the endogenous and exogenous molecules would be at or within physiological limits. If the presence of residues were to be regulated on the basis of "no-detectable-residues" (which was impossible because the endogenous hormones alone were already above the limit of detection), there would nonetheless be more than 1000 billions molecules of the administered exogenous hormone present in one kilogramme of meat at the limit of detection. In other words, the "no-detectable- residue" concept was a false-zero-residue“ concept. For any of the six hormones, it was impossible to estimate the withdrawal times required to reach true "zero-residue" levels even with the aid of the most sensitive analytical methods.
WT/DS26/R/USA
Page 119
Many foods of both plant and animal origin contained natural hormonal substances in such amounts that, depending on the actual diet, the dietary intake could be lower, equal or higher than the amount ingested with beef from hormone-treated cattle. If the endogenous production was taken together with the dietary intake, it would always be much higher. Given their natural presence in food, he stated that one cannot escape eating these hormonally active substances, or similar substances with related biological potential.
Differences between natural occurring hormones at a natural occurring level and the same hormones at ahigher levelor synthetichormones (whatever their levelof residue) resulted systematically in different potential risks to human health, due either to the level of residue or the nature of the residue (metabolites). According to the route of administration of additional hormones, some differences in metabolites patterns might appear. Metabolites could be species-specific: metabolites not formed in cows might form in humans and result in potential risks for human health. The potential induced risk
WT/DS26/R/USA
Page 120
could be due to the hormonal activity of residues. The potential adverse effect could also be a carcinogenic effect.
Within the physiological range these hormones had no different biological effects on human beings. But if one ate only meat from treated animals, there would be a very small enhancement of the mean of food hormonal intake and this would not be transformed in classical biological effects for human beings. The problem was that some other biological effects had needed to be studied over a long time. Such effects included, for example, changes in human fertility or change in the sex ratio, which was currently changing in some countries. But there was no currently known relationship between these developments and the fact whether these countries banned or not these hormones.
In terms of biological activity, in principle, the potential biological effects of the "nature-identical" exogenous hormones were the same as for the "natural" endogenous hormones. While the concentrations of endogenous hormones were, however, regulated depending on e.g., sex, age, developmental stage etc., the exogenous hormones could, in principle, be added in such amounts as to either gradually disturb or - at high doses - even overwhelm the internal regulatory mechanism. Single high doses (e.g., resulting from the ingestion of a "full" injection site) would normally cause transient effects. Repeated high doses (not reached through ingestion of residues resulting from uses such as growth promotion, therapy, or zootechnical treatment) could significantly change the hormonal balance of the individual.
Dr. Arnold summarized that the potential risks to human health arising from the hormonal activity of the xenobiotic hormones was slightly different in quantitative terms. The potential risks arising from other than hormonal actions were qualitatively different. These risks were assessed during the review and approval process, and the approved conditions of use eliminated all unacceptable risks.
Some biological effect could occur. For example, a normal woman might have 30 per cent of her oestrogen receptors occupied at some given point in time. At that same point in time, if she was consuming meat that contained an additional burden of oestrogens because of the use of growth promoters, that receptor occupancymight be something like 30.01 or 30.001, a very very small increase. This increase would not be detectable, not even close to being detectable, by any experimental tools available today. Thus, a biological effect could be occurring; if it was occurring it would not be detectable; and finally the relationship between that biological effect (receptor occupancy) and a toxic effect, say cancer or birth defect or something like that, would be unknown. But if such an effect were occurring, it would be extraordinarily small, close to zero.
WT/DS26/R/USA
Page 121
would also be different. In the case of the naturally occurring hormones, the residues in meat and humans would be very similar. Any difference would not pose a potential risk to health.
In relation to the naturally occurring hormones, Dr. McLean noted that it was not possible to differentiate between the effects of meat from treated animals against meat from untreated animals because essentially they were in the same biological range. In the case of the non-naturally occurring hormones, in establishing the ADI very sensitive end points were derived from studies in non-human primates, and then the situation of sensitive members of the population was taken into account when establishing safety factors. The levels in meat were substantially below those causing any effects in primates, and there was a reasonably good correlation in effects of hormonal levels in primates against humans. Therefore the levels present in meat from animals that were treated with the two non-natural growth promotants would not cause effects in humans consuming the meat.
WT/DS26/R/USA
Page 122
When individual drugs were used in common veterinary practice for therapeutic purposes, the probability of the simultaneous use of two drugs was low; the probability of an interference between them was lower, and lower still was the probability that one could influence the elimination rate of another. For the hormones used as growth promoters, the problem was different. They were used on a very large scale. Many animals were treated so that the probability to have simultaneous administration of other drugs and of these hormones became much higher. A good risk assessment should take this into account, but he was not aware of that having been done either in the countries where these hormones were in use or at the international level.
In the Community, the competent advisory body for the safety evaluation of , progesterone and testosterone was the Committee for Veterinary Medicinal Products (CVMP), a section of the European Medicines Evaluation Agency (EMEA). EMEA was also, in principle, the competent authority for the review of applications for marketing authorization of veterinary medicinal products used for growth promotion. A centralised review process would be mandatory under the provisions laid down in Regulation 2309/93. The use of any hormonal substance for growth promotion was, however, currently prohibited.
Under the provisions of Articles 7 and 14 of Council Regulation 2377/902185, Maximum Residue Limits (MRLs) had to be established for the therapeutic and zootechnical uses of the three "nature-identical" substances before the end of 1996. The EMEA reviewed (in accordance with regulation 2377/90) the therapeutic uses of in cattle and horses. The public summary report states:
"3. Toxicological effects found after oestradiol administration comprise hyperplasia of the endometrium, behavioural changes and effects of metabolic processes. Oestradiol does not induce gene mutations in vitro, but conflicting results are found in chromosomal aberration assays. Following long term exposure the incidence of tumours in tissues with a high level of hormone receptors is increased (e.g. mammary tumours). It is concluded that the toxic effects including carcinogenicity occur as an extension of the physiological effects of oestradiol ...
7. The conclusion of the FAO/WHO Expert Committee on Food Additives (JECFA) that no ADI and MRLs for oestradiol need to be established is adopted. Milk and plasma residue levels after treatment with oestradiol benzoate and oestradiol valerate have shown to be at or within physiological limits. Although it is likely that tissue residue levels will also be within physiological limits, this cannot be guaranteed, given the results with oestradiol hexahydrobenzoate. Still, compared to the lowest human daily production rate of oestradiol in prepubertal boys (6 (µg/d)) and compared to the amount of oestradiol in other food stuffs that are part of the human diet, the amount of exogenous oestradiol that humans will be exposed to through ingestion of tissue from treated animals is biologically insignificant, and will be incapable of exerting a hormonal effect in human beings".
185Council Regulation (EEC) 2377/90, EC Official Journal L224/1 of 18 August 1990.
WT/DS26/R/USA
Page 123
EMEA also evaluated progesterone in accordance with regulation 2377/90. The public summary report consists of 17 paragraphs and shows that the following aspects had been considered by the Committee: natural occurrence, veterinary uses, target animal safety, biological effects, endogenous production in humans, normal levels in food animals and in food of animal origin, pharmacokinetics and metabolism, acute and short-term toxicity in laboratory animals, reproductive toxicity, teratogenicity/embryotoxicity, and long-term effects including carcinogenicity. The evaluation by JECFA of this substance was also considered.
With respect to carcinogenicity it was stated:
"9. According to the International Agency for Research on Cancer (IARC), progesterone does not exhibit mutagenic activity in most in vitro and in vivo tests performed, but is known to increase the tumour incidence in endocrine target tissues (ovaries, uterus, mammae) after continuous (parenteral) doses clearly above the physiological levels. Progesterone is not carcinogenic per se, but acts via an epigenetic mechanism associated with its endocrine activity,
i.e. its ability to cause a hyperproliferative effect at cellular levels mediated by steroid-hormone receptor interaction. Hence, tumours will not result from ingestion of progesterone at levels that do not produce any hormonal effects".
The evaluation process of testosterone in accordance with regulation 2377/90 was still ongoing.
So far no complete risk assessment had been conducted in the Community; the most competent assessment was made by the Scientific Working Group on Anabolic Agents in Animal Production chaired by Professor Lamming. This Committee comprehensively reviewed the available scientific evidence, includingindustry data, concerning five of the substancesunder dispute. Its recommendations were endorsed by the Scientific Committee on Animal Nutrition, the Scientific Committee on Food and the Scientific Veterinary Committee of the EC. However, when the interim report was published, long-term toxicity data on the xenobiotic substances was available only for trenbolone acetate. The report also does not mention that "synergistic effects" and exposure to combinations had been considered. The "Pimenta Committee" did not itself perform a scientific risk assessment but rather expressed views on public perception of scientific opinion, the role of science in society, consumer preference, other socio-economic factors and various aspects of implementation and control of a ban.
The 1995 EC Scientific Conference had no legal mandate and the Steering Committee was entirely free to design the programme and to nominate invited participants. A legal notice in the report clarified that neither the European Commission nor any person acting on behalf of the Commission was responsible for the use which might be made of the information. The Conference did not assess the hazards referred to by the EC in this dispute. With regard to the Codex process, Dr. Arnold noted that it was true that the EC had objected to move these MRLs to the next step in the procedure, but without raising any health issues. The EC had opposed because it had specific legislation prohibiting the use of, and consumers did not wish to eat meat from, treated animals.
Sufficient time had elapsed to allow an assessment of the long-term potential effects on the health of the target animals. If long-term effects of these substances were to be studied, the effects of the endogenously produced amounts of these substances and of the exogenously administered human doses in therapy etc. were most relevant. Lifestyle and ingestion as normal food constituents could also be important. In the unlikely event that the minute amounts added from residues would have any effect on human health, it would not be possible to discover this effect against this background.
In response to the follow-up questions (in particular, question 29), Dr. Arnold noted that the primary mode of exposure to food additives was through ingestion of the substance in the food supply. JECFA considered, where applicable, the occurrence of the substances as normal body constituents; the natural occurrence in foods; exposure from other uses; and interactions with other food additives.
WT/DS26/R/USA
Page 124
For example, when canthaxanthin was evaluated, the Committee considered also the uses of this substance as feed additive in animal nutrition and as an orally administered pigmenting agent for human skin in both pharmaceutical and cosmetic applications. Early in its deliberations about safety margins the Committee had concluded that the margin of safety should allow for, inter alia, the possibility of synergistic action among food additives. The ADI for food additives generally included the natural occurrence and the deliberate addition to food of the substance. Testing of "mixtures" was not normally required. However, some food additives were by their very nature mixtures.
With regard to pesticides, testing of "mixtures" was also not required, neither in the framework of the activities of the FAO/WHO Joint Meeting on Pesticide Residues nor in the legislation of the European Communities as laid down in Directive 91/414/EEC Annex II, Part A (O.J. L 230 pp.15-18).
For veterinary drugs, testing of "mixtures" was not required. However, the comprehensive basic pharmacological screening required for all drugs could provide information requiring further follow-up in specifically designed tests. JECFA had evaluated a number of substances where the ADI was based on a pharmacological No Observed Effect Level. The legislation of the European Union also did not require testing of "mixtures" (see Directive 92/18/EEC amending the Annex to Directive 81/852/EEC, O.J. L 97 of 10 April 1992, pp.1-23). However, there was a difference between "mixtures" and fixed combinations. Such combinations required justification on the basis of adequate testing. Repeated dose toxicity testing could be modified (i.e. reduced) in certain instances, for example, if new combinations of known substances were proposed and no potentiating effects had been observed during initial screening.
Dr. Arnold noted that testing of "mixtures" of hormones was not normally required, unless an unknown fixed combination was administered. He noted that the experimental animals in toxicity testing (or humans, if there were uses in human therapy) were typically exposed to all hormones existing in the body. Virtually all cells were targets of one or several of the approximately 50 known hormones and their metabolites. The same hormone could have several targets; the same cell might have several responses to a single hormone. The battery of tests included exposure at all stages of development from embryonic to fetal to neonate to reproductive state until the end of the whole lifespan, and this way all situations of hormonal control were covered. This meant also that all xenobiotic hormonally active substances had been tested with respect to possible interactions with the hormone system. No specific toxicity testing had been performed with the fixed combinations authorized in the European Communities.
JECFA developed its strategies concerning the evaluation of anabolic hormones in its 25th and 26th Report. These indicated that the toxicological evaluation of residues of anabolic agents that are present in human food obtained from animals treated with these agents must take into account whether the residue was identical to a human endocrine hormone. In the latter case, the possible endocrinological effects and carcinogenic potential of the residue must be closely examined. In addition, it was noted that chemically modifiedhormones, hormonally active agents from plants, and synthetic anabolic agents presented the following specific problems:
extreme potency and consequently the need to ensure minimal residues;
potential tumorigenic activity; and
the presence of their metabolites in animal products that might be of endocrinological or toxicological consequence.
The evaluation for acceptance of the use of xenobiotic anabolic agents in animal food production resembled in many respects the evaluation of pesticides, since the two essential elements required were:
WT/DS26/R/USA
Page 125
adequate, relevant toxicological data; and
comprehensive data about the kinds and levels of residues when the substances are used in accordance with good animal husbandry practice, which required evidence as to the efficacy of the anabolic agent, the amounts used to produce the effect, the residue levels based on field trials, and information about methods of analysis of residue levels that could be used for control or monitoring purposes.186
The number of moleculesthat remain following appropriate use of these agents was very small, particularly in relation to the amount of naturally occurring oestrogens or androgens. So the accompanying risk that would be associated by consuming meat containing residues would be extraordinarily small. It would be very hard on scientific grounds to say that the risk was zero, but it was likely to be very, very small. It could be zero.
In terms of the carcinogenic activity of the hormones in question, it was already known that naturally occurring physiological levels of androgens and oestrogens were carcinogenic. Therefore, the issue of threshold was irrelevant to the toxicological evaluation of these agents. If a hundred thousand molecules of something already exist in the body and some of those molecules were producing a DNA damage event, there was some possibility that the same damage event would occur if another molecule was added to it because it was the same molecule. It would not be possible to distinguish that additional event from the ones that occurred from the thousand molecules. But it would not be possible to say that none of those events were related to that additional one molecule. This would not be detectable. The chances were very very small, but it would be impossible to say that the event could not occur.
With regard to the increased incidence of some tumours occurring in various countries in the world referred to by Dr. Epstein, it was probably true that there was a real increase in testicular cancer and this increase appeared to be predominant in young men which was especially disturbing. Why this was occurring, no one really knew its cause. There was no reason to think, however, that it was associated with oestrogen. There were many other factors that could be causing that. Breast cancer was also on the rise. However, one could not necessarily attribute that increase to exposure to exogenous external oestrogens. There were many reasons why exposure to genotoxic agents could be accounting for the rise in breast cancer rates; one needed to be cognisant of other factors and not just blame the exogenous oestrogens for everything. Regarding the oral contraceptive issue, Dr. Lucier indicated that Dr. Epstein was probably right that there would be an increase in breast cancer risk for women who started taking the pill very young because that extended the period of time in which they were exposed to high levels of oestrogen and this was a known risk factor for breast cancer. The same exposures a little bit later probably would not have an increase, so when averaged out, there was no statistically significant increase in breast cancer from oral contraceptives. No one challenged the fact that oestrogens were carcinogenic.
With regard to Dr. Metzler comments, Dr. Lucier noted that the adducts that may arise from synthetic materials were likely different and of more concern, in terms of a risk assessment, than the ones that arise from the naturally occurring ones, because there already was a given body burden of
186Environmental Health Criteria 70, WHO, Geneva, 1987.
WT/DS26/R/USA
Page 126
the naturally occurring oestrogens. The question of genotoxic or non-genotoxic was essentially irrelevant here since the dose response relationships for oestrogen increases in cell division (one possible mechanism of carcinogenesis) was likely to be linear because normal levels of oestrogen were causing cell replication.
Dr. McLean replied that the 1988 and 1989 JECFA reports addressed the issues of carcinogenicity and genotoxicity and their findings were accepted by Codex. The compounds were not genotoxic carcinogens and the responses seen were related to their hormonal effects. These hormonal effects were evaluated using sensitive end-points in non-human primates. In the case of synergistic effects, it was recognized that the levels resulting from treatment with the naturally occurring hormones were significantly less than the natural levels already present in the treated animals. In addition, it must be recognized that all three hormones occurred in male and female animals and humans. With regard to mutagenicity, data for the naturally occurring substances was reviewed and referenced in the JECFA 1988 Report. The report was not silent on the matter but it referred to mutagenicity data that was in the published literature. In the case of progesterone, testosterone and , there were a number of published documents that JECFA did look at, and they are referenced in its report.
There was also human data available from therapy with various natural and synthetic hormones. This therapy was continued in humans for many years and the risks have been evaluated. The levels found in meat after treatment are many orders of magnitude lower than therapeutic levels, are poorly absorbed orally and so do not represent a hazard. The compounds had been used in many countries for long periods of time and there was no human epidemiological data which suggests any hazard. With regard to Dr. Epstein's observations related to increasing numbers of tumours, Dr. McLean observed that the increase in tumours occurred also in those countries where the use of growth promotants was not widespread and the increase began before the hormonal growth promotants received widespread use.
With relation to oestradiol, progesterone and testosterone, JECFA concluded that the increased tumour incidence was attributed to the hormonal activity of these hormones associated with the very high doses utilized in the experiment. These effects were not considered relevant to an evaluation of the safety of food residues. The Report was silent on the results of mutagenicity studies, which were available to the Committee for review. With relation to trenbolone acetate, JECFA similarly concluded that the carcinogenic effects detected were a direct consequence of the hormonal effect associated with the high doses utilized in the study and not of direst relevance to the interpretation of the safety of the food residues. JECFA also concluded that the tumorigenic effect of zeranol was associated with its oestrogenic properties at the high doses utilized in the cancer study, and therefore a safe exposure level to humans could be determined.
The issue of potential synergy was addressed, at least in part, by JECFA through the conduct of biochemical studies directed at the effect of excretion of hormone combinations when compared to single hormones alone. Although these studies did not specifically address issues related to altered toxicity due to synergy, these limited data did provide evidence that while administration of hormone combinations could be expected to alter excretion kinetics quantitatively, there was no evidence of important qualitative changes. It was, however, clear that definitive studies relating to genotoxicity or carcinogenicity of hormone combinations had not been carried out, even though this was frequently the preferred method of use.
WT/DS26/R/USA
Page 127
Dr. Ritter observed that the use of gonadal type hormones as anabolic agents had been in agricultural practice for approximately 40 years through many parts of the world. Regarding the synthetic hormones, both zeranol and MGA had been used as anabolic drugs, at least in the United States, for approximately thirty years while trenbolone acetate had been in use for approximately ten years. At least in the case of the natural hormones and both zeranol and MGA, broad scale use had now been in place for more than a generation with only few and isolated reports of potential adverse effects. The primary potential concern with regards to human health effects had been that of carcinogenicity. While the use of the growth promoting hormones for periods as long as 40 years should have been sufficient to result in an increased cancer risk in the human population, an increase which could be attributable to the use of these hormones had not become apparent. Noting that both breast cancer, particularly in post-menopausal people, and prostatic cancer had increased sharply in affluent countries. Dr. Ritter pointed out that this increase was thought to be primarily associated with genetic predisposition and significant alterations in lifestyle (Houghton and Ritter, 1995).
Dr. A. Randell commented on the relationship between the Acceptable Daily Intake (ADI) and the Maximum Residue Limit (MRL) and whether or not these were measures of acceptable risk. The maximum residue limit was definitely not a health-based limit. It was a limit established for the control of veterinary drugs in actual practice. The establishment of a maximum residue limit was such that the maximum residue would never lead to residues which would, in a normal ingestion of the product, cause any consumer to exceed the acceptable daily intake. Therefore there was an upper boundary to the maximum residue limit imposed by toxicology considerations. The lower limit was normally derived from residue trials in practice according to the proposed good veterinary practices in the use of these drugs. However, in the case of zeranol, this limit fell so far below not only the toxicologically derived limit but also the limit that one could determine with normal methods of analytical chemistry, that it was increased to take into account the fact that if a decent control programme were not going to be overly expensive, than the limit must be controllable by analytical procedures rather than by residue trials. The acceptable daily intake itself also was not a direct risk assessment in the sense that it provided a statement of quantitative risk. The ADI concept stated that there would be no appreciable risk as a result of exposure to the chemicals concerned. This was true within the JECFA and JMPR framework, for food additives, residues of veterinary drugs and pesticide residues. (Dr. Lucier's comment that the risk was probably somewhere between zero and one in a million, provided some sort of quantitative framework, but JECFA had never established a quantitative evaluation of risk in the application of the ADI.)
Dr. André noted that recent reports of the effects on animal health from the use of the six hormones in dispute as growth promoters confirmed and finalized previous results.187 Based on anatomical or histopathological observations, these results demonstrated the effects of commercially available implants on animal behaviour, development and reproduction. Young bull behavioural modifications had been confirmed: young bulls treated with zeranol "spent more time idling, eating
187Renaville et al. (1987), Dehaan et al. (1990), Moran et al. (1988), (1990), Herenda (1987), Beal et al. (1988).
WT/DS26/R/USA
Page 128
and ruminating than controls". 188 Animal libido was systematically altered. The changes in the behaviour of bulls were usually correlated with histopathological modifications, well described in the 1980s as a tool for treatment diagnosis and more recently confirmed in bulls.189 Similar phenomenon had been recently studied in other species, as the effect of trenbolone in male pigs.190 The influence of these hormones used as growth promoters on reproductive tract development and reproduction had been confirmed in beef heifers: "The lower pregnancy rate in zeranol implanted heifers 100 days after exposure to bulls was caused by failure to cycle early in life and, in those that were cycling, failure to conceive and abortions between 25 and 45 days of gestation".191
First results of recent research work with mice on the effects of hormones on animal sexual development and reproduction were published in 1995. They concluded that "DES and Zeranol administered during mid pregnancy leads to decrease fetal weight and size and lower numbers of male offspring at birth".192 In 1996, the same authors demonstrated "that prenatal exposure to zeranol or DES induced abnormal testicular differentiation in the mouse".
Concerning human health, there was a discussion on the decrease of human sperm counts193, an apparent increase of the incidence of hormonally-mediated diseases like breast cancer and endometriosis, and a decrease in the male/female ratio, which were thought to be due to oestrogens in the environment. Whether the use of hormones for growth promotion was contributing to this could not be scientifically proven at the present time.
Extrapolations from animal to human was an official tool for toxicological and efficiency evaluation of drugs under development. In general, they were regarded as conservative even if they did always contain a certain level of uncertainty. Moreover, these tests on animal models did not take into account the multiple exposure to many different compounds (characteristic for human diet). Humans were sometimes more sensitive for a certain effect than animals, and they were not always in as good health as laboratory animals. The action of food processing on residues was not checked in animal experiments. For some effects of chemicals on humans, such as long term effects over generations, no animal models existed. In other cases, results on animal models have sometimes been considered as not representative. For these reasons, many drugs which had been fully validated in animal models had been banned after several years of use in humans and the discovery of an unknown hazard.
Dr. Arnold indicated that since the last (JECFA) evaluation, science had made further progress in several areas, including the molecular mechanisms of action and the epidemiology of cancer. The newly available information did not, however, substantially change the basis for the evaluation of the five substances by the Codex/ JECFA system. Although animal studies, in general, could provide useful information on possible adverse effects on human health resulting from exposure to chemical substances, the possible health effects resulting from the oral ingestion of trace amounts of residues of the three nature-identical hormones, (oestradiol-17 progesterone and testosterone) could not be reliably and quantitatively predicted from the information contained in available studies.
188Legoshin et al. (1994). 189Tipirdamaz 1991, Ciftci (1990). 190Lopez-Bote et al. (1994).
191King et al. (1995).
192Perez-Martinez et al.
193Nimrod and Benson (1996).
WT/DS26/R/USA
Page 129
Early studies in mice, which were conducted between 1940 and 1973, and which were referenced in the IARC Monographs, were in no way designed to assess the human cancer risk resulting from oral ingestion of low doses of . None of these studies used the oral route of exposure in any animal species. None of them was conducted under GLP-like conditions or according to contemporary test guidelines. Many of them were poorly documented. The principal aim of these studies was to experimentally produce tumours under a variety of condition. The test substance was usually subcutaneously administered (frequently as implant). The doses were often extremely high and not infrequently caused a high number of early deaths. Such dosing was inadequate for carcinogen testing in relation to the safety evaluation of residues. The studies demonstrated that subcutaneous injection or implantation of high doses of
or of its esters alone or in the presence of other carcinogenic factors resulted in increased incidence of mammary, pituitary, uterine, cervical, vaginal, lymphoid and interstitial-cell tumours in mice. The IARC working group concluded on the basis of all animal data thatthere was sufficient evidence for the carcinogenicity of
in experimental animals.
Similarly the animal studies with progesterone which were reviewed by IARC were also limited to subcutaneous and intramuscular injections and subcutaneous implantations of high doses; in rats and rabbits it was always given in combination with other sex hormones. Progesterone was administered together with carcinogenic substances such as dimethylbenzanthracene, methylcholanthrene, diethylstilboestrol or N-2-fluorenyl-diacetamide in the majority of the studies reviewed. From these studies IARC concluded: "There is limited evidence for the carcinogenicity of progesterone in experimental animals. In the absence of epidemiological data, no evaluation of the carcinogenicity of progesterone to humans can be made."
Similarly to the other two nature-identical steroid hormones, testosterone and its esters were tested in experimental animals by subcutaneous injection and/or implantation, and in rabbits by intramuscular injection. In this case IARC concluded: "There is sufficient evidence for the carcinogenicity of testosterone in experimental animals. In the absence of adequate data in humans, it is reasonable, for practical purposes, to regard testosterone as if it presented a carcinogenic risk to humans. The only related data in humans, although insufficient for an evaluation, concern the possible long-term effects of androgenic anabolic-steroids."
The contribution of information obtained from these animal studies was of little relevance for the quantitative assessment of the potential effects on human health of orally ingested, very low residual amounts of these substances having no effects on the physiological hormonal balance. In particular no additional human cancer risk could be deduced from these animal studies in the absence of other relevant information. With reference to studies conducted by other scientists, namely Dr. Liehr and Dr. Cavalieri regarding different types of DNA damage or genotoxicity established for oestrogen, Dr. Arnold agreed that with chemical methods all these things could be produced. The question was whether it occurred in living cells, at what concentrations, what enzymes were involved, what was the compartmentalization of these enzymes, etc. In his view, there were many gaps in the evidence. A scientist could not exclude that at the end somebody might show that oestrogens acted directly on the gene. For the moment, the evidence was not convincing and did not invalidate the basic conclusions of JECFA 1988.
In response to the follow-up questions, Dr. Arnold indicated that hormonal activity was a necessary requirement of hormonal carcinogenesis caused by the substances in dispute. He was not aware of any evidence which would invalidate the ADIs and MRLs established by JECFA/Codex.
WT/DS26/R/USA
Page 130
been learned about (i) the mechanisms whereby hormones triggered responses in target cells including the interplay or cross talk between receptor systems, (ii) the development of credible approaches for dose-response evaluation, (iii) the spectrum of toxic effects produced by hormonally-active agents, and (iv) the role of metabolic activation of hormones. There was a growing body of knowledge that some hormones could be converted to genotoxic metabolites. However, it was not clear whether these metabolites are involved in the carcinogenic process.
Dr. McLean responded that he did not believe that data developed after 1989 would justify a revision of the present use of hormones for growth promotion. For example, studies of in vitro receptor binding assays in the presence of pesticides had not produced evidence of concern. The extrapolation in the case of the hormones was arguably more reliable than with any other class of compounds because their mode of action and their metabolism was similar in both animals and humans. Non-human primate and human data was also available in many cases. In the case of trenbolone and zeranol, the usual safety factors for inter- and intra-species variation were still applied. The extrapolations from animal studies to humans was reliable in this case. The current ADI and MRL would only be invalidated if new data established a lower NOEL. In the case of trenbolone and zeranol, the NOEL used was conservative and therefore low. In any event, if any national body or group believed that it was time for re-evaluation of some of the hormones previously evaluated by JECFA it should approach the JECFA as had been done on a number of occasions with the number of compound of varying chemical classes.
The validity of extrapolations from animal studies to humans was the pre-eminent issue in contemporary toxicological hazard and risk assessment and was the basis of the hazard evaluation for drugs, pesticides, food additives and other contaminants worldwide. Given the reliance by both national regulatory authorities and international agencies on animal studies for predicting possible adverse effects in humans, it would seem de facto that the international scientific community had placed a great deal of faith in the validity of these extrapolations. At the same time, however, it was also clear that the validity of animal models in predicting adverse human effects was influenced by many factors including selection of the appropriate animal model, dose selection, duration and route of exposure and selection of appropriate risk models designed to estimate human risks resulting from the use of a substance under realistic and practical use conditions.
Recognizing the inherent limitations of any predictive model, Dr. Ritter indicated that the scientific community had elaborated a hazard and risk assessment paradigm that involved evaluation of a broad range of toxicological endpoints, selection of the most sensitive for determination of a NOEL, application of an additional uncertainty factor for derivation of an ADI and models for estimation of dietary exposure which invariably substantially exaggerated intake levels. The net effect of this series of overly conservative calculations was most likely very conservative estimates of risk, intended to compensate at least in part, for the limitations inherent in the reliability of animal to human
WT/DS26/R/USA
Page 131
extrapolations. Indeed, in its concluding commentary, the Steering Committee of the 1995 EC Scientific Conference noted that the nature of the calculations described above was such that adherence to the ADI provided a high degree of confidence that adverse effects would not become apparent in people. In other words the European Conference essentially reaffirmed and strengthened the earlier conclusions reached by JECFA that in accordance with operating procedures that were provided for the use of these substances, their use did not constitute a risk to consumers of food commodities produced with the hormones at question, at least for five of the six substances.
Dr. Ritter noted that Dr. Liehr's work and the work of many investigators was to produce an effect. This was to the most fundamental issue in pharmacology and in toxicology, the concept of dose response. Dr. Liehr was very interested in understanding the induction of cancer as a result of exposure to these hormones. Therefore, his protocol would necessarily be designed in such a way so as to produce the desired effect, in his case the tumour. The intent in the case of food residues, was obviously to avoid a dose level which might constitute a human risk. Therefore, the levels that were present as residues in food were thousands or hundreds of thousands or millions of times lower than they would be in an experiment which was specifically designed to induce a tumour. These two experiments were completely at cross purposes with each other; they were from their very initiation intended to produce entirely different results. To compare a protocol which had been designed to produce a tumour to a food residue the use practice of which was set up in such a way so as to minimize the presence of the residue, was a relatively meaningless comparison. Moreover, no one could reasonably assure that the presence of these residues albeit at low levels constitutes zero risk; in fact scientifically it would be impossible to ever test to a certainty. But the point was that if these low levels did constitute a risk, they constituted a risk of a magnitude which approached zero.
WT/DS26/R/USA
Page 132
The EC 1995 Scientific Conference concluded that the effects on animals in terms of disease incidence, performance, general mobility and meat producing efficiency were either unchanged or improved. The effects on animal welfare were negligible in terms of health, performance and effects on behaviour, although it was noted that there might be a transient increase in sexual activity for 2-10 days in some adult steers receiving oestradiol.
Dr. André indicated that previous results on the carcinogenesis of hormones had recently been confirmed. This was true for single hormones as well as for mixtures of hormones, administered in vivo and in vitro. Concerning the mechanismby which the hormones exerted their carcinogenic effects, the most important model was itself. It had been proven that
, which was involvedin the development of breast cancer, stimulated the development of malignant cells. Recent results showed that
enhanced genomic instability in malignant cells, inducing deletions or additions in DNA. It was reported that natural oestrogens (mainly oestradiol) induced cell transformation via over expression and synthesis of oncoproteins. Some new information about the potential role and mechanism of action of progesterone in breast cancer induction and about the biochemical aspects of pituitary oestrogen induced carcinogenesis were published; however, they did not demonstrate achieved molecular induction process.
Recent data on the mechanism of the carcinogenic effects of hormones led to the conclusion that hormones or their metabolites exerted a direct effect on tumour initiation by DNA damage; for oestrogen-dependent tumours, catechol oestrogens and their quinone forms would appear to beinvolved. Further, hormones stimulated the growth of tumours in tissues in which they had specific receptors. They had a complex carcinogenic and genotoxic effect. The hormonal effect stemmed from receptor binding, translocation to the nucleus and gene activation. Powerful oestrogens were strongly bounded and concentrated in the nucleus. If their hydroxylation and oxidation induced potential genotoxicity and carcinogenicity, it was not surprising that the more hormonally efficient they were, the more genetoxic and carcinogen they could be.
Dr. André concluded that hormones had to be considered as genotoxic compounds and for such carcinogens, no threshold level could be defined; no ADI and consequently no MRL could be established. This was the general opinion of toxicologists (Kuiper, EC 1995 Scientific Conference).
WT/DS26/R/USA
Page 133
This opinion had also been adopted by the United States Environment Protection Agency (EPA) for a carcinogenic non genotoxic dioxin, TCDD.
Steroid hormones primarily influence the expression of genetic information at the level of transcription by binding to, and activating, transcription factors. The activated transcription factors interact with the regulatory or promoter region of the genes. The hormone receptors are these transcription factors in the case of e.g., steroid and thyroid hormones. The complete DNA sequence of the human oestrogen receptor was cloned for the first time in 1986 using the breast cancer cell line MCF-7. In the mean time, the genomic genes for all three receptors (oestrogens (ER), progestins (PR) and androgens (AR)) had been cloned. Dr. Arnold characterized the biochemical hormone-receptor-interactions as follows: receptors were present in low concentrations; their binding sites were saturable at physiological concentrations. The binding sites exhibited high specificity and high binding affinity; binding was reversible. Specific receptor binding was the initial and necessary step in the sequence of events leading to a hormonal effect although it was not the only reaction the hormones could undergo. The effects terminated upon dissociation of the hormone-receptor complex.
The involvement of receptors in hormonal signal transduction was thus explained by the fact that in addition to the recognition domain for the hormone they possessed a second functional domain (about 70 amino acids long) through which they could bind to DNA. Steroid-hormone-regulated genes had at least two different regulatory elements, a "generic" promoter element and (a) hormone response element(s) (it was not clear how many response elements actually existed in naturally occurring genes). It bound the hormone-receptor complex more avidly than the surrounding DNA did. The hormone-receptor might also regulate several genes in the same cell and different genes in different target cells. Metabolites and related exogenous synthetic hormones might work on a different set of response elements. The details of how the interaction with DNA promoted transcription was still an area of investigation. However, the basic principle was well established.
Carcinogenesis could be operationally described as a multistage process involving at least three stages: initiation, promotion and progression. The intermediate stage of promotion did not appear to involve structural alteration of the genome of the cell but rather depended on altered gene expression. Both oestrogens and androgens had been shown to be effective promoters in their target cells as well as in liver. In contrast to initiation, promotion was a reversible stage depending on continued administration of the promoting agent. Besides tissuespecificity, thegeneral characteristics of hormonal carcinogenicity were: long induction periods and prolonged exposure at high levels with concomitant severe derangements in homeostasis. The dose-response relationship of promoting agents (as well as the receptor-hormone binding) exhibited sigmoidal curves with an observable threshold and a maximal effect.
Hormonal carcinogenesis in experimental animals had mainly been studied in rats, mice and hamsters. The promotor model of hormonal carcinogenesis had been developed through studies of a variety of tissues of these animals. The question under dispute was not whether these models were still valid but rather whether genotoxic mechanisms could additionally exist or would need to be considered in order to fully explain the carcinogenic potential of hormones. In a recent review ("Molecular Mechanisms of Oestrogen Carcinogenesis", J.D. Yager and J.G. Liehr, Annual Review of Pharmacology and Toxicology 1996, Vol. 36, pp. 203-32) it was confirmed that the hormonal effects of oestrogens were necessary but not sufficient to induce tumours. One metabolite of oestradiol (16-Hydroxyoestrone) was claimed to be capable of binding to DNA in vitro but this could not be
WT/DS26/R/USA
Page 134
confirmed in vivo (unpublished observations by one of the authors). It was furthermore postulated that the catechol-oestrogen formed by hydroxylation in position 4 of the A-ring of the molecule could be involved in the generation of reactive oxygen species through redox cycling. Catecholoestrogens were also able to bind to DNA in vitro, a finding which was not confirmed in vivo.
Dr. Arnold further responded that according to current scientific knowledge, threshold levels below which residues of the hormones had no adverse health effects did exist for a given individual. The individual threshold levels varied and this has to be considered when setting exposure limits for the entire population.
Dr. Arnold indicated that the ADI was derived from a "No Observed Effect Level" (NOEL) using appropriate safety factors. While it did not itself represent a threshold dose, it was indirectly related to a not precisely known threshold dose. Therefore, the ADI concept as defined by JECFA could not be applied in order to establish an exposure limit for direct-acting genotoxic carcinogens for which, at least theoretically, no threshold dose could exist. MRLs could be established, in principle, to regulate genotoxic carcinogens. Basically there existed two possibilities: if the residues were genotoxic carcinogens, alternative models, e.g., risk extrapolation, could be used to find an MRL; in the case of carbadox, JECFA was unable to allocate an ADI because the parent drug was considered to be a genotoxic carcinogen. In this case, however, the residues were demonstrably not carcinogenic under approved conditions of use. Thus, an MRL could be established on the basis of regulating an innocuous metabolite.
Dr. Arnold noted that MRLs primarily facilitated fair international trade and were not directly related to health effects. Only the total dietary intake of the residues of concern, taking into account all MRLs established for the same substance and all relevant commodities and their consumption by the population, could be compared with an exposure limit.
17and testosterone, there was very good evidence that physiological levels were carcinogenic. For example, breast cancer would strike one in nine women in the United States over their lifetime and there was compelling evidence that physiological levels of
were necessary for this effect. The evidence came from both animal experiments and human studies and was derived from a large base of peer-reviewed published information. In the case of oestrogen, early menarche and late menopause increased risk whereas ovariectomy was protective. Also, endometrial cancer was dramatically increased by oestrogen replacement therapy if unopposed by progesterone. In the case of breast cancer, consumption of beef with elevated levels of
could increase risk slightly if the carcinogenic risk was not already saturated by physiological levels of oestrogen. The question of threshold was irrelevant. It was known that existing levels were already carcinogenic.
However, Dr. Lucier noted that evaluation of carcinogenic risk in people consuming meat from growth-promoted animals was complex. For example, progesterone might protect against endometrial cancer. From a different perspective, one knew that diet was a critical determinant of breast cancer risk and that fat in the diet was a risk factor. Therefore, consumption of meat containing residues of growth-promoting agents might actually decrease risk because that meat contained less fat than meat from non-growth promoted animals. On balance, breast cancer and prostate cancer risk could possibly be decreased by eating meat from growth-promoted animals. Moreover, exogenous oestrogens protected against osteoporosis and cardiovascular disease.
Dr. Lucier indicated that simple categorization of agents as genotoxic or non-genotoxic, alone, had little value in determining if an ADI approach or a linear approach was more valid for risk assessment. This opinion was supported by analysis of 500 cancer bioassays conducted by the National Toxicology Programme (NTP).
WT/DS26/R/USA
Page 135
It was also widely recognized that toxic effects which might typically manifest at considerably lower levels then those required for receptor stimulation and attendant tumorigenic effects were a more appropriate basis for evaluation of potential adverse effects in humans following exposure to dietary residues. Notwithstanding, some authors (Liehr, 1996, unpublished; Adlercreutz, unpublished; Arnold, et al., 1996) had recently advanced alternative hypotheses regarding metabolism and mechanism of action of the gonadal hormones.
Historically, for chemicals which displayed carcinogenic properties in laboratory animals, it was assumed that a threshold dose did not exist. This view was due, in large part, to the fact that early studies focused on very potent carcinogens which manifestedtheir effect by direct covalent binding to DNA; agents of this type which acted through direct covalent binding to DNA had also referred to as genotoxic. The subsequent hypothesis argued that even a single molecule of a substance might cause a heritable change in the DNA structure which would ultimately lead to tumour formation. In practice, it was now recognized that such a scenario was not likely, and that in any case many carcinogens acted through a non-genotoxic mechanism and hence did not target cellular macromolecules such as DNA. This particular point had been extensively elaborated with trenbolone, which was the subject of elaborate studies which demonstrated the absence of covalent binding potential associated with this hormone (JECFA, 1988). It was now apparent that NOELs could be determined for many chemicals that may show some evidence of carcinogenicity under experimental conditions. Under conditions where the effect was clearly non-genotoxic it might be possible to derive NOELs, establish an ADI and propose an MRL. Chemicals which produced carcinogenicity through a clear genotoxic mechanism should be regulated as if a threshold dose could not be derived and hence a NOEL, ADI and MRL could not be proposed. In the case of non-genotoxic carcinogens, there was little scientific validity to the belief that these chemicals should be treated or regulated differently than chemicals which produced other adverse effects through non-genotoxic pathways.
WT/DS26/R/USA
Page 136
Dr. Ritter concluded that there was no compelling evidence to suggest that these compounds should be immediately re-evaluated, on the basis not only of the historical reviews which had been carried out by JECFA and other organizations, but also on much more recent reviews, including those published as a result of the European Conference on Growth Promotion in late 1995, early 1996, and indeed on the basis of the statements presented by Dr. Liehr, Dr. Metzler and others. Dr. Liehr's recent work and the work which others recently cited suggested that there were circumstances under which adverse effects could be demonstrated in association with a multitude of these compounds. However, scientists were compelled to look at the totality of the evidence as it was available. In Dr. Ritter's view, the totality of evidence, recognizing the information presented by Dr. Liehr and other scientists, as well as the historical information, suggested that the assessments that were provided in JECFA 1988 continued to assure a reasonable degree of safety to consumers of these commodities. This was not only his opinion, he stressed, but indeed also the consensus conclusion of the 1995 EC Scientific Conference.
WT/DS26/R/USA
Page 137
WT/DS26/R/USA
Page 138
Olaquindox was genotoxic in a number of tests. However the parent drug was extensively metabolised and was not present as a residue if "good practices" were applied. Ivermectin was extremely potent against parasites, but it was safely used in humans for the treatment of e.g. "river blindness" at much higher than residue levels. Carazolol was a hazardous substance for humans with chronic bronchitis which was a substantial part of the general population. In addition, carazolol was administered to the target animals by injection. Benzylpenicillin was absolutely harmless for the general population but could provoke allergic reactions in some sensitized people even at the level of the MRL. Avoparcin was not registered as feed additive on the American continent; its use in the EU had recently been withdrawn. Strict adherence to the identified "good practices" was an absolute requirement for the use of some of these substances. Nearly all of these substances left detectable residues in meat of treated animals under conditions of "good practice in the use of veterinary drugs". At the oral hearing Dr. Arnold added thatthere is a commercially available alternative for carbadox mainly oxytetracycline.
Dr. McLean further noted that olaquindox was considered to be genotoxic and no toxicity studies were available on its metabolites, therefore JECFA was unable to determine an ADI or MRL. Benzylpenicillin was widely used as an antimicrobial agent in animals and humans. Allergenic potential was the major toxic effect and it was recommended that the total intake be kept below 30 micrograms per day for a human. Benzylpenicillin showed no carcinogenic potential. While carazolol showed no genotoxic or carcinogenic potential, it was a potent beta-adrenoceptor blocking agent. MRLs had been established for pig tissues. However, if carazolol was used in pigs to prevent stress during transport to slaughter, there was concern that residues at the site of injection could result in consumers receiving a pharmacologically active dose of the drug. The toxicity and residue data base for ivermectin and related compounds was extensive and the drug was also used in humans. The compound did not show genotoxic or carcinogenic potential and an ADI and an MRL had been established.
WT/DS26/R/USA
Page 139
The organophosphorus compounds (OP) were described by Dr. McLean as powerful neurotoxins, and most of the untoward effects were related to the direct exposure of operators and others. There was a new syndrome of organophosphate - induced delayed polyneuropathy - which was of concern when there was acute exposure.194 The OP had been associated with carcinogenicity and there were reports of acute poisoning associated with the consumption of treated food. Very low levels of the parent compound and/or metabolites might remain, although they could not be detected by sensitive analytical techniques. Dr. McLean did not comment on monensin or avoparcin because of the limited amount of registration data available which was in public domain.
The establishment of an ADI and recommendation of an MRL reflected scientific confidence in the expressed residue levels being essentially devoid of risk to humans exposed to dietary residues on a lifetime basis. Therefore, neither carbadox nor the growth promoting hormones should pose potential adverse human health effects as a result of exposure to dietary residues at or below the MRLs specified by JECFA. Notwithstanding, it was noteworthy that carbadox and at least one of its metabolites was both carcinogenic and mutagenic, while the hormones were not considered carcinogenic or mutagenic at biologically relevant doses. In addition, reliable residue monitoring and risk estimation of carbadox residues was somewhat complicated by the presence of bound residues, while residues of the hormones either fell within normal physiological range or could be reliably and easily measured.
194M. Lotti, Toxicology 21, (1992) p.465.
WT/DS26/R/USA
Page 140
Residues found in milk could be higher, equal or lower, depending on which tissues were compared and depending on many other factors (dose, route, age, sex, withdrawal time etc.)
Dr. André responded that an appropriate assessment of the potential adverse effects on human health stemming from the use of the hormones in dispute "should be more rigorous than common veterinary drugs because of the duration of their use and the very limited, if any, health benefits to the target species".195 Scientists should consider all data available at the present time in dealing with the adverse effects of each compound (see response to question 8), their metabolites and their mechanisms of action, including their action at the level of exposure. Species-specific metabolite patterns should be compared, using (as far as possible) a combination of in vivo and in vitro experiments. Studies should include long-term feeding trials to address life time exposure. All studies should be performed with combinations of the hormones in dispute prior to registration, to obtain valuable data. The effects of the treatment on the kinetics of other common drugs had to be assayed. Epidemiological studies in humans concerning cancer incidence and other hormone related diseases had to be initiated.
195Bridges, 1995 EC Scientific Conference Proceedings, p.250.
WT/DS26/R/USA
Page 141
Of special concern were potential risks for cancer and non-cancer effects (reproduction, development, cardiovascular disease, etc.). Additionally, information on effects in sensitive sub-population should be considered. Sensitivity could be conferred by genetic predisposition (presence or absence of cancer susceptibility genes), age (fetus, children or the aged), gender, existing disease status, nutrition, and co-exposure to other chemicals. These factors needed to be considered, although it was unlikely that information on them would ever be complete for any given substance.
In addition, Dr. Ritter noted that it might also be relevant to establish if the nature of the residue was distinguishable from normal endogenous levels - this was particularly relevant for the use of the natural gonadal hormones which resulted in residue levels typically in the range of untreated animals; whether appropriate and reliable methods existed for the estimation and monitoring of residues of the hormones following their use as growth promoting substances; and knowledge of the uncertainty inherent in the hazard and risk assessment paradigm.
WT/DS26/R/USA
Page 142
The anabolic effect of the hormones in question was proportional to the dosage, with a maximum effect. Farmers tried to obtain larger effects through administration of more implants than recommended, although it was not evident that a double dose, even at one time, gave a better response. In some cases, a parenteral administration (e.g. intramuscular injection) produced a more rapid effect than implants, and farmers tried by this mean to obtain faster effects. Disincentives for farmers to use larger quantities would be a decrease of the benefit/cost ratio, but implants were usually cheap. In countries where they were in use for growth promotion purposes, penalties due to higher residue levels detected in meat products compared to the MRLs might also have dissuasive effects on farmers. When five out of the six hormones in dispute were allowed in France, the experience was that farmers tried to give a second dose later. They did not respect the withdrawal period and they injected another dose at half way through the theoretical withdrawal period. Clearly they saw benefit to do this because the effect was longer at the time. Dr. André further reported that during the four to five years when these hormones were allowed as growth promotants in France, the misuse of other hormones continued and the black market was also present.
Dr. Arnold noted that there were considerable safety margins built into the established MRLs for the synthetic residues. As long as "good practices" were observed, the calculated theoretical maximum daily intakes of residues would range around 5 per cent of the ADI for both zeranol and trenbolone acetate. It had been shown that the implantation of approximately five times the recommended dose of zeranol would have almost no effect on maximum residue levels in the muscle and liver of steers slaughtered only five days after implantation. When a total dose more than 100-times higher than the recommended dose was intravenously administered as 6 split doses over three days to steers and the animals were slaughtered on the third day after the last dose, consumption of the meat of these animals would still not have caused an above-ADI intake of residues of zeranol.196 The studies available to the 32nd and 34th JECFA showed that residues of trenbolone were highest if heifers were implanted shortly (15 days) before slaughter. With respect to the calculated theoretical maximum intakes, there was no difference whether this implant was the first or the second which the animal had received. From these data, it seemed unlikely that even the frequent use of higher doses, or repeat doses or slaughtering the animals at earlier than recommended times after implantation, could result in residue intakes in excess of the ADI by the consumer. For the three nature-identical hormones, the safety margins (theoretical residue intakes compared with endogenous production rates of the most sensitive sub-population) were much higher than for the synthetic substances.
Dr. Arnold had no information with regard to melengestrol acetate, nor regarding incentives or disincentives for farmers to use combinations or illegal "cocktails" of these hormones. Although these pellets and another devices had been developed to give optimum results when the dose was respected, this did not necessarily prevent some farmers using more implants. But this did not necessarily cause higher residue levels in the carcass. On the other side, it was clear that if twice the amount was injected directly, all levels increased in plasma and in tissues, not necessarily in a linear way, but significantly. That was the difference between a slow release device (like ear implant) and a direct injection. On the other hand, the long term release of high doses might influencethe pattern of hormone excretion in the animal's body and this was the intention of using such compounds.
196Food and Nutrition Paper 41, pp.44-45, FAO (1988).
WT/DS26/R/USA
Page 143
and there was no advantage in exceeding this dose. The dose suggested by the sponsor gave the optimum response and therefore, within limits, there was no need or benefit in administering more than was suggested. The regulatory authorities also examined the effects of overdosing during the registration process. Most regulatory authorities, including JECFA, required that data for residues setting did take into account dose rates that were in excess of what was normally used (generally at least twice and sometimes more) so that the effect of overdosing or variations in uptake could be seen. In countries where the use of these compounds was permitted, there were good educational campaigns for farmers as to the correct use and the reasons why the prescribed dose should not be exceeded, and the penalties that existed if one did. The results from residues surveys showed that, by and large, there was no exceeding of the MRL. In Australia, trenbolone and zeranol had been targeted in residue surveys very specifically looking for violations, and to all intents and purposes violations did not occur. Even if there was misuse, and that was difficult to prove, there were still no residues that exceeded the MRL. A similar situation existed with the naturally occurring hormones, where it was not possible to determine whether or not a violation had occurred because the levels that were found in the carcass fall within the normal range.
Dr. McLean observed, however, that in countries where the use was not controlled and there was no farmer education campaign and where it was difficult to apply a penalty, then the MRL was significantly exceeded. One of the important factors of legalizing these compounds was to conduct an education campaign and to put in place monitoring procedures to ensure that the MRL is not exceeded. What was important with the maximum residue limit is to understand that it was a legal limit and not a health limit. In other words, exceeding of the MRL did not represent a hazard to health but rather a limit at which the authorities took action. However, to exceed the MRL would not be considered to be good practice.
Dr. Ritter noted that it was very difficult, if not impossible, to estimate potential hazards which might be associated with the administration of the hormones contrary to good practice, as the magnitude of the potential hazard would be related to the nature, extent, frequency and magnitude of the inappropriate administration. It had been reported that in the case of the anabolic hormones and using commercially available implants, repetitive implantation had only little influence on the residue profile.197 Similarly, full and half-dosages led to similar hormone levels in edible tissues which, in the case of the endogenous hormones, were within the physiological range. Further, unlawful and improper use of oestradiol might result in residue levels some 300-fold in excess of established tolerance limits, and yet it was virtually impossible to visualize any hazard to humans ingesting meat from animals treated with zeranol.198 Similarly, when steer calves were implanted with zeranol in accordance with recommended procedures, the margin of safety for consumption of edible products was greater than 25,000 and 150,000 for liver and muscle, respectively. While this work did not specifically address the issue of potential abuse, the large margins of safety postulated by the authors suggested that even under limited circumstances of abuse, it was unlikely that consumers would be exposed to unacceptable risks.199 In countries where use of these hormones, the six hormones in particular, had been permitted for a very extended period of time, most notably Canada and the United States, monitoring and compliance programmes which had been conducted for many years consistently demonstrated that residue levels were entirely within recommended limits and that instancesof violative residues, that was residues which would indicate abuses taking place, had almost never been reported. It seemed that at least in those jurisdictions where use was lawful, the practicality of abuse had never become a reality. There were few and isolated examples of violative residues in those countries where use had been permitted.
197Hoffman and Evers (1986).
198Truhaut et al. (1985).
199Sundlof and Stickland (1986).
WT/DS26/R/USA
Page 144
Dr. André replied that specific consequences stemming from the use of legally marketed combinations were not known. If they existed potential hazards could be due to synergistic and/or additive effects of the residues. This answer was also valid for illegal "cocktails" of the six hormones in which more than two components could be present. However there was another important potential hazard from the systematic use of these combinations: their use could result in modifications of kinetic parameters of other drugs. For example, trenbolone and testosterone were shown to dramatically decrease the elimination rate of sulfamethazine, trimethoprim and antipyrine in goats, with opposite experimental results in rats. As a consequence, residue levels could be higher than the MRLs at slaughter, even when legal withdrawal periods were respected.200 Recently, the increase of the level of beta agonist residues in liver owing to concomitant oestradiol treatment had been demonstrated in calves.201
Dr. André was aware of only one combination (TBA + oestradiol) having been assayed by the JECFA in the context of residue determination in steers. No other combinations had been tested whereas individual components had been tested for toxicological effects. Evidence of synergistic effects of combinations of hormones existed and were well studied mechanisms. For example, concerning endocrinological therapy in humans, combinations of oestrogen and progestin hormones were used; oestrogens induced the synthesis of specific receptors for progestins, which could then exert their specific action. Legally marketed combinations had been developed on the basis of synergistic effects of their components on growth promotion.
Incentives for farmers to use combinations (legally marketed) or illegal "cocktails" of these hormones were frequently reported by official control bodies, since farmers tended to think that the more hormones used, the better the anabolic effect. Moreover, interactions between these hormones and other endogenous biochemical parameters (e.g. corticoids, IGF, etc.) were of great importance for human health and could be documented.
Dr. Arnold reported that the pharmacodynamic and toxic effects of practically all possible combinations of oestrogens, progestins and androgens had been studied in whole animals, organs, tissue and cell cultures, and other in vitro systems under a great variety of conditions. There was ample evidence of both synergistic and antagonistic effects. Such effects physiologically played important
200Van Miert (1988).
201Kuiper, 1995 EC Scientific Conference proceedings, p.377.
WT/DS26/R/USA
Page 145
roles in endocrinology and metabolism. The current understanding of these phenomena clearly indicated that concerns over human health hazards arising from combinations of minute amounts of residues in meat were not justified.
The flourishing grey market in other veterinary medicines within, for example, Germany showed that farmers tried bypassing official distribution channels when purchasing otherwise licensed high quality products, mainly in order to save costs and prevent having to pay the veterinarian. To what extent "good practices" in the use were otherwise followed remained subject to speculation. It seemed, however, that not all rules were violated at one time. The situation with prohibited substances (e.g. hormones, chloramphenicol) and with abused substances (e.g., clenbuterol) was completely different. There was clearly a black-market where the "dirty products" prevailed. The availability of reasonably priced legal alternative products could be an incentive for many farmers to legalise their practices. Whether this would eliminate illegal cocktails now that the black-market has been well established, remained entirely speculative.
In general, safety (toxicology) evaluations of drugs, pesticides and other food contaminants were carried out on single compounds only, rather than as combinations. There were several reasons for this approach which included prior pharmacological and biochemical knowledge that toxicity associated with combinations was not likely to produce effects greater than the sum of the individual components. On a more practical level, it was difficult to contemplate the broad range of possible combinations which would be the subject of testing. Finally, as different commercial interests might be involved in the production and sale of various growth promoting hormones, proprietary interests would make it unlikely that combinations could be evaluated toxicologically. In some cases, the efficacy of hormone combinations had been evaluated. In the case of combinations of trenbolone with zeranol or oestradiol, for example, it was found that (i) the combinations were equivalent in young bulls and steers, (ii) anabolics containing oestradiol were more effective in veal calves, and (iii) the composition
WT/DS26/R/USA
Page 146
and quality of the meat was not modified by the use of combinations when compared to single use.202 As with any drug or unlawful drug combination, farmers sometimes were under the incorrect impression that combinations which had not been evaluated might provide significantly enhanced efficacy over those which had been approved for use. Such improper use of illegal cocktails could affect withdrawal times and residue limits and might, in some jurisdictions, result in compliance and enforcement as disincentives to the unlawful use of non-approved cocktails. Dr. Ritter noted that it must also be recognized that the quest for ever improved yield could lead some producers to the use of unapproved drug combinations.
202Bouffault and Willemart (1983).
WT/DS26/R/USA
Page 147
WT/DS26/R/USA
Page 148
withdrawal period involved. The implication for human health would in turn be related to these two variables.
Dr. Ritter indicated that illegal injection of oestradiol preparation could produce injection site residues which largely exceeded tolerances. In the case of feed additives, improper use might imply use of approved drugs in non-approved species, improper dose, improper combinations, improper duration of use and improper withdrawal periods. However, the provision of human safety was assured through the establishment of maximum residue levels, levels considered to be safe if consumed by humans in the diet for an entire lifespan. In the event that any of the improper use conditions noted above occurred, and to the extent that such improper use would impact on final residue levels, human safety was assured through appropriate monitoring of the food supply for compliance with approved MRLs.
WT/DS26/R/USA
Page 149
In addition, Dr. André stressed that the future of the animals treated either for therapeutic or zootechnical purposes or for growth promotion purposes was very different. In the first case the animals usually remained on the farm for months or years after treatment and the problem of residues was not of concern. The only comparable situation was for oestradiol benzoate, for which, in case of unsuccessful treatment in a cow, a withdrawal period of two months had to be observed. In the second case, the animals were slaughtered weeks or months after the treatment. Furthermore, the delivery conditions differed. In the EC, the therapeutic use of these hormones was particularly regulated. Finally, Dr. André noted that none of these therapeutic/zootechnical uses would change the level of residues of hormones in meat on large scale as occurred from growth promotion use. No scientist had said that it was a bad thing to use these hormones for therapeutical use; their use on a large scale for growth promotion was not the same thing.
Dr. Arnold replied that the total dose contained in an implant of was much higher than the dose injected for therapeutic purposes because the implant contained the quantity needed over a long time (e.g., 200 days). On a daily basis, however, the dose injected was much higher, and therefore, tissue residues (excluding the injection site) after treatment were much higher if compared with the residues measured at any time after implantation. No representative data were available for progesterone or testosterone that would permit comparisons. With the exception of oestrus synchronisation, individual identified animals were treated.
WT/DS26/R/USA
Page 150
veterinarian. Notwithstanding, at least in the case of the natural hormones, the approach adopted by the Community for therapeutic/zootechnicaluses provided for residue levels which fell within the range of levels normally found in untreated animals. While the basis for use of these hormones for growth promotion purposes was fundamentally different, the resulting residues appeared, nevertheless, to also fall well within the normal physiological range.
WT/DS26/R/USA
Page 151
Dr. Arnold responded that melengestrol acetate was only used as a feed additive. Zeranol and trenbolone acetate were used as implants only. , progesterone and testosterone were administered by different routes including injection and implantation. It was difficult to make comparisons about potential effects due to means of administration because there were too many variables in the equation (chemical nature, dose, route, formulation, target species). If all uses were carried out in compliance with the established conditions of use, the remaining residues could be regarded as safe. The safety margins might vary from one condition to the other and the consequences of non-compliance might also be different.
WT/DS26/R/USA
Page 152
from country to country (within certain limits). Only few methods had been internationally collaboratively studied according to internationally harmonised protocols. Only few methods were fully validated in accordance with the criteria first defined by the Codex Alimentarius and later adopted by the EEC in Directive 85/592.
Dr. McLean explained that there were two basic types of assays used. Screening methods were used to detect drugs for metabolites at the level of interest. These had high sample throughput and aimed to avoid false negative results. Confirmatory methods provided unequivocal identification of the drug and/or metabolites at the level of interest. The cost of analysis was considerably greater than for screening methods. These methods were constantly being reappraised.
Dr. Ritter responded that methodology for MGA had been reported203 with a limit of detection of 5 ppb. and a limit of quantitation of 10 ppb., both in fat, the target tissue for this residue. LC methodology for the trenbolone acetate had been reported204 with a limit of detection of 2 ppb in muscle and liver for -trenbolone, and a limit of quantitation of 2 ppb in muscle and 4 ppb in liver for
trenbolone in liver and kidney. GC/MS methodology for zeranol had been reported by Covey et al (1988) with a limit of detection of 0.1 ppb. and a limit of quantitation of 0.2 ppb. in liver or muscle. The Joint FAO/WHO Expert Committee on Food Additives (WHO, 1988) reported that radioimmunoassays could detect free and conjugated
and
-trenbolone at levels of 75ng/kg in tissues. Radioimmunoassays were generally regarded as appropriate screening methodsonly, while conventional analytical procedures such as gas chromatography/liquid chromatography/mass spectrometry are generally regarded as suitable for confirmatory analysis.
203Anderson and Fesser (1996).
204Shih-Hsien Hsu et al . (1988).
WT/DS26/R/USA
Page 153
The scientific concept behind sampling, however, was not clear and was subject to different interpretations. It seemingly mixed elements of a statistically based monitoring system with other concepts. The reporting format was only semi-quantitative with respect to the reported residue contents of positive samples. In consequence, the results could not be used to compare the situation in the EC member States or, for example, to conduct a crude assessment of exposure of European consumers to residues of anabolic substances.
Dr. Arnold further explained that a network of reference laboratories had recently been set up with an outstanding European reference laboratory on top. However, the moderate financial contribution to the European reference laboratories from the Commission was given only on an annual basis, leading to difficulties in ensuring experienced permanent staff in these institutes. The quality of the national reference laboratories co-ordinating the activities of the routine laboratories carrying out the daily work could not be judged. Although it was not known how many routine laboratories were accredited nationally, it was obvious that systems of quality management had not yet been established everywhere in the Community.
The endogenous production in the animals of , testosterone and progesterone causes difficulties in the control of their use. MRLs have not been set, and cannot be set, for the therapeutic and zootechnical uses, because the distribution of the three natural hormones in the three categories of animals (untreated animals, animals illegally treated for growth promotion, and animals legally treated for the permitted purposes) largely overlap. Because withdrawal times could not be enforced on a broad scale and distribution chains were insufficiently controlled, only limited attempts
requiring a lot of man-power and other resources - could be made to enforce compliance. If there was a suspicion, of course, the animal involved can be traced back and the producer questioned about, e.g., the prescription of the drug, the veterinarian involved etc. The main difficulty was to identify suspect animals in order to start further investigations and measures. Only the discovery of an injection site or an implant containing illegal substances could provide proof of a treatment. Decision limits for the levels of had provisionally been established in order to identify suspect animals.
Enforcement was difficult if people did not want to comply with the legal rules - which otherwise would perfectly guarantee consumer protection. It appeared that illegal anabolic hormones were readily available. The expectation of extra profits was an incentive for the continued the use of these substances including beta agonists, not only for farmers but also for participants in the illegal network of distribution.
WT/DS26/R/USA
Page 154
cent violation rate 95 per cent of the time. The number of samples that were required to do that varied from country to country and from commodity to commodity becauseit was a function of the use practices for the substances involved.
Control of veterinary drugs used in food production by farmers was generally achieved by national authorities through imposition of regulations requiring that drugs of this type be dispensed only on the written authority of a properly qualified veterinarian and in compliance with relevant use information established by appropriate national regulatory authorities. Compliance was further assured through rigorous monitoring and, where appropriate, through enforcement activity. Cost implications generally related to the increased cost associated with the requirement that such drugs only be
WT/DS26/R/USA
Page 155
administered under the authority of a properly qualified veterinarian. Costs would also be proportional to the rigour with which a monitoring and compliance programme was designed and implemented.
For products for which particular conditions of use were indicated, the control was usually in the hands of practitioners. New recording systems were under development (see response to question 1). Controls by official bodies (e.g. Veterinary Inspection Services) were effective in some countries. In other countries regulations were changing in order to apply EC Directive 96/23, allowing for more controls to be performed on the farm itself. This was particularly true for the hormones in dispute, as for other banned compounds (chloramphenicol, rhonidazol, etc.).
WT/DS26/R/USA
Page 156
In the European Communities, no new active principle had been placed on the market since 1992 unless a Community MRL had been established under the provisions of regulation 2377/90. New animal drug applications had to be accompanied by a proposed regulatory method suitable to enforce tolerances/MRLs. The main performance characteristics (accuracy, precision, limit of detection, limit of quantification, specificity) and the ruggedness, practicability, applicability, susceptibility to interference of the methods were reviewed together with the other documentation relating to the quality, safety and efficacy of the drug.
The review of the old substances, however, was still ongoing. The evaluation of the old substances had also begun in the pesticides sector. The majority of these substances had a long history of uses. Although they had never been evaluated according to contemporary safety requirements, they could probably be regarded as safe in the light of the experience with these substances. Should new evidence indicate that human health hazards could arise from the permitted uses of these substances, the competent authorities would withdraw the respective products until the problems were clarified. There were some examples, however, where highly efficacious substances with a long history of use had to be withdrawn from the market of veterinary drugs for food animals when more contemporary standards were applied in their re-evaluation (e.g., chloramphenicol, nitrofuranes, nitroimidazoles).
When old substances were reviewed in the Community to establish MRLs, the Committee for Veterinary Medicinal Products always considered the available analytical methodology. An MRL was not proposed if no analytical method was available. The review of the methods and international harmonisation of methods suitable for the enforcement of Codex Standards was conducted by the FAO/WHO Codex Committee for Residues of Veterinary Drugs in Foods.
The more difficult problem was to monitor so many substances in all foods of animal origin. The chemical properties of these substances were so different that it was impossible to include all residues in a few multi-residue methods. There were substances requiring special equipment for their detection. It was not practically feasible to monitor all potential residues in all food commodities every year. The most reasonable approach was to categorise compounds according to potential hazard and the likely exposure of consumers to the residues. Some compounds required inclusion in residue-monitoring programmes every year; others might be selected for longer cycles. For a few compounds, perhaps no monitoring was necessary. Directive 86/469/EEC covered veterinary drug residues in meat. Starting in 1988, EC member States and countries exporting meat to the Community had to submit residue control plans also for these substances. EC member States had more flexibility to adapt the plans to the actual needs (e.g., areas of risk). The new legislation adopted in 1996 enlarged the scope to cover other foods than meat. An increasing number of screening tests were now available which were based,
e.g. on microbiologicalinhibition or immunochemistry. These methods were suitable to identifysuspect "positives" (samples violating MRLs) and exhibit a low rate of "false-negative" results. However, an important aspect of the control of "good practices" was the control of distribution; distribution of veterinary drugs was not harmonised in the Community.
WT/DS26/R/USA
Page 157
Control of use was generally accommodated through appropriate control of sale (restriction of use generally associated with “prescription” drugs) and rigorous monitoring and compliance programmes, further supported by appropriate enforcement action. In regard to analytical methodology available for detection and monitoring purposes, both screening and confirmatory methods were utilized. Availability of appropriate and reliable analytical methodology for other production aids, such as veterinary drugs, varied somewhat with the specific agent and might range from difficulties of bound residues with carbadox and nitrofuran, to the special and important considerations which were necessary when assessing the microbiological risk due to residues of antimicrobial drugs in food.
WT/DS26/R/USA
Page 158
since physiological levels of the natural hormone exceeded any threshold that might exist (see response to question 8). The synthetic hormones could exert toxic effects not only related to their hormone action but also because of other structural/functional properties.
WT/DS26/R/USA
Page 159
For example, a recent outbreak of Escherichia coli food poisoning in Scotland made 400 peopleseriously ill and killed 17, while a similar outbreak in Japan infected 10,000 and killed 11 (Coghlan, A (1997) New Scientist 153 (2066)7).
In the case of the synthetic hormones utilized in growth promotion, at least two had been subjected to international review and safe residue levels had been recommended. The purpose of such a labelling programme would be unclear given that an international review had already concluded that residues at or below proposed levels did not constitute a risk to consumers, even under the exaggerated calculations and assumptions utilized in the development of MRLs.
Dr. Arnold replied that no such insurance can be given. The meat would always contain residues even if withdrawal times were observed. The concept of "no residue" had been largely abandoned some 20 years ago when it became evident that "no residue" was a function of the limit of detection of the analytical method. All permitted substances, including the hormones, were regulated on the basis of an "acceptable residue" in the EEC. The "no residue" concept still applied to banned substances, but this was trivial. In the case of hormones it was acceptable if the sum of endogenous plus exogenous hormones was at or within physiological limits.
The response of Dr. Ritter to this question is contained in his response to question 20 above.
INTERIM REVIEW
On 21 May 1997, the European Communities and the United States requested the Panel to review, in accordance with Article 15.2 of the DSU, the interim report that had been issued to the parties on 7 May 1997. The European Communities also requested the Panel to hold a further meeting with the parties to discuss the points raised in its written comments and other points which they would develop during that meeting.
We decided to hold concurrent interim review meetings with the parties for both this dispute and the parallel panel requested by Canada. This decision was, inter alia, based on the similarities of both cases and the fact that the interim reports in both cases only differ in the description of the arguments of the parties, whereas the sections dealing with the scientific experts and the legal findings
WT/DS26/R/USA
Page 160
in the two cases are almost identical. In light of that decision, we also decided, on Canada's request and after consultations with the parties involved, to make a copy of the sections of our interim report dealing with our consultation with scientific experts and our findings, together with the comments submitted in that regard by the European Communities and the United States, available to Canada. By letter of 3 June 1997, the European Communities objected to both decisions, arguing that they affected due process and its rights of defense and, consequently, its rights and obligations under the WTO Agreement. It made, however, no specific claims of prejudice. Since we could not see how the European Communities could be prejudiced by these decisions, we rejected this objection.
In a letter dated 28 May 1997, the United States requested that, during the meeting with the parties, the Panel direct the European Communities not to raise any comments that were not provided in the EC's letter of 21 May 1997. It further requested that the Panel disregard any such comments, including the comments that the European Communities has said in part C of its letter that it will be submitting with respect to some unspecified number of paragraphs in excess of 49 paragraphs of section 7 of the interim report. The United States argued that to do otherwise would create serious prejudice to the United States, which could hardly be expected to be prepared to rebut such comments.
Article 15.2 of the DSU provides the following:
"... Within a period of time set by the panel, a party may submit a written request for the panel to review precise aspects of the interim report prior to circulation of the final report to the Members. At the request of a party, the panel shall hold a further meeting with the parties on the issues identified in the written comments..." (emphasis added).
It appeared to us that the European Communities, by only enumerating the numbers of the paragraphs of the findings section of our interim report in relation to which it has concerns and not stating in its written comments the precise aspects it wishes the Panel to review, did not respect the wording of Article 15.2 of the DSU. We considered, however, that the main object and purpose of Article 15.2 is to make sure that the other party is aware of the issues which will be raised at the interim review meeting in order to allow it to prepare its rebuttal. In light of the fact that the EC's written request for review of the findings section only related to factual aspects, i.e., the correct reflection of the EC arguments and the names of and references to individual scientific experts, we considered that the United States would not be unduly prejudiced by allowing the European Communities to present these factual points at the interim review meeting. We, therefore, decided that the European Communities could raise these points on the conditions that it would limit itself to factual issues and to the paragraph numbers it had enumerated in its written comments. We note that this decision corresponds to what the United States requested in its letter of 28 May 1997.
The Panel met with the parties on 4 June 1997 in order to hear their arguments concerning the interim report. We carefully reviewed the arguments presented by the European Communities and the United States and the responses offered by both sides.
The European Communities requested us to find that Articles 15.2 and 15.3 of the DSU and the general principle of due process prevent the Panel from modifying aspects of the interim report on which the parties did not submit comments. However, at the end of the interim review meeting the European Communities appeared to modify this request by asking the Panel to review its findings in light of the factual comments made. We considered that no provision in the DSU limits us to only modify those paragraphs commented upon by the parties. Article 15.3 of the DSU only provides that "[t]he findings of the final panel report shall include a discussion of the arguments made at the interim review stage...".
WT/DS26/R/USA
Page 161
The European Communities made two types of comments on the findings section of the interim report. The first concerned 20 paragraphs in which the European Communities is stated to have argued or agreed to something which, according to the European Communities, does not or not completely reflect the EC's position taken during the proceedings. These comments related to the following paragraphs: 8.105, 8.109, 8.111, 8.112, 8.115, 8.131, 8.175, 8.176, 8.187, 8.190, 8.193, 8.204, 8.205, 8.213, 8.220, 8.224, 8.228, 8.232, 8.243 and 8.274. Since these proposed changes concerned the representation of the EC's own legal or factual arguments, we accepted most of them.
The second type of comments made by the European Communities related to paragraphs where the phrase "the scientific experts advising the Panel" is used without, according to the European Communities, citing the names of the scientists nor the place where they have made the statements the Panel is invoking. It also argued that frequently the reference provided does not reflect the views of all the scientists. The European Communities also requested us to review the accuracy of the factual information contained in several paragraphs. We carefully considered all the factual comments thus made and, where we agreed with them, modified those paragraphs accordingly.
The European Communities also requested us to include in the final report the procedural decisions taken by the Panel during the course of its work. We added these decisions in a new Section B on organizational issues.
The United States commented on the section of our findings related to the scope of the measures in dispute. In response, we expanded our reasoning in these paragraphs. It also requested the Panel to make additional findings on Articles 2.2, 2.3 and 5.6 of the SPS Agreement. For the reasons set out in the report, we did not consider it necessary to address these claims in order to resolve the matter in issue in this dispute.205 The United States also requested the deletion of paragraphs 8.72 to 8.77. Since we considered these paragraphs to be vital to our consideration of this case, we rejected this request. The United States further submitted corrections of its own arguments and factual elements and suggested several drafting changes, most of which have been taken into account in our final report. It argued, for example, that our findings under Article 5.5 of the SPS Agreement with respect to the endogenous production of the hormones, the hormones used for therapeutic or zootechnical purposes and the hormones present in other foods, inaccurately ascribe to its arguments under Article 5.5 which it made under Articles 2 and 5.6. This remark is reflected in paragraph 8.171. Nevertheless, we considered it appropriate, inter alia, for the overall structure of our report, to address these situations under Article 5.5.
Both parties also suggested further changes or additions in respect of the interim report's descriptive sections which we took into account in re-examining that part of the report. In this context, the European Communities requested us to append the transcripts of the joint meeting with the experts advising the Panel to the descriptive part of the report, arguing that many important statements made by the experts in these meetings are not reflected in Section 6 of the interim report. In order to increase the transparency of our work and to take into account most of the comments made by the European Communities on the descriptive part of the interim report, we decided to annex the transcripts of the joint meeting with the experts of 17-18 February 1997 to our final report.
205See Appellate Body Report on "United States - Measures Affecting Imports of Woven Wool Shirts and Blouses from India", adopted on 23 May 1997, WT/DS33/AB/R, p.19.
WT/DS26/R/USA
Page 162
FINDINGS
This dispute arises essentially from the following facts. In 1981 the Council of the European Communities ("EC Council") adopted Directive 81/602/EEC206, inter alia, requiring the EC member States of the European Communities to prohibit the administration to farm animals of substances having a thyrostatic, oestrogenic, androgenic or gestagenic action. Directive 81/602/EEC further provided that pending adoption of a decision of the EC Council on the administration to farm animals for growth promotion purposes207 of , testosterone, progesterone, zeranol and trenbolone208 EC member States could continue to apply the national regulations in force concerning those substances.209 In 1988 the EC Council adopted Directive 88/146/EEC210 which brought the administration to farm animals for growth promotion purposes of these five hormones within the general prohibition imposed by Directive 81/602/EEC. The 1988 Directive also required the prohibition of importation from third countries of animals and of meat from animals to which substances with thyrostatic, oestrogenic, androgenic or gestagenic action have been administered. 211 Two exceptions to this general ban are provided in Directive 88/299/EEC212: (i) the administration for therapeutic treatment213 of
, testosterone, progesterone and some of their derivatives; and (ii) the administration for zootechnical treatment214 of substances having an oestrogenic, androgenic or gestagenic action which are authorized in accordance with EC Directives on veterinary medicinal products.215 On 29 April 1996, the EC Council adopted Directive 96/22/EC216 (repealing and replacing Directives 81/602/EEC, 88/146/EEC and 88/299/EEC) which confirms and extends the above-mentioned prohibitions. This 1996 Directive will enter into force on 1 July 1997.217
The United States claims that the European Communities, by banning the importation of meat and meat products from animals to which any of six specific hormones have been administered for purposes of promoting the growth of the animals, has acted inconsistently with the Agreement on the Application of Sanitary and Phytosanitary Measures ("SPS Agreement"), in particular Articles 2, 3
206EC Official Journal, L 222, 7 August 1981, p.32.
207Article 5 of Directive 81/602/EEC uses the term "for fattening purposes". However, during the Panel proceedings and in this report the term "for growth promotion purposes" is used.
208The five hormones are the subject of this dispute. A sixth hormone in dispute, melengestrol acetate or MGA, falls under the general prohibition of Directive 81/602/EEC and is addressed in paragraph 8.4.
209See para. 2.2.
210EC Official Journal, L 70, 16 March 1988, p.16.
211See para. 2.3.
212EC Official Journal, L 128, 21 May 1988, p.36.
213Therapeutic treatment means treatment of a disease or other health problem.
214Zootechnical treatment for the purposes of the EC measures in dispute means, inter alia, treatment for the synchronization of oestrus, termination of unwanted gestation, the improvement of fertility and the preparation of donors and recipients for the implantation of embryos (Article 2, paragraph 1(b) of EC Directive 88/299/EEC).
215See para. 2.4.
216EC Official Journal, L 1125, 23 May 1996, p.3.
217See para. 2.5.
WT/DS26/R/USA
Page 163
and 5; the Agreement on Technical Barriers to Trade ("TBT Agreement"); and the General Agreement on Tariffs and Trade 1994 ("GATT"), in particular Articles I and III.
The European Communities rejects these claims.
The six hormones in dispute are: , testosterone, progesterone, zeranol and trenbolone (the five hormones mentioned above which were brought within the general prohibition required by Directive 81/602/EEC by Directive 88/146/EEC) and melengestrol acetate ("MGA"; a sixth hormone falling under the general prohibition of Directive 81/602/EEC).
is a natural hormone with oestrogenic action (i.e., responsible for female characteristics); testosterone is a natural hormone with androgenic action (i.e., responsible for male characteristics); progesterone is a natural hormone with gestagenic action (i.e., responsible for maintaining pregnancy); zeranol is a synthetic hormone with oestrogenic action (which mimics the action of
); trenbolone is a synthetic hormone with androgenic action (which mimics the action of testosterone); and MGA is a synthetic hormone with gestagenic action (which mimics the action of progesterone).218 Natural hormones are hormones which are produced endogenously in animals and humans. Synthetic hormones are hormones which are artificially produced.
, testosterone and progesterone are hereafter also referred to as the three natural hormones; zeranol, trenbolone and MGA are hereafter also referred to as the three synthetic hormones.
In the course of these proceedings, we considered several issues related to the gathering and submission of scientific evidence. These concerned the appointment of scientific experts to advise the Panel, the deadline for submission of scientific evidence by the parties to the dispute and a request by the European Communities to ask specific national and international authorities to provide the Panel with scientific studies and data.
Article 11.2 of the SPS Agreement provides the following:
"In a dispute under this Agreement involving scientific or technical issues, a panel should seek advice from experts chosen by the panel in consultation with the parties to the dispute. To this end, the panel may, when it deems it appropriate, establish an advisory technical experts group, or consult the relevant international organizations, at the request of either party to the dispute or on its own initiative" (emphasis added).
Article 13 of the DSU reads as follows:
"1. Each panel shall have the right to seek information and technical advice from any individual or body which it deems appropriate ...
2. Panels may seek information from any relevant source and may consult experts to obtain their opinion on certain aspects of the matter. With respect to a factual issue concerning a scientific or other technical matter raised by a party to a dispute, a panel may request an advisory report in writing from an expert review group. Rules for the establishment of such a group and its procedures are set forth in Appendix 4" (emphasis added).
218See para. 2.8.
WT/DS26/R/USA
Page 164
As outlined above219, we decided to request the opinion of experts on certain scientific and other technical matters raised by the parties to this dispute. We note that both parties to this dispute assumed that in seeking advice from scientific experts, we would request an advisory report from an "expert review group" as provided for in Article 13.2 and Appendix 4 of the DSU. We considered, however, that neither Article 11.2 of the SPS Agreement nor Article 13.2 of the DSU limits our right to seek information from individual experts as provided for in Article 11.2, first sentence, of the SPS Agreement and Articles 13.1 and 13.2, first sentence, of the DSU. For our examination of this dispute, we considered it more useful to leave open the possibility of receiving a range of opinions from individual experts on specific scientific and technical questions, rather than to establish an expert review group which would have been required to reach a consensus view on the basis of general terms of reference given to it by the Panel.
We chose the scientific experts in consultation with the parties to the dispute in accordance with Article 11.2 of the SPS Agreement. A list of names of individuals expert in the subject matter before the Panel was provided by the Codex Commission secretariat as well as by the International Agency for Research on Cancer ("IARC"). The parties were given the opportunity to comment on the names stated in this list and in particular to submit any compelling objections they might have with regard to any of the individuals appearing on this list. The parties were invited to nominate one expert each, not necessarily from the list provided by the Panel. The Panel then selected three additional individuals from the list taking into account the comments of the parties. We also sought information from the Codex Commission secretariat which answered our questions in writing and also sent an expert to our oral hearing with the experts.220
The procedures we adopted for our consultation with the experts and the views they expressed are set out in paragraphs 6.1 and following. It is of particular importance that we made clear to the experts advising the Panel that we were not seeking a consensus position among the experts but wanted to hear all views.221 However, we also pointed out at the joint meeting with experts that, in order to gain time, where an expert agreed with a statement made or answer provided by another expert, the former expert did not have to take the floor.222 Any reference made in our findings to "scientific experts advising the Panel" (or "experts" or "scientific experts") refers to one or more of the five individual experts we thus appointed and, as the case may be, the expert sent by the Codex Commission secretariat. This phrase does not refer to nor includes the scientists who were part of the delegations of the parties to this dispute. These are referred to in this report by name, followed by the name of the delegation of which they were part.
With respect to the submission of scientific evidence by the parties to this dispute, we decided that they could submit written material on new scientific evidence to support their arguments by no later than 8 February 1997. We took this decision in order to ensure that both the parties and the scientific experts advising the Panel would get an opportunity to examine the scientific evidence before the 17-18 February joint meeting with the experts.
The European Communities also requested that the United States provide the Panel the originals of the studies and other relevant data on which its competent authorities based the decision to authorize the use of the hormones at issue. It also considered that the originals of the studies and other data on which the 1988 JECFA Report based its recommendations should be provided to the Panel. As
219See paras. 6.1 ff.
220The parties' arguments with respect to the appointment of the experts are set out in paras. 6.3 ff.
221See Transcripts of the joint meeting with experts of 17 February 1997, para. 45.
222Ibid., para. 88.
WT/DS26/R/USA
Page 165
far as the studies and data used by US authorities is concerned, we did not consider this information to be relevant to address the EC measures in dispute. Similarly, we did not consider it necessary to request the studies and data on which the 1988 JECFA Report is based since it was our understanding that both parties involved in this dispute participated in the elaboration of this report.
In the course of our work, we also considered several issues related to the parallel existence of this Panel, requested by the United States, and the panel requested by Canada. Although different panels, both panels relate to the same EC measures, were dealt with by the same panel members and were assisted by the same scientific experts. These issues relate to the organization of joint meetings and the access by parties in one of these panel proceedings to materials submitted to the other panel.
In this respect we note, as a general guideline, Article 9.3 of the Understanding on Rules and Procedures Governing the Settlement of Disputes ("DSU"), which reads as follows:
"If more than one panel is established to examine the complaints related to the same matter, to the greatest extent possible the same persons shall serve as panelists on each of the separate panels and the timetable for the panel process in such disputes shall be harmonized" (emphasis added).
Prior to our meeting with scientific experts, we decided to hold that meeting jointly for both this Panel and the parallel panel requested by Canada. This decision stemmed from the similarities of the two cases (the same EC measures are at issue and both cases are dealt with by the same panel members) and our decision to use the same scientific experts in both cases. In addition, we considered that, from a practical perspective, there was a need to avoid repetition of arguments and/or questions at our meetings with the scientific experts. The European Communities objected to this decision arguing that one joint meeting with the experts, instead of two separate meetings, was likely to affect its procedural rights of defence. Where it made precise claims of prejudice to its rights of defence, we took corrective action.223
In view of our decision to hold a joint meeting with experts, we also decided to give access to all of the information submitted under each panel proceeding to the parties in the other panel proceeding, including the parties' second written submissions, written versions of oral statements, questions raised by the Panel and the parties in each case and the answers we received, as well as all scientific documentation submitted by the parties. By doing so, we understood that, in this Panel, we could also consider, where appropriate, the materials submitted before the panel requested by Canada.224 The European Communities objected to this decision arguing that it was most likely going to affect its substantive and procedural rights of defence. It made, however, no specific claims of prejudice. We considered that providing all information to all parties involved in both panels would increase the transparency of our work, without depriving parties of their substantive or procedural
223The European Communities argued that a joint meeting with experts deprived it of its right to present its legal and scientific positions twice (a first time before this Panel and a second time before the panel requested by Canada). In light of this objection we decided that the European Communities would be allowed to address the joint meeting twice (a first time after the United States and a second time after Canada). See Transcripts of the joint meeting with experts of 17 February 1997, paras.1-2.
224This does, of course, not mean that, in this Panel, the United States has incorporated or agreed with the arguments submitted by Canada in the parallel panel it requested or that we have to deal, in this Panel, with the arguments and claims submitted by Canada in the other panel. It only means that, where we considered it appropriate, we would be able to invoke or address, in this Panel, factual elements, scientific evidence or arguments submitted by the European Communities or Canada to the parallel panel requested by Canada.
WT/DS26/R/USA
Page 166
rights, and is in line with the object and purpose of Article 9.3 of the DSU calling for a harmonization of timetables of both panels. On these grounds we rejected the EC's objection.
The United States contests the EC ban on imports of meat and meat products from cattle, treated with any of six specific hormones (oestradiol-17 , testosterone, progesterone, zeranol, trenbolone and MGA) for growth promotion purposes. The United States does not challenge the EC ban on imports of meat and meat products from cattle treated with hormones other than the six specified or treated with any hormones for purposes other than growth promotion.225
At the end of the Panel proceedings, the United States argued that its complaint was not limited to meat and meat products of bovine origin (i.e., originating from cattle) but also extended to, for example, lamb meat. In this regard the United States referred to an Annex to its first submission where it is stated that zeranol is approved in the United States for use in lambs. We note, however, that, although technically within the terms of reference of the Panel226, at no point in the Panel proceedings did the United States raise specific arguments or submit factual or scientific evidence against a ban imposed by the European Communities on the use of hormones in farm animals other than cattle. We further note that all scientific studies invoked by both the United States and the European Communities relate to risks to human health or cattle from the ingestion or administration of hormones; not risks to any other animals. We find, therefore, that the EC ban in so far as it relates to meat or meat products from farm animals other than cattle falls outside the scope of this dispute.
At the end of the Panel proceedings, the United States also seemed to argue that its complaint was not limited to meat and meat products but also extended to live animals.227 However, our terms of reference direct us to examine "the matter referred to the DSB by the United States in [document WT/DS26/6, i.e., the request for the establishment of a panel by the United States]". In that document the United States only claimed that the contested EC measures "adversely affect imports of meat and meat products".228 We therefore consider that the matter referred to in that document is limited to meat and meat products and thus find that the EC ban on imports of live animals falls outside our terms of reference.
Finally, we note that the European Communities argues that its import ban on live animals to which any of the six hormones have been administered, is necessary for the protection of both human and animal health.229 However, the European Communities does not make this argument with respect to its import ban on meat or meat products. Specifically, the European Communities has not argued that its import ban on meat or meat products is necessary for the protection of animal health either inside or outside the EC territory. Since the animal health arguments invoked by the European Communities exclusively relate to its import ban on live animals and considering the finding reached
225See para. 3.1.
226The request for the establishment of a panel by the United States of 25 April 1996 (WT/DS26/6), refers to "meat and meat products" without explicit limitation to meat and meat products of bovine origin (see para. 1.4).
227See para. 4.9.
228See para. 1.4.
229See para. 4.208.
WT/DS26/R/USA
Page 167
above that the EC ban on imports of live animals treated with hormones does not fall within the scope of this dispute, we find that within the scope of this dispute we need not take into account the arguments made by the European Communities which relate to animal health.230
The United States invokes arguments relating to three different agreements: the SPS Agreement, the TBT Agreement and GATT. The European Communities, in turn, invokes the same three agreements in its defense. We next examine which of these agreements apply to the present dispute.
With respect to the SPS Agreement, both parties agree that the EC measures in dispute are sanitary measures in the sense of Paragraph 1(b) of Annex A of the SPS Agreement.231 Paragraph 1(b) of Annex A defines a sanitary measure as
"any measure applied to protect human or animal life or health within the territory of the Member from risks arising from additives, contaminants, toxins or disease-causing organisms in foods, beverages or feedstuffs".
Footnote 4 to Annex A specifies that "contaminants" include, for the purposes of Annex A, "pesticide and veterinary drug residues and extraneous matter". Since the six hormones in dispute are veterinary drugs, the parties agree that the alleged risks at issue arise from contaminants.
We agree with the parties that the EC measures in dispute are "applied to protect human ... lifeor health" within the territory of the European Communities from risks arising from "contaminants", namely residues of six specific hormones, in foods (according to paragraph 1(b) of Annex A). That the contested EC measures are, inter alia, "applied to protect human ... life or health" can be inferred from the preambles to, and legislative history of, Directives 81/602/EEC and 88/146/EEC.232 Since both parties agree that the contested EC measures are "sanitary measures", we see no need to further examine in this dispute the definition of measures "applied to protect human ... life or health".
Both parties also agree that, according to Article 1.1 of the SPS Agreement, the SPS Agreement is applicable to this dispute.233 Article 1.1 provides that the SPS Agreement
"applies to all sanitary and phytosanitary measures which may, directly or indirectly, affect international trade".
We agree with the parties that the EC measures "may, directly or indirectly, affect international trade". It cannot be contested that an import ban affects international trade.
With respect to the application ratione temporis of the SPS Agreement to the EC measures in dispute, we note that the SPS Agreement entered into force on 1 January 1995 (the date of entry into force of the WTO Agreement of which, according to Article II:2 of that Agreement, the SPS Agreement is an integral part). The EC measures in dispute, however, were enacted before 1
230Our finding that animal health does not fall within the scope of this dispute does, of course, not mean that we cannot take into account scientific evidence relevant for risks to human life or health which is derived from studies or tests applied to animals.
231See paras. 4.7-4.8.
232See paras. 2.2-2.3.
233See paras. 4.5 and 4.8.
WT/DS26/R/USA
Page 168
January 1995 (namely 31 July 1981 and 7 March 1988), thus raising the issue of whether the SPS Agreement applies to these measures.
Article 3.2 of the DSU directs us to clarify the provisions of the SPS Agreement "in accordance with customary rules of interpretation of public international law". According to established practice, the fundamental rules of treaty interpretation set out in the Vienna Convention on the Law of Treaties ("Vienna Convention") form part of these customary rules of interpretation.234 The general principle in international law, as embodied in Article 28 of the Vienna Convention, is that "[u]nless a different intention appears from the treaty or is otherwise established, its provisions do not bind a party in relation to ... any situation which ceased to exist before the date of the entry into force of the treaty ..." (emphasis added). The EC measures can, in this context, be considered as continuing "situations" which were enacted before the entry into force of the SPS Agreement but which did not cease to exist after that date (contrary to the situation envisaged in Article 28). In line with Article 28 of the Vienna Convention, the SPS Agreement should, therefore, in principle apply to these EC measures, unless an intention to the contrary can be established. 235
An examination of the SPS Agreement reveals no intention to the contrary. Indeed, several provisions of the SPS Agreement confirm the general principle that the SPS Agreement should also apply to sanitary measures which were enacted before its entry into force but which remain in force thereafter. Except for Article 14 which authorizes delays in the application of some or all of the provisions of the SPS Agreement for least-developed and other developing countries, no transition periods are provided for. The fact that Article 14 explicitly provides for a two-year transition period for developing countries with respect to some of their existing sanitary and phytosanitary measures, confirms that the SPS Agreement generally applies to measures enacted before the entry into force of the SPS Agreement but which are maintained in force after that date. This is also confirmed in several provisions of the SPS Agreement which explicitly address situations where Members "maintain" a sanitary or phytosanitary measure, such as Article 2.2 ("Members shall ensure that any sanitary ... measure ... is based on scientific principles and is not maintained without sufficient scientific evidence
..."), Article 3.3 ("Members may introduce or maintain sanitary ... measures ... if ..."), Article 5.6 ("... when establishing or maintaining sanitary ... measures ... Members shall ensure that ...") and Article 5.8 ("... a specific sanitary ... measure introduced or maintained by another Member ...").
We finally note that according to Article XVI:4 of the WTO Agreement, each Member "shall ensure the conformity of its laws, regulations and administrative procedures with its obligations as provided in the annexed Agreements [including the SPS Agreement]". This provision confirms that measures which already existed as of the date of entry into force of the SPS Agreement also need to be consistent with the requirements imposed by that Agreement.
Thus, we find that the SPS Agreement is applicable to this dispute.
In respect of the applicability of the TBT Agreement to this dispute, we note that Article 1.5 of the TBT Agreement reads as follows:
234See Appellate Body Reports on "United States - Standards for Reformulated and Conventional Gasoline", adopted on 20 May 1996, WT/DS2/AB/R, pp.16-17 and "Japan - Taxes on Alcoholic Beverages", adopted on 1 November 1996, WT/DS8/AB/R, pp.10-12.
235We refer, in this respect, to the Reports of the Panel and Appellate Body on "Japan - Taxes on Alcoholic Beverages", adopted on 1 November 1996 (WT/DS8/R and WT/DS8/AB/R), where both the Panel and the Appellate Body applied GATT (which entered into force on 1 January 1995) to the Japanese Liquor Tax Law (of 1953 and last amended on 1 May 1994), even though that Japanese measure had been enacted and most recently been amended before the entry into force of GATT, on the implicit ground that the Japanese measure remained in force after that date. The same reasoning was applied in the Reports of the Panel and Appellate Body on "United States - Standards for Reformulated and Conventional Gasoline", op. cit.
WT/DS26/R/USA
Page 169
"The provisions of this Agreement do not apply to sanitary and phytosanitary measures as defined in Annex A of the Agreement on the Application of Sanitary and Phytosanitary Measures".236
Since the measures in dispute are sanitary measures, we find that the TBT Agreement is not applicable to this dispute.
We finally note that this dispute relates to trade in goods (in casu imports of meat and meat products) and that on its face GATT applies.237 In this context, we note that the United States only invokes GATT after having addressed the SPS Agreement and that the European Communities does not invoke any GATT provision other than Article XX(b) as a justification for the EC measures in dispute.
Since both the SPS Agreement and GATT apply to this dispute, we next examine the relationship between these two agreements.
The parties to the dispute present diverging views with respect to whether we should first address GATT or the SPS Agreement. However, neither of the parties claims that the relevant provisions of the SPS Agreement and GATT are in conflict. Therefore, we do not need, as a preliminary matter, to address the General Interpretative Note to the Multilateral Agreements on Trade in Goods which only applies "[i]n the event of conflict between a provision of [GATT] and a provision of another Agreement in Annex 1A [inter alia, the SPS Agreement]".
The European Communities makes a distinction between the "substantive" and "procedural" provisions of the SPS Agreement. According to the European Communities, the substantive provisions only interpret Article XX(b) of GATT238, without adding any new obligations, while the procedural provisions contain requirements additional to GATT. Therefore, the European Communities concludes, the "substantive" provisions of the SPS Agreement can only be addressed if recourse is made to GATT Article XX(b), i.e., if, and only if, a violation of another provision of GATT is first established. The additional "procedural" provisions, on the other hand, can be examined directly and independently of a prior GATT violation.239
The United States argues that the SPS Agreement is the lex specialis for a review of sanitary measures and should, therefore, be addressed first. The United States claims that the application of the SPS Agreement does not require a prior violation of GATT since the SPS Agreement is a "free-
236Similarly, but less explicitly, Article 1.4 of the SPS Agreement provides that "[n]othing in this Agreement shall affect the rights of Members under the Agreement on Technical Barriers to Trade with respect to measures not within the scope of this Agreement".
237With respect to the application ratione temporis of GATT to this particular case, the same reasoning and findings apply as those developed for the application ratione temporis of the SPS Agreement in paragraphs 8.24 and 8.25.
238Article XX(b) of GATT reads as follows: "Subject to the requirement that such measures are not applied in a manner which constitute a means of arbitrary or unjustifiable discrimination between countries where the same conditions prevail, or a disguised restriction on international trade, nothing in this Agreement shall be construed to prevent the adoption or enforcement by any contracting party of measures: ... (b) necessary to protect human, animal or plant life or health".
239See para. 4.4.
WT/DS26/R/USA
Page 170
standing" agreement which applies to all sanitary measures and imposes requirements additional to those in Article III or the exceptions in Article XX of GATT.240
In examining the relationship between GATT and the SPS Agreement, we recall the fundamental rules of treaty interpretation set out in the Vienna Convention.241 Article 31 of the Vienna Convention prescribes that a treaty has to be interpreted "in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose".242
We first consider the wording of Article 1.1 of the SPS Agreement which reads as follows: "This Agreement applies to all sanitary and phytosanitary measures which may, directly
or indirectly, affect international trade. Such measures shall be developed and applied
in accordance with the provisions of this Agreement".
According to Article 1.1 of the SPS Agreement, two requirements need to be fulfilled for the SPS Agreement to apply: (i) the measure in dispute is a sanitary or phytosanitary measure243; and (ii) the measure in dispute may, directly or indirectly, affect international trade.244 There are no additional requirements. The SPS Agreement contains, in particular, no explicit requirement of a prior violation of a provision of GATT which would govern the applicability of the SPS Agreement, as asserted by the European Communities.
We further note that the distinction proposed by the European Communities between "substantive" and "procedural" provisions of the SPS Agreement has no basis in the text of that Agreement and would, in any event, seem to be difficult to apply to most provisions contained therein. For example, the obligation to base a sanitary measure on a risk assessment in accordance with Article 5 of the SPS Agreement includes both substantive and procedural elements.245
Moreover, we find the EC claim that the SPS Agreement does not impose "substantive" obligations additional to those already contained in Article XX(b) of GATT not to be persuasive. It is clear that some provisions of the SPS Agreement elaborate on provisions already contained in GATT, in particular Article XX(b). The final preambular paragraph of the SPS Agreement provides, indeed, that the Members desired "to elaborate rules for the application of the provisions of GATT 1994 which relate to the use of sanitary or phytosanitary measures, in particular the provisions of Article XX(b)". Examples of such rules are, arguably, some of the obligations contained in Article 2 of the SPS Agreement. However, on this basis alone we cannot conclude that the SPS Agreement only applies, as Article XX(b) of GATT does, if, and only if, a prior violation of a GATT provision has been established. Many provisions of the SPS Agreement impose "substantive" obligations which go
240See para. 4.5.
241The legal basis for the Panel to invoke the Vienna Convention is outlined above in paragraph 8.25 and footnote 234.
242According to Article 32 of the Vienna Convention, recourse may only be had to supplementary means of interpretation "in order to confirm the meaning resulting from the application of Article 31" or in case Article 31 leaves the meaning "ambiguous or obscure" or leads to "a result which is manifestly absurd or unreasonable".
243As defined in Paragraph 1 of Annex A of the SPS Agreement, quoted and discussed in paras. 8.21 and 8.22.
244See para. 8.23.
245See paras. 8.112 ff.
WT/DS26/R/USA
Page 171
significantly beyond and are additional to the requirements for invocation of Article XX(b).246 These obligations are, inter alia, imposed to "further the use of harmonized sanitary and phytosanitary measures between Members"247 and to "improve the human health, animal health and phytosanitary situation in all Members".248 They are not imposed, as is the case of the obligations imposed by Article XX(b) of GATT, to justify aviolation of another GATT obligation (such as a violation of the non-discrimination obligations of Articles I or III).
We note in this respect that the general approach adopted in Article XX(b) of GATT is fundamentally different from the approach adopted in the SPS Agreement. Article XX(b), which is not limited to sanitary or phytosanitary measures, provides for a general exception which can be invoked to justify any violation of another GATT provision. The SPS Agreement, on the other hand, provides for specific obligations to be met in order for a Member to enact or maintain specific types of measures, namely sanitary and phytosanitary measures.
The conclusion that the SPS Agreement contains obligations which are not already imposed by GATT is confirmed in Article 2.4 of the SPS Agreement which provides that "[s]anitary or phytosanitary measures which conform to the relevant provisions of this Agreement shall be presumed to be in accordance with the obligations of the Members under the provisions of GATT 1994 which relate to the use of sanitary or phytosanitary measures, in particular the provisions of Article XX(b)". Indeed, to presume that one set of obligations (in casu GATT) is met because another set of obligations (in casu the SPS Agreement) has been fulfilled, seems to imply that the latter set of obligations imposes at least as many as, and probably more obligations than, the former. Support for this conclusion is also found in Article 3.2 of the SPS Agreement which provides that "[s]anitary or phytosanitary measures which conform to international standards, guidelines or recommendations shall be deemed to be necessary to protect human, animal or plant life or health, and presumed to be consistent with the relevant provisions of this Agreement and of GATT 1994" (emphasis added). While both agreements may apply in a given factual situation, the foregoing provision nonetheless establishes the SPS Agreement as an agreement which imposes obligations which are different from those imposed by GATT.
We therefore find that, in accordance with the ordinary meaning to be given to the terms of the SPS Agreement in their context and in the light of its object and purpose (in conformity with Article 31 of the Vienna Convention), there is no requirement, in any of the provisions of the SPS Agreement, that a prior violation of a GATT provision need be established before the SPS Agreement applies.
Having reached the conclusion that we are not per se required to address GATT claims prior to those raised under the SPS Agreement, we must then decide which of the two agreements we should examine first in this particular dispute. The SPS Agreement specifically addresses the type of measure in dispute. If we were to examine GATT first, we would in any event need to revert to the SPS Agreement: if a violation of GATT were found, we would need to consider whether Article XX(b) could be invoked and would then necessarily need to examine the SPS Agreement; if, on the other hand, no GATT violation were found, we would still need to examine the consistency of the measure with the SPS Agreement since nowhere is consistency with GATT presumed to be consistency with the SPS Agreement. For these reasons, and in order to conduct our consideration of this dispute in the most efficient manner, we shall first examine the claims raised under the SPS Agreement.
246One example is the obligation contained in Article 3.1 of the SPS Agreement which provides that "[t]o harmonize sanitary and phytosanitary measures on as wide a basis as possible, Members shall base their sanitary and phytosanitary measures on international standards, guidelines or recommendations, where they exist ...".
247Preambular paragraph 6 of the SPS Agreement.
248Preambular paragraph 2 of the SPS Agreement.
WT/DS26/R/USA
Page 172
The United States claims violations of Articles 2, 3 and 5 of the SPS Agreement. Article 2 elaborates on the basic rights and obligations of Members under the SPS Agreement. Article 3 deals, more specifically, with the objective of harmonization of sanitary measures on the basis of international standards, guidelines or recommendations. Article 5 deals, in turn, with the obligation of risk assessment and the determination and application by Members of their appropriate level of sanitary protection.
Article 3.1 requires Members to base their sanitary measures on international standards, guidelines or recommendations except as otherwise provided for in the SPS Agreement, and in particular in Article 3.3. We note, therefore, that even if international standards may not, in their own right, be binding on Members, Article 3.1 requires Members to base their sanitary measures on these standards.
According to Article 3.2 sanitary measures which conform to internationalstandards, guidelines or recommendations are presumed to be consistent with both the SPS Agreement and GATT. We shall therefore, as a first step, examine whether there are international standards, guidelines or recommendations with respect to the EC measures in dispute and, if so, whether the EC measures are based on these standards, guidelines or recommendations in accordance with Article 3.1.
If there are international standards, guidelines or recommendations and the European Communities has not based its measures thereon, we will need, as a second step, to examine whether the European Communities can justify its measures under Article 3.3 since Article 3.1, which imposes the requirement to base sanitary measures on international standards explicitly refers to Article 3.3 as providing for an exception to this requirement.
Finally, if there are no international standards, guidelines or recommendations with respect to the EC measures in dispute, or to some of them, there would be no standards, guidelines or recommendations for these measures to be based on in line with Article 3.1. However, even in that case the consistency of the EC measures in dispute with Articles 2 and 5 of the SPS Agreement would still need to be examined.
Given the nature of disputes under the SPS Agreement, which imposes substantive and procedural requirements raising various, and in this case complex, issues of fact, the allocation of the burden of proof is of particular importance. It involves consideration of the wording, general outline and object and purpose of the SPS Agreement as a whole. We, therefore, first examine this issue in general before addressing the specific burden of proof for each of the provisions in dispute in more detail below.
In stating its claims under the SPS Agreement, the United States seems to presume that the European Communities bears the burden of proof. The United States argues that the SPS Agreement, inter alia, requires the European Communities to base its sanitary measures on a risk assessment and prohibits the European Communities from maintaining such measures without scientific evidence. According to the United States, the SPS Agreement does not allow for measures to bemaintained without scientific evidence until such time as science proves "beyond doubt" that there is no risk (against which a sanitary measure can protect, for example, consumers). The United States seems, therefore, to conclude that it is up to the European Communities to provide evidence that there is a risk to be protected
WT/DS26/R/USA
Page 173
against and that there has been a risk assessment. It is not up to the United States to prove that there is no risk or that the European Communities did not carry out a risk assessment. 249
The European Communities argues that the burden of proof should rest on the party challenging the consistency of sanitary measures with the SPS Agreement (in casu the United States). The European Communities claims, inter alia, that it is up to the United States to provide evidence that the use of the hormones in dispute for growth promotion is safe and without risk.
In addressing the burden of proof under the SPS Agreement, we consider that, as is the case in most legal proceedings, the initial burden of proof rests on the complaining party in the sense that it bears the burden of presenting a prima facie case of inconsistency with the SPS Agreement. It is, indeed, for the party that initiated the dispute settlement proceedings to put forward factual and legal arguments in order to substantiate its claim that a sanitary measure is inconsistent with the SPS Agreement. In other words, it is for the United States to present factual and legal arguments that, if unrebutted, would demonstrate a violation of the SPS Agreement. Once such a prima facie case is made, however, we consider that, at least with respect to the obligations imposed by the SPS Agreement that are relevant to this case, the burden of proof shifts to the responding party.250
In our view, the allocation of evidentiary burden under the SPS Agreement to the Member imposing a sanitary or phytosanitary measure flows directly from the wording of many of the provisions contained in that Agreement and in particular the first three words thereof:
"Members shall ensure that..." (e.g. Articles 2.2, 2.3, 5.1 and 5.6 of the SPS Agreement; emphasis added).
Moreover, the wording of Article 5.8 (although this provision relates more to transparency than to any requirement of legal justification) further supports our reading of this assignment of burden of proof to the party imposing the measure:
"When a Member has reason to believe that a specificsanitary or phytosanitary measure introduced or maintained by another Member is constraining, or has the potential to constrain, its exports and the measure is not based on the relevant international standards, guidelines or recommendations, or such standards, guidelines or
249See para. 4.101.
250Two provisions of the SPS Agreement do, however, explicitly confer the burden of proof upon the exporting Member (i.e., the Member contesting the sanitary or phytosanitary measure), namely Article 4.1 on equivalence and Article 6.3 on pest- or disease-free areas or areas of low pest or disease prevalence. The fact that in these instances the burden of proof is explicitly conferred upon the exporting Member confirms that under other SPS provisions the burden of proof may shift to the Member imposing the sanitary or phytosanitary measure. We note, in this respect, the Report of the Appellate Body on "United States - Measures Affecting Imports of Woven Wool Shirts and Blouses from India" (adopted on 23 May 1997, WT/DS33/AB/R) which addressed the issue of burden of proof under the Agreement on Textiles and Clothing (the "ATC") as follows:
"... a party claiming a violation of a provision of the WTO Agreement by another Member must assert and prove its claim. In this case, India claimed a violation by the United States of Article 6 of the ATC. We agree with the Panel that it, therefore, was up to India to put forward evidence and legal argument sufficient to demonstrate that the transitional safeguard action by the United States was inconsistent with the obligations assumed by the United States under Articles 2 and 6 of the ATC. India did so in this case. And, with India having done so, the onus then shifted to the United States to bring forward evidence and argument to disprove the claim. This, the United States was not able to do and, therefore, the Panel found that the transitional safeguard action by the United States "violated the provisions of Articles 2 and 6 of the ATC" (pp.16-17).
WT/DS26/R/USA
Page 174
recommendations do not exist, an explanation of the reasons for such sanitary or phytosanitary measure may be requested and shall be provided by the Member maintaining the measure" (emphasis added).
Finally, we note that this assignment of burden of proof to the party imposing the measure is also supported by Article 3.2 which introduces a presumption of consistency with the SPS Agreement for sanitary measures which conform to international standards, guidelines or recommendations. Article 3.2 states the following:
"Sanitary or phytosanitary measures which conform to international standards, guidelines or recommendations shall be deemed to be necessary to protect human, animal or plant life or health, and presumed to be consistent with the relevant provisions of this Agreement and of GATT 1994".
Introducing a general presumption of consistency with an agreement in favour of a party (in casu the party imposing the measure) in the event that certain conditions are met, seems, indeed, to presuppose that the burden of proof under that agreement in principle (i.e., in cases where these specific conditions are not met) rests on that party.
We thus find that, for the purposes of this dispute, the United States bears the burden of presenting a prima facie case of inconsistency with the SPS Agreement, after which the burden of proof shifts to the European Communities to demonstrate that its measures in dispute meet the requirements imposed by the SPS Agreement.
Article 3.1 of the SPS Agreement reads as follows:
"To harmonize sanitary and phytosanitary measures on as wide a basis as possible, Members shall base their sanitary and phytosanitary measures on international standards, guidelines or recommendations, where they exist, except as otherwise provided for in this Agreement, and in particular in paragraph 3".
The first question we must address is whether there exist any "international standards, guidelines or recommendations" with respect to the administration of any of the six hormones in dispute for growth promotion purposes. For food safety, the health concern at issue in this dispute, paragraph 3(a) of Annex A of the SPS Agreement defines "international standards, guidelines or recommendations" as
"the standards, guidelines and recommendations established by the Codex Alimentarius Commission relating to food additives, veterinary drug and pesticide residues, contaminants, methods of analysis and sampling, and codes and guidelines of hygienic practice" (emphasis added).
In line with Article 3.1, we consider that if such Codex Alimentarius Commission standards, guidelines or recommendations ("Codex standards") exist with respect to the administration of any of the six hormones in dispute for growth promotion purposes, a sanitary measure taken by a Member should either be based on these standards or be justified under Article 3.3 of the SPS Agreement.
WT/DS26/R/USA
Page 175
Within the scope of the measures in dispute, we note that Codex standards exist for five of the six hormones at issue (i.e., for all hormones at issue other than MGA).251 We will accordingly examine the definition and scope of application of these Codex standards and determine whether they apply to the EC measures in dispute.
The Codex Alimentarius Commission ("Codex"), an international body of which most WTO Members (including the United States and the EC member States of the European Communities) are members, establishes, inter alia, Acceptable Daily Intakes ("ADIs"), Maximum Residue Limits ("MRLs") and other recommendations for veterinary drugs. It does so on the basis of the advice of the Codex Committee on Residues of Veterinary Drugs in Foods and the recommendations of the Joint FAO/WHO Expert Committee on Food Additives ("JECFA"). While Codex is composed of government representatives of EC member States, JECFA is composed of independent scientists. JECFA makes scientific evaluations and recommendations; Codex takes the decision whether or not to adopt these recommendations. However, once adopted Codex recommendations are, according to the General Principles of Codex, not binding upon Codex members. They are only of an advisory nature. The procedures to be followed to adopt a Codex recommendation have been outlined above.252
The goal of JECFA's evaluation of veterinary drugs is
"to establish safe levels of intake by setting Acceptable Daily Intakes (ADIs) and to develop maximum residue limits when veterinary drugs are used in accordance with good veterinary practice" (emphasis added).253
The term "good veterinary practice" (as well as the term "good animal husbandry practice" often used in JECFA Reports) is for Codex purposes, according to the Codex expert advising the Panel254, a synonym for what is known in Codex terminology as "Good Practice in the Use of Veterinary Drugs" ("GPVD"; hereafter also referred to as "good practice"), in turn defined as
"the official recommended or authorized usage including withdrawal periods, approved by national authorities, of veterinary drugs under practical conditions". 255
An ADI set by Codex is "an estimate by JECFA of the amount of a veterinary drug, expressed on a body weight basis, that can be ingested daily over a lifetime without appreciable health risk (standard man = 60 kg)" (emphasis added).256 This ADI is derived from the experimental no observable effect level in the most appropriate animal species, by applying an appropriate safety factor. A Codex MRL, on the other hand, if implemented in national law, determines the amount of residues which is legally permitted or recognized as acceptable in food and is primarily a regulatory tool to ensure that intake does not exceed the ADI and that good practice is observed. A Codex MRL is frequently set at levels
251See paras. 2.20, 2.21 and 2.23.
252See paras. 2.15-2.16.
253See para. 2.14.
254See para. 2.19.
255Ibid.
256See para. 2.17.
WT/DS26/R/USA
Page 176
below (even far below) the theoretical safe levels determined from an ADI. Codex MRLs for veterinary drugs are normally expressed in µg/kg on a fresh weight basis in meat.257
With respect to the three natural hormones in dispute, , progesterone and testosterone (classified by Codex as "veterinary drugs"), similar Codex standards apply. For all three hormones, when used for growth promotion purposes, it was considered "unnecessary" to establish an ADI or MRL. For all three hormones the following footnote explained the word "unnecessary":
"Establishing an ADI and an [MRL] for a hormone that is produced endogenously at variable levels in human beings was considered unnecessary by the Committee. Residues resulting from the use of this substance as a growth promoter in accordance with good animal husbandry practice are unlikely to pose a hazard to human health".258
The 32nd JECFA Report of 1988259, on which the Codex standards are based, concluded for all three natural hormones administered for growth promotion purposes that the residue levels of each of these hormones when found in meat from animals treated with implants according to good animal husbandry practice are extremely low when compared with the amounts endogenously produced daily in human beings or normally present in the dairy products or tissues of untreated animals or other foods. According to JECFA, the potential toxic effect of residues of these hormones is directly related to their hormonal effect. Since the additional residue levels in treated animals have no hormonal effect, the Report concluded that these residue levels are not capable of exerting any toxic effect. JECFA further noted that the total residue levels in treated animals fall well within the normal range of levels found in untreated animals of different types and ages. On the basis of this safety assessment and in view of the difficulty of determining the levels of residues attributable to the use of this hormone as a growth promoter in cattle (residues of endogenous natural hormones in meat cannot, according to JECFA, be practically distinguished from those exogenously administered), JECFA concluded that it was "unnecessary" to establish an ADI or MRL for these hormones.260
With respect to two of the three synthetic hormones at issue, zeranol and trenbolone (classified by Codex as "veterinary drugs"), the following Codex standards apply: an ADI of 0-0.5 and 0-0.02
µg/kg body weight, respectively, and an MRL of 2 µg/kg -trenbolone in bovine muscle and 10 µg/kg
in bovine liver.261
The 32nd JECFA Report of 1988, on which the Codex standard for zeranol is based, noted that zeranol was shown to be a weak oestrogen which mimics the action of . The Report concluded that the toxic (in casu tumorigenic) effect of zeranol is associated with its hormonal (i.e., oestrogenic) properties and that an ADI could thus be established on the basis of a no-hormonal-effect level. Adopting what it considered to be a conservative approach by using as a basis studies on ovariectomized female cynomolgus monkeys (highly sensitive to oestrogenic substances) and using a safety factor of 100, JECFA set an ADI for human beings of 0-0.5 µg/kg of body weight. For a 70 kg person consuming 500 g of meat daily over an entire lifetime, the maximum permissible or safe level of zeranol residues in meat would then, according to JECFA, be 70 µg/kg of edible tissue. However, the Report noted that when zeranol is administered to cattle according to the proposed good
257See para. 2.18.
258See para. 2.21.
259See para. 2.22.
260Ibid.
261See para. 2.23.
WT/DS26/R/USA
Page 177
practice, the maximum mean residue levels would not exceed 0.2 µg/kg in muscle and 10 µg/kg in liver at any time after implantation (irrespective of the withdrawal period respected). These residue levels obtained on the basis of good practice are thus far below what JECFA determined to be the safety level of 70 µg/kg. However, in order to set a level which is detectable by using methods available for routine residue analysis, the official Codex MRL was increased to 2 µg/kg in muscle and set at 10 µg/kg in liver.262
Trenbolone acetate is the chemical form or ester used for the administration of trenbolone. Trenbolone, or trenbolone acetate ("TBA"), an androgen which mimics the action of testosterone, is rapidly hydrolysed after administration to cattle, the major metabolites (i.e., compound into which TBA breaks down by chemical activity when entering the body) being occurring, inter alia, in liver, and
-trenbolone, present in muscle.263 With respect to TBA, the 32nd JECFA Report of 1988 concluded that its potential toxic effects only arise as a consequence of its hormonal activity and concluded that, therefore, an ADI could be established on the basis of a no-hormonal-effect level.264 Adopting what it considered to be a conservative approach by using as a basis studies on castrated male rhesus macaque monkeys which are highly sensitive to compounds with antigonadotropic activity and pigs which are a sensitive model for assessing hormonal effects of TBA and using a safety factor of 100, JECFA later set an ADI for human beings of 0-0.02 µg/kg of body weight (34th JECFA Report of 1989). The maximum ADI for a 60 kg person would thus be 1.2 µg of TBA residues. JECFA then set MRLs for
-trenbolone in muscle and
-trenbolone in liver of 2 µg/kg and 10 µg/kg respectively, based on average residue levels in heifers at 15-30 days after implantation of 300 mg TBA, noting that concentrations would even be lower at the proposed good practice. According to JECFA, the MRLs thus obtained on the basis of conservative estimates would not exceed the Codex ADI or safe level at any time after implantation of the drug (irrespective of the withdrawal period respected).
The European Communities argues that the Codex standards outlined above are not relevant to this dispute. It argues that there are no Codex standards for the use of hormone growth promoters, only Codex standards for maximum residue levels and that since the EC measures in dispute do not set maximum residue levels, there exist no Codex standards on which the EC measures need to be based. Moreover, the European Communities argues, the Codex standards invoked are levels of protection, not measures, and since there is no obligation in the SPS Agreement to adopt Codex recommended levels of protection, the standards invoked are irrelevant for the EC measures in dispute.265
The European Communities also notes that the decision by Codex (of July 1995) to formally adopt the five Codex standards at issue was taken by a majority of only 33 votes in favour, 29 votes against and 7 abstentions; a very close vote which is unusual in Codex practice where proposals are normally adopted by consensus, indicating that the issue of hormone growth promoters has been and continues to be very controversial. 266
The European Communities finally argues that the process which led to the adoption of the Codex standards started long before the entry into force of the SPS Agreement and was only completed six months after that date. At the time the standards were discussed, Codex members were, therefore, according to the European Communities, unaware of the fact that the Codex standards, which within
262See para. 2.24.
263See para. 2.25.
264Ibid.
265See para. 4.79.
266See para. 4.77.
WT/DS26/R/USA
Page 178
the Codex system are only of an advisory nature, would in the future become "binding" by virtue of the SPS Agreement.267 The European Communities seems to consider this element as a reason to disregard these Codex standards in this dispute.
In considering these EC arguments, we note that Article 3.1 unambiguously prescribes that "... Members shall base their sanitary ... measures on international standards ... where they exist ..." (emphasis added). Paragraph 3 of Annex A of the SPS Agreement states equally clearly that the international standards mentioned in Article 3:1 are "for food safety, the standards ... established by the Codex Alimentarius Commission relating to ... veterinary drug ... residues ..." (emphasis added). No other conditions are imposed in the SPS Agreement on the relevance of international standards for the purposes of Article 3. Therefore, as a panel making a finding on whether or not a Member has an obligation to base its sanitary measure on international standards in accordance with Article 3.1, we only need to determine whether such international standards exist. For these purposes, we need not consider (i) whether the standards reflect levels of protection or sanitary measures or the type of sanitary measure they recommend, or (ii) whether these standards have been adopted by consensus or by a wide or narrow majority, or (iii) whether the period during which they have been discussed or the date of their adoption was before or after the entry into force of the SPS Agreement.268
We note that the five Codex standards outlined above are standards relating to veterinary drug residues as required in paragraph 3(a) of Annex A269 and apply exclusively with respect to cattle and meat and meat products of bovine origin and with respect to five of the six hormones in dispute when these hormones are used for growth promotion purposes.270 We recall the scope of the EC measures in dispute, in particular that they are limited to the EC ban on imports of meat and meat products of bovine origin from cattle treated with any of six specific hormones if the treatment with any of these substances is carried out for growth promotion purposes.271 We find, therefore, that international standards exist with respect to the EC measures in dispute, to the extent they relate to five of the six hormones at issue (all but MGA), in the sense of Article 3.1 and paragraph 3(a) of Annex A. We must next determine whether the EC measures are based on these international standards in terms of Article 3.1.
The United States argues that the European Communities ignored the five Codex standards outlined above. Rather than establishing an ADI or MRL for any of these hormones, the United States submits, the European Communities has chosen to prohibit the sale of meat from any animal to which these hormones have been administered for growth promotion purposes, whether or not there is any residue of these hormones found in that meat.272 The European Communities does not submit that
267See para. 4.78.
268We recall in this respect that the Codex standards in dispute have in any event been adopted in July 1995, i.e., subsequent to the entry into force of the SPS Agreement on 1 January 1995. With respect to the timeframe for the application of the SPS Agreement in general, we refer to paragraphs 8.24-8.27.
269See para. 8.56.
270See para. 2.20.
271See para. 8.16.
272See para. 4.64 and 4.67.
WT/DS26/R/USA
Page 179
its measures are based on the Codex standards, but rather argues that, as discussed above273, there are no relevant Codex standards on which its measures in dispute need to be based.
The SPS Agreement does not explicitly define the words based on as used in Article 3.1. However, Article 3.2, which introduces a presumption of consistency with both the SPS Agreement and GATT for sanitary measures which conform to international standards, equates measures based on international standards with measures which conform to such standards. Article 3.3, in turn, explicitly relates the definition of sanitary measures based on international standards to the level of sanitary protection achieved by these measures. Article 3.3 stipulates the conditions to be met for a Member to enact or maintain certain sanitary measures which are not based on international standards.274 It applies more specifically to measures "which result in a higher level of sanitary ... protection than would be achieved by measures based on the relevant international standards" or measures "which result in a level of sanitary ... protection different from that which would be achieved by measures based on international standards". One of the determining factors in deciding whether a measure is based on an international standard is, therefore, thelevel of protection that measure achieves. According to Article 3.3 all measures which are based on a given international standard should in principle achieve the same level of sanitary protection. Therefore, if an international standard reflects a specific level of sanitary protection and a sanitary measure implies a different level, that measure cannot be considered to be based on the international standard.
We find, therefore, that for a sanitary measure to be based on an international standard in accordance with Article 3.1, that measure needs to reflect the same level of sanitary protection as the standard.275 In this dispute a comparison thus needs to be made between the level of protection reflected in the EC measures in dispute and that reflected in the Codex standards for each of the five hormones at issue.
Without limiting the possibilities of how a level of protection may be expressed for a particular substance, we consider that in the specific field of veterinary drugs (including the six hormones at issue), a level of protection can be directly linked to the amount of residues of that drug allowed either to be ingested by humans on a daily basis or to be present in a particular food.276 A level of protection can thus, inter alia, be expressed by way of setting a maximum amount of residues allowed for daily intake by humans over a lifetime (often defined as acceptable daily intake or ADI277) and (or) by way of adopting a maximum amount of residues allowed to be present in a particular food (often defined as maximum residue limit or MRL278). However, the fact that an ADI or MRL can reflect a level of protection (without stricto sensu itself being a level of protection), does not exclude, as the European
273See paras. 8.66-8.69.
274Article 3.1 explicitly refers to Article 3.3 as providing for exceptions to the general obligation to base sanitary measures on international standards: "... Members shall base their sanitary ... measures on international standards ..., except as otherwise provided for in this Agreement, and in particular in paragraph 3" (emphasis added).
275We recall, however, that, given the exceptions provided for in Article 3.3, the requirement that a sanitary measure reflects the same level of protection as the relevant international standard is in no way absolute. This requirement is only imposed for a measure to be based on such international standard in accordance with Article 3.1.
276The concept of "appropriate" level of protection is examined in paragraph 8.79.
277See paras. 8.59 ff.
278Ibid.
WT/DS26/R/USA
Page 180
Communities has argued, that an ADI or MRL can also be a sanitary measure in the sense of the SPS Agreement.
In this dispute, two of the international standards applicable, namely the Codex standards with respect to zeranol and trenbolone (two synthetic hormones), provide for an ADI of 0-0.5 and 0-0.02
µg/kg of body weight, respectively, and an MRL of 10 µg/kg for bovine liver and 2 µg/kg for bovine muscle for zeranol and an MRL of 10 µg/kg for bovine liver and 2 µg/kg of
-trenbolone for bovine muscle.279 These ADIs and MRLs reflect the level of protection set by the Codex standards.280 To determine whether the EC measures in dispute with respect to zeranol and trenbolone are based on these Codex standards, we need to examine whether the level of protection reflected in the EC measures is the same as the level of protection expressed by the Codex standards.281 Since the EC measures in dispute do not allow the presence of any residues of these two hormones in any meat or meat product or any of these residues to be ingested by humans (imposing what it calls a "no residue" level), the level of protection reflected in the EC measures is significantly different from the level of protection set by the Codex standards (a "no residue" level as opposed to an ADI of maximum 0.5 and 0.02 µg/kg of body weight and an MRL of 2 and 10 µg/kg for, respectively, bovine muscle and bovine liver). The EC measures in dispute, in as far as they relate to zeranol and trenbolone, are, therefore, not based on existing international standards as specified in Article 3.1.
When establishing the other three Codex standards applicable to the EC measures in dispute, Codex considered it "unnecessary" to set an ADI or MRL for residues of , testosterone and progesterone (the three natural hormones).282 The amount of residues of these hormones administered for growth promotion purposes allowed by these Codex standards is, therefore, in any event higher than zero (a maximum level of such residues has not even been prescribed; this level is hereafter referred to as an "unlimited residue level"). The EC measures in dispute, on the other hand, do not allow the presence of any residues of these three hormones administered for growth promotion purposes (again imposing what the European Communities calls a "no residue" level). The level of protection reflected in the EC measures is, therefore, significantly different from the level of protection reflected in the Codex standards (a "no residue" level as opposed to an unlimited residue level). The EC measures in dispute, in so far as they relate to
, testosterone and progesterone, are, therefore, not based on existing international standards as specified in Article 3.1.
We thus find that the EC measures in dispute (except to the extent they relate to the hormone MGA) result in a different level of sanitary protection than would be achieved by measures based on the relevant Codex standards and are, therefore, not based on existing international standards as specified in Article 3.1.
We next examine whether the EC measures with respect to five of the six hormones in dispute, which are not based on existing international standards, otherwise are consistent with the requirements of the SPS Agreement (sections 4 and 5). We then address the EC measures which relate to the sixth hormone, MGA, for which no international standard exists (section 6).
279See paras. 8.63-8.65.
280See para. 8.74.
281See para. 8.73.
282See para. 8.62.
WT/DS26/R/USA
Page 181
The fact that the EC measures for , testosterone, progesterone, zeranol and trenbolone are not based on existing international standards does not necessarily mean that those measures are inconsistent with the requirements of the SPS Agreement. Article 3.3 reads as follows:
"Members may introduce or maintain sanitary or phytosanitary measures which result in a higher level of sanitary protection than would be achieved by measures based on the relevant international standards, guidelines or recommendations, if there is a scientific justification, or as a consequence of the level of sanitary or phytosanitary protection a Member determines to be appropriate in accordance with the relevant provisions of paragraphs 1 through 8 of Article 5. Notwithstanding the above, all measures which result in a level of sanitary or phytosanitary protection different from that which would be achieved by measures based on international standards, guidelines or recommendations shall not be inconsistent with any other provision of this Agreement".
A footnote to Article 3.3, first sentence, then specifies:
"For the purposes of paragraph 3 of Article 3, there is a scientific justification if, on the basis of an examination and evaluation of available scientific information in conformity with the relevant provisions of this Agreement, a Member determines that the relevant international standards, guidelines or recommendations are not sufficient to achieve its appropriate level of sanitary or phytosanitary protection".
The concept of an "appropriate level of sanitary protection" is defined in paragraph 5 of Annex A of the SPS Agreement as:
"The level of protection deemed appropriate by the Member establishing a sanitary
... measure to protect human, animal or plant life or health within its territory".
A Note to this paragraph adds the following:
"Many Members otherwise refer to this concept as the 'acceptable level of risk' ".
For a sanitary measure to be justified under Article 3.3 the measure needs, first of all, to "result in a higher level of sanitary protection than would be achieved by measures based on the relevant international standards, guidelines or recommendations". We recall the comparison made above between the level of protection reflected in the EC measures and that implied in the Codex standards for each of the hormones at issue, in particular that the level reflected in the EC measures is different from that implied in the Codex standards.283 For purposes of our analysis under Article 3.3, we assume that the former level is higher than the latter, in line with the first sentence of Article 3.3. In addition, the sanitary measure needs to fulfil one of the following two conditions:
there is a "scientific justification" for imposing the measure, i.e., the Member imposing the measure has determined "on the basis of an examination and evaluation of available scientific information in conformity with the relevant provisions of [the SPS] Agreement, ... that the
283See para. 8.77.
WT/DS26/R/USA
Page 182
relevant international standards, guidelines or recommendations are not sufficient to achieve its appropriate level of sanitary ... protection" ("the first exception"); or
the measure is "a consequence of the level of sanitary ... protection a Member determines to be appropriate in accordance with the relevant provisions of paragraphs 1 through 8 of Article 5" ("the second exception").
However, according to the second sentence of Article 3.3, even if one of these conditions is fulfilled, the party imposing the measure must still comply with the other provisions of the SPS Agreement.
We will consider first whether either the first or the second exception outlined above is met. In doing so, we first address the relationship and difference between these two exceptions. The United States argues that both exceptions have the same effect since both refer to a situation where the basis for departing from the relevant international standard is that the international standard is not sufficient to achieve the Member's appropriate level of protection.284 The European Communities argues that the first exception is fulfilled when the international standard is inadequate, faulty or obsolete from a scientific point of view and that, according to the second exception, a Member is in any case entitled to introduce or maintain measures which aim at achieving its appropriate level of protection, to be determined in accordance with Article 5 of the SPS Agreement.285
We note that both exceptions explicitly refer to other provisions of the SPS Agreement. The first exception contains the following reference: "... on the basis of an examination and evaluation of available scientific information in conformity with the relevantprovisions of [the SPS] Agreement ..." (emphasis added). The second exception refers to "... the relevant provisions of paragraphs 1 through 8 of Article 5" (emphasis added). Article 3.3, second sentence, in turn, explicitly states that even if the sanitary measure at issue falls under one of the two exceptions of Article 3.3, first sentence, the sanitary measure in question still needs to be consistent with all provisions of the SPS Agreement other than Article 3.
We find, therefore, that, whatever the differencemight be between thetwo exceptions, a sanitary measure can only be justified under Article 3.3 if it is consistent with the requirements contained in Article 5. If we were to find that the EC measures in dispute are inconsistent with the requirements imposed by Article 5, these measures cannot be justified under Article 3.3. However, even if we find that the EC measures at issue are consistent with the requirements imposed by Article 5, this will still not be sufficient for these measures to be justified under Article 3.3 since to reach that conclusion we also need to find that the EC measures in dispute fulfil all provisions of the SPS Agreement other than Articles 3 and 5 (in casu Article 2).
We recall the findings made above on the burden of proof under the SPS Agreement286, in particular that for the obligations imposed by the SPS Agreement that are relevant to this case, the party contesting a sanitary measure (in casu the United States) bears the initial burden of proof in that it has to present a prima facie case of inconsistency with the SPS Agreement, after which the burden of proof shifts to the party imposing the measure (in casu the European Communities).
284See para. 4.91.
285See para. 4.90.
286See paras. 8.48-8.55.
WT/DS26/R/USA
Page 183
We consider that this allocation of burden to the party imposing a sanitary measure is, for the reasons set out above287, applicable to Article 3.3 and particularly justified under the first sentence thereof which contains specific requirements to be fulfilled for a Member to justify a sanitary measure which is not based on an international standard.
One purpose of the SPS Agreement, as explicitly recognized in the preamble, is to promote the use of international standards, guidelines and recommendations. To that end, Article 3.1 imposes an obligation on all Members to base their sanitary measures on international standards except as otherwise provided for in the SPS Agreement, and in particular in Article 3.3 thereof. In this sense, Article 3.3 provides an exception to the general obligation contained in Article 3.1. Article 3.2, in turn, specifies that the complaining party has the burden of overcoming a presumption of consistency with the SPS Agreement in the case of a measure based on international standards. It thereby suggests by implication that when a measure is not so based, the burden is on the respondent to show that the measure is justified under the exceptions provided for in Article 3.3.
We find, therefore, that once the complaining party provides a prima facie case (i) that there is an international standard with respect to the measure in dispute, and (ii) that the measure in dispute is not based on this standard, the burden of proof under Article 3.3 shifts to the defending party.288
Since in this dispute we have already found that there exist international standards289 and that the EC measures at issue are not based on these standards290, we find that the burden of justifying the measures in dispute under Article 3.3, and in particular under the first sentence thereof, rests on the European Communities.
In summary, in sections 3 and 4 we have found that: (i) there exist international standards, as defined in Article 3.1 and paragraph 3(a) of Annex A of the SPS Agreement, with respect to the EC measures in dispute to the extent they relate to five of the six hormones at issue (all but MGA);
(ii) the EC measures in dispute, in as far as they relate to these five hormones, are not based on these international standards, as required in Article 3.1; and (iii) the EC measures, to the extent they are not based on these international standards, can only be justified under Article 3.3 if these measures meet, inter alia, the requirements imposed by Article 5.
In the next section we will, therefore, examine whether the EC measures in dispute with respect to the five hormones at issue for which international standards exist are consistent with the requirements imposed by Article 5.
287See paras. 8.51-8.54.
288This approach is in line with established practice under GATT 1947 and GATT 1994 where, for example, the burden of proof to justify an inconsistency with another GATT provision under Article XX also rests on the defending party. See Panel Report on "Canada - Administration of the Foreign Investment Review Act", adopted on 7 February 1984, BISD 30S/140, p.164, para. 5.20; on "United States - Section 337 of the Tariff Act of 1930", adopted on 7 November 1989, BISD 36S/345, p.393, para.5.27 and on "United States - Standards for Reformulated and Conventional Gasoline", op. cit., p.38, para. 6.20. In this respect, we also note that the Report of the Appellate Body on "United States - Measures Affecting Imports of Woven Wool Shirts and Blouses from India" stated the following with respect to the burden of proof under Articles XX and XI:2(c)(i) of GATT: "Articles XX and XI:(2)(c)(i) are limited exceptions from obligations under certain other provisions of the GATT 1994, not positive rules establishing obligations in themselves. They are in the nature of affirmative defences. It is only reasonable that the burden of establishing such a defence should rest on the party asserting it" (op. cit., p.16).
289See para. 8.70.
290See para. 8.77.
WT/DS26/R/USA
Page 184
Article 5 of the SPS Agreement deals mainly291 with two separate aspects of a Member' s decision to enact or maintain a sanitary measure. These two aspects are separated in the SPS Agreement, which provides for specific rights and obligations in respect of each of them.
The first aspect relates to the exercise of assessing the risks to human, animal or plant life or health against which a sanitary measure is intended to protect. This is referred to in the SPS Agreement as risk assessment.292 With respect to food safety, the potential adverse effects (if any) related to a specific substance are established together with the probability of occurrence of any such effects.293
According to Article 5.1, a Member needs to ensure that its sanitary measures are based on an assessment of risks. The obligation to base a sanitary measure on a risk assessment may be viewed as a specific application of the basic obligations contained in Article 2.2 of the SPS Agreement which provides that "Members shall ensure that any sanitary ... measure is applied only to the extent necessary to protect human, animal or plant life or health, is based on scientific principles and is not maintained without sufficient scientific evidence ..." (emphasis added). Articles 5.1 to 5.3 sum up factors a Member needs to take into account in making this assessment of risks.294
As will be outlined below295, an assessment of risks is, at least for risks to human life or health, a scientific examination of data and factual studies; it is not a policy exercise involving social value judgments made by political bodies.
The second aspect of a Member's decision to enact or maintain a sanitary measure relates, inter alia, to the determination and application of the appropriate level of sanitary protection by that Member against the risks to human, animal or plant life or health which have been assessed in accordance with Articles 5.1 to 5.3. This aspect is commonly referred to by the parties to this dispute as an essential part of risk management.296 The Member wishing to impose a sanitary measure must decide the extent to which it can accept the potential adverse effects related to a specific substance which have been identified in the risk assessment.
Articles 5.4 to 5.6 are particularly relevant to the risk management decision. Article 5.4 establishes the objective of minimizing negative trade effects in the determination by a Member of its appropriate level of protection. Article 5.5 aims at achieving consistency in the application of the
291Except for Article 5.8 which has not been invoked in this dispute.
292See the title of Article 5 of the SPS Agreement ("Assessment of risk ...") and Annex A of the SPS Agreement providing a definition for "risk assessment".
293See the definition of "risk assessment" contained in paragraph 4 of Annex A of the SPS Agreement, discussed below in para. 8.98.
294Article 5.7 deals with cases where relevant scientific evidence is insufficient at which point a Member may, under certain conditions, take provisional sanitary measures. The European Communities has explicitly stated that this provision does not apply to the EC measures in dispute.
295See paras. 8.98 and 8.104-8.107.
296See paras. 4.97 and 4.101.
WT/DS26/R/USA
Page 185
concept of appropriate level of protection. Article 5.6, in turn, provides that the sanitary measure which is finally adopted shall not be more trade-restrictive than required to achieve the appropriate level of protection of the Member concerned. Articles 5.4 to 5.6 may be viewed as specific applications of the basic obligations provided for in Article 2.2 which, inter alia, states that "Members shall ensure that any sanitary or phytosanitary measure is applied only to the extent necessary to protect human, animal or plant life or health" (emphasis added) and Article 2.3 which provides that "Members shall ensure that their sanitary and phytosanitary measures do not arbitrarily or unjustifiably discriminate between Members where identical or similar conditions prevail ..." and that "Sanitary and phytosanitary measures shall not be applied in a manner which would constitute a disguised restriction on international trade" (emphasis added).
As will be outlined below297, the risk managementphase involves non-scientific considerations, such as social value judgments.
According to Article 5.1:
"Members shall ensure that their sanitary or phytosanitary measures are based on an assessment, as appropriate to the circumstances, of the risks to human, animal or plant life or health, taking into account risk assessment techniques developed by the relevant international organizations".
Paragraph 4 of Annex A of the SPS Agreement defines "risk assessment" with respect to contaminants (including residues of the hormones at issue) as
"the evaluation of the potential for adverse effects on human or animal health arising from the presence of ... contaminants ... in food, beverages or feedstuffs" (emphasis added).
Guided by the wording of these provisions, we consider that, in this dispute, a risk assessment carried out in accordance with the SPS Agreement should (i) identify the adverse effects on human health (if any) arising from the presence of the hormones at issue when used as growth promoters in meat or meat products, and (ii) if any such adverse effects exist, evaluate the potential or probability of occurrence of these effects.
Article 5.1 provides in general terms, without any limitation in time, that "Members shall ensure that their sanitary or phytosanitary measures are based on an assessment, as appropriate to the circumstances, of the risks ...". It does not prevent that with respect to a sanitary measure enacted before the entry into force of the SPS Agreement, the risk assessment is carried out or invoked after the entry into force of that Agreement (and thus after the enactment of the sanitary measure in question). However, the fact that a sanitary measure may be enacted before the entry into force of the SPS Agreement does not mean that, once the SPS Agreement entered into force, there is no obligation for the Member in question to base that measure on a risk assessment. 298 Moreover, the more general obligation contained in Article 2.2 of the SPS Agreement explicitly provides that "Members shall ensure that any sanitary or phytosanitary measure ... is based on scientific principles and is not maintained without sufficient scientific evidence ..." (emphasis added).
297See paras. 8.160 ff.
298Our reasoning and general finding with respect to the application ratione temporis of the SPS Agreement developed in paras. 8.24-8.27 also applies to Article 5.1 of that Agreement.
WT/DS26/R/USA
Page 186
We also recall our finding reached above on the specific burden of proof under Article 3.3299, in particular that the burden of proving that the requirements imposed by Article 3.3 (inter alia, consistency with Article 5) are met, rests with the Member imposing a sanitary measure which deviates from an international standard. Since the EC measures examined in this section (relating to all hormones in dispute other than MGA) are not based on existing international standards and need to be justified under the exceptions provided for in Article 3.3, the European Communities has the burden of proving that its measures are based on a risk assessment in accordance with Article 5.
In this respect we consider at the outset that it is for the European Communities to submit evidence before the Panel that its measures are based on a risk assessment; it is not for the Panel itself to conduct its own risk assessment on the basis of scientific evidence gathered by the Panel or submitted by the parties during the Panel proceedings.
We next examine: (i) the techniques and factors to be taken into account in a risk assessment in accordance with Article 5; (ii) whether the European Communities has demonstrated the existence of such a risk assessment; and (iii) assuming that such risk assessment exists, whether the European Communities has demonstrated that its measures at issue are based on this risk assessment.
None of the parties suggest that there are "risk assessment techniques developed by the relevant international organizations" in the sense of Article 5.1 which have to be taken into account in a risk assessment for the hormones at issue.300 We note, however, that, even though no formal decision has as yet been taken by Codex with respect to risk assessment techniques, Codex, and more particularly JECFA, has a long-standing practice with respect to the assessment of risks related to veterinary drug residues (including hormone residues). The techniques thus developed have been outlined above.301 We also note the Report of the Joint FAO/WHO Expert Consultation on the Application of Risk Analysis to Food Standards Issues convened at the request of Codex in March 1995. In this Report "risk assessment" is defined as follows:
"The scientific evaluation of known or potential adverse health effects resulting from human exposure to foodborne hazards. The process consists of the following steps:
(i) hazard identification, (ii) hazard characterization, (iii) exposure assessment, and
(iv) risk characterization. The definition includes quantitative risk assessment, which emphasizes reliance on numerical expressions of risk, and also qualitative expressions of risk, as well as an indication of the attendant uncertainties" (p.6).302
Article 5.2 provides for the factors which a Member has to take into account in the assessment of risks:
"In the assessment of risks, Members shall take into account available scientific evidence; relevant processes and production methods; relevant inspection, sampling and testing methods; prevalence of specific diseases or pests; existence of pest- or
299See paras. 8.84-8.88.
300See para. 4.113.
301See paras. 8.59 ff.
302A revised version of this definition has been accepted at the 12th session of the Codex Committee on General Principles held in November 1996 and reads as follows: "A scientifically based process consisting of the following steps: (i) hazard identification, (ii) hazard characterization; (iii) exposure assessment, and (iv) risk characterization" (Codex Alimentarius Commission, CX/GP96/3).
WT/DS26/R/USA
Page 187
disease-free areas; relevant ecological and environmental conditions; and quarantine or other treatment".
None of the parties involved in this dispute has argued that any factors listed in Article 5.2 other than the following three are relevant for an assessment of risks related to the hormones at issue in this case:
available scientific evidence;
relevant processes and production methods; and
relevant inspection, sampling and testing methods.
In particular, we note that none of the parties has argued that factors not listed in Article 5.2, such as consumer preferences, can be taken into account in a risk assessment in accordance with Article 5.
Article 5.3 sums up relevant economic factors to be taken into account in assessing risks to animal or plant life or health. Since the scope of this dispute is limited to issues of human life or health303, Article 5.3 has no application to the matter under consideration.
We finally note that the parties agree that, for the purposes of the EC measures in dispute, a risk assessment in accordance with Article 5 is a scientific process aimed at establishing the scientific basis for the sanitary measure a Member intends to take.304
The European Communities has referred to the following scientific evidence concerning the five hormones at issue:305
the 1982 Report of the EC Scientific Veterinary Committee, Scientific Committee for Animal Nutrition and the Scientific Committee for Food on the basis of the Report of the Scientific Group on Anabolic Agents in Animal Production306 ("Lamming Report");
the 1983 Symposium on Anabolics in Animal Production of the Office international des epizooties
("OIE")307 ("1983 OIE Symposium");
the 1987 Monographs of the International Agency for Research on Cancer ("IARC") on the Evaluation of Carcinogenic Risks to Humans, Supplement 7308 ("1987 IARC Monographs");
the 1988 and 1989 JECFA Reports309;
303See para. 8.19.
304See paras. 4.111 and 4.123.
305See para. 4.28.
306See para. 2.28.
307See paras. 4.16 and 4.132.
308See para. 4.129.
309See paras. 2.22-2.25.
WT/DS26/R/USA
Page 188
the 1995 European Communities Scientific Conference on Growth Promotion in Meat Production310 ("1995 EC Scientific Conference");
articles and opinions by individual scientists relevant to the use of hormones (three articles in the journal Science, one article in the International Journal of Health Service, one report in The Veterinary Record and separate scientific opinions of Dr. H. Adlercreutz, Dr. E. Cavalieri, Dr. S.S. Epstein, Dr. J.G. Liehr, Dr. M. Metzler, Dr. Perez-Comas and Dr. A. Pinter, all of whom were part of the EC delegation at our joint meeting with experts).311
The European Communities also referred to several reports of the European Parliament (the Nielsen Report of 1981, the first Collins Report of 1985, the second Collins Report of 1989 and the Pimenta Report of 1989) and opinions of the EC Economic and Social Committee (of 1981 and 1984).312 However, we recall the findings we reached above, in particular that risk assessment is a scientific process. Thus, we consider that the non-scientific reports and opinions of the European Parliament and the EC Economic and Social Committee, which evaluate the scientific and other reports submitted to them, are not part of the risk assessment process, but of the risk management process, further elaborated below.313
We next consider whether the scientific evidence and attendant evaluation referred to by the European Communities constitutes a risk assessment in the sense of Article 5. We recall that under the SPS Agreement a risk assessment should, for the purposes of this dispute, identify the adverse effects on human health arising from the presence of the specific hormones at issue when used as growth promoters in meat or meat products and, if any such adverse effects exist, evaluate the potential or probability of occurrence of these effects.314 We further recall that a risk assessment should be a scientific examination of data and studies315 and that the SPS Agreement sets out factors which need to be taken into account in a risk assessment. 316 We finally recall that no risk assessment techniques, as referred to in Article 5.1, have as yet been formally adopted by Codex.317 The Agreement does not further specify the requirements of what constitutes a risk assessment in accordance with Article 5.
We note that the European Communities has invoked several scientific reports which appear to meet these minimum requirements of a risk assessment (in particular the Lamming Report and the 1988 and 1989 JECFA Reports) and that the scientists advising the Panel seemed to consider these
310See para. 2.33.
311With respect to the scientific evidence referred to by the European Communities in relation to the potential adverse effects of the hormones at issue on the health or life of live animals, we recall our findings reached in paragraph 8.19 that the animal health arguments invoked by the European Communities exclusively relate to the EC import ban of live animals (which does not fall within the scope of the EC measures in dispute) and that these arguments need, therefore, not to be taken into account within the scope of this dispute.
312See paras. 4.21; 4.29; 4.36 and 4.40.
313See paras. 8.160 ff.
314See para. 8.98.
315See para. 8.107.
316See paras. 8.104 and 8.105.
317See para. 8.103.
WT/DS26/R/USA
Page 189
reports, from a scientific and technical point of view, to be risk assessments. 318 We shall, therefore, for the purposes of this dispute, assume that the European Communities has met its burden of demonstrating the existence of a risk assessment carried out in accordance with Article 5.
Article 5.1 requires Members to "ensure that their sanitary ... measures ... are based on an assessment ... of the risks to human ... life or health". It does not, however, specify how to determine whether a measure is based on a risk assessment. In our view, this determination has both a procedural and a substantive aspect.
Notwithstanding the fact that Article 5 does not contain specific procedural requirements for a Member to base its sanitary measures on a risk assessment, we consider that, according to the ordinary meaning of the words based on put in their context and in light of the object and purpose of Article 5319, there is a minimum procedural requirement contained in Article 5.1. In our view, the Member imposing a sanitary measure needs to submit evidence that at least it actually took into account a risk assessment when it enacted or maintained its sanitary measure in order for that measure to be considered as based on a risk assessment.
We note that in this dispute the European Communities, which has the burden of proving that it based its measures on a risk assessment, has not provided any evidence that the studies it referred to (in so far as they can be considered as part of a risk assessment) or the scientific conclusions reached therein, have actually been taken into account by the competent EC institutions either when it enacted these measures (in 1981 and 1988) or at any later point in time. We note, in this respect, that none of the preambles to the EC measures at issue mention any of the scientific studies referred to by the European Communities. These preambles only refer to the non-scientific reports and opinions of the European Parliament and the EC Economic and Social Committee, which cannot be considered as part of a risk assessment. 320
In particular with respect to the articles and opinions of individual scientists on which the European Communities focused our attention, we note that the European Communities has argued that some of these articles and opinions (in particular those published in 1995 and 1996) are what it called "new evidence" of which it was only informed during the Panel process.321 In so far as that is the case, these articles and opinions cannot for the purposes of this dispute be considered as part of a risk assessment on which the European Communities based its measures unless there would be some evidence that the competent EC institutions actually considered these articles and opinions or reexamined the potential risks related to the specific substances at issue in light of these articles and opinions. In that event, the European Communities would then have confirmed or changed its earlier conclusions on the basis of such consideration or reexamination. According to the terms of reference given to us as a dispute settlement panel, we have no mandate to reexamine the risk assessment referred to by the
318See, in general, answers by experts to Panel Questions 5 and 13, paras. 6.51 ff. and 6.142-6.147 and, in particular, Dr. Arnold at para. 6.57, Dr. McLean at para. 6.145 and opinion of Dr. Ritter, Transcripts of the joint meeting with experts of 17 February 1997, paras. 20 and 352.
319In accordance with the rules of treaty interpretation contained in Article 31 of the Vienna Convention.
320See para. 8.109.
321See paras. 4.143-4.148 and 4.192. However, later on in the Panel process, the European Communities confirmed that many of these articles and opinions constitute evidence which was already taken into account in the other studies referred to by the European Communities. In so far as this is correct, see paras. 8.132 ff.
WT/DS26/R/USA
Page 190
European Communities in light of this "new evidence", nor to make our own risk assessment. These articles and opinions, in so far as they constitute "new evidence" gathered by the European Communities during the Panel process can, therefore, from a procedural point of view, not be considered as part of a risk assessment on which the European Communities based its measures in dispute.322
For these reasons, we find that the European Communities has not met its burden of proving that it met the minimal procedural requirement contained in Article 5.1 and that, therefore, the EC measures in dispute are inconsistent with the requirements of Article 5.1.
Even if the European Communities would have fulfilled these minimum procedural requirements, there would still be a need to examine the substantive requirements contained in Article 5.1. From a substantive point of view, we consider that in this dispute we should, in accordance with the ordinary meaning of the words based on put in their context and in light of the object and purpose of Article 5323, proceed as follows to determine whether the EC measures at issue are based on a risk assessment:
(i) we need to identifythe scientific conclusions reached in each of the studies referred to by the European Communities; (ii) we need to identify the scientific conclusion reflected in the EC measures in dispute; and (iii) we need to determine whether the scientific conclusion reflected in the EC measures can be considered as being in conformity with any of those reached in the studies referred to by the European Communities.
For purposes of this analysis, we first address the studies referred to by the European Communities which specifically address one or more of the hormones in dispute when used for growth promotion purposes before examining the studies which generally relate to one or more of these hormones.
Scientific conclusions reached in the studies referred to by the European Communities which specifically address one or more of the hormones in dispute when used for growth promotion purposes
The Lamming Report came to the following conclusions with respect to the potential for adverse effects on human health arising from the presence in meat of residues of the hormones in dispute administered for growth promotion purposes:
"5.1. The Scientific Working Group is of the opinion that the use of , testosterone and progesterone [the three natural hormones in dispute] and those derivatives which readily yield the parent compound on hydrolysis after absorption from the site of application, would not present any harmful effects to the health of the consumer when used under the appropriate conditions as growth promoters in farm animals.
Evaluation of the data on 'trenbolone' and 'zeranol' [two of the three synthetic hormones in dispute] revealed that some data on the hormonal no-effect-level and the toxicology of these compounds and their metabolites are still missing.
322These articles and opinions are, from a substantive point of view, further addressed in paras. 8.130 ff.
323In accordance with the rules of treaty interpretation contained in Article 31 of the Vienna Convention.
WT/DS26/R/USA
Page 191
The Scientific Working Group considers it necessary that additional information be provided before a final conclusion can be given on trenbolone and zeranol" (emphasis added).324
The Scientific Working Group continued its review of trenbolone and zeranol up to 1985 but was dismantled by the EC Commission just before it was to submit its final report.
Some of the members of the group published, in their personal capacity, the unofficial report in "The Veterinary Record". The conclusions reached with respect to the potential for adverse effects on human health arising from the presence in meat of residues of trenbolone and zeranol administered for growth promotion purposes, were the following:
"(1) We have examined the extensive data available concerning the toxicology of trenbolone and zeranol.
"(2) We believe there is adequate evidence from both short term and long term tests that these compounds and their metabolites found as residues do not show significant genotoxic potential.
...
"(5) The levelsof trenboloneand zeranoland their major metabolites found in edible tissue, following accepted husbandry practices, are substantially below the hormonally effective doses in animal test systems and therefore do not present a harmful effect to health" (emphasis added).325
With respect to the 1983 OIE Symposium, we note that the report of the symposium consists of a book with a series of articles by individual scientists or groups of scientists. No formal scientific conclusion was agreed by all participants or by the OIE itself. The book contains, however, a foreword to the articles which includes the following statement with respect to the potential for adverse effects on human health arising from the presence in meat of residues of the three natural hormones in dispute administered for growth promotion purposes:
"The myth that all anabolics are dangerous to human health is still very much alive in many countries. It must be discredited. There is common agreement with the proof presented at this meeting that the endogenous anabolics (natural hormones) such as
17-estradiol, progesterone, and testosterone, when administered as implants in animals, are not hazardous to man" (emphasis added).
The European Communities invokes two specific quotes from individual articles submitted at the OIE Symposium in support of its measures.326 However, these quotes, which only stress the inherent uncertainties of scientific evidence and the evolving character of science, do not seem to invalidate the "common agreement" outlined in the foreword quoted above.
The 1988 JECFA Report came to the following conclusion with respect to the potential for adverse effects on human health arising from the presence in meat of residues of all three natural hormones in dispute administered for growth promotion purposes:
324See para. 2.28.
325See para. 4.35.
326See paras. 4.16 and 4.128.
WT/DS26/R/USA
Page 192
"The Committee therefore concluded that the amount of exogenous [oestradiol-17 , testosterone and progesterone] ingested in meat from treated animals would be incapable of exerting a hormonal effect, and therefore any toxic effect, in human subjects.
The Committee considered an ADI unnecessary for a hormone that is produced endogenously in human beings and shows great variation in levels according to age and sex. The Committee concluded that residues arising from the use of [any of the three natural hormones in dispute] as a growth promoter in accordance with good animal husbandry practice are unlikely to pose a hazard to human health ...
On the basis of its safety assessment of residues of [any of the three natural hormones in dispute], and in view of the difficulty of determining the levels of this hormone as a growth promoter in cattle, the Committee concluded that it was unnecessary to establish an Acceptable Residue Level" (emphasis added).327
As outlined above328, the 1988 and 1989 JECFA Reports set ADIs and MRLs for zeranol and trenbolone (two of the three synthetic hormones in dispute). JECFA reached the conclusion that their toxic effects are linked to their hormonal effects and that, therefore, a no-hormonal-effect level could be established which would ensure that residues up to such level are safe. JECFA also concluded that the safety level or ADI it thus adopted would not be exceeded at any time after proper implantation (irrespective of the withdrawal period respected).
The Steering Committee of the 1995 EC Scientific Conference came to the following conclusions with respect to the potential for adverse effects on human health arising from the presence in meat of residues of the hormones in dispute administered for growth promotion purposes:
"When used as growth promoters, the naturally occurring gonadal steroid hormones (oestrogen, testosterone and progesterone) can increase the growth rate, the proportion of lean meat to fat and the food-conversion efficiency of some species of animals ... while the residues of hormone in the resulting meat are within the normal physiological range ... To the extent that improved food-conversion efficiency reduces excretion of nitrogen and phosphorus per unit of meat production, the environmental benefits (while small) are likely to be positive. That is the basis on which the Steering Group endorses the conclusions of Working Group II that the conditions defined in countries where the use of these hormones as growth promoters is permitted are a reasonable safeguard of public health.
Several materials similar in their physiological effects to the natural sex hormones (such as trenbolone and zeranol) are also used as growth promoters in meat production ... The Steering Committee endorses the opinion of Working Group II that the definition of the MRL for trenbolone and its metabolites reached by various international committees [inter alia, JECFA] provide reasonable protection of public health" (emphasis added).329
The relevant conclusions of Working Group II of the Conference on "Assessment of Health Risk", which specifically address the safety of the hormones at issue when used as growth promoters, read as follows:
327See para. 2.22.
328See paras. 8.63-8.65.
329See para. 2.33.
WT/DS26/R/USA
Page 193
"Natural sex hormones
At present, there is no evidence for possible health risks to the consumer due to the use of natural sex hormones for growth promotion, since:
Residue levels of these substances measured in meat of treated animals fall within the physiological range observed in meat of comparable untreated animals.
The daily production of sex hormones by humans is much higher than the amounts possibly consumed from meat, even in the most sensitive humans (prepubertal children and menopausal women).
Due to an extensive first-pass metabolism, the bioactivity of ingestedhormones is low, thus providing a further safety margin.
Zeranol and trenbolone
...
At the doses needed for growth promotion, residue levels are well below the levels regarded as safe (the MRLs). There are, at present, no indications of a possible human health risk from the low levels of covalently-bound residues of trenbolone" (emphasis added).330
As can be deduced from all conclusions outlined above, none of the scientific evidence referred to by the European Communities which specifically addresses the safety of some or all of the hormones in dispute when used for growth promotion, indicates that an identifiable risk arises for human health from such use of these hormones if good practice is followed. All of the scientific studies outlined above came to the conclusion that the use of the hormones at issue (all but MGA, for which no evidence was submitted) for growth promotion purposes is safe; most of these studies adding that this conclusion assumes that good practice is followed. We note that this conclusion has also been confirmed by the scientific experts advising the Panel.331
Scientific conclusions reached in the studies referred to by the European Communities which generally relate to one or more of the hormones in dispute
The European Communities puts particular emphasis on the 1987 IARC Monographs and the articles and opinions of individual scientists outlined above.
330Ibid.
331See answers by experts to Panel Question 7, paras. 6.93 ff, including the answers by Dr. André (at para. 6.93) and, albeit slightly qualified, Dr. Lucier (at para. 6.95: "Dr. Lucier responded that, to his knowledge, there was no piece of scientific evidence to indicate that any of the six hormones in question had unequivocally caused adverse effects in humans when administered and used properly. However, there was some information available which raised concern for a slight effect on the incidence of human disease"). In this respect, we note Dr. Lucier's statement that, according to his tentative estimates, between zero and one person in a million who eat 500 grams of meat, treated with oestrogens for growth promotion purposes in accordance with good practice, per day over their lifetimes, get cancer (see Transcripts of the joint meeting with experts of 18 February 1997, paras. 742 and 819). This 0-1 in a million risk is caused by the total amount of oestrogens in treated meat (the amount of endogenous oestrogens being highly variable and, according to Dr. Lucier, already being carcinogenic), not by the small fraction thereof which is added for growth promotion purposes and which is relevant for the purposes of this dispute. Moreover, this estimate only represents a statistical range of 0 to 1 in a million, not a scientifically identified risk. The concept of "zero risk", to which this statement is closely related, is further dealt with in paras. 8.149 ff. We note, finally, that all experts confirmed that there is no evidence that particular or more significant health problems exist in countries where the hormones at issue are allowed for administration as growth promoters as compared to countries where such use is prohibited (see answers by experts to Panel Question 9, paras. 6.118-6.122).
WT/DS26/R/USA
Page 194
The 1987 IARC Monographs, in so far as they relate to human health332, contain evidence with respect to three general categories of hormones, namely oestrogens, androgens and progestins, without distinguishing the specific hormones falling within each of these categories or the natural hormones from the synthetic hormones.333 The Monographs classify oestrogens as agents which are carcinogenic (meaning that there is sufficient evidence of carcinogenicity in humans); androgens as agents which are probably carcinogenic; and progestins as agents which are possibly carcinogenic.334
We note that the scientific evidence included in these Monographs relates to the carcinogenic potential of entire categories of hormones or the hormones at issue in general. The Monographs do not consider the carcinogenic potential of these hormones when used specifically for growth promotion purposes or with respect to residue levels comparable to those present after such use.335 Moreover, the Monographs do not specifically evaluate, as is required on the basis of paragraph 4 of Annex A of the SPS Agreement336, the potential for adverse effects arising from the presence in food (in casu meat or meat products) of residues of the hormones in dispute or from residue levels comparable to those present in food.
We further note that, according to the scientific experts advising the Panel, the data and studies contained in these Monographs with respect to the carcinogenic potential of the hormones in dispute have been fully taken into account in the 1988 and 1989 JECFA Reports which, at several occasions, explicitly refer to these Monographs.337 Nowhere do the 1988 and 1989 JECFA Reports reject the conclusions reached in the 1987 IARC Monographs. On the contrary, the Monographs constitute part of the evidence on which the JECFA Reports are based. JECFA recognized that all five hormones at issue have a carcinogenic potential but concluded that this potential was linked to the hormonal effect of these hormones.338 Since JECFA considered that the additional residues of the three natural hormones present in treated meat are not capable of exerting any toxic effect, it decided that it was unnecessary to set ADIs or MRLs for these hormones.339 With respect to zeranol and trenbolone, JECFA identified a no-hormonal-effect level and adopted, on that basis, ADIs and MRLs which, if respected, would ensure the safeuse of these hormones.340 The IARC Monographs and JECFA Reports did not, therefore, reach contradictory but rather complementary scientific conclusions.
332In this respect, we recall our finding that within the scope of this dispute we need not to take into account the arguments made by the European Communities which relate to animal health (see para. 8.19).
333As outlined in paras. 2.8 and 2.9, and zeranol are oestrogens; testosterone and trenbolone are androgens and progesterone is a progestin.
334See para. 4.129.
335See also answer by Dr. Arnold to Panel Question 6, para. 6.83.
336See para. 8.98.
337See answers by experts to Panel Question 5 where all experts confirm that the JECFA Reports took into account cancer risks (para. 6.51 ff.). See, in particular, statements made by Dr. Randell, the Codex expert, at the joint meeting with experts of 17 February 1997 (Transcripts, paras. 239, 297 and 436 and by Dr. McLean at the joint meeting with experts of 18 February 1997 (Transcripts, para. 823).
338The link made by JECFA between hormonal and toxic effect does not contradict the conclusions in the IARC Monographs; on the contrary, since the latter state the following: "There is a basic incongruity between the human data and the animal carcinogenicity data. As noted earlier, however, the effects of these chemicals [inter alia, the general categories to which the hormones in dispute belong] in humans appear, at least in most cases, to be linked to the hormonal milieu" (emphasis added).
339See para. 8.62.
340See paras. 8.63-8.65.
WT/DS26/R/USA
Page 195
For these reasons, we consider that the 1987 IARC Monographs, in so far as they can be regarded as part of a risk assessment for the specific hormones at issue when used as growth promoters in the sense of Article 5.1, have been taken into account in, and do not contradict, the other studies referred to by the European Communities (in particular the 1988 and 1989 JECFA Reports) which explicitly conclude that the specific use of these hormones as growth promoters in accordance with good practice is safe.
The European Communities finally invokesseveral articles andopinionsof individualscientists. These articles and opinions deal with the carcinogenic or genotoxic potential of hormones and criticize the scientific methodology or conclusions of the other studies referred to by the European Communities. A summary of the content of some of these articles and opinions is contained in paragraphs 4.131-4.136 and 4.180. The scientific evidence included in these articles and opinions relates to the carcinogenic or genotoxic potential of entire categories of hormones or the hormones at issue in general; not when used specifically for growth promotion purposes or with respect to residue levels comparable to those present after such use.341 Moreover, these articles and opinions do not specifically evaluate, as is required on the basis of paragraph 4 of Annex A of the SPS Agreement, the potential for adverse effects arising from the presence in food (in casu meat or meat products) of residues of the hormones in dispute or from residue levels comparable to those present in food.
Of the articles and opinions invoked, the European Communities has put much emphasis on the opinion of Dr. Liehr, a scientist with the EC delegation, on "Potential Genotoxicity of Hormones". We note the statement by Dr. Liehr himself that the evidence he submitted is only based on tests carried out at elevated doses of oestrogens and that the relevance of these high dose effects to potential risks related to the low levels of oestrogens in meat from growth promoted animals has not yet been evaluated.342 Moreover, we recall that this opinion only applies to one of the hormones in dispute, namely, and does not address the use of this hormone as a growth promoter.
We further note that, according to the Codex expert advising the Panel, most of the evidence contained in these articles and opinions and the potential risks addressed therein were already evaluated and taken into account in the 1988 and 1989 JECFA Reports.343 Indeed, in the event these articles and opinions should be considered as evidence which was, as the European Communities itself argued at the end of the Panel proceedings, not "new" but was already taken into account in the 1988 or 1989 JECFA Reports, the Lamming Report or the 1995 EC Scientific Conference, these articles and opinions would then not invalidate or contradict the scientific conclusions reached in these other studies, which specifically address the use of the hormones in dispute for growth promotion purposes, but rather constitute part of the evidence on which these studies are based.
We also note that, even if these articles and opinions could be considered as scientific evidence which is part of a risk assessment for the specific hormones at issue when used as growth promoters in the sense of Article 5.1 and which was not yet considered in the other studies invoked by the European Communities, the scientific experts advising the Panel were of the view that this evidence, as it stands today, does not invalidate or contradict the scientific conclusions reached in the other studies invoked
341See para. 2.22.
342See para. 4.143. In this respect, we also note the following statements by Dr. Liehr: "... we have no genotoxicity data at low levels" and "... I agree ... that at the moment this is an interesting hypothesis and I have never labelled it more than a hypothesis and I also agree ... that many pieces are missing ..." (Transcripts of the joint meeting with experts of 17 February 1997, para. 330). The fact that Dr, Liehr repeatedly called for a more thorough scientific examination of the risks related to oestrogens and for more data, reveals that the studies he provided to us do not yet contain conclusive evidence of an identifiable risk.
343See Transcripts of the joint meeting with experts of 17 February 1997, para. 297. See also answers by experts to Panel Question 5, paras. 6.51 ff.
WT/DS26/R/USA
Page 196
by the European Communities which specifically relate to the safety of the hormones at issue when used for growth promotion purposes:
Dr. McLean:
"Some of the new data that has been submitted particularly relies upon in vivo and in vitro carcinogenicity testing and also some of the mutagenicity testing but I do not believe that it is any more significant than the sort of data that was available at the time the original appraisal was made [in the 1988 and 1989 JECFA Reports]".344
Dr. Arnold:
"... from my point of view, the significant new evidence which has been produced since that Committee [the 32nd and 34th meetings of JECFA] did not invalidate the basic conclusions and therefore I still am feeling very comfortable with the conclusions although I admit that a lot of new evidence has been produced by the scientific community". 345
Dr. Ritter:
"The nature of scientific interpretation is that legitimate bona fide knowledgeable scientists may reach different conclusions from the same set of data. But I think the consensus opinion of that Conference [the 1995 EC Scientific Conference] was that the weight of evidence used then continues to prevail now, and that the assessments and conclusions drawn then are still consistent with the available information now".346
"I agree that there is further work that is indicated. I agree that statements made by scientists, such as Dr. Liehr, will continue to contribute to our understanding, but I also agree that the totality of evidence, re-evaluated as recently as December 1995 [by the 1995 EC Scientific Conference], suggests that the way in which these substances are used and the residues which they produce, do not constitute a risk to human health".347
Dr. Lucier:
"There is a group of scientists [including Dr Liehr, a scientist with the EC delegation] who are looking at the role of oxidated damage and genotoxicity of oestrogens ... But for this narrow purpose that we are talking about today, about the influence of this on additional risk from oestrogens from eating, consuming meat containing oestrogens from growth promoted animals, it doesn't have too much consequence. ... I would come up with the same answer either case; there would be no difference in the risk. So I think in that respect whether or not oestrogen is genotoxic, has less consequence then what we talked about up to this point in time".348
344Transcripts of the joint meeting with experts of 17 February 1997, para. 4
345Ibid., para.356. 346Ibid., para.352. 347Ibid., para.424.
348The other scientist advising the Panel, Dr. André did not explicitly express his view on this issue but did not contest the statements made by the other scientists. Dr. McLean has furthermore answered in the affirmative when asked whether he believed that the Codex standards are fully adequate to address the problem (Transcripts of the joint joint meeting with experts of 17 February 1997, para. 9).
WT/DS26/R/USA
Page 197
For these reasons, we find that the European Communities has not demonstrated that the scientific evidence it referred to, which generally addresses the safety of some or all of the hormones in dispute, would indicate that an identifiable risk arises for human health from the use of these hormones for growth promotion purposes if good practice is followed. In this respect we recall that all scientific experts advising the Panel confirmed this conclusion and stated that, as of today, no scientific evidence is available which concludes that an identifiable risk arises from the use of any of the hormones at issue for growth promotion purposes in accordance with good practice.349
The finding we thus make does, of course, not exclude that future scientific developments could require modifications to the scientific conclusions reached in the studies referred to by the European Communities.
Scientific conclusion reflected in the EC measures
The European Communities bans the use for growth promotion purposes of any of the hormones in dispute, including the use of these hormones in accordance with good practice. During the Panel proceedings it has made clear that it considers any residue level of these hormones to be unsafe for human health, setting its level of protection at a "zero residue" level. The scientific conclusion reflected in the EC measures in dispute is thus that the use of the hormones in dispute for growth promotion purposes, even in accordance with good practice350, poses an identifiable risk to human health.
The conformity of the scientific conclusion reflected in the EC measures with the scientific conclusions reached in the studies referred to
In our view, the scientific conclusion reflected in the EC measures in dispute, i.e., that the use of the hormones in dispute for growth promotion purposes, even in accordance with good practice, is not safe, does not conform to any of the scientific conclusions reached in the evidence referred to by the European Communities. All the evidence referred to by the European Communities which specifically relates to the use of the hormones at issue for growth promotion purposes concludes that the use of these hormones as growth promoters in accordance with good practice is safe.351 Moreover, none of the evidence referred to by the European Communities which generally deals with one or more of the hormones in dispute contradicts this conclusion. 352 The EC import ban of meat and meat products from animals treated with any of the five hormones at issue for growth promotion purposes, allegedly necessary to protect human health, in so far as it also applies to meat and meat products from animals treated with any of these hormones in accordance with good practice, is, therefore, not based on the scientific evidence submitted to the Panel.
The European Communities, however, submits the following additional arguments (sections 5 and 6). We note that these arguments have not been supported by scientific evidence other than the evidence examined above. We consider it nonetheless appropriate to examine whether these arguments demonstrate that the EC measures in dispute are, from a substantive point of view, based on a risk assessment in accordance with Article 5.1.
349See footnote 331.
350Since, according to the experts advising the Panel (answers by experts to Panel Question 3, paras. 6.32 ff), any use of the hormones in dispute will always leave some residue level, albeit a very small one, the administration of these hormones in accordance with good practice will also leave some residue and thus not achieve the EC "zero residue" level of protection.
351See para. 8.124.
352See para. 8.134.
WT/DS26/R/USA
Page 198
General categories of risks invoked by the European Communities
The European Communities argues that it has based its ban on the existence of the following categories of risks related to the hormones at issue: (i) risks arising from the nature and mode of action of the hormones; (ii) risks arising from the action of metabolites; (iii) risks arising from the action of combinations (or cocktails) of hormones and from multiple exposure of humans; (iv) risks arising from problems related to detection and control of hormones; (v) risks arising from the administration and use of hormones; and (vi) risks arising from various other parameters, in particular the inherent limits to science.353
The United States argues that the European Communities has never performed an appropriate assessment of these alleged risks and has, in any event, not relied on, nor put forward, any assessment of these risks that could serve as a basis for the EC ban.354
We recall that the European Communities has not referred to any scientific evidence, other than that examined above355, in which the categories of risks put forward by the European Communities have been assessed and that none of the scientific evidence referred to by the European Communities reached the conclusion that any of the hormones in dispute when administered for growth promotion purposes in accordance with good practice has an adverse effect on human health.356
Moreover, with respect to the alleged risks related to the nature and mode of action of the hormones in dispute (including carcinogenicity and long-term effects) and the action of metabolites or combinations of and multiple exposure to these hormones (i.e., the first three categories of risks invoked by the European Communities), we note the statements of the scientific experts advising the Panel that the available data relating to these risks have all been taken into account by the JECFA Reports and/or the Lamming Report and/or the 1995 EC Scientific Conference.357 All three reports concluded that these risks do not materialize if the hormones in dispute are used as growth promoters in accordance with good practice. The European Communities has not provided any evidence to the contrary. The EC import ban of meat and meat products from animals treated with any of the five hormones at issue for growth promotion purposes, in so far as it also applies to meat and meat products from animals treated with any of these hormones in accordance with good practice, is, therefore, not based on an assessment of these categories of risks.
In addition, with respect to the alleged risks arising from problems related to the detection, control, administration and use of the hormones in dispute (i.e., the fourth and fifth category of risks invoked by the European Communities), we note that the European Communities has not referred to evidence, other than that outlined above358, in which an assessment is made of the possible adverse health effects related to the potential abuse of these specific hormones when used for growth promotion purposes. The European Communities has restricted itself to pointing out the condition contained in many of the scientific conclusions mentioned above, namely that the safety of the hormones is to a
353See para. 4.126.
354See para. 4.110. 355See paras. 8.119 ff. 356See para. 8.137.
357See answers by experts to Panel Question 5, paras. 6.51 ff., in particular the answers by Dr. Arnold, Dr McLean and Dr Ritter and the opinions of Dr. Randell, paras. 6.166 (see also Transcripts of the joint meeting with experts of 17 February 1997, paras. 239, 297, 374 and 436 and of 18 February 1997, paras. 730 and 823).
358See paras. 8.119 ff.
WT/DS26/R/USA
Page 199
certain extent conditional upon their administration in accordance with good practice359, without further providing an assessment of the potential adverse effects related to non compliance with such practice.
We further note that the European Communities argues that if it were to allow the use of the three natural hormones at issue it would encounter special problems in inspecting, sampling or testing for residues of these hormones in meat. These problems relate, according to the European Communities, to the inherent characteristics of the natural hormones which make them indistinguishable from natural hormones present endogenously in meat or present in meat subsequent to therapeutic or zootechnical use of these hormones.360
We recall that, in this dispute, the factors which can be taken into account in a risk assessment under Articles 5.1 and 5.2 are limited to "available scientific evidence" and "relevant inspection, sampling and testing methods".361 To the extent that the problems in inspecting (sampling and testing) natural hormones would actually pose a risk and could thus, arguably, be taken into account as a risk arising from "relevant inspection, sampling and testing methods" in the sense of Article 5.1, we consider that the European Communities encounters the same problems in inspecting for natural hormones under its current regime. The EC ban of natural hormones used for growth promotion purposes, combined with its tolerance for these hormones when used for therapeuticor zootechnical purposes or when present endogenously in meat and other foods, would seem to cause more problems in inspecting for banned natural hormones than a regime where the use of all natural hormones would be allowed in combination with, for example, a maximum residue or tolerance level for all natural hormones in any meat regardless of the origin and use of these hormones. Indeed, only under the EC current regime does the problem of how to distinguish between endogenous and added natural hormones arise; under a regime with an MRL or tolerance level for all natural hormones there would be no need to distinguish endogenous from added natural hormones.
With respect to the alleged risks related to the control (or, in other words, the abuse) of the hormones at issue (both natural and synthetic), we further note that even though a Member would seem to be able to take into account risks arising from difficulties of inspecting, sampling or testing which are specific to a particular substance in a particular food, the "relevant inspection, sampling and testing methods" referred to in Article 5.2, do not seem to cover the general problem of control (such as the problem of ensuring the observance of good practice) which can exist for any substance. The risks related to the general problem of control do not seem to be specific to the substance at issue but to the economic or social incidence related to a substance or its particular use (such as economic incentives for abuse). These non-scientific factors should, therefore, not be taken into account in risk assessment but in risk management. Moreover, even if these factors could be takeninto account in a risk assessment, we note that the European Communities has not provided convincing evidence that the control (or the prevention of abuse) of the hormones in dispute is more difficult than the control of other veterinary drugs the use of which it allows. It has neither provided evidence that control would be more difficult under a regime where the hormones in dispute were allowed under specific conditions than under the current EC regime where the hormones in dispute when used as growth promoters are banned. The experts advising the Panel made clear that the potential for abuse under both regimes would be comparable, some noting that abuse would probably occur more frequently under a regime where the
359However, we note, in this respect, the statements made by some of the scientific experts advising the Panel that even if good practice is not followed the use of the five hormones at issue will in many cases still be safe (see answers by experts to Panel Question 17, paras. 6.172-177, in particular the answers by Dr. Arnold, Dr. McLean and Dr. Ritter and opinions of Dr. McLean, Dr. Randell and Dr. Arnold, Transcripts of the joint meeting with experts of 17 February 1997, respectively at paras. 3, 26 and 166). We also recall the conclusions reached by JECFA that the MRLs (i.e., the levels which are set on the basis of good practice) it established for zeranol and trenbolone are far higher than the safety levels for these hormones.
360See para. 4.113.
361See para. 8.105.
WT/DS26/R/USA
Page 200
hormones are banned compared to one allowing the controlled use of prescribed products in predetermined dosages with well-defined educational programmes, good communication between the different actors involved and appropriate penalties for misuse.362 In this context, we note, therefore, that banning the use of a substance does not necessarily offer better protection of human health than other means of regulating its use.
In this respect, we finally note that for the three natural hormones in dispute, the European Communities has, for control purposes, adopted MRLs and thereby accepted tolerance levels which are higher than the "zero residue" level reflected in the measures in dispute. In so doing, the European Communities itself seemingly confirms the scientific conclusions reached in all the scientific evidence examined above, namely that residues of these hormones, including when used as growth promoters, are safe below a certain level, and contradicts the conclusion reflected in the EC measures in dispute, namely that only a "zero residue" level ensures the protection of EC consumers.
For all the reasons outlined above363, we find that the EC import ban of meat and meat products from animals treated with any of the five hormones at issue for growth promotion purposes, in so far as it also applies to meat and meat products from animals treated with any of these hormones in accordance with good practice, is not based on an assessment of the fourth or fifth category of risks invoked by the European Communities.
In the sixth general category of risks invoked by the European Communities (risks arising from various other parameters), the European Communities argues that none of the studies it referred to as part of a risk assessment proves beyond doubt or concludes in an unqualified manner that the presence of residues of the hormones in dispute in meat or meat products present no risk whatsoever. The European Communities refers, inter alia, to the conclusions of the 1988 JECFA Report which state that residues arising from the hormones at issue used as growth promoters are only unlikely to pose a hazard to human health and to the basic premise of JECFA recommendations which aim at establishing standards which correspond to a no appreciable or no significant risk increase due to the exposure to the substances in question and not to a zero risk increase. The European Communities apparently considers, therefore, that this residual risk, albeit minute and not appreciable, constitutes the risk (derived from a risk assessment) on which the EC ban is based in accordance with Article 5.1, arguing that, according to EC risk management, risk other than zero is not acceptable.364
The United States argues that science can never prove beyond doubt that there is no risk and can only be used to determine whether there is a risk associated with the use of a particular substance; it cannot eliminate the possibility that a potential risk may be found in the future. According to the United States, the SPS Agreement does not allow measures to be maintained without scientific evidence until such time as science proves "beyond doubt" that there is no risk.365
We recall the conclusions we reached above on burden of proof, in particular that the European Communities has, with respect to its measures which deviate from international standards, the burden of proving the existence of a risk assessment (and, derived therefrom, an identifiable risk) on which
362See, for example, answers by Dr. André and Dr. Ritter to Panel Question 14, paras. 6.149 and 6.155 and opinions of Dr. Arnold, Dr. Lucier and Dr. André, Transcripts of the joint meeting with experts of 17 February 1997, paras. 269 and 274 and of 18 February 1997, paras. 829 and 830
363See paras. 8.143-8.147.
364See para. 4.200.
365See para. 4.46.
WT/DS26/R/USA
Page 201
the EC measures in dispute are based. It is not, in this dispute, for the United States to prove that there is no risk.
We further note that, according to scientists advising the Panel, science can never provide a certainty, i.e. exclude once and for all that a specific substance can ever have adverse health effects.366
In this respect we also note that the sixth category of risks invoked by the European Communities is, as stated by the scientific experts advising the Panel367 and admitted by the European Communities368, not identifiable and that, therefore, these risks can a priori not be assessed by scientists (as required in Article 5.1). In this sense, these potential risks, which are present for any substance (also for substances or uses of substances allowed in the European Communities), are only the consequence of science not being capable of assuring that no risks will ever arise from a substance.
We finally note that the EC objective of "zero risk" cannot be achieved in practice; not even under the EC ban itself since the European Communities cannot guarantee that there is a zero probability that illegal use of the hormones at issue will occur. Moreover, this "zero risk" objective cannot, as further examined below369, in any case be achieved for the three natural hormones in dispute since the European Communities allows the ingestion of these same hormones occurring endogenously in meat and other foods as well as the use of these hormones for therapeutic or zootechnical purposes.
The EC ban on the use of the hormones in dispute for growth promotion purposes is, therefore, not based on an assessment of the sixth and final category of risks invoked by the European Communities.
For these reasons, we find that the EC import ban of meat and meat products from animals treated with any of the five hormones at issue for growth promotion purposes, in so far as it also applies to meat and meat products from animals treated with any of these hormones in accordance with good practice, is not based on an assessment of any of the six general categories of risks invoked by the European Communities.
The precautionary principle
The European Communities also invokes the precautionary principle in support of its claim that its measures in dispute are based on a risk assessment. To the extent that this principle could be considered as part of customary international law and be used to interpret Articles 5.1 and 5.2 on the assessment of risks as a customary rule of interpretation of public international law (as that phrase is used in Article 3.2 of the DSU), we consider that this principle would not override the explicit wording of Articles 5.1 and 5.2 outlined above, in particular since the precautionary principle has been incorporated and given a specific meaning in Article 5.7 of the SPS Agreement. We note, however, that the European Communities has explicitly stated in this case that it is not invoking Article 5.7.
We thus find that the precautionary principle cannot override our findings made above, namely that the EC import ban of meat and meat products from animals treated with any of the five hormones
366See, for example, opinions of Dr. Arnold, para. 6.25 and Dr. Ritter, para. 6.92. In this respect, we note that the SPS Agreement explicitly deals with situations where there is scientific uncertainty regarding risks related to a substance, in Article 5.7 (discussed in paras. 8.248 ff.), but that the European Communities has not invoked this provision in this case.
367Ibid.
368See para. 4.197.
369See paras. 8.186 ff.
WT/DS26/R/USA
Page 202
at issue for growth promotion purposes, in so far as it also applies to meat and meat products from animals treated with any of these hormones in accordance with good practice, is, from a substantive point of view, not based on a risk assessment.
In summary, in this section we have found that, even assuming that the European Communities has demonstrated the existence of a risk assessment in accordance with Article 5, it has not fulfilled the minimal procedural requirements contained in Article 5.1 to base its sanitary measures on a risk assessment. We have also found that, even if it would have fulfilled these minimal procedural requirements, the European Communities has not met its burden of proving that its measures in dispute, in so far as they also ban the import of meat and meat products from animals treated with any of the five hormones at issue for growth promotion purposes in accordance with good practice, are, from a substantive point of view, based on a risk assessment. The EC measures in dispute, in so far as they relate to five of the six hormones at issue for which international standards exist, are, therefore, inconsistent with the requirements of Article 5.1. The fact that these measures are not based on existing internationalstandards (contrary to Article 3.1)370 cannot, therefore, be justifiedunder Article 3.3 which includes as one of the requirements for justification, consistency with Article 5.1. The EC measures, in so far as they relate to five of the six hormones at issue for which international standards exist, are, therefore, also inconsistent with the requirements of Article 3.1.
We recall that there is a distinction between risk assessment which is a scientific examination and risk management which involves social value judgments.371 Once the risks have been assessed, i.e., once the risks and their probability of occurrence identified, a Member will need to decide, on the basis of its own value judgments, whether it can accept these risks. In so doing a Member sets its "appropriate level of sanitary protection". The determination and application of the appropriate level of protection by a Member is part of risk management.
We recall the definition of "appropriate level of sanitary protection", namely: "The level of protection deemed appropriate by the Member establishing a sanitary
... measure to protect human, animal or plant life or health ..." (paragraph 5 of Annex
A of the SPS Agreement; emphasis added).
We also note the wording of Article 5.5, further examined below:
"... in the application of the concept of appropriate level of sanitary ... protection against risks to human life or health, or to animal and plant life or health, each Member shall avoid ..."
Guided by the wording of these provisions and the object and purposes of the SPS Agreement, we consider that if there is no scientific evidence of an identifiable risk, there is no basis on which to adopt a measure to achieve a level of sanitary protection under the SPS Agreement, except as provided in Article 5.7. If this were not the case, i.e., if a Member could adopt a level of protection and implementing sanitary measures even if it did not provide scientific evidence of an identifiable risk, no effect would be given to the obligation contained in Article 5 to base sanitary measures on an assessment of risks. This approach would undermine the wording and object and purpose of the SPS Agreement.
370See para. 8.77.
371See paras. 8.91 ff.
WT/DS26/R/USA
Page 203
We have found above372 that the European Communities has not provided evidence of an identifiable risk related to the presence of five of the six hormones at issue for which international standards exist when these hormones are used for growth promotion purposes in accordance with good practice. Accordingly, the European Communities has not established the existence of any identifiable risk against which the EC measures at issue, in so far as they also ban the use of the five hormones when used as growth promoters in accordance with good practice, can protect human life or health. Since we considered above373 that the adoption of a sanitary measure presupposes the existence of an identifiable risk (except as provided in Article 5.7), it is not possible for the European Communities to ban the use of these hormones as growth promoters in accordance with good practice.
However, even if we would have found that the European Communities met its burden of proving that its measures are based on an assessment of risks in accordance with Articles 5.1 and 5.2 and even if, for that reason, the European Communities could have adopted a measure to achieve its appropriate level of protection against these risks, there would still be a need to examine whether the determination and application of this level of protection is consistent with Articles 5.4 to 5.6. We will, therefore, next examine these provisions.
The parties to this dispute seem to agree that the establishment of an "appropriate level of sanitary protection" by a Member is a sovereign act, namely, as the definition in paragraph 5 of Annex A of the SPS Agreement provides, the level of protection "deemed appropriate by the Member establishing a sanitary ... measure" (emphasis added). As outlined above374, we note, however, that Members have agreed, in exercising their sovereign right to set their appropriate levels of protection, to observe the provisions of the SPS Agreement, in particular Articles 5.4 and 5.5 thereof. Furthermore, in choosing a measure to achieve that appropriate level of protection Members have agreed to observe the provisions of Articles 2, 5.1 to 5.3 and 5.6.
We finally recall our findings reached above on the specificburden of proof under Article 3.3.375 In particular, we found that the burden of proving that the requirements imposed by Article 3.3 (inter alia, consistency with Article 5) are met, in order to justify a sanitary measure which deviates from an international standard, rests with the Member imposing that measure. Since the EC measures examined in this section (relating to all hormones in dispute other than MGA) are not based on existing international standards and need to be justified under the exceptions provided for in Article 3.3, the European Communities bears the burden of proving that the determination and application of its level of protection is consistent with Articles 5.4 to 5.6.
Article 5.4 provides the following:
"Members should, when determining the appropriate level of sanitary or phytosanitary protection, take into account the objective of minimizing negative trade effects" (emphasis added).
Guided by the wording of Article 5.4, in particular the words "should" (not "shall") and "objective", we consider that this provision of the SPS Agreement does not impose an obligation. However, this
372See para. 8.137.
373See para. 8.161. 374See paras. 8.161 ff. 375See paras. 8.85 ff.
WT/DS26/R/USA
Page 204
objective of minimizing negative trade effects has nonetheless to be taken into account in the interpretation of other provisions of the SPS Agreement.
Article 5.5 provides the following:
"With the objective of achieving consistency in the application of the concept of appropriate level of sanitary or phytosanitary protection against risks to human life or health, or to animal and plant life or health, each Member shall avoid arbitrary or unjustifiable distinctions in the levels it considers to be appropriate in different situations, if such distinctions result in discrimination or a disguised restriction on international trade. Members shall cooperate in the Committee, in accordance with paragraphs 1, 2 and 3 of Article 12, to develop guidelines to further the practical implementation of this provision. In developing the guidelines, the Committee shall take into account all relevant factors, including the exceptional character of human health risks to which people voluntarily expose themselves" (emphasis added).
We note, in this respect, the basic obligations contained in Article 2.3:
"Members shall ensure that their sanitary and phytosanitary measures do not arbitrarily or unjustifiably discriminate between Members where identical or similar conditions prevail, including between their own territory and that of other Members. Sanitary and phytosanitary measures shall not be applied in a manner which would constitute a disguised restriction on international trade" (emphasis added).
Article 2.3 deals, in general terms, with sanitary measures which discriminate between Members or which are applied in a manner which would constitute a disguised restriction on international trade. Article 5.5, on the other hand, deals more specifically with distinctions in levels of protection (which will normally be reflected in one or more sanitary measures) which result in discrimination or a disguised restriction on international trade.
We consider that the first part of the first sentenceof Article 5.5 ("With the objective of achieving consistency in the application of the concept of appropriate level of sanitary or phytosanitary protection against risks to human life or health, or to animal and plant life or health ..."; emphasis added), unlike the second part, does not impose an obligation upon Members. Consistency is not imposed as an obligation but as an objective which nonetheless has to be taken into account in the interpretation of Article 5.5.
We further note that the Committee on Sanitary and Phytosanitary Measures, established by Article 12 of the SPS Agreement to "provide a regular forum for consultations", has been given a mandate by Article 5.5, second sentence, to "develop guidelines to further the practical implementation of this provision" and, in so doing, needs to "take into account all relevant factors, including the exceptional character of human health risks to which people voluntarily expose themselves". No such guidelines have to date been developed. However, considering the mandatory wording of the second part of the first sentence of Article 5.5 ("each Member shall avoid arbitrary or unjustifiable distinctions in the levels it considers to be appropriate in different situations ..."; emphasis added) and the existence of the basic obligations contained in Article 2.3376, we find that the lack of guidelines by the Committee in no way limits the legally binding nature of the second part of the first sentence of Article 5.5.
376See para. 8.168.
WT/DS26/R/USA
Page 205
The United States argues that the European Communities fails to justify the following differences in regulatory treatment: (i) a ban on natural and synthetic hormones when used for growth promotion purposes as opposed to not setting any limit for residues of the natural hormones present endogenously in untreated meat and other foods (such as milk, cabbage, broccoli or eggs) and residues of these hormones when used for therapeutic or zootechnical purposes; and (ii) a ban on the hormones in dispute when used for growth promotion purposes as opposed to allowing the use of the veterinary drug carbadox as a growth promoter in swine production. Only with respect to the last mentioned difference in treatment does the United States invoke and address Article 5.5.377
The European Communities rejects these claims, arguing that it does not make distinctions in its levels of protection for different situations and that, even if it were to make such distinctions, these distinctions are justified and do not result in discrimination or a disguised restriction on international trade.378
We next examine the elements that must be assessed to determine if a Member' s sanitary measure does not conform to the requirements of the second part of the first sentence of Article 5.5. The relevant part of Article 5.5 reads as follows:
"each Member shall avoid arbitrary or unjustifiable distinctions in the levels it considers to be appropriate in different situations, if such distinctions result in discrimination or a disguised restriction on international trade".
The first element contained in Article 5.5 is that the Member concerned adopts different appropriate levels of sanitary protection in "different situations". The second element is that the distinction in levels of protection for the different situations is "arbitrary or unjustifiable". The third element is that the distinction in levels of protection results in "discrimination or a disguised restriction on international trade". In order to find a sanitary measure to be inconsistent with Article 5.5 all three elements need to be present.
As to the first element, the words "different situations" have been interpreted by the parties as follows. The European Communities argues that"different situations" only covers different situations for the same residue or for different residues where the adverse health effect is the same. According to the European Communities, "different situations" cannot mean that the same level of protection must be applied to similar health hazards, whatever their nature or severity, coming from similar substances. The United States argues that the "different situations" referred to in Article 5.5 of necessity must be comparable situations. It argues, for example, that the purported health risk from carbadox and the hormones in dispute, when used for growth promotion, is in both instances, cancer in humans and that, therefore, the different situations invoked by the United States are comparable.379
We note that both parties in dispute agree that the scope of "different situations" contained in Article 5.5 includes situations which deal with the same substance as well as situations which involve the same adverse health effect. For this reason, considering the lack of guidelines by the Committee on Sanitary and Phytosanitary Measures and without further defining or limiting the scope of "different situations", we find that, for the purposes of this dispute, we can compare situations where the same substance or the same adverse health effect is involved as "different situations" in the sense
377See para. 4.220.
378See para. 4.218.
379See para. 4.220.
WT/DS26/R/USA
Page 206
of Article 5.5. For the sake of clarity in this particular case, we will hereafter refer to such "different situations" as "comparable situations" since these situations need to be compared for the purposes of Article 5.5 and are, therefore, "comparable".
The second element contained in Article 5.5 is that the distinction in levels of protection for comparable situations is "arbitrary or unjustifiable".
The United States argues that, in the absence of any principle or criterion that accounts for the selection of differing levels of sanitary protection, the distinction in the levels of protection is arbitrary and unjustifiable.380 The European Communities argues that Article 5.5 clearly states that "arbitrary or unjustifiable" distinctions are to be avoided if, and only if, they result in discrimination or a disguised restriction on trade. If they do not result in discrimination or a disguised restriction on trade, the European Communities concludes, they are not prohibited by Article 5.5.381
The third element contained in Article 5.5 is that the distinction in level of protection results in "discrimination or a disguised restriction on international trade".
The United States has not presented a claim with respect to the term "discrimination"; only with respect to the term "disguised restriction on international trade". The United States argues that "a disguised restriction on international trade" is present in the context of Article 5.5 where a Member claims a legitimate basis for the difference in the chosen levels of protection being compared, but where instead the differing levels of protection are being employed for commercial reasons to restrict trade.382 The European Communities argues that the measures in dispute do not result in discrimination and that the fact that sanitary measures affect imports is not a sufficient reason to claim that they restrict trade, or even less, that they discriminate.
We note, first of all, the relation between this third element of "discrimination or a disguised restriction on international trade" in Article 5.5 and the basic obligations contained in Article 2.3 of the SPS Agreement providing that:
"Members shall ensure that their sanitary and phytosanitary measures do not arbitrarily or unjustifiably discriminate between Members where identical or similar conditions prevail, including between their own territory and that of other Members. Sanitary and phytosanitary measures shall not be applied in a manner which would constitute a disguised restriction on international trade" (emphasis added).
We also note the relation between these two provisions and the language of the chapeau of Article XX of GATT which reads as follows:
"Subject to the requirement that such measures are not applied in a manner which would constitute a means of arbitrary or unjustifiable discrimination between the countries where the same conditions prevail, or a disguised restriction on international trade, nothing in this Agreement shall be construed to prevent the adoption or enforcement by any contracting party of measures: ...." (emphasis added).
With respect to the meaning of "discrimination" and "a disguised restriction on international trade" we recall the conclusions reached by the Appellate Body in its Report on "United States -
380Ibid.
381See para. 4.218.
382See para. 4.221.
WT/DS26/R/USA
Page 207
Standards for Reformulated and Conventional Gasoline" where the terms "arbitrary or unjustifiable discrimination" and "a disguised restriction on international trade", contained in the chapeau of Article XX of GATT, were examined as follows:
" 'Arbitrary discrimination', 'unjustifiable discrimination' and 'disguised restriction' on international trade may, accordingly, be read side-by-side; they impart meaning to one another. It is clear to us that 'disguised restriction' includes disguised discrimination in international trade. It is equally clear that concealed or unannounced restriction or discrimination in international trade does not exhaust the meaning of 'disguised restriction'. We consider that 'disguised restriction', whatever else it covers, may properly be read as embracing restrictions amounting to arbitrary or unjustifiable discrimination in international trade taken under the guise of a measure formally within the terms of an exception listed in Article XX. Put in a somewhat different manner, the kinds of considerations pertinent in deciding whether the application of a particular measure amounts to 'arbitrary or unjustifiable discrimination', may also be taken into account in determining the presence of a 'disguised restriction' on international trade. The fundamental theme is to be found in the purpose and object of avoiding abuse or illegitimate use of the exceptions to substantive rules available in Article XX" (original emphasis).383
We further recall the Appellate Body Report on "Japan - Taxes on Alcoholic Beverages" where the Appellate Body found that for an internal tax measure to be inconsistent with the second sentence of Article III:2 of GATT, three separate issues must be addressed so as to give full meaning to the text and context of this provision: (i) the products need to be "directly competitive or substitutable";
(ii) they need to be "not similarly taxed"; and (iii) the dissimilar taxation needs to be "applied ... so as to afford protection to domestic production".384 The Appellate Body found that the panel had erred in blurring the distinction between the second and third issue by equating dissimilar taxation (i.e., tax difference above a de minimis level) with the separate and distinct requirement of demonstrating that the tax measure "affords protection to domestic production".385 The Appellate Body then concluded the following:
"As previously stated, a finding that 'directly competitive or substitutable products' are 'not similarly taxed' is necessary to find a violation of Article III:2, second sentence. Yet this is not enough. The dissimilar taxation must be more than de minimis. It may be so much more that it will be clear from that very differential that the dissimilar taxation was applied "so as to afford protection". In some cases, that may be enough to show a violation . In this case, the Panel concluded that it was enough. Yet in other cases, there may be other factors that will be just as relevant or more relevant to demonstrating that the dissimilar taxation at issue was applied 'so as to afford protection' ... And, in every case, a careful, objective analysis, must be done of each and all relevant facts and all relevant circumstances to determine 'the existence of protective taxation'. Although the Panel blurred its legal reasoning in this respect,
383Appellate Body Report on "United States - Standards for Reformulated and Conventional Gasoline", adopted on 20 May 1996, WT/DS2/AB/R, p.25.
384Appellate Body Report on "Japan - Taxes on Alcoholic Beverages", adopted on 1 November 1996, WT/DS8/AB/R, p.24.
385Panel Report on "Japan - Taxes on Alcoholic Beverages", adopted on 1 November 1996, WT/DS8/R, paras. 6.33-34 and 7.1(ii).
WT/DS26/R/USA
Page 208
nevertheless we conclude that it reasoned correctly that in this case, the Liquor Tax Law is not in compliance with Article III:2" (emphasis added).386
We consider the reasoning in both Appellate Body Reports to be equally relevant to the relationship between the three elements contained in Article 5.5. All three elements impart meaning to one another. Nevertheless, in order to give effect to all three elements contained in Article 5.5 and giving full meaning to the text and context of this provision, we consider that all three elements need to be distinguished and addressed separately. However, we also agree that in some cases where a Member enacts, for comparable situations, sanitary measures which reflect different levels of protection, the significance of the difference in levels of protection combined with the arbitrariness thereof may be sufficient to conclude that this difference in levels of protection "result[s] in discrimination or a disguised restriction on international trade" in the sense of Article 5.5 (in line with the argument that the magnitude of the very differential of a dissimilar taxation may be enough to conclude that a dissimilar taxation is applied so as to afford protection, as provided for in the second sentence of Article III:2 of GATT).
We next examine, in light of the three elements of Article 5.5 outlined above, the distinctions in levels of sanitary protection allegedly made by the European Communities which have been invoked by the United States. In order to conduct our consideration of this dispute under Article 5.5 in the most efficient manner, we first address the alleged differences in treatment provided by the European Communities for the natural hormones in dispute. In this examination we compare the treatment of these hormones when used as growth promoters with both the treatment of these hormones occurring endogenously in meat and other foods (such as milk, cabbage, broccoli or eggs) and when used for therapeutic or zootechnical purposes. In a second step, we address the alleged differences in treatment provided by the European Communities for the natural hormones in dispute as opposed to that of the synthetic hormones at issue. In a third step, we address the alleged differences in treatment provided by the European Communities for all hormones in dispute (other than MGA) when used as growth promoters as opposed to that for carbadox, an antimicrobial growth promoter.
Comparable situations with different levels of sanitary protection
This examination involves a comparison of the levels of protection for the same substance, namely, respectively, , testosterone and progesterone, in different situations depending on the origin or use of that substance. Since we have found above that we can compare situations where the same substance is involved as "different" situations (which we refer to as "comparable" situations for the purposes of this dispute) in the sense of Article 5.5387, we find that the treatment of the three natural hormones in dispute when used for growth promotion purposes as opposed to the treatment of these hormones which (i) occur endogenously in meat and other foods and (ii) which have been administered for therapeutic or zootechnical purposes, constitute comparable situations in the sense of Article 5.5.
The European Communities argues that the origin of these hormones (whether endogenously produced or exogenously administered) causes these hormones to be different, claiming that the hormones present endogenously in meat and other foods have formed part of the human diet for centuries. We note, however, that the European Communities did not submit any evidence in support of its claim that these hormones have different effects. Moreover, all scientific experts advising the Panel have
386Appellate Body Report on Japan - "Taxes on Alcoholic Beverages", adopted on 1 November 1996, WT/DS8/AB/R, p.30.
387See para. 8.176.
WT/DS26/R/USA
Page 209
concluded that residues of the three natural hormones present endogenously in meat and other foods or administered for therapeutic or zootechnical purposes are qualitatively the same as the residues of these hormones administered for growth promotion and that if any differences between these hormones could exist (e.g., differences in pathways taken or metabolites), these differences would in any event not have consequences for the potential adverse effects of these hormones.388 Therefore, even if these hormones would not be totally identical substances, they pose, in any event, the same adverse health effect and can, therefore, according to our finding made above389, be considered as comparable situations for the purposes of Article 5.5.
We next examine whether the European Communities has adopted a different level of protection for these comparable situations.
With respect to the three natural hormones administered for growth promotion purposes, the European Communities argues that its level of sanitary protection is concerned only with added hormones; in other words, the European Communities does not consider that it is acceptable to expose consumers to any hormones in their food over and above the levels which occur in nature, as any such additional exposure could be a hazard to health.390
The appropriate level of protection set by the European Communities for natural hormones present endogenously in meat and other foods or administered for therapeutic or zootechnical purposes is an unlimited residue level.391 In other words, the European Communities has not adopted any maximum residue level for these categories of natural hormones.392 With respect to oestradiol-17 when used for therapeutic or zootechnical purposes, this unlimited residue level of protection has recently been confirmed by the European Communities when it adopted the conclusions reached in the 1988 JECFA Report and classified this hormone, when used for therapeutic or zootechnical purposes, as a substance for which MRLs are unnecessary.393 With respect to the two other natural hormones in dispute, progesterone and testosterone, when used for therapeutic or zootechnical purposes, no final decision has as yet been taken by the competent EC authorities. 394
We thus find that the level of protection adopted by the European Communities for the three natural hormones in dispute when used for growth promotion and that adopted for the same hormones
(i) occurring endogenously in meat and other foods and (ii) used for therapeutic or zootechnical purposes, is different ("no residue" level as opposed to an unlimited residue level) and that, therefore, distinctions in levels of protection for these comparable situations exist in the sense of the first element of Article 5.5.
388See answers by experts to Panel Questions 2 and 4, paras. 6.22-6.31 and 6.40-6.50 and opinions of all experts advising the Panel to an oral question asked by the US representative at the joint meeting with experts of 17 February 1997, Transcripts, paras. 77, 79, 84, 85 and 87.
389See para. 8.176.
390Ibid. 391Ibid.
392With respect to hormones administered for therapeutic treatment, the European Communities argues that in practice no residues of these hormones will be ingested by consumers because animals undergoing such treatment are not allowed to be slaughtered. However, the fact is that the European Communities has decided that the adoption of MRLs for these hormone residues is unnecessary and has thus not set any residue limit (see para. 8.190 in fine).
393See answer by Dr. Arnold to Panel Question 5, para. 6.55.
394Ibid., para. 6.56.
WT/DS26/R/USA
Page 210
"Arbitrary or unjustifiable" difference in levels of sanitary protection
We next examine whether these two distinctions in levels of protection are "arbitrary or unjustifiable". We first address the distinction made between the three natural hormones when used as growth promoters and the same hormones occurring endogenously in meat and other foods. We then examine the distinction made between the three natural hormones when used as growth promoters and the same hormones when used for therapeutic or zootechnical purposes.
Natural hormones used as growth promoters as opposed to those occurring endogenously in meat and other foods. The European Communities has not provided any reasons, other than those addressed above, why it has adopted a different level of protection for the residues of these two categories of natural hormones. The European Communities has, in particular, not provided any evidence that the risk related to the natural hormones used as growth promoters is in any way higher than the risk related to natural endogenous hormones. We also recall that the experts advising the Panel concluded that both categories of hormones (either exogenously administered to animals or endogenously present in animals, meat, other foods or human beings) pose the same potential adverse effects.395
In this respect we further recall the conclusion reached in the 1988 JECFA Report that the total residue level of natural hormones in meat from treated animals (i.e., the combination of natural hormones endogenously present and those added for growth promotion) falls well within the physiological range of levels found in meat from untreated animals, which levels vary according to the sex and age of the animal.396 We also note that, according to data submitted to the Panel, the residue level of natural hormones in many natural products (such as eggs and soya oil) is much higher than the level of residues of these hormones administered for growth promotion as well as the total residue level of these hormones in treated meat.397 In this respect, we further note that Codex has also established MRLs for substances which endogenously occur in natural products.398
With respect to the potential difficulties in detecting the presence of natural hormones used as growth promoters, we refer to our conclusions reached above399, namely that only under the current EC regime (a ban) does the problem arise of how to distinguish between endogenous and added natural hormones whereas under a regime where one would allow the use of these hormones (with, for example, an MRL or tolerance level for all natural hormones) there would be no need to distinguish endogenous from added natural hormones. We also note that, in any event, the problem of detection would be the same for both natural hormones used as growth promoters and those occurring endogenously in
395See para. 8.187 and, in particular, the footnote thereto.
396See para. 8.62. This conclusion has not been contested by either the parties or experts advising the Panel. For example, according to data submitted to the Panel (the accuracy of which has not been disputed by the parties), the residue level of testosterone in 500 grams of untreated bull meat is 1,560 nanograms as opposed to 35 nanograms in 500 grams of meat from a heifer implanted with testosterone.
397For example, according to data submitted to the Panel (the accuracy of which has not been disputed by the parties), the residue level of equivalents in a 50 to 60 grams hen's egg is 1,750 nanograms (in 10 ml of soybean oil, 20,000 nanograms) as opposed to 11.4 nanograms in 500 grams of steer meat implanted with
and 75 nanograms in 500 grams of untreated cow meat. In other words, the oestrogen content of 1 hen's egg is equivalent to 76.5 kg of implanted steer beef. In this respect, we also note that a pre-puberty male child naturally produces 41,000 nanograms of oestrogens in 24 hours, an adult man 136,000 nanograms and a pregnant women 20,000,000 nanograms.
398For example, MRLs have been adopted by Codex for (naturally occurring) cyanide in cassava flour and in Gari (a cassava product).
399See para. 8.145.
WT/DS26/R/USA
Page 211
meat and other foods (since the scientific experts stated that both are qualitatively the same400) and can, therefore, not justify a different treatment.
We finally note that even if some form of justification could be deduced from the arguments submitted by the European Communities, such could not, in any event, justify so significant a difference in levels of protection between a "no residue" level for natural hormones administered for growth promotion and an unlimited residue level for natural hormones endogenously present in meat and other foods.
We thus find that the European Communities has not met its burden of proving that the distinction it makes in levels of protection for residues of the three natural hormones in dispute when administered for growth promotion purposes and residues of the same natural hormones present endogenously in meat and other foods is justifiable and that, therefore, this particular distinction in levels of protection is "arbitrary or unjustifiable" in the sense of the second element contained in Article 5.5.
Natural hormones used as growth promoters as opposed to those used for therapeutic or zootechnical purposes. The European Communities argues that the use of the natural hormones for therapeutic and zootechnical purposes occurs on a small scale, is subject to very strict conditions (such as administration by a veterinarian and strict withdrawal periods) and normally only involves cattle intended for breeding, not for slaughter; whereas the use of these hormones as growth promoters occurs on a much larger scale and is more difficult and costly to control.401 These differences in use and control, the European Communities argues, ensure that any risk related to the therapeutic or zootechnical use of these hormones is prevented and that in practice a level of "no residue" is achieved, as is the case for the use of these hormones as growth promoters. For these reasons, the European Communities concludes, the distinction in levels of protection is justified.402
We note that, according to scientific experts advising the Panel, zootechnical use of these hormones can occur on a large scale and at regular intervals, namely each year for oestrus synchronization of entire herds.403 Moreover, even when these hormones are used for therapeutic or zootechnical purposes, all parties and scientific experts advising the Panel agree that some residue level, albeit a very small one, will always remain in the meat when the treated animal is eventually slaughtered.404 Therefore, a "no residue" level cannot in practice be achieved when these hormones are used for therapeutic or zootechnical purposes.
However, since we have already concluded that the difference in levels of protection imposed in the European Communities for the three natural hormones when used for growth promotion purposes as opposed to those present endogenously in meat and other foods cannot be justified, we consider it unnecessary to decide whether or not the distinction made by the European Communities between natural hormones used as growth promoters and those used for therapeutic or zootechnical purposes is justified.
400See para. 8.187.
401See para. 4.71.
402See para. 4.69.
403See answers by experts to Panel Questions 19 and 20, paras. 6.183-6.193.
404Ibid. See also opinions of all experts advising the Panel on a question by the US representative at the joint meeting with experts of 17 February 1997, Transcripts, paras. 90, 91, 93 and 95 and answers by experts to Panel Question 3,
paras. 6.32-6.39 and para. 4.68.
WT/DS26/R/USA
Page 212
Difference which results in "discrimination or a disguised restriction on international trade"
We next examine whether the difference in levels of protection between residues of the three natural hormones in dispute when administered for growth promotion purposes and residues of the same natural hormones present endogenously in meat and other foods, results in discrimination or a disguised restriction on international trade within the meaning of the third element of Article 5.5.405
We recall the considerations made above on the relationship between the three elements contained in Article 5.5.406 We recall, in particular, that in some cases the significance of the difference in levels of protection for comparable situations combined with the arbitrariness thereof may be sufficient to conclude that this difference in levels of protection results in "discrimination or a disguised restriction on international trade".
In this case, we note, firstly, the significance of the difference in levels of protection for the three natural hormones in dispute when administered for growth promotion purposes and residues of the same hormones present endogenously in meat and other foods, namely a "no residue" level as opposed to an unlimited residue level. We recall, secondly, that the European Communities has not provided any plausible justification for this significant difference. We note, finally, that this difference in levels of protection results in an import ban (on meat and meat products treated with any of the three natural hormones in dispute for growth promotion purposes) which restricts international trade. For these reasons, we find that the difference in levels of protection imposed by the European Communities for the three natural hormones in dispute when administered for growth promotion purposes and those present endogenously in meat and other foods, results in "discrimination or a disguised restriction on international trade" in the sense of Article 5.5.
We consider that this finding is further supported by two additional factors. Firstly, we recall some of the objectives (other than the protection of human health) that the European Communities had in mind when enacting or maintaining the EC ban on the use of the natural hormones for growth promotion purposes, as stated in the preambles of the EC measures in dispute407 and in the reports of the European Parliament and the opinions of the EC Economic and Social Committee referred to by the European Communities408, namely harmonizing the regulatory schemes of the different EC Member
405Since we made no finding on the justifiability of the difference in levels of protection for the natural hormones in dispute when administered as growth promoters and those administered for therapeutic or zootechnical purposes, we do not address, for that additional difference in levels of protection, the third element of Article 5.5.
406See paras. 8.182-8.184.
407Such as preambles 5 and 6 to Directive 88/146/EEC which state, inter alia, the following:
"Whereas the administration to farm animals of certain substances having a hormonal action is at present regulated in different ways in the Member States; ... whereas this divergence distorts the conditions of competition in products that are the subject of common market organizations and is a serious barrier to intra-Community trade;
"Whereas these distortions of competition and barriers to trade must therefore be removed by ensuring that all consumers are able to buy the products in question under largely identical conditions of supply and that these products correspond to their anxieties and expectations in the best possible manner; whereas such a course of action is bound to bring about an increase in consumption of the product in question".
408Such as the Nielsen Report of 1981, the Collins Reports of 1985 and 1989 and the Pimenta Report of 1989 of the European Parliament and the opinions of the EC Economic and Social Committee of 1981 and 1984, outlined in paras. 2.28-
2.33. See also Judgment of the Court of Justice of the European Communities in the case C-331/88, "The Queen v. The Minister for Agriculture, Fisheries and Food and the Secretary of State for Health, ex parte: Fedesa and Others", 1990,
ECR I-4023, at p. I-4065, para. 25: "... the material made available to the Court ... shows that the possibility of a reduction
(continued...)
WT/DS26/R/USA
Page 213
States, thereby removing competitive distortions and barriers to intra-Community trade in beef, and bringing about an increase in the consumption of beef, thereby reducing the internal beef surpluses and providing more favourable treatment to domestic producers.
Secondly, we note that before the EC ban came into force, the percentage of animals treated with any of the hormones in dispute was significantly lower in the European Communities than in the United States. At that time, according to the European Communities, only four or five EC member States allowed the use of some of these hormones. One member State (the United Kingdom) which has replied to the EC's request for information on this issue, indicated that "anecdotal evidence suggests that growth promoting hormones may have been used [in] up to 40% of UK cattle prior to the ban".409 On the other hand, according to a table provided by the United States, an average of 70 per cent of all US cattle were, at that time, treated with one or more of these hormones.410 By banning the internal sale and import of meat treated with natural hormones for growth promotion purposes (which represents a significantly higher proportion of the total US meat supply than of the total European Communities meat supply) but continuing to allow any level of residues of these natural hormones present endogenously in meat, the European Communities favoured the consumption of domestic meat and, therefore, de facto discriminates against US meat in favour of EC meat. In this sense, the difference in levels of protection in the European Communities for residues of hormones present endogenously in meat and other foods and residues of the same natural hormones when administered for growth promotion purposes could be said to result in "discrimination or a disguised restriction on international trade".
We thus find that the European Communities has not met its burden of justifying the distinction it makes in levels of protection for residues of the three natural hormones in dispute administered for growth promotion purposes and residues of the same natural hormones present endogenously in meat and other foods, in light of the three elements contained in Article 5.5, and that, therefore, the EC measures in dispute, in so far as they relate to the three natural hormones at issue, are inconsistent with the requirements imposed in Article 5.5.
We next examine the alleged different treatment provided by the European Communities for, on the one hand, two of the three synthetic hormones in dispute for which international standards exist (zeranol and trenbolone)411 and, on the other hand, the natural hormones in dispute occurring endogenously in meat and other foods.412
408(...continued)
in surpluses was indeed taken into consideration during the process leading to the adoption of the directive [in casu, Directive 88/146/EEC] ...".
409See para. 4.15, footnote 35.
410See para. 4.9, footnote 25.
411As mentioned above, the hormone MGA, for which no international standard exists, will be dealt with in a separate section in paragraphs 8.250 ff.
412Since we made no finding on the difference in levels of protection for natural hormones administered as growth promoters and those administered for therapeutic or zootechnical purposes (see para. 8.200), we do not address the alleged difference in levels of protection for synthetic hormones administered as growth promoters and natural hormones administered for therapeutic or zootechnical purposes.
WT/DS26/R/USA
Page 214
Comparable situations with different levels of sanitary protection
In this examination we compare different substances, namely, respectively, zeranol and oestradiol-
17and trenbolone and testosterone. As outlined above413, both synthetic hormones at issue are produced to mimic one of the natural hormones in dispute (zeranol mimics
and trenbolone mimics testosterone). However, both parties in this dispute and the experts advising the Panel agree that the situations thus compared involve at least the same adverse health effect, namely carcinogenicity. 414
Since we decided above that we can compare situations where the same adverse health effect is involved as "different" situations (which we refer to as "comparable" situations for the purposes of this dispute) in the sense of Article 5.5415, we find that the treatment of zeranol and trenbolone and the treatment of the natural hormones in dispute which occur endogenously in meat and other foods, are comparable situations in the sense of the first element of Article 5.5.
We next examine whether the European Communities has adopted different levels of protection for these comparable situations.
With respect to zeranol and trenbolone, the European Communities adopted a "no residue" level as its appropriate level of protection.416 As outlined above417, the level of protection in the European Communities for the natural hormones present endogenously in meat and other foods is an unlimited residue level.
We thus find that the levels of protection adopted by the European Communities for residues of zeranol and trenbolone and that for residues of the natural hormones in dispute which occur endogenously in meat and other foods are different ("no residue" level as opposed to an unlimited residue level) and that, therefore, a distinction in levels of protection for these comparable situations exists in the sense of the first element of Article 5.5.
"Arbitrary or unjustifiable" difference in levels of sanitary protection
We next examine whether this difference in levels of protection is "arbitrary or unjustifiable". The European Communities has not provided convincing evidence that the synthetic hormones (which mimic the natural hormones) are inherently more dangerous than the natural hormones.418 Most of the evidence referred to by the European Communities to prove potential risks relates to the natural hormones, in particular .419 According to the scientists advising the Panel, synthetic
413See para. 8.4.
414See answers by experts to Panel Question 4, paras. 6.38-6.48. See also Transcripts of joint meeting with experts of 17 February 1997, paras. 342-348.
415See para. 8.176.
416See para. 4.93.
417See para. 8.190.
418See also answers by experts to Panel Question 4, paras. 6.40-6.50 and Transcripts of the joint meeting with experts of 17 February 1997, para. 348, where Dr. Lucier stated that in his opinion residues of synthetic hormones are of more concern than those of natural hormones because the risks related to the natural hormones are already there due to those occurring endogenously in the body, whereas residues of the synthetic hormones are new to the body. This difference does not, however, relate to the inherent characteristics of both categories of hormones, but to the fact that one category occurs endogenously in humans, whereas the other does not.
419See paras. 4.143 and 4.145.
WT/DS26/R/USA
Page 215
hormones can also be better detected and controlled than natural hormones.420 Moreover, the fact that ADIs and MRLs exist for zeranol and trenbolone and not for the natural hormones does not, according to the experts advising the Panel, per se mean that the latter are inherently safer than the former since the international standards for both synthetic and natural hormones reflect essentially the same level of protection, namely a "no appreciable risk" level.421 Therefore, even if there could be valid reasons to subject the natural hormones to a treatment different from the synthetic hormones422, the European Communities has not provided justification for so significant a difference in levels of protection as between a "no residue" level (for the synthetic hormones at issue) and an unlimited residue level (for the natural hormones endogenously present in meat and other foods). We recall, in particular, that the European Communities has not provided evidence that the use of zeranol or trenbolone for growth promotion purposes in accordance with good practice (for example, the Codex MRLs) is unsafe.423 In other words, it has not submitted any justification for adopting a "no residue" level, instead of the Codex MRLs.
We thus find that the European Communities has not met its burden of justifying the distinction it makes in levels of protection for zeranol and trenbolone and the natural hormones in dispute which occur endogenously in meat and other foods. For these reasons, we find that the difference in levels of protection thus made by the European Communities is "arbitrary or unjustifiable" in the sense of the second element contained in Article 5.5.
Difference which results in "discrimination or a disguised restriction on international trade"
We recall the considerations made above on the relationship between the three elements contained in Article 5.5.424 We recall, in particular, that in some cases the significance of the difference in levels of protection for comparable situations combined with the arbitrariness thereof may be sufficient to conclude that this difference in levels of protection results in "discrimination or a disguised restriction on international trade".
In this case, we note, firstly, the significance of the difference in levels of protection for zeranol and trenbolone and that for the natural hormones in dispute which occur endogenously in meat and other foods, namely a "no residue" level as opposed to an unlimited residue level. We recall, secondly, that the European Communities has not provided any plausible justification for this significant difference. We note, finally, that this difference in levels of protection results in an import ban (on meat and meat products treated with zeranol or trenbolone) which restricts international trade. For these reasons, we find that the difference in levels of protection imposed by the European Communities for zeranol and trenbolone and that for the natural hormones in dispute which occur endogenously in meat and other foods, results in "discrimination or a disguised restriction on international trade" in the sense of Article 5.5.
420See answers by experts to Panel Question 22, paras. 6.199-6.202.
421See, for example, opinion of Dr. Randell, para. 6.76. We also note, in this respect, the opinion of Dr. Arnold, para. 6.45: "The potential risks arising from other than hormonal actions [related to the synthetic hormones] were qualitatively different. These risks were assessed during the review and approval process, and the approved conditions of use eliminated all unacceptable risks".
422See, for example, the Codex standards which have adopted MRLs for the synthetic hormones and not for the natural hormones.
423See para. 8.15, 137.
424See paras. 8.182-8.184.
WT/DS26/R/USA
Page 216
We consider that this finding is further supported by the two additional factors outlined above425, which are equally valid for the distinction in levels of protection made by the European Communities for zeranol and trenbolone and the natural hormones in dispute which occur endogenously in meat and other foods.
We thus find that the European Communities has not met its burden of justifying the distinction it makes in levels of protection for zeranol and trenbolone and the natural hormones in dispute which occur endogenously in meat and other foods, in light of the three elements contained in Article 5.5, and that, therefore, the EC measures in dispute, in so far as they relate to zeranol and trenbolone, are inconsistent with the requirements imposed in Article 5.5.
We next examine the alleged different treatment provided by the European Communities for five of the six hormones in dispute (all but MGA) when used for growth promotion purposes and carbadox. We recall that this agent is an antimicrobial growth promoter used as a feed additive in swine production.
Comparable situations with different levels of sanitary protection
In this examination we compare different substances. However, both parties in this dispute and the experts advising the Panel agree that the situations thus compared involve the same adverse health effect, namely carcinogenicity. 426
Since we have found above that we can compare situations where the same adverse health effect is involved as "different" situations (which we refer to as "comparable" situations for the purposes of this dispute) in the sense of Article 5.5427, we find that the treatment of the five hormones at issue when used as growth promoters as opposed to that of carbadox are comparable situations in the sense of the first element of Article 5.5.
We next examine whether the European Communities has adopted a different level of protection for these comparable situations.
The United States argues that the EC level of protection for the hormones at issue when used for growth promotion is different from that for carbadox. The United States submits that with respect to the hormones at issue when used for growth promotion purposes the European Communities has suggested that its appropriate level of protection is "zero risk". With respect to carbadox, the United States argues that it is clear from the fact that the European Communities permits the use of carbadox and the sale and consumption of meat from animals to which carbadox has been administered, that the European Communities is choosing an appropriate level of protection that is less stringent than for the hormones involved in this dispute.428
With respect to the hormones in dispute when used for growth promotion purposes, the European Communities adopted a "no residue" level as its appropriate level of protection.429 With respect to carbadox, the European Communities argues that, even though these substances are allowed, strict
425See paras. 8.204 and 8.205.
426See answers by experts to Panel Question 11, paras. 6.128-6.136 and para. 4.220.
427See para. 8.176.
428See para. 4.220.
429See para. 4.93.
WT/DS26/R/USA
Page 217
controls, specific characteristics of these substances and the way they are administered, ensure that no residue levels will remain in treated pigs when slaughtered and that, therefore, in the European Communities in practice the same level of protection applies to carbadox as the level adopted for the hormones in dispute, namely a "no residue" level.430
We note that the European Communities allows the use of carbadox as a growth promoter in pigs and has not set any MRL for that substance. The European Communities thus, in principle, accepts an unlimited residue level of these substances in pork meat. Moreover, we recall that, contrary to what the European Communities argues, a "no residue" level cannot be achieved in practice when use of the substance concerned is allowed (even under strict conditions) since there will always be some residue level of the substance or a metabolite, albeit a very small one, left in the meat, even after a long period of time.431 We consider, for these reasons, that the European Communities cannot reasonably claim that its level of protection for carbadox is a "no residue" level.
We thus find that the level of protection adopted by the European Communities for the hormones at issue when used for growth promotion purposes as opposed to that adopted for carbadox is different (a "no residue" level as opposed to an unlimited residue level) and that, therefore, a distinction in the levels of protection for these comparable situations exists in the sense of the first element of Article 5.5.
"Arbitrary or unjustifiable" difference in levels of sanitary protection
We next examine whether this distinction in levels of protection is "arbitrary or unjustifiable" in the sense of the second element of Article 5.5.
The United States argues that the risks related to carbadox are at least as serious as those related to the use of the hormones in dispute as growth promoters.432 It refers to the 36th JECFA Report of 1991 which could not set an ADI for carbadox but did adopt MRLs for one of its metabolites, as opposed to the 32nd and 34th JECFA Reports of 1988 and 1989 which only adopted ADIs and MRLs for zeranol and trenbolone and considered MRLs for the three natural hormones to be unnecessary. The United States submits that none of the arguments put forward by the European Communities justifies a stricter level of protection for the hormones in dispute (which, according to the United States, are safe) than the level of protection for carbadox (a substance which, according to the United States, may pose serious risks).433
The European Communities claims that the distinction in levels of protection is justified on the following grounds: (i) carbadox is not a hormone; (ii) carbadox only indirectly acts as a growth promoter by combating the development of bacteria and by aiding the intestinal flora of piglets, thereby also exerting preventive therapeutic effects (whereas the hormones directly act as growth promoters and have no preventive therapeutic action when used as growth promoters); (iii) carbadox is only commercially available in prepared feedstuffs (not as injections or implants) in predetermined dosages;
(iv) there are no alternatives to carbadox available which have the same therapeutic action; (v) carbadox cannot be abused since it only has growth promotion effects in piglets up to four months old and a
430EC second submission to the Panel on EC Measures Concerning Meat and Meat Products (Hormones) - Complaint by Canada, pp.9-10, paras. 35-36.
431See opinions of all experts advising the Panel in Transcripts of the joint meeting with experts of 17 February 1997, paras. 90, 91, 93 and 95 and answers to Panel Question 3, paras. 6.32-6.39.
432See para. 4.220.
433See paras. 4.220 and 4.221.
WT/DS26/R/USA
Page 218
withdrawal period of at least 28 days is fixed; and (vi) carbadox is used in such small quantities and is hardly absorbed that it leaves practically no residues at all in meat destined for human consumption. 434
We note, first of all, that the European Communities has not submitted scientific evidence in support of these alleged justifications. We next examine the six arguments put forward by the European Communities in light of the opinions of the experts advising the Panel and the arguments submitted by the United States.
With respect to the first EC argument, i.e., the fact that carbadox is a antimicrobial agent and not a hormone, the European Communities has not submitted any reason why this difference could in itself justify a different regulatory treatment in the light of their potential carcinogenic effect. We thus find that this argument does not justify the distinction in levels of protection for the five hormones at issue when used as growth promoters and carbadox.
The European Communities next argues that the five hormones at issue when used as growth promoters have no therapeutic effect on animals as opposed to carbadox which combats the development of bacteria and aids the intestinal flora of piglets. We note that, according to scientific experts advising the Panel, the hormones at issue when administered as growth promoters may also have beneficial effects (such as improved composition of the carcass upon treatment in terms of more lean meat and less fat).435 Moreover, we recall that the hormones at issue are, effectively, used for therapeuticpurposes and that such use of the three natural hormones in dispute is allowed in the European Communities. For these reasons, we consider that both the hormones in dispute and carbadox may have therapeutic effects and thus find that the second EC argument does not justify the distinction in levels of protection for the five hormones at issue when used as growth promoters and carbadox.
With respect to the third EC argument, i.e., the fact that carbadox is only commercially available in prepared feedstuffs (not as injections or implants) in predetermined dosages and is, therefore, allegedly less open for abuse, we note that one of the scientific experts advising the Panel stated that injections or implants are more accurate and reliable methods to administer growth promoters than additives in feedstuffs (because of carry-over risks from treated to untreated feed).436 The experts also stated that additives in feedstuffs pose additional risks in that they may harm the persons handling the feedstuff.437 We also recall that, according to the experts advising the Panel, the commercially available products containing any of the five hormones at issue for implantation or injection also contain predetermined dosages of these hormones.438 We thus find that the third EC argument does not justify the distinction in levels of protection for the five hormones at issue when used as growth promoters and carbadox.
Addressing the fourth EC argument that there are no alternatives to carbadox available which have the same therapeutic action, we note that one of the experts advising the Panel stated that there are readily available alternatives, such as oxytetracycline.439 We thus find that this EC argument does
434EC second submission to the Panel on EC Measures Concerning Meat and Meat Products (Hormones) - Complaint by Canada, p.9, para. 35. We considered it appropriate to also address these EC arguments in this Panel report (for our reasoning on this issue see para. 8.15, 137).
435See answers by experts to Panel Question 2, paras. 6.22-6.31, in particular answers by Dr. Arnold, Dr. McLean and Dr. Ritter. See also opinion of Dr. Lucier, Transcripts of the joint meeting with experts of 18 February 1997, para. 742 where he feels unable, as a scientist, to compare the risks related to the hormones with their potential benefits.
436See answers by experts to Panel Question 21, paras. 6.194-6.198, in particular answer by Dr. McLean at para. 6.197.
437See answers by experts to Panel Question 21, paras. 6.194-6.198, in particular answers by Dr. André and Dr. McLean at paras. 6.194 and 6.197.
438See answers by experts to Panel Question 15, paras. 6.156-6.166.
439See opinion of Dr. Arnold, para. 6.130.
WT/DS26/R/USA
Page 219
not justify the distinction in levels of protection for the five hormones at issue when used as growth promoters and carbadox.
We recall the fifth EC argument, i.e., that the potential for abuse is allegedly smaller for carbadox than for the hormones at issue since the former only exert growth promotion effects in piglets up to four months and are subject to a strict withdrawal period. We note that, according to the experts advising the Panel, there is no guarantee that the piglets treated with carbadox will not be slaughtered. Residues of this substance or its metabolites may thus enter the food chain. We also note that, as is the case for the use of carbadox in the European Communities, the use of the hormones at issue as growth promoters may also be made subject to strict conditions. We thus consider that the European Communities has not submitted evidence proving that carbadox can be more easily controlled than the five hormones at issue and find, therefore, that the fifth EC argument does not justify the distinction in levels of protection for the five hormones at issue when used as growth promoters and carbadox.
The European Communities further argues that carbadox is used in such small quantities and is hardly absorbed so that it leaves practically no residues at all in meat destined for human consumption. We recall that, according to the experts advising the Panel, once a substance has been administered to an animal there will always be some residue level of this substance or a metabolite left, albeit a very small one, in the meat of that animal.440 We further note that, according to the 36th JECFA Report of 1991 which assessed the risks related to carbadox, not only carbadox itself (for which no ADIs could be established) but also one of its metabolites, quinoxaline-2-carboxylic acid (for which an MRL was adopted) may present a health risk. We finally note that, according to the scientific experts441, residue levels of the hormones at issue will also rapidly decline after administration to an animal or ingestion by humans. For these reasons, we find that the sixth EC argument does not justify the distinction in levels of protection for the five hormones at issue when used as growth promoters and carbadox.
The European Communities finally submits that it authorizes the use of about 10,000 to 15,000 veterinary medicinal products and that the fact that the United States limits its claim under Article 5.5 to only one substance, proves that the European Communities has already achieved a remarkable degree of consistency in its levels of sanitary protection. The European Communities also informed the Panel that the EC Council, by decision of 26 February 1996, already took action on its own initiative to review carbadox. We consider that these arguments do not justify the distinction which is currently still made by the European Communities in levels of protection for the five hormones at issue when used as growth promoters and carbadox. On the contrary, these arguments suggest an acknowledgment by the European Communities that the distinction in levels of protection it currently makes may not be justified and will be reviewed.
For the above reasons, we find that the European Communities has not met its burden of justifying the distinction it makes in levels of protection for the five hormones at issue when used as growth promoters and carbadox and that the European Communities has, a priori, not met its burden of justifying so significant a distinction between a "no residue" level for the hormones at issue when used as growth promoters and an unlimited residue level for carbadox. We find, therefore, that the distinction in levels of protection thus made by the European Communities is "arbitrary or unjustifiable" in the sense of the second element contained in Article 5.5.
440See para. 8.199 and, particularly with respect to carbadox, opinion of Dr. Lucier, Transcripts of the joint meeting with experts of 18 February 1997, para. 275.
441See, for example, opinion of Dr. Lucier, para. 6.37 and Dr. Arnold, Transcripts of the joint meeting with experts of 17 February 1997, para. 91.
WT/DS26/R/USA
Page 220
Difference which results in "discrimination or a disguised restriction on international trade"
The United States submits that the distinction made by the European Communities in levels of protection for the hormones at issue when used as growth promoters and carbadox results in a disguised restriction on international trade in the sense of the third element contained in Article 5.5. According to the United States, the European Communities swine industry (where growth promoters are allowed) is relatively more efficient and market-oriented than the European Communities beef industry (where growth promoters are banned), a sector where efficiency is not that important because of domestic price support measures, import protection and export subsidies. When banning hormones, the United States concludes, the European Communities clearly wanted to reduce beef supplies, a concern not present in the pork sector where the European Communities wanted, on the contrary, to preserve competitiveness and maintain export markets.442
We recall that the three elements contained in Article 5.5 all impart meaning to one another and that in some cases the significance of the difference in levels of sanitary protection for comparable situations combined with the arbitrariness of thereof, may be sufficient to conclude that this difference in levels of protection results in "discrimination or a disguised restriction on international trade".443
In this case, we note, firstly, the significance of the difference in levels of protection for the five hormones at issue when used as growth promoters and carbadox, namely a "no residue" level as opposed to an unlimited residue level. We recall, secondly, that the European Communities has not provided any plausible justification for this significant difference. We note, finally, that this difference in levels of protection results in an import ban (on meat and meat products treated with any of these five hormones at issue) which restricts international trade. For these reasons, we find that the difference in levels of protection imposed by the European Communities for the five hormones at issue when used as growth promoters and carbadox, results in "discrimination or a disguised restriction on international trade" in the sense of Article 5.5.
We consider that this finding is further supported by the two additional factors outlined above444, which are equally valid for the distinction in levels of protection made by the European Communities for the five hormones at issue when used as growth promoters and carbadox.
We finally note that there is another factor which indicates that the distinction in treatment made by the European Communities for the hormones at issue when used as growth promoters and carbadox results in "discrimination or a disguised restriction on international trade". That is the fact that the hormones at issue, which are banned in the European Communities, are used for growth promotion in the bovine meat sector where the European Communities seemingly wants to limit supplies445 and is arguably less concerned with international competitiveness, whereas carbadox, which is allowed in the European Communities, is used for growth promotion in the pork meat sector where the European Communities has no domestic surpluses and where international competitiveness is a higher priority.
We thus find that the European Communities has not met its burden of justifying the distinction it makes in levels of protection for five of the six hormones at issue (all but MGA) when used as growth promoters and carbadox, in light of the three elements contained in Article 5.5, and that, therefore,
442See para. 4.221.
443See paras. 8.182-8.184.
444See paras. 8.204 and 8.205.
445See also para. 8.204.
WT/DS26/R/USA
Page 221
the EC measures in dispute, in so far as they relate to these five hormones in dispute, are inconsistent with the requirements imposed in Article 5.5.
In summary, in this section we have found that the EC measures in dispute, both in so far as they relate to the two synthetic hormones (zeranol and trenbolone) and the three natural hormones at issue for which international standards exist, are inconsistent with the requirements contained in Article 5.5. The fact that the EC measures in dispute are not based on existing international standards (contrary to Article 3.1) can, for that reason, not be justified on the basis of Article 3.3. The EC measures, in so far as they relate to five of the six hormones at issue for which international standards exist, are, therefore, also inconsistent with the requirements of Article 3.1.
Article 5.6 reads as follows:
"Without prejudice to paragraph 2 of Article 3, when establishing or maintaining sanitary or phytosanitary measures to achieve the appropriate level of sanitary or phytosanitary protection, Members shall ensure that such measures are not more trade-restrictive than required to achieve their appropriate level of sanitary or phytosanitary protection, taking into account technical and economic feasibility".
A footnote to Article 5.6 states the following:
"For purposes of paragraph 6 of Article 5, a measure is not more trade-restrictive than required unless there is another measure, reasonably available taking into account technical and economic feasibility, that achieves the appropriate level of sanitary or phytosanitary protection and is significantly less restrictive to trade".
Since we found above that the EC level of protection reflected in the EC measures in dispute has been adopted in violation of Article 5.5, we do not consider it necessary to further examine whether these measures are also more trade restrictive than required to achieve that level in the sense of Article 5.6.
Article 5.7 reads as follows:
"In cases where relevant scientific evidence is insufficient, a Member may provisionally adopt sanitary or phytosanitary measureson thebasis of available pertinent information, including that from the relevant international organizations as well as from sanitary or phytosanitary measures applied by other Members. In such circumstances, Members shall seek to obtain the additional information necessary for a more objective assessment of risk and review the sanitary or phytosanitary measure accordingly within a reasonable period of time".
We recall our finding reached above on the role of the precautionary principle in the SPS Agreement, in particular that this principle would not override the explicit wording of that Agreement, inter alia, because it has been incorporated in a specific form in Article 5.7.446 In this dispute, the European Communities has explicitly stated that its measures are not provisional measures in the sense of Article 5.7. We do, therefore, not need to further examine this provision.
446See para. 8.158.
WT/DS26/R/USA
Page 222
We recall that with respect to the third synthetic hormone in dispute, MGA, no international standard exists.447 As outlined above, the European Communities is, therefore, not under an obligation to base its sanitary measure in respect of this hormone on an international standard in accordance with Article 3.1.448
However, even though no international standard exists for MGA, the EC measures in dispute relating to MGA still need to comply with the other provisions of the SPS Agreement. The United States has invokedviolations of Articles 2 and 5. Since Article 2 provides for basic rights and obligations which are further specified in Article 5, we first examine the consistency of the EC measures in dispute relating to MGA with the requirements of Article 5.449 The consistency of the EC measures relating to all hormones in dispute (including MGA) with the requirements of Article 2 will be dealt with below.450
We recall our finding reached above on the general burden of proof under the SPS Agreement451, in particular that for the obligations imposed by the SPS Agreement that are relevant to this case, the party contesting a sanitary measure (in casu the United States) bears the burden of presenting a prima facie case of inconsistency with the SPS Agreement, after which the burden of proof shifts to the party imposing the measure(in casu the European Communities). We consider that, for the reasons mentioned above452, this allocation of evidentiary burden is applicable to the obligations imposed on Members under Article 5. We recall, in particular, the wording of Article 5.1, especially the first three words thereof:
"Members shall ensure that their sanitary ... measures are based on an assessment ... of the risks ..." (emphasis added)
and the wording of the second part of the first sentence of Article 5.5 :
"... each Member shall avoid arbitrary or unjustifiable distinctions in the levels it considers to be appropriate in different situations, if ..." (emphasis added).
Therefore, in this dispute the United States has to present a prima facie case that the EC measures in dispute relating to MGA are inconsistent with the requirements of Article 5, after which the burden shifts to the European Communities to prove that it has complied with these requirements.
447See para. 8.70.
448See paras. 8.56 ff.
449We only examine the consistency of the EC measures relating to MGA with the requirements contained in Articles 5.1 to 5.3 and 5.5. With respect to the other provisions of Article 5 we refer to our considerations set out in section 5 which equally apply to the EC measures relating to MGA.
450See para. 8.271. 451See paras. 8.48 ff. 452See paras. 8.52-8.54.
WT/DS26/R/USA
Page 223
With respect to Articles 5.1 to 5.3 dealing with the requirement of a risk assessment, the United States argues that the European Communities has not submitted any risk assessment for the hormone MGA and that the European Communities has, therefore, a priori, not based its measures with respect to MGA on any risk assessment as required by Article 5.1.453 The United States further submits that its Food and Drug Administration has fully evaluated the safety of MGA and concluded that the use of this hormone as a growth promoter does not present a risk to human health.454 We find that the United States thus meets its burden of presenting a prima facie case of inconsistency with Article 5.1.
We recall our reasoning outlined above on the requirement of the existence of a risk assessment in accordance with Articles 5.1 to 5.3455, in particular that with respect to the five other hormones in dispute we assumed that the European Communities met its burden of demonstrating the existence of a risk assessment since it referred to several scientific reports which appear to meet the minimum requirements of a risk assessment.
With respect to MGA, we note, however, that the European Communities has not submitted any scientific evidence in which the potential for adverse effects on human health of MGA residues is evaluated. Moreover, the scientists advising the Panel have at several occasions stated that they are not aware of any publicly available scientific study which evaluates the safety of MGA456; the studies carried out by the United States are proprietary studies which remain confidential.
The European Communities argues that the EC measures in dispute regulate hormones on the basis of their physiological action, not on the basis of individual substances and that the administration of any substance having an oestrogenic, androgenic or gestagenic action is covered by the EC ban, including MGA which has a gestagenic action.457
We note, however, that with respect to all five other hormones in dispute, JECFA, Codex and the European Communities itself have conducted or invoked risk assessments for each individual substance. We further note that one of the basic principles of a risk assessment appears to be that it needs to be carried out for each individual substance.458 As was stated in the 1995 EC Scientific Conference:
453See paras. 4.110-4.112.
454See para. 4.131 and 4.136.
455See paras. 8.108-8.111.
456See, for example, statements by Dr. Ritter and Dr. McLean, Transcripts of the joint meeting with scientific experts of 17 February 1997, paras. 352 and 354.
457See para. 4.93.
458See, for example, answer by Dr. André to Panel Question 13 (para. 6.142) and the practice in JECFA which only examines specific substances and this, mostly, when these substances are used for specific purposes. In particular, the Joint FAO/WHO Expert Consultation on Residues of Veterinary Drugs in Foods (29 October - 5 November 1984; FAO Food and Nutrition Paper No. 32) recommended that a " ... scientific body should rely on the advice of experts in veterinary medicine, animal science, toxicology, microbiology, immunology, analytical chemistry and related sciences, and establish criteria for the safety of individual veterinary drug residues as appropriate, taking into account their public health significance, good animal husbandry and drug use practices, the likelihood of residues and the availability of adequate analytical methodology" (p.15, emphasis added).
WT/DS26/R/USA
Page 224
"It must be emphasised that a separate risk assessment must be conducted on each growth promoting substance. It is not appropriate to attempt to produce a detailed generic risk assessment for a class of growth promoters".459
We thus find that the European Communities has not met its burden of demonstrating the existence of a risk assessment with respect to MGA and that, therefore, the EC measures in dispute, in so far as they relate to the hormone MGA, are not based on an assessment of risks in accordance with Article 5.
We recall, in this respect, that the European Communities has explicitly stated that Article 5.7, which deals with cases where relevant scientific evidence is insufficient and allows a Member to take provisional sanitary measures, does not apply to the measures in dispute, including those relating to MGA.460
We further recall our reasoning and findings reached above (with respect to the five other hormones in dispute) on the procedural and substantive requirements a Member must satisfy in order to base its sanitary measures on a risk assessment in accordance with Articles 5.1461. We recall, in particular, that the European Communities has, from a procedural point of view, not provided any evidence that the studies it referred to have actually been taken into account by the competent EC institutions or are reflected in the EC measures in such a way that these measures could be said to be based on these studies. We further recall that the European Communities has not met its burden of proving that its measures in dispute, in so far as they also ban the use of the five hormones at issue for growth promotion purposes in accordance with good practice, are, from a substantive point of view, based on a risk assessment.
The same reasoning applies a priori to the EC measures with respect to MGA since the European Communities has not submitted any study in which the risks related to MGA are assessed. We thus find that the European Communities has not met its burden of providing evidence to the Panel that its measures in dispute, in so far as they relate to the hormone MGA, are, either from a procedural or a substantial point of view, based on a risk assessment and that, therefore, these measures are inconsistent with the requirements of Article 5.1.
Even if we had found that the European Communities met its burden of proving that its measures relating to MGA are based on an assessment of risks in accordance with Articles 5.1 and 5.2 and even if, for that reason, the European Communities could have adopted an appropriate level of protection against these risks, there would still be a need to examine whether the determination and application of this level of protection is consistent with Article 5.5.462 In this respect, the United States argues that the European Communities fails to justify the following differences in regulatory treatment: (i) a ban on MGA when used for growth promotion purposes as opposed to not setting any limit for residues of the natural hormones present endogenously in untreated meat and other foods (such as milk, cabbage, broccoli or eggs) or used for therapeutic or zootechnical purposes; and (ii) a ban on MGA when used for growth promotion purposes as opposed to allowing the use of carbadox as a growth promoter in swine production. Only with respect to the last mentioned difference in treatment does the United States explicitly invoke Article 5.5.
4591995 EC Scientific Conference Proceedings, p.250.
460See para. 8.249. 461See paras. 8.112 ff. 462See para. 8.163.
WT/DS26/R/USA
Page 225
We refer to paragraphs 4.209-4.211 for the arguments submitted by the United States with respect to these distinctions in light of the three elements contained in Article 5.5 and find that the United States meets its burden of presenting a prima facie case of inconsistency with Article 5.5.
We recall our reasoning and findings reached above with respect to the EC measures in dispute relating to the hormones at issue other than MGA, in particular our finding that the European Communities has not met its burden of justifying the distinction it makes in levels of protection for residues of zeranol and trenbolone (two of the synthetic hormones in dispute) and residues of the natural hormones in dispute which occur endogenously in meat and other foods, in light of the three elements contained in Article 5.5, and that, therefore, the EC measures in dispute, in so far as they relate to zeranol and trenbolone, are inconsistent with the requirements imposed in Article 5.5.463
We consider that this reasoning and these findings equally apply to the EC measures in dispute relating to MGA (the third synthetic hormone in dispute). Firstly, the European Communities has adopted different levels of protection (a "no residue" limit464 as opposed to an unlimited residue level) for comparable situations, in casu situations posing the same adverse health effect (i.e., carcinogenicity), namely for MGA used as a growth promoter and the natural hormones in dispute which occur endogenously in meat and other foods in the sense of the first element of Article 5.5. Secondly, the European Communities has not submitted evidence that this difference in levels of protection is justified and has thus not met its burden of proving that this difference is not "arbitrary or unjustifiable" in the sense of the second element of Article 5.5. Thirdly, the European Communities has not met its burden of rebutting the arguments and evidence submitted by the United States that this difference in levels of protection results in "discrimination or a disguised restriction on international trade" in the sense of the third element of Article 5.5.
We thus find that the European Communities has not met its burden of justifying the distinction it makes in levels of protection for MGA used as a growth promoter and the natural hormones in dispute which occur endogenously in meat and other foods, in light of the three elements contained in Article 5.5, and that, therefore, the EC measures in dispute, also in so far as they relate to MGA, are inconsistent with the requirements imposed by Article 5.5.
We further recall our reasoning and findings reached above with respect to the EC measures in dispute relating to the hormones at issue other than MGA, in particular our finding that the European Communities has not met its burden of justifying the distinction it makes in levels of protection for residues of the hormones at issue (other than MGA) when used for growth promotion purposes and residues of carbadox in light of the three elements contained in Article 5.5 and that, therefore, the EC measures in dispute, in so far as they relate to the hormones in dispute (other than MGA), are inconsistent with the requirements imposed by Article 5.5.465
We consider that this reasoning and these findings equally apply to the EC measures in dispute relating to MGA. Firstly, the European Communities has adopted different levels of protection (a
463See para. 8.218.
464See para. 4.104.
465See para. 8.244.
WT/DS26/R/USA
Page 226
"no residue" limit466 as opposed to an unlimited residue level) for comparable situations, in casu situations posing the same adverse health effect (i.e., carcinogenicity) 467, namely for MGA used as a growth promoter and carbadox in the sense of the first element of Article 5.5. Secondly, the European Communities has not submitted any evidence that this difference in levels of protection is justified and has thus not met its burden of proving that this difference is not "arbitrary or unjustifiable" in the sense of the second element of Article 5.5. Thirdly, the European Communities has not met its burden of rebutting the arguments and evidence submitted by the United States that this difference in levels of protection results in "discrimination or a disguised restriction on international trade" in the sense of the third element of Article 5.5.
We thus find that the European Communities has not met its burden of justifying the distinction it makes in levels of protection for MGA used as a growth promoter and carbadox, in light of the three elements contained in Article 5.5, and that, for this reason, also the EC measures in dispute which relate to MGA are inconsistent with the requirements imposed by Article 5.5.
In summary, in this section we have found that the EC measures in dispute relating to MGA are inconsistent with the requirements contained in Articles 5.1 and 5.5.
Since we have found that the EC measures in dispute are inconsistent with the requirements of Articles 3 and 5 of the SPS Agreement and considering that Articles 3 and 5 provide for more specific rights and obligations than the "basic rights and obligations" set out in Article 2, we see no need to further examine whether the EC measures in dispute also violate Article 2.
Since we have found that the EC measures in dispute are inconsistent with the requirements of the SPS Agreement, we see no need to further examine whether the EC measures in dispute are also inconsistent with Article I or III of GATT.
As noted above in paragraph 8.42, if we were to find an inconsistency with Article I or III of GATT, we would then need to examine whether this inconsistency could be justified, as argued by the European Communities, under Article XX(b) of GATT and would thus necessarily need to revert to the SPS Agreement under which we have already found inconsistencies. Since the European Communities has not invoked any defence under GATT other than Article XX(b), an inconsistency with Article I or III of GATT would, therefore, in any event, not be justifiable.
466See para. 4.104.
467See para. 4.108.
WT/DS26/R/USA
Page 227/228
In order to avoid any misunderstanding as to the scope and implications of the findings above, we would like to stress that it was not our task to examine generally the desirability or necessity of the EC Council Directives in dispute. The ability of any Member to take sanitary measures which do not affect international trade was not at issue in the present case. Our examination was confined to those aspects of the EC measures that have been raised by the United States, namely the EC import ban on meat and meat products of bovine origin treated with any of six specific hormones for growth promotion purposes. It was further limited to the specific provisions of GATT and the SPS Agreement which have been invoked by the European Communities in support of this import ban. That is the necessity of the import ban, which the European Communities strictly construed as a sanitary measure, for the protection of human life or health. Likewise, the ability of any Member to enact measures which are intended to protect not consumer health but other consumer concerns was not addressed. In this regard, we are aware that in some countries where the use of growth promoting hormones is permitted in beef production, voluntary labelling schemes operate whereby beef from animals which have not received such treatment may be so labelled.
CONCLUSIONS
In light of the findings above, we reach the following conclusions:
The European Communities, by maintaining sanitary measures which are not based on a risk assessment, has acted inconsistently with the requirements contained in Article 5.1 of the Agreement on the Application of Sanitary and Phytosanitary Measures.
The European Communities, by adopting arbitrary or unjustifiable distinctions in the levels of sanitary protection it considers to be appropriate in different situations which result in discrimination or a disguised restriction on international trade, has acted inconsistently with the requirements contained in Article 5.5 of the Agreement on the Application of Sanitary and Phytosanitary Measures.
The European Communities, by maintaining sanitary measures which are not based on existing international standards without justification under Article 3.3 of the Agreement on the Application of Sanitary and Phytosanitary Measures, has acted inconsistently with the requirements contained in Article 3.1 of that Agreement.
We recommend that the Dispute Settlement Body requests the European Communities to bring its measures in dispute into conformity with its obligations under the Agreement on the Application of Sanitary and Phytosanitary Measures.
WT/DS26/R/USA
Page 229
I would like to welcome the scientific experts and the parties to this expert meeting. Let me start by informing you that the proceedings of this meeting are being recorded, therefore, when taking the floor, representatives and experts should please use their microphones. As indicated in my letter of 7th of February 1997 the Panel decision to join the two meetings with scientific experts stems from the similarities of the two cases, as does the Panel's decision to use the same scientific experts in both cases and to invite Canada and the United States to participate on equal footing in the meetings in both cases. In addition, there is the consideration that from a practical perspective there will be a need to avoid the repetition of arguments and of questions. I recall that as to the proceedings in legal terms, the cases are not formally joined. They are joined for this session today and tomorrow as a hearing of the experts and the debate with the experts. I tend to say that there can only be one truth in this matter; we cannot have two truths. That is the philosophy we should address at this meeting and take advantage of full mutual information here. The purpose of this meeting is that the experts can expand on their written briefs and that our questions can be put to them, views can be challenged. I would like to take the opportunity at this stage to thank the experts very deeply that they responded in such a short period of time to the request of the panel. As you know, we are operating under very stringent time constraints, we have to produce reports with certain delays and this puts an enormous pressure not only on us but also on you. And I very much would like to thank you at the outset that you agreed mostly to work on this, I assume, over Christmas holidays to produce your substantive reports. These reports are very substantive and it is not a matter of repeating them line by line but really to highlight the main points and the focus and also to put the Panel into a position to be fully informed about controversial issues which may exist here which would allow us to make a legal assessment here.
I also would like to stress that the proceedings are confidential. Everything which is being said in this room is subject to the rules of dispute settlement. So it's confidential unless release is permitted by one of the parties. We had a request this morning from the European Community at an informal meeting to grant the opportunity to have the parties' experts from EC side making upfront statements in order to fully inform the parties and the Panel in these proceedings on their views. The Panel has discussed this request and has come to the conclusion that we would deny this opportunity for the following reasons. This meeting of two days was set up as a hearing of experts and the purpose is to hear the experts and put questions to the experts by making use of scientific experts within the delegations. It is not the purpose of this meeting to hear new evidence which has not been submitted by February 8th. And it is not the purpose of this meeting to have formal statements by the parties, but just from the experts. Now, this will not exclude that the experts which are on the delegations
WT/DS26/R/USA
Page 230
of the parties can speak to the point, that they would do so when time is given to the delegations and they are in a position to expand on their views during these interventions. These interventions will then again give the experts the opportunity to react to what was said from the delegations. In the light of this, I would like to explain how we would like to proceed today. We would first have statements by the experts and I would like to invite you to focus on your main points, your main arguments, your main areas where you see the issues, the main areas of contention and also where you see the problems in your colleagues' reports. After the statements, followed by the expert of the Codex, there will be discussion within the panel of experts. When this is concluded, the United States will have the opportunity to put questions and to comment on the experts' views. This will be followed by the intervention of the European Community, with the possibility, as I said, to have statements by your own experts and it will then go to Canada for the same proceeding. And since we have two proceedings here mixed, the EC will again have an opportunity to take the floor in the very end. After this process, we would intend to give the experts again individually the floor for a final statement where they could stress their views again and their conclusions. The Panel would then sit in the evening and we would then try to focus on more specific questions which from a legal point of view may still not be clear and in the meeting of tomorrow afternoon we as a Panel would then come in with very specific questions to the experts which we still may have to clarify. So we have a general debate which in the beginning hopefully that which will allow us to focus on the main issues and tomorrow there might be very specific questions here. This does not exclude that today the parties may raise specific questions to the experts based on their written submissions, based on the answers to the questions here. And I assume that delegations are prepared to do so. My last point is please try to be short and to the point in all your interventions. Please try to avoid repetitions, repetitious statements, so that we can go along with the time. This is, I think, the end of my introduction. I would now like to declare open the hearings. I repeat that all that is said in this room remains confidential under the rules of Panel proceedings and I would propose the following order: Professor McLean, Dr. Arnold, Professor André, Dr. Ritter, Dr. Lucier and then Dr. Randell as the Codex expert. I am informed that the slide machine has not been installed, so if somebody would like to show slides it would have to be done in the afternoon. I apologize for this. So may I then give the floor to Professor McLean.
Thank you Mr. Chairman. I will be brief and highlight the parts of my submission that I think are most germane. In relation to good practice and good veterinary and good animal husbandry practice, I think it is important to realize that this good practice is the practice that is negotiated between the sponsor of the product and the registering authorities to define the conditions under which the compound will be used. And once those conditions are defined, then they make up part of the registration approval and in that way the maximum residue limit can be established. The other thing that is important with the maximum residue limit is to understand that it is a legal limit and not a health limit. In other words, the exceeding of the MRL does not represent a hazard to health but is rather a limit at which the authorities take action. However, I must say that to exceed the MRL would not be seen to be good practice. In relation to the meat that is produced, to all intents and purposes, the meat is similar. In fact it may not be possible to tell the difference between meat produced using growth-promoting agents and meat that is produced without them. That when we look at those two hormones for which there is a maximum residue limit, then a minute residue may remain that cannot be detected by commonly-used regulatory methods. I think it is important to realize that the methods used are regulatory methods rather than some of the more sophisticated methods that can go to orders of magnitude below the regulatory method. If we were relying upon those sorts of methods then the surveillance process would be cumbersome and costly and actually preclude satisfactory surveillance. One of the keys to the use of these sorts of compounds is to combine the use of the compounds with surveillance to ensure that good practice is followed and the MRL is not exceeded.
JECFA has looked at these compounds particularly in its 1988 meeting and again in 1989 and it was at that point that the ADI and MRL was established, or in the cases of three naturally-occurring hormones it was deemed that an MRL was not necessary. These two meetings were particularly
WT/DS26/R/USA
Page 231
important, because it was determined for the three naturally-occurring hormones that there was no way of showing that meat treated with the hormones or untreated meat were substantially different and so therefore to regulate them by an analytical technique would not be possible or practical and indeed in many cases the levels that were seen after treatment were extremely small in relation to the naturally-occurring substances. It is interesting to note that JECFA since 1988 and 1989 has reviewed a number of compounds on a second occasion. These compounds have been referred to the JECFA meetings by sponsors or by regulatory authorities and governments. But on no occasion has the matter of the determination the '88 and '89 JECFA meetings in relation to the five hormones that it approved been, on no occasion has any organization sought to revisit the matter of oestradiol 17 beta, progesterone, testosterone, trenbolone and zeranol. I think it is very important because on a very significant number of occasions JECFA has at the request of various groups revisited the matter of the toxicology and/or the residues of products or compounds that it set MRL and ADIs for. And that the opportunity exists and has existed since 1988 and 1989 to revisit the matters of the five hormones under consideration and that opportunity has not been taken. It is interesting to see at this stage the amount of material that is alleged to show that the compounds should be revisited, that is being produced for this particular meeting yet no one has taken the opportunity to refer to JECFA. Yet on number of occasions, as I said, the opportunity to do that is being taken with other compounds. When JECFA approved the use of the three naturally-occurring hormones it accepted the fact that indeed we live virtually in a sea of hormones. These hormones came from endogenous levels produced in humans, and it came in material that we take in in the diet either from eating meat naturally or from eating other material that contains hormonally active substances; things that are commonly found in the diet including eggs, soya products, a number of plant materials and fungal products such as those from Fusarium from which zeranol is derived. Indeed it is very difficult to determine exactly where our hormone burden comes from but the facts of the matter are that humans are constantly being exposed to very significant levels of hormones and that the incidence of tumours associated which are with hormones in humans such as, for example, the breast or the prostate, have not significantly increased since the surveying has been carried out. When I say that we must take into account that the diagnostic technology available for diagnosing some of these tumours and also the level of education of the public has substantially increased but, if you correct for that sort of data and correct it for age, then there has been very little change. Some of the new data that has been submitted particularly relies upon in vivo and in vitro carcinogenicity testing and also some of the mutagenicity testing but I do not believe that it is any more significant than the sort of data that was available at the time the original appraisal was made.
The hormones in question are used in a number of countries and have been used for many years and epidemiological data do not suggest that this has had an untoward effect on human health. There are a number of compounds that have been used in veterinary medicine or in association with crops from which we have seen untoward effects in humans, but the residues present in the five hormones in question have not shown any deleterious effects in humans that have been accurately documented. Many of the hormones are lipid soluble and therefore pass into the fat portion of milk and that is a significant intake for humans, particularly for young children and also for adults in the intake of butter and cheese; I'm referring to the three naturally-occurring hormones. The hormones in question are not well absorbed orally, and indeed the uptake there is extremely small with the exception of course of melengestrol acetate. As well, there is data that shows that anything up to approaching 80 per cent of some of the hormones are actually destroyed upon cooking.
In relation to the illegal cocktails, in the countries where the compounds are registered then the use of the illegal cocktails is virtually non-existent. There are registered products which are combinations of zeranol and trenbolone, oestradiol and progesterone or trenbolone or testosterone and combinations like that and the registration of the products permits control of use. In relation to the use of the compounds, they are packaged in such a way that they are in an inert matrix and that matrix is injected into the animal, in the ear, by specially developed injection apparatus, and that therefore does control the injection into the animal. In countries where the drugs are not registered and used illegally then the administration by an acceptable technique that controls the release and is removed on slaughter, is not available.
WT/DS26/R/USA
Page 232
Coming back again to the MRL, I stress that it is a legal limit based on daily consumption of a product for a lifetime corresponding to the uptake of the ADI, however, it is not a health limit and occasional consumption is of no concern and I think I will stop there Mr. Chairman.
Thank you very much, Professor. Could your statement be summarized that you believe that the current standards we have are fully adequate to address the problem?
Yes, Mr. Chairman, I do believe on having examined countries where the compounds have had a long history of use, having examined the packages of data submitted to JECFA and I think it is particularly important that people that are talking about these hormones see the full packages of data submitted to JECFA by the sponsors. These packages of material contain the individual animal data and are much more satisfactory if one is going to pass judgement than some of the material in the open literature. Where these compounds have been used, then I believe that evidence shows that they are used responsibly as compared with countries where the use is not permitted but they are still used, and under those circumstances I believe that there should be some concern but used in accordance with the agreed protocols, then I believe that they are safe.
Thank you very much for your statement. I give the floor to Dr. Arnold.
Thank you Mr. Chairman. Since I share many views expressed by Professor McLean I will not repeat arguments put forward by him. I will also not try to summarize my written answers which I have submitted in due time. I will try to be very brief and only make a few very general statements at this time.
My general view of the problem is very similar to what has been expressed most recently by the Steering Committee and a specialist working party of a scientific conference on growth promotion in meat production organized by the European Community. The experts which had been invited by Commissioner Fischler have summarized, and this is also my view, concerning the natural sexhormones and I read from their summary: "At present there is no evidence for possible health risks to the consumer due to the use of natural sex hormones for growth promotion since residue levels of the substances measured in meat of treated animals fall within the physiological range observed in meat of comparable untreated animals. The daily production of sex hormones by humans is much higher than the amounts possibly consumed from meat. Even in the most sensitive humans, pre-pubertal children and menopausal women, due to an extensive first pass metabolism, the bio-availability of ingested hormones is low, thus providing a further safety margin." And on zeranol and trenbolone this conference has stated, and I share these views: "At the doses needed for growth promotion, residue levels are well below the levels regarded as safe, the MRLs. There are at present no indications of a possible human health risk from the low level of covalently-bound residues of trenbolone."
This is my first very general statement. Concerning the not-disputed carcinogenic properties of these substances, my personal views very similar to what has most recently been expressed by Professor Jonathan Li, who is a leading expert in the field of kidney carcinogenesis. He has summarized on the basis of the evidence we have today: "The inescapable conclusion is that hormones, particularly oestrogens, are non-genotoxic carcinogens. This latter term, however, is somewhat misleading since non-genotoxic carcinogens by definition ultimately affect permanent genetic changes leading to neoplastic transformation. Perhaps a more appropriate term would be epigenotoxic, defined as referring to an
WT/DS26/R/USA
Page 233
agent that is not involvedin direct covalent or indirect interactions with geneticmaterial, but nevertheless is able to elicit heritable changes by alternative mechanisms." I am inclined to share these views. I think a lot of confusion has emerged in recent years because at the time when the policies towards genotoxic carcinogens have been established, genotoxicity has meant clear mutagenicity. In the meantime, over the past decades, many tests have been developed which picked up very early end points of damage to the genome. Just to mention one, single-strand breaks, I could mention many others, in the DNA. Now the problem is, it is clearly a genotoxic effect a single-strand break observed, for example, in vitro. However this Panel should know that typically in a cell this event occurs some 5,000 times per hour and the capacity to repair such a damage is in the order of 200,000 such events per hour. Now for example, I give only one example, Dr. Liehr was able to demonstrate such breaks at extremely high doses and he was able to demonstrate a ten per cent increase it remains to be confirmed that such an event is of any significance in vivo. I have similar problems with Dr. Metzler's paper who reports about genotoxic potential. For example, in his paper on page 12, he says with respect to agents causing numerical chromosomal abberations (aneuploidy) a biochemical threshold can be expected on theoretical grounds and can be measured in in vitro systems such as culture cells. However, it is impossible to determine this threshold in an intact organism in vivo, because it is not known which tissue is most susceptible at which stage of development and in different individuals. For residues in food, exposure occurs for people of all ages and values of susceptibilities. I agree, first of all, that on theoretical grounds there should be a threshold. I also agree that we do not know where this threshold exactly lies but what we perfectly know is that the doses which have provoked such effects are many orders of magnitude above the concentrations we are talking about today, a minimum of 1,000, sometimes 10,000 or 100,000 times. There exist many papers and literature which have demonstrated that the no-effect levels for such effects are in the order of one micro-molar or slightly below and there we are still far above the concentrations we are talking about today. So, my conclusion is a lot of evidence has been put together about genotoxicity of the substances, mostly in vitro, but the relevance to the in vivo situation still remains unclear. Particular activities have been focused on the metabolism of the so-called catechol oestrogens. This is a category of compounds which is known since several decades. I had already to learn these structures when I was a student and also more than 20 years ago their metabolic activation has already been proposed. However, the real relevant work has been done in the past years and the idea is that these substances could be activated in metabolism to act as carcinogens. The problem which I have with the papers written by Dr. Liehr, not only on the occasion of this Panel but I have very carefully examined, a review article he has published last year in Annual Review of Pharmacology and Toxicology, that sometimes I have the idea that he does not clearly discriminate between what has been done as pure chemical synthesis or in tests in in vitro and what has been shown in in vivo situations. I cannot find any evidence supporting the view that the metabolic pathways elucidated to sometime by in vitro studies are of relevance in in vivo. So, I think a lot of misunderstanding is created by the different uses of the term "genotoxic carcinogen" as it has evolved in history and as it is used nowadays. I think I would not like to add much more at this moment, but I will be open to answer questions.
Thank you very much Dr. Arnold. Do you have from the Panel any questions at the moment? Not now? Okay. Thank you very much for your statement. Now I give the floor to Professor André, please.
Thank you, Mr. Chairman. I will be a very short for the moment because I did not prepare a specific assessment. You don't ask before, but maybe you don't know how to manage this meeting today.
WT/DS26/R/USA
Page 234
Yes, and I didn't ask you, so that the speeches will be short indeed.
Okay, thank you. So it's a nice surprise. But I think also that many things have been said before and written in the answers to the questions. I would just like to add some comments, personal comments. As an expert I am here as a veterinarian first, I am a biochemist secondly. But I am now working more in the analytical field of residue control of these hormones in dispute. And when I received the questions it looked that many of these questions more or less outside my field of strict competence, because I am more concerned with toxicology as well as with some veterinary or animal husbandry practices. However, I accept to answer and I decided at the beginning to work as a student would do, which means looking at first at recent scientific available data and synthesizing them. I was very impressed to discover the number of recent publications, between 1990 and now, about these topics of hormones and specifically on hormone's action in animal health or about their toxicology. I also try always to stay outside the political debate and to be strictly an honest scientific. About some points here, I just comment. For example, about good animal husbandry practices as well as good veterinary practices, it has been said just before my talk that when drugs in general are registered, this registration is with some withdrawal periods and rules of impairment. In fact problems, the real problem, I think, is the problem of control, how to be sure that these practices are really respected and what we can say now is that official control services are in charge to control these. But also that now more and more people themselves, I would like to say, people as veterinarians, as farmers, as farmer organizations, are taking the control and are ensuring that they will only use registered drugs as it is possible to officially use them. About residues and remaining residues, there is a question about the persistence of residues in meat and animals. It is clear that this depends of any drug but it sometimes very surprising to discover that when you develop new sophisticated methods, you observe that residues are still remaining a very very very longer time than it is commonly known. And we have personal experiences for certain drugs, near hormones in this field. About toxicity of these compounds, and more precisely carcinogenicity or genotoxicity, I was very impressed to read many papers about these compounds, recent scientific papers (and I sent you the original papers) and just looking fast on summaries, conclusions of these papers, it seems for everybody now clear that they have very dramatic toxicity and that most of the natural as well as synthetic hormones have an action on DNA as said by Dr. Arnold. The problem is just the problem of dose response and the problem of threshold. I think that when a compound is usually recognized now as genotoxic, the problem of threshold is not of concern. And I would like to compare in simple words as if, Mr. Chairman, you were student, and I am sorry for the comparison. If you drive, for example, 30,000 miles a year, you have a probability, a risk, to have an accident. If you drive only 1 kilometre a year, you have also a probability to have an accident. You cannot say that under one thousand kilometres or under a thousand miles a year, you will never have an accident. And that is, I think, the same thing for these kind of compounds. The more you use, the more you have risk to have an accident. But you have no limit under which it is possible to be sure that you will never have an accident. Sorry for the comparison, if you think it is too simple, but I think it is sometimes useful. Concerning the very first problem of JECFA and Codex and these institutions, it is clear that the jobs they do is very useful for everybody and that nobody, no scientist, will think that it is a bad assessment done by these organizations. I just have a personal problem about the definition of drugs as published by the Codex, for the reason that drugs in these definitions involve all familiar known drugs but also physiological modifiers, as these hormones are when they are used for growth promotion. And to my personal opinion, there is very big difference between a drug, I mean zootechnical and therapeutical purposes of these hormones, and the use as growth promoter, for the reason that you may admit that you have a small risk when use a drug because in this case, you cannot avoid to use it and it is very more difficult to admit that you take a risk when using just to modify. Maybe also the consequence is that a risk assessment for such modifiers must be done with other rules than the classical rules for drugs. This has been said in the previous conference in Brussels in 1995. Coming back to the comparison, you can admit that you
WT/DS26/R/USA
Page 235
could have an accident going to the hospital. So for the moment I think I will stop the speech. I apologize too for my English, which is bad, it is not my mother tongue. But we can later come back on many points, if you wish.
Thank you very much for your excellent introduction to your paper and assessment. May I now turn to Dr. Ritter.
Thank you, Mr. Chairman. This communication system, I must comment, is a wonderful device in itself for arbitration as it only permits one person to speak at a time. I should say that coming fourth following three distinguished scientists leaves me little opportunity to add anything of any consequence additionally. But having said that of course I will take the opportunity to say a few things. I should begin by really thanking the Panel for the opportunity to contribute to the debate. This is a debate which has gone on for sometime now and I suspect one which may continue for sometime yet, and it is of course an extraordinary privilege to be given the opportunity to contribute to the debate with my understanding of the issue. So, I am grateful for that understanding, for that opportunity.
I will attempt to very quickly add a few points to those that have already been made by those who came before me. You know of course, Mr. Chairman, that myself and my three colleagues were provided with a very limited period of time in which to provide comments and so at least in my case I have attempted to do so to the best of my ability within the time constraints that were provided. There will be unquestionably information that some may cite that I omitted in my comments and that is a matter, I think of practical necessity rather than an error, if you wish. In formulating my opinions I have, to the extent available to me, attempted to use as large an international database as possible. That is why I have made extensive use of ... [data](?) [cassette side finished] ... by WHO, by JECFA, by Codex and in the published scientific literature. I have not referred to proprietary data, even to the extent to which I may have had it available to me. I have, of course, made extensive reference to the proceedings of the European Conference on Growth Promotion for two reasons. First and foremost, because I believe that that Conference which was held in Brussels in late 1995, represented the views of an outstanding group of scientists from the international community and brought some considerable expertise to the debate. And secondly, because I believe that this 1995-96 Conference was a very important and very recent updating of information which had already been reviewed internationally by organizations such as JECFA in 1988 and in 1989. So certainly it is my opinion that this whole issue has been re-evaluated as recently as early 1996. And it is my view that the European Conference has essentially reaffirmed, and indeed I'll go as far as to say strengthened, the earlier conclusions which were reached by JECFA, and that is that in accordance with operating procedures that are provided for the use of these substances, that their use does not constitute a risk to consumers of food commodities produced with the hormones at question, at least for five of the six. The Panel would be aware that MGA has not had the benefit of an international review. My colleague Professor McLean has already articulated the fact that an MRL is not a health limit, but rather a regulatory limit and that the limit may well be different in different jurisdictions for different operational reasons, because of climate, because of geography, because of use practice and otherwise. But the ADI is an international value. And I would draw the attention of the Panel particularly to the point that for the three natural hormones, it was the opinion of JECFA, and indeed the view re-enforced by the international conference in December of 1995, that these natural hormones do not require the establishment of an acceptable daily intake, because in fact their use reflects levels which are entirely within the range that one might expect in animals that had not been treated at all. That is, their use would result in a residue which would be indistinguishable from animals which had never been treated. In the case of xenobiotic hormones it goes without saying of course that these hormones are not natural by definition, and hence any level which can be detected is outside of what one might expect in an animal which has not have been treated. But even in that case, the conclusion that has been drawn
WT/DS26/R/USA
Page 236
internationally, is that again in accordance with the recommended use procedures that these levels which do remain constitute no risk whatsoever for consumers of the product.
I should say, Mr. Chairman, that there has been some considerable debate as to the potential for abuse of these products and the health risks that this abuse may constitute. I would add my personal conviction that I do not believe that the potential for abuse in any way speaks to the safety of appropriate use. I make that point because the juxtaposition that is created often in the debate is that because there is potential for abuse of anything, whether we are talking about these hormones or whether we are talking about motor vehicles or whether we are talking about alcohol, I think we need to remember that that potential has no impact whatsoever on the appropriate use. I would also suggest to the Panel that because the potential for abuse exists, in my view, is not a reason to impute the safety of appropriate use. And if I may be permitted to draw a very simplistic example, not unlike the one that my colleague Professor André drew a moment ago, I hardly think that we would be considering the banning of the automobile because some people speed. It is a question of risk management and not risk assessment. And the issue that I think we have to deal with is how do we control those people who speed so that we can all enjoy the benefit of the automobile and similarly how do we appropriately manage the potential for abuse with these hormones rather than imply that this abuse somehow or another makes the appropriate use unsafe. I should add perhaps before I leave the issue of abuse just very quickly that it is noteworthy that in countries where use of these hormones, the six hormones in particular, has been permitted for a very extended period of time, and I refer most notably to Canada and the United States, monitoring and compliance programmes which have been ongoing now from many many years have consistently demonstrated that residue levels are entirely within recommended limits and that instances of violative residues, that is residues which would indicate abuses taking place, have almost never been reported. So that it seems that at least in those jurisdictions where use is lawful, the practicality of abuse has never become a reality. There are, as I say, few and isolated examples of violative residues in those countries where use has in fact been permitted.
It is perhaps important, I think, to bring to the attention of the Panel the fact that we have an enormously rich experience in the human case with these hormones, and I draw the Panel's attention specifically to the use of oestrogen, and most notably oestrogen in combination with progesterone, as an oral contraceptive which has now been in practice in the human population for perhaps 35 to 40 years and that there have been tens of millions if not hundreds of millions of women who have taken these steroids on a daily basis, in many instances for their entire reproductive life, periods of time that may have ranged from 30 to 50 years. And that these levels have typically reflected an intake and exposure that ranges somewhere between 20 and 100 micrograms, depending on the pharmaceutical preparation, on a daily basis for the entire reproductive cycle of the hundreds of millions of women in question. It is interesting to note that these levels are thousands, if not hundreds of thousands, of times greater than the levels, Mr. Chairman, to which you or I might be exposed in consuming meat which results from the use of these hormones in production. And although the issue of breast cancer most notably is certainly one of considerable interest these days, I think it is fair to say that to date, there is no compelling evidence whatsoever to demonstrate that the use of oral contraceptives containing two of the very hormones at issue before this panel, over at least an entire generation in hundreds of millions of patients, has not resulted in any compelling evidence to suggest that the use of these drugs in fact constitutes any increased risk to these women of cancer whatsoever. Finally, Mr. Chairman, I would only conclude by saying that certainly in the case of the natural hormones, and as I have already indicated at least in the case of trenbolone and zeranol for the xenobiotic hormones, I think the five have all been subjected to both national and international reviews which have been re-reviewed, if you like, as recently as December 1995. It would certainly be my view, I would certainly share the views expressed both by JECFA and by this international forum that the use of these compounds, in accordance with provisions provided on a regulatory level for their use in countries that in fact permit the use of these products, does not constitute an increased risk or harm to those consumers who consume food commodities produced with the aid of these hormones. Thank you, Mr. Chairman.
WT/DS26/R/USA
Page 237
Thank you very much for your statement. May I turn now to Dr. Lucier.
Thank you, Mr. Chairman and I also thank the Panel for inviting me to participate in this review. My comments really evolve from my research activities in the field of receptor biology. I have had a laboratory activity in the area of receptor biology for over 20 years now, and more recently I have been involved in coordinating a national toxicology programme within the Department of Health and Human Services of the United States, that is responsible for providing a toxicological evaluation on agents of public health concerned whether they be in the environment, in the workplace, in foodstuffs, physical agents and so forth. I have also been involved over the years working with the International Agency on Research on Cancer on a number of their efforts in their hazard identification, their risk identification, chemical carcinogens and in particular those in which hormonal activity seems to be at the root of the potential carcinogenic activity. Having come last, I probably won't say too many things that haven't been said before, but let me go through some of the key issues at least from my perspective and those that I feel comfortable in commenting on. One is that the residues that occur from the natural hormones are certainly going to be indistinguishable from those which occur from the natural endogenous materials. This is not true, of course, for the synthetics. With appropriate analytical methods and appropriate residue levels one is capable of detecting residues of the synthetic materials here in question. I should also point out that the half-life, the biological persistence, of the agents in question is for the most part rather short. They do not stick around the body very long. But having said that, half life being just what it says is the time in which it takes for half the material to disappear. So, if something has a half life of one day if you start out with ten units of it, one day later you have five, two days later you have two and a half and so forth. So even though several months may elapse following an exposure to an agent even with a relatively short half life, a few molecules may remain. The number of molecules may be remarkably close to zero, probably not detectable by analytical methods, but a few molecules will likely remain. I should also point out in terms of the carcinogenic activity of the hormones in question, we already know that physiological levels, naturally-occurring levels of androgens and oestrogens are carcinogenic. One out of ten women in the industrialized world gets breast cancer. There is compelling mechanistic biological information, human studies and toxicological studies to show that oestrogens are at the root of this. That their cancer is a multi-step process but that oestrogens appear to be responsible for the majority of the carcinogens that are seen in the site. So we already know that physiological levels are carcinogenic. We already know that early menarche increases risk, late menopause increases risk. We know that when a woman loses her ovaries and her oestrogen production diminishes, the risk to breast cancer also diminishes. Charles Huggins in his noteworthy work back in the early sixties show that ovariectomy (removal of the ovaries) prevented breast cancer in experimental animals. The same is true of endometrial cancer. We know from human studies that endometrial cancer is elevated significantly by exposure to oestrogen replacement therapy when this is unopposed by progesterone. Conjugated oestrogens have been called a known human carcinogen by the International Agency for Research on Cancer. DES of course has, but DES is not a question here and the mechanism by which DES produces cancer may not be relevant to the issues here. Tamoxifen, an anti-oestrogen in breasts, produces cancer in the endometrium because it acts like an oestrogen in the endometrium, the uterus. It acts like an anti-oestrogen in the breast so it blocks ... chemotherapeutic agent in women who have breast cancer and the size of the tumours frequently diminish after this exposure. Oral contraceptives are also a known human carcinogen but not because of breast cancer. They are a known human carcinogen because of increased risk of liver cancer. The increased risk is quite small, but nevertheless there has been a consistent risk shown across studies. After having said that, I need to add that the number of molecules that remain following appropriate use of these agents is very small. They are very small in relation to the amount of naturally-occurring oestrogens or androgens. So the accompanying risk that would be associated by consuming meat containing residues would be extraordinarily small. It would be very hard on scientific grounds to say that the risk was zero. But it is likely to be very, very small. It could be zero. It
WT/DS26/R/USA
Page 238
could be as high as by estimates as one cancer in a million people exposed to them over their lifetime. So the risk, to sum up those comments, is somewhere between zero and somewhere around one in a million. And that one in a million is a very difficult number to pin down. A lot of assumptions go into it. What I would say would be a conservative estimate, an upper bound estimate, and I can go through as to how I came up with that number if those details are required. A couple of other points that I would like make. One is that the issue of threshold really is not relevant here. As I already said, the amount of material that is added to the existing burden is very very small. The amount of naturally-occurring oestrogens and likely androgens are already carcinogenic, so you are already adding to a carcinogenic burden. So the issue of threshold is totally irrelevant to the toxicological evaluation of these agents. It may also be irrelevant to distinguish between genotoxic and genotoxicity, when one gets right down to it. I would say that there is a great deal of evidence to support the notion that receptor mediated events, i.e. normal hormone pathways, are the main reason why hormones and hormonal agents are carcinogenic. However, there is a body of knowledge that suggests that they do genotoxic activity, this genotoxic activity could possibly contribute to the carcinogenicity I mean there are examples of some in vivo experiments that have shown this in whole animals, and I can go over those if requested. Butmy bottom line is an extensive discussion of genotoxicity versus non-genotoxicity really isn't relevant to the discussion and to the outcome because we already know that the existing levels of hormone are carcinogenic. The only reason for distinguishing genotoxicity from non- genotoxicity in my understanding, and it may be a naive understanding, is to determine whether or not to apply a linear or a threshold model for risk assessment. A threshold model would assume that there is a dose below which no effect could occur. Since we are already into the effect region, again we are adding an existing burden by adding an existing number of chemicals to that burden, again keeping in mind that this is a very small addition and may be remarkably close to zero. We also, in the course of the national toxicology programme, done carcinogenicity evaluations on approximately 500 agents. Many of these agents are carcinogenic by virtue of their genotoxic activity, many are carcinogenic by virtue of their non-genotoxic activity. When we looked at those response relationships for genotoxic and non-genotoxic carcinogens, we couldn't distinguish one from the other. In other words, non-genotoxic carcinogens were no more likely to exhibit linear behaviour than genotoxic ones. So again that goes back to my point that the genotoxic versus non-genotoxic issue, I don't think that is a real one. A couple of other points to make. One is, when one looks at the natural versus the synthetic compounds, the synthetic compounds clearly have hormonal activities, that's why they are used in the situation. They may also have other kinds of activity in addition to that because molecular structure is different than that for the naturally-occurring oestrogens so there may be different metabolites, possibly a different spectrum of adverse effects. I assume that when one looks at the toxicological data that is available on these, and not all of it was available to me because some of it is proprietary, that the appropriate studies would be done to determine that other effects are not occurring. Effects that could be detected by sub-chronic 90 days studies or three year studies and by appropriate molecular screening techniques. Another point regarding the cocktails, the synergy. There was a recent paper indicating possible synergy of oestrogens in science, using yeast. At this point, the relevance of that system to even risk has not been shown, and those results have not been repeated in human cells. So I think the issue of synergy, more than you might expect from adding two weak oestrogens together, the issue of additivity or antagonism, of one inhibiting the other, really is an open question. In my mind, additivity is the appropriate way to establish risk. I should also reiterate again that cancer is a multi-step process. We can define the different steps operationally, in some cases, somewhat mechanistically. It involves DNA damaging steps, creation of mutation, the ability of cells that harbour this mutation to proliferate more rapidly than normal cells. It is generally thought that oestrogens act primarily at that step that stimulates thegrowth rate of genetically altered cells, cells that already harbour mutation, and there is a wealth of scientific information and literature to show that. The work by Dr. Liehr, Dr. Metzler, Dr. Cavalieri and other folks indicate that there may be some genotoxic activity for these agents, suggesting that they may act to some extent on that first step, which would be the creation of that mutation. So it appears possible that within the framework of a multi-step model for cancer, that hormones and these hormones in question by act on more than one step. And the last point that I want to make is we talk a lot about the toxicology of these agents, however, it is important to point out that oestrogens have been used in therapy for a long time for many purposes. Certainly
WT/DS26/R/USA
Page 239
oestrogen replacement therapy is one. There is a convincing body of knowledge to show that oestrogens can prevent bone loss in women who are susceptible to that, so it can reduce the risk of osteoporosis. There are also studies to show that oestrogens can decrease heart disease. So when one starts looking at an overall risk assessment, the small amount of oestrogens that are added by the consumption of meat containing these residues, it seems appropriate also to look at the whole balance of effects, what are the benefits versus risks. And I think that is going to be a hard question to answer. There may be a very very small risk from adding these smaller molecules , but there may also be a small benefit. I don't think we have the tools in which to adequately determine that risk benefit ratio, and I think that is going to be one of the difficulties that the Panel has in coming to grips with what recommendations, what decision that they make. Let me stop there, and again I will be glad to add more details to my comments or respond to questions as the meeting is on.
Thank you very much for your introduction. May I give the floor to Dr. Randell, the Codex expert.
Thank you, Mr. Chairman. And thank you for the opportunity of meeting with you and your fellow Panel members. And also for the opportunity of meetingso many old colleagues from my JECFA days; I haven't seen some of them for quite a number of years. JECFA is the FAO-WHO Joint Expert Committee on Food Additives. It was founded in 1955 by FAO and WHO with the intention of giving guidance to these organizations and their member governments on the matter of chemical substances added to food, essentially for whatever reason. In its early days, JECFA established what was known as the ADI (Acceptable Daily Intake) concept, initiated in fact by Professor Rene Truhaut, who is no longer with us. The ADI concept is built around a concept of adding no appreciable risk when you add chemicals to food. That concept has been with us, as I say, since the late 1950s and with very minor modifications it stands as it stands today. Chemicals that have been given an ADI by JECFA, or by national authorities because the same concept has been adopted by many national authorities, an intake less than that ADI, over a lifetime, produce no appreciable risk to the person consuming it. In 1960-61, FAO and WHO then turned their attention to pesticide residues and the joint meetings on pesticide residues formulated the concept of the Maximum Residue Level. This is different to the ADI. Maximum Residue Limit was established to ensure that pesticides used in field conditions under good agricultural practice did what they were supposed to do and did it within a framework of safety. If I could use a relatively simple explanation of the MRL in the pesticide context. If we apply a pesticide to a field we do it for certain reasons, we do it to kill a pest. If you apply more than that then you have exceeded what you need to do and you have surplus residue, which is thought to be an unwise thing. If you spray less pesticide on the field you get the worst of both worlds: you get the pest and you get residues which you didn't want to have in the first place. So to a certain extent the maximum residue level is a measure of how much of the chemical is needed to achieve the objective of pest control and at the same time falls within the framework of the safety analysis. JECFA and JMPR were expert committees and they still are expert committees. And it was in 1962 that FAO and WHO formed the intergovernmental Codex Alimentarius Commission, which was established to protect the health of consumers and to facilitate, to remove unnecessary barriers to trade in food products. The work of Codex, the work of JECFA and the work of JMPR went on, and still goes on, with increased interest from many parties and certainly a lot of increased interest since the Marrakesh Agreement. In the early 1980s, a number of Codex committees began to raise questions, governments at these committee meetings, began to raise questions about how do we deal with the residues of the substances used for growth promotion. It is of interest to note that these were brought up in the Pesticide Residues Committee, Committee on Food Additives, Committee on Meat Hygiene, Committee on Poultry Products. There was in fact some sort of competition between the Codex committees as to who was going to handle this hot potato. In 1984, FAO and WHO convened an expert consultation to advise the two organizations on how to handle this particular problem. It was an expert consultation on process rather
WT/DS26/R/USA
Page 240
than on scientific detail. However, the consultation did come to the conclusion that it was an appropriate area for Codex to work in; that the most significant problem in the field was not the legitimate use of growth promoters or any other veterinary drugs, under good veterinary practice conditions, but rather the abuse or uncontrolled use; that it would be appropriate for the Codex Commission to establish a separate committee to handle these matters. And the expert consultation also had the first stab at providing a definition of what was meant by veterinary drugs. It was the intention of the experts of this consultation, and I was privileged to be the secretary of that consultation, to make the definition as broad as one could possibly imagine and this is why we even get references in the definition to apiculture and bee growing. The expert consultation resulted in the establishment of a committee by the Codex Alimentarius for residues of veterinary drugs in food, and it also gave the stimulus for JECFA to begin work in this area in earnest and to obtain the first really comprehensive data packages on the substances about which we are speaking today. They were evaluated in 1987, and a follow-up evaluation for trenbolone acetate in 1989. The recommendations of the JECFA were in the form of maximum residue limits (MRLs) although of course for the endogenous hormones the rather peculiar statement that the MRL was that an MRL not necessary, an internal tautology which I think we are going to have to live with for sometime. These MRLs were processed through the Codex system and were first discussed at the 19th Codex Alimentarius Commission session in 1991, when it was decided not to adopt the MRLs as they currently stood. When the complementary information on trenbolone acetate had passed through the system, the MRLs were re-presented for adoption at 21st session of the Codex Alimentarius Commission, and were adopted. That is the historical level.
On the scientific side, I would like to come back to questions which you asked the Codex and I assume JECFA Secretariat in your written questions, particularly about the relationship between the acceptable daily intake and the maximum residue limit and whether or not these are measures of the acceptable risk. The maximum residue limit is definitely not a health-basedlimit. It is a limit established for the control of veterinary drugs in actual practice. The thinking which goes into the establishment of a maximum residue limit is such that the maximum residue limit will never lead to residues which would, in a normal ingestion of the product, cause any consumer to exceed the acceptable daily intake. Therefore there is an upper boundary to the maximum residue limit imposed by toxicology considerations. The lower limit is normally derived from residue trials in practice according to the proposed good veterinary practices in the use of these drugs. However, in certainly one case, in the case of zeranol, this limit falls so far below not only the toxicologically derived limit but it even falls well below the limit that one can determine with normal methods of analytical chemistry, and therefore has been increased to take into account the fact that if you are going to operate a decent control programme, that is not going to be overly expensive, than you have a limit which is controlled by the analytical procedure rather than by the residue trials. The acceptable daily intake itself also is not a direct risk assessment in the sense that it provides a statement of quantitative risk. The ADI concept states that there will be no appreciable risk as a result of exposure to the chemicals concerned. This is true of within the JECFA and JMPR framework. This is true of food additives, residues of veterinary drugs and pesticide residues. I appreciate very much your comment Dr. Lucier that the risk is probably somewhere between zero and one in a million (one in ten to the minus six), because this does put some sort of quantitative framework on that. But JECFA has never established a quantitative evaluation of risk in the application of the ADI. I think I will leave my introductory comments there if that is satisfactory to you, but I would be very pleased to answer additional questions on the evaluations. Thank you.
Thank you very much, Mr. Randell. According to our procedures, I would now like to give the floor to the United States delegation for comments or questions to the experts. May I give you the floor, Mr. Brinza?
WT/DS26/R/USA
Page 241
Thank you, Mr. Chairman. First with respect to comments on the responses that we received from the experts, we do have some written comments that we would like to make available to you all now that would facilitate us getting through what I suspect would be a lengthy exercise - it would not be very fruitful to just read them. So let me provide those to you. I think we should have enough copies for the Panel and the experts as well as for the other delegations. We thought it would be useful to ask some questions that would go to all five of the experts. The representative from Codex of course is welcome to respond as well, but we assume that you are serving more in your ex-officio capacity with respect to the Codex institutional issues. What we would like to do would be to ask some questions of all five, that I think would be fairly simple questions, just to make sure that we understood the areas of the responses where there was consensus since there were different responses - there were in some cases different wordings - so we just want to make sure we understood correctly all the responses and the foundations for some of those responses.
For example, with respect to the three natural hormones - oestradiol 17 beta, progesterone and testosterone - we assumed that all five would agree that these three compounds are all present in the meat-producing animals as well as in humans naturally (endogenously produced, I think, would be the better way of explaining that).
Secondly, assuming that is true, do they also agree that these three hormones are all naturally present in the meat and the milk from those animals? I believe we have heard some of the experts already refer to that but it could be useful to make sure that all the five of them agree with that. Assuming there is agreement on that, when food products containing the natural levels of these hormones, in other words levels within the normal physiological range, do these hormones have a biological effect on consumers? We have a few more like this and I will defer to you on whether it is appropriate to get responses now on those and then we go on from there or whether you prefer to do them all at once in this particular series and then have the responses. It may be difficult for them to keep track of the different questions.
I would propose that you directly put the questions to the Panel and enter into a dialogue and then go on through your programme.
Thank you, Mr. Chairman. If that is the instance then, if I could ask the five experts with respect to each of those first three questions, I just want to make sure we understood correctly your responses, firstly are these three hormones naturally present in the meat-producing animals and in humans? It may be useful to just go around the table starting with Dr. Ritter.
The extent to which they are presented of course as you will know is a function to some extent of the status of the human in question and the age, but to provide a general answer, I think, in my view the general answer is "yes", noting that there will be some exceptions in variance on the degree. I understand that the levels are different but that they are all present to some extent.
Yes. There will be variations in levels but the three naturally-occurring hormones are found in both animals and humans.
WT/DS26/R/USA
Page 242
I agree, Mr. Chairman, and we perfectly know the levels occurring from fetal life to until very late age.
Yes. The hormones are the same but it is not clear if the metabolites sometimes could be different. For example, if you give a natural hormone as (.?.) to calves or to any meat producing animals, they can metabolize and transform the hormone and in these conditions the food intake for human is not the same. You see animals have been treated as if they are not treated, qualitatively.
If I could just clarify with Professor André, are we talking about levels of that magnitude with respect to the administration of any of these three for growth promotion purposes in accordance with good animal husbandry practice?
The metabolictransformation, the biological transformation, is totally independentof the level. It is not a problem of level. Maybe I don't understand well your question.
I thought your earlier answer had been that there is a possibility that these could be administered at such a level that they cannot be metabolized by the animal. And my question was, are the doses we are talking about for growth promotion so high that they would be at a level that the animal cannot metabolize them.
There will be no answer to this.
Yes, these materials would be present at all stages of life, in both males and females, as my colleagues have said, in varying amounts.
Then, the next question was, would the hormones also be present in the meat of the animals and milk from those animals? Again recognizing that the levels will vary.
Mr. Chairman, I think there are different pathways of metabolism for these substances. And certainly at high levels, it depends somewhat on the dose which pathway is used, but essentially metabolism is similar in animal and in man. For example, in cattle the 17 alpha oestradiol plays a certain greater role than in human beings, but this metabolite is less active as an oestrogen or practically not active as an oestrogen. So there are slight differences.
WT/DS26/R/USA
Page 243
This goes to the first question. I am very glad I could understand this. So let's go to the second question, then. Just as a matter of procedure, we don't go for consensus. We just take and record what the experts say, but we are not trying to assess some consensus here.
I understand, Mr. Chairman. It is just that to the extent that we are all talking from a common understanding of the facts, then it is useful for us to perceive that. Yes, the second question is then will each of these three hormones be present in the meat and the milk from these animals - the oestradiol 17 beta, testosterone and progesterone.
Subject to the same variances that I indicated a moment ago, the answer is yes.
They will be in the milk?
Yes, again subject to the same variances that we discussed a moment ago, although the levels and the extent to which they are present and the ratios and so on and so forth will be a function of the particular status of the animal in question, but to avoid going into a lengthier response, than is really necessary in the general case the answer Mr. Chairman, I think would be yes. They would be present.
Mr. Chairman, it may be useful as part of these questions, to save some time, we are also asking whether the levels that we are talking about would be within the normal physiological range with respect to these hormones when they are administered for growth promotion purposes.
To answer that question, again the answer is yes. The position, certainly that the that I have offered, Mr. Chairman, is at least that when these substances are used in accordance with the prescribed use conditions, the levels that are present are in fact indistinguishable from those which would be present in an animal which had not been treated at all. I should say, to clarify that just a little bit further, there is a range of concentrations which is what I have been attempting to allude to that one might normally detect in untreated animals. It's not a finite number, but rather there is a distribution of values over which one typically might find these hormones in an untreated animal. The levels in animals that have been treated falls within the range of those numbers that might be detected in animals that have otherwise [cassette ended ...]
I don't want to impart the impression that there is a precise number for testosterone levels, there is a range of values that one finds. As one might expect, these are biological living systems and there is a distribution of values, but treatment results in a residue level which falls within that range.
WT/DS26/R/USA
Page 244
Mr. Chairman, the answer is yes and to facilitate an explanation, I suggest that if you look at Attachment 1 of my submission, then there are a number of figures in there that relate to progesterone to oestradiol 17 beta, and to testosterone in various tissues and in particular the excretion in the milk, and when you come to the other than naturally-occurring hormones, for the other hormones trenbolone and zeranol, because they are lipid soluble, one would expect them to appear in the tissues and in the milk.
Chairman, we have a rather complete data base showing us, as the previous speakers have already said, there is a large variability according to physiological status and there can be a factor for certain hormones up to 1,000 and I have presented the data in my written responses. Therefore I agree that if the substances are used according to good veterinary practices, either for growth, promotion or for the therapeutic uses authorized in the EC, you have initially a slight difference but at the time of slaughter there is no means to discriminate between treated and non-treated animals, irrespective of whether for growth promotion or for therapeutic and zootechnical purposes.
Yes, both hormones are present in meat and milk, it's clear, and the level is a very variable level is clear also. I don't agree with Professor Arnold just on the last point, when he says that it is impossible to distinguish between hormones, injected hormones and natural produced hormones. In some cases it is possible to distinguish the two by physical means of isotope relations, but maybe not at a very low levels, it is very difficult to do.
Thank you.
Yes, these materials are present in milk at varying levels. Progesterone is much higher than the other two, as much as a 100 to 1,000 times higher, but they all are present.
Thank you. You wanted to reply.
A broad comment on Professor André, I totally agree this is a theoretical possibility, if you label the hormones with stable isotopes, you can discriminate between added hormones and endogenous hormones, but this has absolutely no relevance in practice.
No, I totally disagree. I was speaking about natural occurring isotopes, C-12, C-13 and its possible to distinguish in this case, some hormones naturally-occurring or injected, as they do in sportsmen. They do this currently in sportsmen now for doping in sport and they can distinguish endogenous testosterone and injected testosterone, even if the compound is really the same, but the isotopic composition is not really the same.
WT/DS26/R/USA
Page 245
Thank you, Sir. We take note of the two views.
I just want to add another thing. I think it is important when one looks at milk, that because these hormones are fat soluble, then the concentration in the two commonly used or common articles of commerce, which are cream and milk, is very substantially greater. It is something of the order of three to four times into cream and about 40 times into butter, and so I think one must take that into account as well.
Thank you. Then the next question was at the levels we are talking about, within the normal physiological range, would these residue hormones have a biological effect on the consumer?
I think that to attempt to answer that question would really be to pre-empt the outcome of this Panel. I think that is the question which the Panel will endeavour to answer. So Mr. Chairman, with respect, I think the issue that you face at the end of these proceedings is to answer exactly that question. At the levels that are present after the use, does this constitute a level which may result in a biological activity? I am going to, if I may Mr. Chairman, modify your question a little bit to the one I would like to answer, rather than the one you asked. I must say I feel somewhat restrained to make an attempt to answer your question directly for the reasons that I have said. Because as I have indicated, at least in my view, that in the case of the natural hormones these levels that are present in the meat of treated animals is essentially within the range that it would be present in the meat of untreated animals, I would suggest that the question becomes moot as to whether or not there is biological activity associated with these levels of the hormone. I think it is a moot point because the level is no greater or lesser than it would be in the meat of an untreated animal. That is a consumer would be exposed to no lesser or greater biological activity as a result of this level from either a treated animal or from an untreated animal because both levels are essentially in the same range. Thank you.
Mr. Chairman, in relation to the naturally-occurring hormones, then it is not possible to differentiate between the effects of produce from treated animals against produce from untreated animals because essentially they sit in the same biological range. In the case of the non-naturally occurring hormones, in establishing the ADI, then there were very sensitive end points derived from studies in non-human primates, and then the situation of sensitive members of the population was taken into account when establishing safety factors. So the levels that one would get in meat are substantially below those causing any effects in primates, and there is a reasonably good correlation in hormonal levels in, or effects of hormonal levels in primates against humans. So therefore the levels that you would see in meat from animals that were treated with the two non-natural growth promotants would not be causing effects in humans if they consume the meat.
Mr. Chairman, the three natural hormones as they endogenously occur at physiological levels certainly have biological effects of human beings. This is their role and their function. These steroids are amongst the most successful molecules which have evolved and you find them already in very low species. So it is quite clear that at physiological levels they have important functions and they interact with many other hormones and they act almost on every cell in the body. Now, the additional amount
WT/DS26/R/USA
Page 246
you could ingest with meat, either from untreated or from treated animals according to what veterinary practices, cannot modulate this endogenous activity of these hormones because the amounts are too small.
I take that this is all subject to the application of good veterinary practices. It is always the big "if", as I read it in your paper too, that is the assumption.
I agree with that fact and I believe with my colleagues that within the physiological range as these hormones have no acute difference biological effects on human beings. But in the physiological range the problem and the question is to know if when you treat animals or whether you don't treat animals you have the same mean in the physiological range. I mean I believe that if you only eat meat from treated animals, you will have as Dieter Arnold said, a very small enhancement of the mean of your food hormonal intake and this will not be transformed in classical biological effects for human beings. You don't see anything, but the problem is that some other biological effects have not been studied and need a very very very long time to be studied. I think for example, as changes in human fertility or change in sex ratio as we know this sex ratio for example is now changing in some countries. But we have no relation between these countries and the fact that they banned or not these hormones, that is not the problem now. But what I only say is that this change in human reproduction for example, has not been studied nor related to hormonal change or not, and I have personally some doubt on this.
Thank you, Dr. Lucier.
The short answer is yes, some biological effect could occur. If you think about in terms of a normal condition, say a normal woman would have 30 per cent of her oestrogen receptors occupied at some given point in time. At that same point in time, if she is consuming meat that contains an additional burden of oestrogens because of the use of growth promoters, that receptor occupancy may be something like 30.01 or 30.001, a very very small increase. This increase would not be detectable, not even close to being detectable, by any experimental tools that we have today. So the answer is that a biological effect could be occurring, if it is occurring it would not be detectable, and finally the relationship between that biological effect and a toxic effect, say cancer or birth defect or something like this, would be unknown. But if such an effect was occurring it would be extraordinarily small, remarkably close to zero.
Thank you. Dr. Randell would you like to take the floor. Now? A procedural question or? OK.
Mr. Chairman, thank you. During the organizational meeting we had this morning, you said that we would review how the procedure is going after a while and now it is our feeling that it would probably be appropriate, after the United States has posed the three or four or five questions, so probably we can stay here, and then Canada will come up with the follow-up questions in this same area, and then we take up the floor ourselves, because that is going too long and we would like if possible to intervene in this area of residue levels and distinctions between the three natural and the three synthetic.
WT/DS26/R/USA
Page 247
I think that it would be more appropriate and will probably improve the quality of the debate and the discussion we are having here. If the United States can continue with one or two more questions and Canada, and then we can come in at this stage with similar questions in this area which you are discussing now.
Well my personal feeling is that we are in a good process. To me it is very educational the way we proceed and I would prefer to stick to the modes we set out and you can come in with a question when it's your time. You can then also focus on the points you want being less reactive than when you have to follow up to what they are doing. I think, it's my assessment that it's efficient and of course as time goes by, less and more specialized question will become necessary and we will focus and target more down and you will bring in other aspects than the other parties. I would have a preference to continue the way we are doing. I give you the floor.
Thank you Mr. Chairman. The next question is related to the one we were just asking. I just want to make sure I understood the response correctly which is, would the residues found in the meat be chemically identical whether it came from application for growth-promotion purposes versus endogenously produced versus applied for zootechnical or therapeutic purposes? Would it be the same chemical residue, recognizing the levels again would be variable?
I presume that your question is directed of course to the natural hormone?
Yes.
In the interest of expediency Mr. Chairman, I will say essentially yes. Essentially yes.
Mr. Chairman, the residues will be similar for zootechnical uses where quite high levels are used, you can get spill-over or inhibition of certain of the metabolic pathways, and in that way you might get qualitatively similar but quantitatively quite different residue profiles.
This is also my answer, Mr. Chairman, qualitatively the same but quantitatively could be slightly different depending on the compound administered, the dose, the route of administration, etc. There are some differences.
Thank you, yes Dr. Ritter.
WT/DS26/R/USA
Page 248
I am sorry Mr. Chairman, just to clarify, perhaps I misunderstood the question, but was the question directed at for the same use, that is what ...
Could you restate?
Yes, the question was would the residue be chemically the same whether it was endogenously produced, for example, pick testosterone. Whether it was endogenously produced testosterone, testosterone administered for growth promotion purposes or testosterone administered for zootechnical or therapeutic purposes.
Then of course I certainly agree with the answer provided by my colleagues. Qualitatively the residue in my view would be indistinguishable. Quantitatively, of course, the amount that could be detected would be a function of the amount that would have been administered either for growth promotion or therapeutic uses.
I think that the residues are for the greatest part of the residues the same, when used for growth promotion natural hormones and for use as therapeutic zootechnical. If differences appear it can be in metabolic profiles, I mean that with the long term effect you induce some different enzyme panels and this enzyme, after a long time, can produce small amounts of different metabolites as when you are just using these hormones one time. But to my knowledge, this is known for other compounds but not for hormones, but to my knowledge also it has not been studied.
Thank you, Dr. Lucier.
Yes.
Thank you. Well, if you agree you don't have to take the floor, in order to win time that is fine.
The next question I think Dr. Lucier will find it an easier time to agree with because I believe I understood a statement he made that given the concept of half-life, that if you administer one of these hormones to an animal, in this case I am talking of all six of them, you administer a hormone to an animal there will be a residue, even far out into the future, even though it may be only so few molecules it could not be detected by any available means that we have at our disposal currently. I just wanted to make sure I had understood that and whether that was the view or what I had understood from the responses of the other experts as well. Please correct me if I have misstated.
WT/DS26/R/USA
Page 249
Are you asking me specifically? That is correct a few molecules would likely remain, not detectable certainly, but there is every likelihood there would be a few of them left around.
Mr. Chairman, hormones administered from exogenous sources have very short half-lives initially after administration, in the range of minutes. But then the half-lives becomes longer and longer because these hormones go to different compartments in the body and after a few hours, the half-life is already in the order of some 50, 100 minutes, initially it is in the order of some 10-15 minutes. So it is very difficult to predict for how long the last molecule will persist in the body, but certainly for a very long time.
Thank you.
For the sake of the record Mr. Chairman, I agree with Dr. Arnold.
You nod as well?
If you would like a complete record, Yes.
Well it is not necessary, thank you. Can we go on?
I would like to turn now to questions that have been raised in some of the documents that have been provided to you all and make sure that I had understood correctly the responses in some of these instances. With respect to the studies that Dr. Liehr had performed, I want to make sure that we are all understanding that the dose comparisons of the studies, the doses that he was using in his study versus the doses that we are talking about for growth promotion purposes. I just want to make sure that we all understood this the same way, that the dosage used in Dr. Liehr's study was 61 micrograms per day in the male Syrian hamster, which weighs approximately 100 grams. Let me just run through these, then we can discuss this a little bit. That would be approximately 610 micrograms per kilogram per day, if we are doing our math correctly, and for a 60 kilogram adult male human that would be equivalent to about 36.6 milligrams per day and that adult male humans produce about 48 micrograms of oestradiol 17 beta per day. Given that math then, what Dr. Liehr was delivering to the hamsters in question was 36,360/48 times more than the comparable average daily production rate in adult men, and that in turn is 15,000 times more than the residue of oestradiol in meat from treated animals. I just wanted to make sure that we had the numbers there correctly, because then that would lead to result of a difference between the dose used in the treatment of the hamsters and the residue in 500 grams of meat, of a significant difference, it would be equivalent to the oestradiol from 11.5 million 500 gram portions of meat being injected into a human male every day to be comparable. I can run through those numbers again, I recognize that was rather fast.
WT/DS26/R/USA
Page 250
Yes, the Community, Mr. Christoforou.
I think it would be probably more useful, if we can get the written version of these statements and the numbers.
Well, I understood its about hamsterizing humans.
Yes, indeed, but there was some other references to other animals, so we would like really to see some other products, so we would appreciate it if we can get the written version with the exact figures for us to have a look, thank you.
Yes, I assume you want to come back to this comparison later on. But are you in a position to answer or would you postpone this need of figures? Dr. Arnold.
Provided that I correctly understood the question, I can perhaps give a preliminary answer. If a pellet is administered to hamsters in the order of several milligrams 20, 50 or more, you can think about 100, 150 micrograms which are released per day. Therefore some authors in some studies have simulated this situation by constantly infusing similar amounts to the animals. So these are in fact very high doses which are released from such implants.
Professor Mc Lean.
If it's helpful Mr. Chairman, I would take the same line as Dr. Arnold. It is quite clear that the dose administered to the hamster in a pellet was a number of milligrams, a hamster weighs a few hundred grams and to extrapolate from the work in the hamster to the human situation I think is fairly dangerous.
Mr. Chairman, without double checking all of the calculations that were just given to us vis-à-vis the comparisons, I think the point of my colleagues Professor McLean and Dr. Arnold make is the correct one. That is that there is orders of magnitude difference here, between the dose levels that were used by Professor Liehr and those that might result as residues for exposure in the question of meat. I think we need to remind ourselves that the fundamental intent here is different. That is Professor Liehr's work and the work of many investigators is to produce an effect. This speaks to the most fundamental issue in pharmacology and in toxicology and that is the concept of dose response. Professor Liehr is very interested in understanding the induction of cancer as a result of exposure to these hormones, so of course his protocol would necessarily be designed in such a way, so as to produce
WT/DS26/R/USA
Page 251
the desired effect, in his case the tumour. The intent obviously in the case of food residues, Mr. Chairman, is to avoid a dose level which may constitute a human risk, so it will come as small surprise that the levels that are present as residues in food are thousands or hundreds of thousands or millions of times lower than they would be in an experiment which is specifically designed to induce a tumour. These two experiments, if you like, are completely at cross purposes with each other. They are from their very initiation intended to produce entirely different results. So to compare a protocol which has been designed to produce a tumour to a food residue the used practice of which is set up in such a way so as to minimize the presence of the residue, is in my view a relatively meaningless comparison. Thank you.
I am not in fundamental disagreement with what has been said, but I think it is important to point out one thing, that physiological levels of oestrogen do result in oxidative damage to DNA and this has been shown in the rat and so this type of damage does occur in physiological situations in experimental animals and this is a particular form of damage to DNA that's been quantified. So the levels that are normally circulating in the body are producing some of these DNA damaging metabolites. Exactly how it does it, is open to question and with the work of Dr. Liehr which would suggest one way, other works may suggest other ways, but I think it is clear that that amount of damage is occurring and would occur following a small dose because it is already added to an additional body burden in which this is already occurring. But again that additional DNA damage would extraordinary small, again remarkably close to zero.
Thank you. When you talk about physiological levels, may I take it that means naturally-occurring levels?
That is right. Of course these change during the normal cycling phases of people as well as experimental animals, so there is a range of values that occur, but within that normal range of values that you see in experimental animals, which is similar to what is seen in people as well.
Thank you. Any further remarks on this point?
Sorry, I just had a follow up to the last intervention. I just want to make sure I understood. Is there evidence that at the levels we are talking about of residue in meat from the use of these hormones for growth-promotion purposes in accordance with good animal husbandry practice, is there evidence being put forward of DNA damage from those levels of residue?
Let me answer the question in another way, perhaps I have caused some confusion. Say you already have a hundred or thousand molecules of something in the body, and that some of those molecules are producing a DNA damage event. If you add another molecule to it there is some possibility that that same event will occur because it is the same molecule. You will not be able to distinguish certainly that additional event from the ones that occurred from the thousand molecules, but you won't be able to say, no none of those events are related to that additional one molecule. That is what I meant that
WT/DS26/R/USA
Page 252
it would not be detectable, very very small, but it would be impossible to say that that event could not occur.
Can I just follow up for my understanding. The comparison between Dr. Liehr's protocols, inducing cancer on the one hand using very high doses and on the other hand using very very limited doses here. Is it your view that despite one uses very very small doses, there still might be some effects?
Yes, because you are already adding to an existing very large number of same molecules, so just instead of 1,000 molecules you have 1,001 is irrelevant here.
And that is why you basically challenge, if I am correct, that the concept of thresholds...
That is exactly right. I should also point out though, to make it clear from my view, that I do not think it has been demonstrated that this particular mechanism is related or proven to be related to the carcinogenicity of oestrogens. It is at this point an interesting hypothesis, there some plausibility to it, but it has not been proven to be associated or a cause of tumours produced by oestrogen.
117. Yes ....
Thank you very much, sorry perhaps to interfere into the discussion. Just one precision: by adding one molecule, does it mean that there is a linear increase in probability of DNA damage or you think that its progressive? Does it mean that it will perhaps put something more or it is just a very linear function?
It is already adding to an existing burden, if that burden that already exists is on a linear part of a dose response curve, then adding an additional molecule would create a small amount of additional damage. It is possible that that mechanism is already saturated maximal at the physiological concentrations, if that is the case, then that damage would not occur, this is the opposite situation of the threshold.
Thank you. Any further comments on this? No, ok. One more.
I said one moment please.
WT/DS26/R/USA
Page 253
Oh one moment, I see.
Mr. Chairman, a few more questions. I just did want to follow up on this last point. This might get a bit technical so I may have to impose on the experts to help translate this into terms that are easily understood by people like myself, who are more lay person than toxicologist or scientist.
Let's make sure I understood. I believe I understood reference to DNA damage. I just want to make clear. Is DNA damage, in other words DNA adducts, always indicative of a genotoxic effect?
No. Do you want a longer answer? There is a wealth of scientific evidence to show that for many carcinogens a DNA adduct step is essential. For many of the polycyclic aromatic hydrocarbons for example, that occur as contaminants and many industrial processes at present in our air, there is a specific DNA adduct that's been identified that has been shown to be related as close as you can come in a causal sense to the tumours that result. In other cases, some DNA adducts have no mutating capability, so there is a wide variation. Not all adducts are alike, some have very strong mutating potential, some have weak mutating potential, others have no mutating potential at all.
Sorry, would you like to take the floor on this point?
I think it could be interesting to the Panel to note that a very large number of adducts to DNA exists at all times in all cells of the body.
Thank you, I was about to invoke Dr. Arnold's name. I understood earlier that we heard that a single strand break in DNA actually occurs something in the order of 5,000 times per hour, I may not have gotten all this correctly, and that the human ability to repair is about 200,000 events per hour. I just was going to ask Dr. Lucier whether the DNA damage that he was referring to would fall within the range that we are referring to here, that Dr. Arnold had referred to as normally occurring.
I think that the kind of damage that would result from oxidation, catechol oestrogen formation of 17 beta oestradiol, would result, and I am not a hundred per cent sure of that and, I am sure I will be corrected when the other folks start asking questions, but I believe that would be something called 8-hydroxyguanosine would be the primary oxidative damage step. You don't remember that, we will not give you a quiz later, but this occurs in very high frequency. I do not have the exact numbers, but it does occur in very high frequency in the human body.
Thank you.
WT/DS26/R/USA
Page 254
Mr. Chairman, I did not want to suggest to the Panel that we should not take such effects seriously. This is the value of these indicator tests, of these early events, that we get conscious that something is happening to DNA and that we further investigate. The only thing I wanted to say is that single-stranded breaks are per se insufficient evidence, particularly if observed in vitro and you need to do follow up studies.
Thank you.
I would like to add one comment to that. I think that that is a very good point. There are a lot of different ways of looking at genetic damage and the science of chemical carcinogenesis evolves or progresses in itty bitty little steps and each piece of evidence adds to the total degree of evidence that a chemical is acting through that pathway. But simply having a positive DNA adduct test or a positive cell-transformation test in itself is not convincing. What becomes convincing when you have a whole body of knowledge from cell systems, from experimental systems, in vivo, from limited studies you may obtain from human samples, and start putting it all together and that adds to the degree of compelling nature of the evidence. So it is not like now we do not have it, now we have it, it sort of goes in a step wise fashion.
Thank you. Would you like to continue Mr. Brinza?
If there are no other comments. I wanted to ask a couple of questions with respect to the concept of an acceptable daily intake level, which was explained I thought quite well by Dr. Randell earlier. I just would like to make sure that I understood, and from the experts' responses that they were using the concept of an ADI in the same way, which is that it is the amount of a substance that can be consumed daily by humans over a lifetime without an adverse effect.
Yes.
The silence means the others were doing it in the same manner, using the same manner? In that case conceptually would it be possible to establish an ADI for the three synthetic hormones involved in this dispute, trenbolone, zeranol and MGA?
Certainly in my view the answer is yes.
The answer from my point of view is yes. I would just like to reiterate that the end points for both of them took into account sensitive subpopulations, sensitive human subpopulations, such as
WT/DS26/R/USA
Page 255
pre-pubertal children and menopausal women and also that they were supported by data generated from studies or administrations to non-human primates.
I agree with the previous speakers.
Could perhaps Dr. Randell explain why we do not have ADIs on these hormones. Why were they not established?
The difference between the three natural hormones and the xenobiotic hormones I think has been discussed earlier today. Firstly, for the xenobiotic hormones it is relatively simple application of the ADI concept and the ADI paradigm if you wish to establish ADIs for these substances. For the endogenous hormones, the JECFA looked at the available information, found that because these were produced in man and that the levels of ingestion were several magnitudes less than endogenous production, then it was unnecessary to set an ADI. On the basis that the end point would achieve the same result, that there would be no additional risk to the humans.
On the necessity, is that view shared by the other experts? Dr. Arnold?
Mr. Chairman, it might be interesting to the Panel that in the EC we have reviewed the three natural hormones independently from JECFA. At least a result of one such evaluation has been published in the Official Journal. We have set an MRL not necessary for oestradiol two years ago for the therapeutic uses and the whole evidence has been reviewed by the Committee for Veterinary Medicinal Products and the European Medicine Evaluation Agency. Since these are just nine lines, maybe I can read the conclusions of the EC Committees and its legally binding now, published in the Official Journal. "The conclusions of the FAO/WHO Expert Committee on Food Additives, JECFA, that no ADI and MRLs for oestradiol will need to be established, is adopted. Neck and plasma residue levels after treatment with oestradiol benzoate and oestradiol valerate have shown to be within physiological limits. Although it is likely that tissue residues levels will also be within physiological limits, this cannot be guaranteed given the results with oestradiol-hexahydrobenzoate. Still compared to the lowest human production rate of oestradiol pre-pubertal boys, 6 micrograms per day, and compared to the amount of oestradiol in other foodstuffs that are part of the human diet, the amount of exogenous oestradiol that humans will be exposed through ingestion of tissue from treated animals is biologically insignificant and will be incapable of exerting hormonal effect in human beings." So it is about the same line of evidence and of arguments put together by the EC Committee, than has been previously done by JECFA.
Thank you. Is this reference in your paper?
Yes, it is.
WT/DS26/R/USA
Page 256
Thank you.
And if you want I can give you a copy of that Council Regulation.
Please Dr. Randell.
I would just like to state that an acceptable daily intake is a numerical estimate of what might be called a no effect level in humans, with built in safety factors. However, in the broad range of substances that JECFA has evaluated over a number of years, it has on many occasions not allocated a numerical ADI. The reason for doing this is that certain substances are essentially safe to ingest at any level and in these cases JECFA will give an ADI which it calls "not specified". It does not specify the acceptable daily intake. Very common safe food additives such as ascorbic acid, vitamin C, fall into this group. On these three substances were the only three substances in which JECFA has used the expression "not necessary". In this particular case JECFA felt that it was not necessary because of the overwhelming endogenous production by humans of these substances and that to set an ADI would not achieve any additional improvement in understanding of the safety of these substances in humans. So "not necessary" was basically an evaluation of what JECFA was trying to achieve in general in terms of trying to protect health of consumers.
Thank you. Further comments on this or would you like to go on?
Thank you Mr. Chairman. I believe that I have understood previously Dr. Randell and would like to make sure and confirm, that in the instance of the two synthetics that were reviewed by JECFA, on trenbolone and zeranol, that the ADI was set using the methodology that you described in general for an ADI. That ADI has the same meaning for those two as it does for other substances.
The ADI for the zeranol and the trenbolone was established on the basis of a no-effect level being identified in test animals. In these two particular cases, the no-effect level which JECFA selected was the no hormonal effect level for both substances.
Thank you. If I could turn to a subject that was touched on by one of the experts earlier today, I will just make sure I understood correctly. With respect to the particular implantation of five of these hormones under discussion, what factors, if any, make the ear as a site of implantation favourable for use over other sites of implantation?
Would you like to take the floor?
WT/DS26/R/USA
Page 257
I suppose, following the order that we now seem to have set, Mr. Chairman.
Well it is unfair that you have to take it, just feel free when you want to answer.
I think the primary impetus, Mr. Chairman, for the selection of the ear is because it is a tissue which is not normally used to enter the food supply, so it makes identification of the source of the material very easy. I do not want to say ensures, but it goes some very considerable distance to ensuring, that injection site residues, which understandably could be much higher than normal tissue residue levels, are extremely unlikely to enter the food supply. It is my understanding that that is the primary motivation for selecting a site which will never enter the food supply.
And one additional advantage of using the ear, it is very easy to palpate a pellet or an implant under the skin of the ear, whereas if it is subcutaneously on the body then you would never find it. So that is a second way of identifying treated animals. In fact for some other treatments with hormones, but not these ones, it is actually possible to remove the implant from the ear.
If I could just follow up on that, I understood a reference earlier that in fact the implants come with a device which is specially designed to implant the pellet in the ear. I believe I have understood the reference that that device would not be suitable for implanting it through the hide, for example. I just want to make sure I have understood that correctly.
I would be a little bit unwilling to extend it that far. It is just that the "gun" used to administer the implants is such that it makes administration convenient, easy and accurate, in that the pellets or the implants are generally in a cartridge and it rotates around a little bit like a six shooter and so it is a convenient, easy and accurate way to administer the prescribed dose.
Thank you. If I could then ask whether there is any specific evidence that the experts are aware of evidence that using a dose of hormone greater than that approved for growth promotion would lead to a proportionately greater response in feed efficiency.
Generally not. When the sponsor suggests the dose, then the dose that they suggest will give the optimum response and so therefore, within limits, there is no need, no return in administering more than is suggested. The second thing is, however, most regulatory authorities, including JECFA, require that data setting residues does take into account dose rates that are in excess of what is normally used, generally twice at least and sometimes more, so that the effect of overdosing, if you like, or variations in uptake, can be seen.
WT/DS26/R/USA
Page 258
Thank you and then if I could just follow on, I am sorry did you ...?
I was only going to add, Mr. Chairman, that I think Dr. Randell in his opening comments articulated in my view very very clearly, the nature of the balance that one tries to strike in establishing an ADI and an acceptable use practice. I think, in my view, he made the point very very well, that to use too little achieves nothing. That is one does not obtain the desired effect but at the same time results in a residue. To do too much, to use too much, goes beyond what one can anticipate in terms of an effect. So that the correct is the one which is ultimately recommended and that is a value which will optimize the balance between the return, the benefit if you and the residue. That is the level to which we are referring.
Mr. Chairman, in the development by industry of these pellets and another devices, they have been optimized with respect to the dose. However this does not necessarily prevent some farmers to use more implants, for example. But this again does not necessarily cause higher residue levels in the carcass. There are some examples that I have shown in my written response, where the use of up to six implants did not necessarily have the effect that the residue levels in the remaining carcass were significantly higher. So it is a question you cannot easily answer, you have to answer on a case-by- case basis. On the other side, it is quite clear that if you inject twice the amount directly, that then all levelsincrease in plasma and in tissues, not necessarily in a linear way, but they increase significantly, if you double the dose, for example by injection. That is the difference between a slow release device and a direct injection. On the other hand the long term release of high doses may have influences on the pattern of hormone excretion in the animal's body and this is the intention of using such compounds.
Thank you very much. May I just make, oh sorry, yes, please.
Just one comment on this point. It is not so evident that a double dose, even at the same time, will give a better response. But what we observed a lot of years ago in our country, when these hormones were allowed by implants in the ear, is that farmers tried to put another of the same doses but later. They do not wait the withdrawal period and they inject another one maybe at the half of the theoretical withdrawal period, and they had benefit to do this, clearly, because the effect was more longer at the time. Consequently it is clear also that respecting the strict withdrawal period at the end, the residue level was higher. Do you understand?
Yes. My question would be, do you think this was typical for European situation or is that a phenomenon which is observed world-wide?
It is not a scientific question, but then, well I think really it is not typical for any countries. It is a problem of farmer, of respect with regulation and of benefits that they have. I can return through the same question, do you think people drive faster than allowed in any country or in another one?
WT/DS26/R/USA
Page 259
It is a problem of respect of regulation, only. But it is clear that it was the same situation at that time in France as it is now in US, for example. Because one implant was allowed in the ear with a withdrawal period, it is quite the same situation it was just 10 years ago. That situation seems to be the same, I think.
Thank you. Would you like to comment on this point?
I just would like to make a comment. Generally, of course, for zeranol and trenbolone the use of excess would be or could be picked up if the MRL was exceeded, and so therefore there would be a penalty if the MRL was exceeded. And so therefore, farmers knowing that you can analyze for quite accurately, or screen for and confirm, the presence of zeranol or trenbolone are not going to do that. And the other interesting thing is that in countries where the use of these compounds is permitted and there are good educational campaigns for farmers as to the correct use and the reasons why you should not exceed the prescribed dose, and the penalties that exist if you do, the results from residues surveys show that by and large there is no exceeding of the MRL. However in those countries where the use is not controlled and that is no farmer education campaign and difficult to apply a penalty, then we know that the MRL is significantly exceeded. So one of the important factors of legalizing these compounds is to be able to conduct an education campaign and to be able to put in place monitoring campaigns to see that the MRL is not exceeded.
Thank you. Dr. Arnold.
Mr. Chairman, in JECFA we use a term called theoretical maximum daily intake. That means we multiply arbitrarily high consumption figures with MRL figures, to calculate what could happen in the worse case scenario. If you assume all animals are treated, all are at the MRL and under these conditions about five per cent of the ADI is used with zeranol and trenbolone. It was very interesting for me to see that in the documentation of the 32nd JECFA there was an example of very big abuse. It was intentionally done this study, and I tell you what people had done. They had administered a dose more than 100 times higher than the recommended dose. They have administered it intravenously, in six doses over three days, so really it is a scenario you cannot more exaggerate. Under these conditions the theoretic maximum daily intake was increased by a factor of three, so that we were in the order of 15-16 per cent of the ADI. This is my comment, to show the potential consequences of misuse or abuse of these substances. We have some data, and this data had been available to this 32nd JECFA.
Thank you very much. We are past one o'clock and I wonder, do you intend to continue with the questions for some time, I suppose.
Well, Mr. Chairman, actually I was about to say, looking at the clock, that it might be of some relief to you and the Panel and others that we only have one or two more questions, relatively short ones ...
WT/DS26/R/USA
Page 260
If we could finish the US questions and then we would break. Would you agree to that or would you need a break? Well, yes. If you agree why don't we finish your questions and then break. Thank you.
Just a quick follow up I believe to the answer we received earlier. Is there any evidence that implantation of the approved implants at a site other than the ear would lead to greater feed efficiency? In other words, is there any reason to implant it somewhere else to get a better effect?
I think the general consensus would be that providing it subcutaneously, probably not.
One further question, there are several hormonal compounds that are used for as oestrus control in the European Union and among these are medroxy progesterone acetate and allyltrenbolone and another compound methyl testosterone is also used for sex reversal in aquaculture, we understand. The question, though, is whether these compounds are chemically related to the synthetic hormones used for growth promotion in animals.
Mr. Chairman, I like the use of the word related because it provides one with enormous scope. Of course they are related. They are structure-related and they have activity and function which is similar to the hormones that we are discussing here at issue. But of course there are notable differences as well, so they are "related", I think is a good word to use.
Yes, I would agree with that comment that chemically they are different. In total, but they very have conserved parts of the molecule and they perform or stimulate cells in much the same way. Nevertheless, since their parts are a little bit different, they will be metabolised different so you will have some different breakdown products, you will have different, they may go to different parts of the body, preferentially because of those differences in the molecule. So they will act pretty much the same way once they get into a tissue, but the amount that gets into a tissue might be quite a bit different, or its biological persistence in a tissue or cell might be quite a bit different.
Thank you. Dr. Arnold.
Two substances named by the US delegation are certainly chemically related and also what concerns the mechanism of action, there are great similarities. Nevertheless, allyltrenbolone is not useful for growth promotion, it has no anabolic properties and therefore there is an exemption in the EC directives concerning this substance.
WT/DS26/R/USA
Page 261
Thank you.
Generally I am in agreement, except I think we also should note that methyl testosterone is more active orally than testosterone and I think allyltrenbolone is active orally as well. But I think we should note that the oral activity of those two compounds is higher than the oral activity of the chemically related compounds. That is the only point I make.
Thank you. Yes, Professor Arnold.
I agree with what has been said about the relation between hormones they are related. It is a good word. But we have just to compare what is comparable. When we speak about these hormones we have to speak only about the therapeutic and zootechnical uses of the hormone in question and not about growth promotion. It is very different, because they are used punctually (?) on any individual or groups, well defined, of animals so it is very different. It is not the same use.
Thank you Mr. Chairman. I just want to thank all of the experts for their assistance and thank the Panel for their patience.
Thank you very much. This brings us to the end of this morning's session. I have the impression that the learning rate this morning was much higher than it usually is in these rooms and I was very impressed. About timing, could we continue at three o'clock with the questions by the European Community? Could we do it three o'clock sharp, otherwise we may run into night. Do you have a procedural point, Mr. Christoforou?
When would we expect the document by the US on these quantities of micrograms?
Could you arrange to hand them over
Mr. Chairman, we do not have copies with us but we can certainly provide that to you when we reconvene.
Thank you very much. The meeting is adjourned.
WT/DS26/R/USA
Page 262
Mr. Chairman and Members of the Panel, Ladies and Gentlemen, Experts advising the Panel
I think it is the normal practice to allow this delegation to present itself, if you please excuse me, because there are so many scientists from so many places. I think five minutes or three minutes presentation by all the delegation might be helpful as these scientists will be intervening in this debate. Since the EC delegation is known to the Panel, it is the normal composition of the delegation, I would only request scientists which are advising the European Community in this case to present themselves very briefly towards on what they have been until now.
I will start then with Doctor Liehr on my left-hand side please.
Mr. Chairman, my name is Dr. Joachim Liehr and I am a chemist and have worked for the last 17 years on the mechanism of oestrogen-induced tumours. This mechanism I have concentrated on, the genotoxicity studies, because everybody knew at the time when I started these studies that oestrogens are hormones and many people accepted the fact that hormonal agents act as hormones and it was at the same time clear that oestrogens are complete carcinogens. There are quite a number of animal model systems where oestrogens are complete carcinogens without any other agents. So in order to examine this carcinogenicity effect, complete carcinogenicity effect, I wanted to know can oestrogens also act as genotoxins in addition to, not separate of, but in addition to, the hormonal effects that they normally exert. This has resulted in a number of positive genotoxicity effects not just in vitro but also many of them in vivo and quite a different class, many different classes of genotoxicity, so that we can talk in the words of Dr. Lucier of compelling evidence that there is genotoxicity by oestrogens.
I am here because also I have done from a different approach what Dr. Liehr has tried to attempt. When I started my research in chemical carcinogenicity I have a strong feeling that oestrogen is the origin of many human cancers, and it took more than 25 years to resolve the problem. Mr. Chairman, that happened because in order to understand how oestrogen induced cancer, could initiate cancer, we have studied a large class of moderate compounds called polycyclic aromatic hydrocarbons which induce cancers and are present, many of them, in our environment, from the combustion of organic material like gasoline, etc. These compounds have been an excellent model for understanding how oestrogen induced cancer and we are now at the point in which we can say that oestrogens are initiating cancer and by the so-called receptor-mediated process in cancer and produce cancer. Therefore my point in contention is that since they initiate cancer for people that they do not have the protective mechanism that normally everybody should have, even a minimal dose that can imbalance our equilibrium could be a factor in inducing cancer to humans. I would be very glad to express this in a more visual way with the slides later during these procedures if I can.
I am professor of toxicology and I sit on the Scientific Committee on Animal Nutrition for the EU and the UK Veterinary Products Committee. My interest is mechanisms of toxicity and their use in risk assessment.
WT/DS26/R/USA
Page 263
I have been trained as a chemist and worked in field of biochemical toxicology for more than 20 years. My major interests over the years have been the biochemical mechanisms of oestrogen and hormonal carcinogenicity and more recently we have focused on the chromosomal effect of oestrogens as carcinogens, which we think is an important aspect of hormonal carcinogenesis.
I am a pathologist by training and I have been involved in mutagenesis oestrogen and carcinogenesis for the last 15 to 20 years. Now I am at the National Institute of Public Health in Budapest, Hungary. Thank you.
I am a pathologist by background. I have worked for some three to four decades in problems of carcinogenesis and to a lesser extent in mutagenesis. My major interests are in public health and preventive medicine. Of relevance in this connection I was a key advisor to the United States Congress in the early 70s in relation to a series of actions leading to the eventual banning of DES, having analysed DES residue data which we obtained from a variety of sources including FDA and USDA. My testimony was critical to the banning of the use of DES as an anabolic.
I am a chemist and 25 years involved in residue analysis of anabolic agents, all types of analysis, research, development, regulatory, forensic, whatsoever. The last years I am focused on the question of the reliability of analysis, for what purpose and for what price. I am employed with the Dutch National Institute of Public Health and the Environment and involved in EC groups and Codex groups and so on.
I am a pediatric endocrinologist, Associate Professor of Pediatrics Endocrinology in the University of Ponce in Puerto Rico and prior with the University of Puerto Rico Medical Science Campus. I am also an advisor for the Centre for Disease Control and for NIA, National Institute of Health, on the area of diabetes. I am an advisor on diabetes and technology for HEW in the United States.
As you see there is one scientist missing and this is Dr. Adlercreutz, who unfortunately could not join us today but he will be with us tomorrow. He informed me that he would only like to make a very short statement, less than five minutes, when he is in the room and be willing to reply to statements and questions since his paper has been reviewed and criticized both by the United States and Canada. Mr. Chairman, while I have the floor one more procedural issue is that the United States has argued against the submission of any evidence, new evidence in this case. We have heard that this morning in the Preparatory Meeting, and yet we have seen today a fairly long document circulating commenting on the replies of the scientists. According to the rules of the DSU, this document should not circulate unless it is read out. The Community would agree that it is circulated if we are given the chance to provide a similar document within a very short time-limit. This document is circulated so I do not know the status of this document. If it is to remain, I have no objection, but it stays with the files of the Panel, but we are given the chance of three days maximum or four to provide similar
WT/DS26/R/USA
Page 264
documents in this respect. I think that would be according to the rules of the Dispute Settlement Understanding.
Mr. Christoforou, would you give us the rule which prescribes that every document has to be read out?
Mr. Chairman, the rule is that every time we meet you say please give the written version of your oral statement otherwise you cannot file documents with the Panel. If you provide a limited interpretation to that practice I can provide documents without reading them. But I remember the argument was made by the United States in our second substantive meeting where they would like that we read literally the documents that we were submitting. This is the rule and a standard practice, and it is the submissions in the annex giving the table of the dispute settlement, it is the written version of the oral statements will be provided. It is in the annex of the DSU. I can provide you with the exact provision.
We do invite, we have invited you in fact, to submit your comments in writing, but this would not mean that you have to read them out because I usually encourage people to summarize what they hand in in writing instead of reading it out. I think it should not be a problem of you having such a document submitted within a short period of time - three days - okay.
Thank you. Mr. Chairman, we have heard carefully the summarizing that was made by the distinguished experts advising of the Panel in this case and we would like to tacklea number of responses that were given to first their initial presentation and then the way the United States and a number of the questions posed by the United States.
The first question would relate to what we discussed this morning in this room. Physiological levels of residues of these hormones in the human body. Of course this relates to the three natural hormones only which occur naturally, it does not relate to the discussion on the synthetic xenobiotic hormones. It is also frequently referred to in the JEFCA report of 1988, where sometimes reference is made that the residues, for example, of oestradiol, would be increased from two to five-fold but this was still considered to be within the physiological normal physiological levels. In this case and in this respect I have Dr. Epstein who would like to make a two-minute explanation then ask a particular question and I think, I understand from the Secretariat, a paper is circulating that gives the exact calculations of those so-called physiological levels. This is important in our view because we later link this to the most sensitive part of the population - the so-called pre-pubertal boys, and the values that they have given for daily 24-hours concentration of oestradiol, because this is an important issue in our view. So I will give the floor to Dr. Epstein please.
Thank you. I first of all would like to thank the Chairman of the Panel for allowing me this opportunity to address you briefly. Before I commence my brief statement I have just been reminded that I neglected to give you my professional affiliations, which is Professor of Environmental and Occupational Medicine at the School of Public Health in Chicago.
WT/DS26/R/USA
Page 265
The JEFCA document and preceding documents clearly give the impression that residues of natural anabolics fall within the normal physiological ranges as indicated previously. There is also a clear statement that levels in muscle are normal. There is also the statement that levels in pre-pubertal boys are about six micrograms a day, which is roughly a thousand times the residue levels of oestradiol.
I would now like to comment on these statements and point out that they appear to be inconsistent with industry data which I believe are available and have been available to JEFCA and other bodies. Let us look first of all at the data for Synovex S which is oestradiol and progesterone and when you examine levels of oestradiol in liver you find they increase six-fold, in kidney nine-fold, in muscle twelve-fold and in fat twenty-three fold. So that doesn't appear to be entirely consistent with what we heard this morning: namely that muscle levels are not increased when in fact they increased twelve- fold and a twenty-three fold excess for fat hardly appears to be trivial. Now these relate of course just to the levels of the parent oestradiol, and not of course its metabolites, so therefore this represents an underestimate of the true hormonal residues in the meat.
Turning now to the question of the factor of six micrograms a day being a thousand times the residue levels of oestradiol for pre-pubertal boys: this pre-pubertal data, however, does not relate to children and enfants from the age of 0 to 8. Now based on IARC documentation, volume 21, 1979, page 42 to 46, you can make the very simple calculations showing what are the levels of oestradiol production daily for children of this age and it turns out that the levels range from 0.16 to
0.8 micrograms a day. So we are now shifting the levels in pre-pubertal boys from six micrograms to 0.16. This is important for two reasons: first of all the figure of six is clearly misleading and secondly our level of concern is very much greater for infants and young children and children under the age of eight, because there is abundant data on the much higher level of sensitivity to carcinogens of young children than to adults. So for these two reasons, I think it is important to focus on this. In fact the mathematics clearly show that the margin of safety, the so-called margin of safety or the excess over background levels, is not a thousand-fold but say for fat it is eight-fold, less than an order of magnitude. It ranges from eight to 40, less than an order of magnitude at the lower level. So these are some very fundamental distinctions and misapprehensions on these conceptions which I would like to focus on.
All this relates purely to the legal application. The application conformance with standard recommended good husbandry practice. If I may turn for the moment to the illegal, I should like to preface my comments on the illegal by quoting from a reference that Dr. Ritter kindly provided us with; namely the reference to Truhaut in 1985 who stated that the unlawful use of the anabolics can result in residues 300-fold in excess of the established tolerances. I repeat, the unlawful use can result in residues 300-fold in excess of the established tolerances. I have not reviewed the original paper of Dr. Truhaut but I rely on Dr. Ritter's reference to it and the statement - the quotation - he makes from it.
The second point in this connection I should make is when we are talking about illegal or misplaced implants, we are not talking about an occasional liberation by an irresponsible farmer. We are talking about two things: first of all about common practice as I will prove in a moment. Secondly we are talking about a situation in which to all intents and purposes there is no reasonable or practicable way of monitoring whether or not illegal practice is common or not. I refer you to a 1986 USDA survey of some 32 major feedlots in the United States where half the cattle were found to have misplaced implants. I repeat, we are not dealing with an occasional abberation of an irresponsible farmer. We are dealing with a common place situation in which there is no way of monitoring, of residue analysis, of practical residue analysis, so how can we even begin to discuss concepts of ADI if we do not have the slightest level, slightest idea, of the kind of residues. We already know that when administered, when the hormones are administered in accordance with recommended practice, the levels are way in excess of what you gentlemen have been led to believe. But when it comes to the misplaced implants we have two sets of problems. One you have the problem at the implant site in which it could be argued that in fact inspectors regularly check every piece of the meat and if they find an implant they would
WT/DS26/R/USA
Page 266
cut out that affected piece of meat and discard it. It is questionable whether in fact that happens and our friends from USDA may perhaps be willing to give us the assurance that every portion of every animal slaughtered is inspected for misplaced implants. But far more to the point the fact is when you implant, put an implant in the avascular, the non-vascular subcutaneous tissue under the skin of the ear, the rate of absorption is extremely low, but when you implant it in muscle which is highly vascular, you can have a much higher rate of absorption. Now I have searched the literature as thoroughly as I could and cannot find any reference or any data at all levels of hormones of oestradiol at sites at a distance from the implant site. The basic knowledge of pharmacology could make it very clear that when you implant a hormone or any agent in a highly vascular tissue then its rates of absorption would be many-foldhigher, whether it's one order of magnitude, two orders of magnitude, three orders of magnitude - I do not have the slightest idea, but I do know that misplaced implants are not uncommon and I do know from the industry data that the levels of the residues, are way in excess of what you have been led to believe.
I would like now to raise two questions, after the statement of Dr. Epstein, and to all our scientists advising the Panel. If it is correct as we have been making the calculations, and the document is circulating, about the level of oestradiol in pre-pubertal boys, which is not six micrograms per 24 hours but it is eight-fold less than the physiological levels, then I would request the scientists to give us their view: Is this really a risk, an additional or an increased risk of these residues causing the risks we have been discussing here among other things carcinogenicity, to this type of sensitive part of human population? This is the first question, thank you.
Thank you. May I invite you just to comment generally on Doctor Epstein's thesis, because we would like to have your reaction on this? Who would like to take the floor? Doctor Ritter, please.
Thank you Mr. Chairman. I offer to, I foolishly offer to lead only because I think I would like to clarify for the Panel a specific reference that Dr. Epstein made to my submission with regard to the Truhaut reference. It is correct indeed that I cited this reference, but I would submit to the Panel that the reference has been somewhat misquoted. As a point of clarity, Dr. Epstein did correctly note that I refer to the possibility of the 300-fold excess in residues with oestradiol.
But if I may, Mr. Chairman, and I will provide specific reference to the Panel and to the EU, I have brought it along will me. But to put into proper context what the authors have said exactly is that"when 50 milligrams of oestradiol in an oil-injectablepreparation is administered intra-muscularly in the neck of a veal calf three weeks prior to slaughter" and then they go on to talk about this illegal practice, "the levels of residues in the vicinity of the injection site may be 300-times greater".
This is a pivotal issue, Mr. Chairman, because in fact there is not an injection at issue for the proper use. So this particular three-fold increase in residues refers to a multiple of abuses, if you like. It would be at the wrong site and using the wrong method of administration. The 300-fold increase in potential residue does not refer to the normal use. This refers to an extraordinarily abnormal use. The same authors in referring to that very conclusion that they have drawn then go on to say in their conclusion: "It is our belief that neither 17 beta oestradiol nor zeranol, poses any toxicity problem when used as an anabolic in animal production. It is widely agreed that in the instance of hormonally mediated toxicological effects including carcinogenesis, a tolerance level can be set". They finally conclude "even in the instancesin which misuse may have occurred, it is virtually impossible to visualize any hazard to humans ingesting meat from animals treated with zeranol". These are their words, not mine.
WT/DS26/R/USA
Page 267
I make this point of clarity because I did not want the Panel left with the impression that this 300-fold increase to which I referred is something that could be anticipated, if you like, under normal circumstances, and indeed as the authors suggest it is not likely that one could anticipate it even under abnormal conditions of abuse. It simply demonstrates than one can create a scenario where at the injection site, which would not be applicable in normal growth promoting uses, it can be in excess.
Who would like to speak to the question submitted by the Community?
Just to make a comment about the reaction to the implant if it is placed in the ear compared with in the muscle. Of course in those vascular tissues like the muscle then you get quite an active tissue reaction and indeed you get the implant walled off. One of the features of injections subcutaneously in the ear is that you do not get the massive reaction at the site and it is specifically designed, and all the residue studies are carried out, with the implant injected in the ear. The aim of injecting it there is to get prolonged and slow release at low levels over a period of time; so that is specifically designed to be that way.
Thank you. Could you offer any comments on the table we were given and in particular the impact on the pre-pubertal situation? You do not have to speak. Let me know. Yes please Doctor.
Chairman, there is one problem when this information comes so quickly, it takes some time to look at the evidence. The only thing I can do at the moment is to confirm that these figures really stand in the IARC report in volume 21. It is important to note that these values differ tremendously according to the source. So I can't say from where this level of six micrograms came. This I do not know currently, but I have some tables from some text books of endocrinology with me showing that there is a tremendous range and if you read carefully the text in which IARC explains the way how they have estimated, they make quite clear that this is an estimate and that the variability can be enormous and they give examples of individual goods where there was a very large range of data. So the figures are okay, but you cannot insist and say this is the one.
Thank you Mr. Chairman. I offer only two points for consideration by the Panel. I think as Dr. Arnold has already noted we are responding somewhat on the fly, even though you suggested that we need not feel any obligations to do so.
In the table we have been provided, Mr. Chairman, the intake factors that have been calculated are based on approximate level of exposure to 500 grams of meat. This table has been calculated on the basis of pre-pubertal boys under eight years of age. I would suggest to you Mr. Chairman that it is not likely that a child of two or three will consume 500 grams of meat a day. This calculation is based on daily intake in a 500 gram sample of food, every day for his entire life. I would submit to you Mr. Chairman that as a member of the JEFCA process now for some years, I feel that the 500 gram estimate is an over-exaggeration even in the case of adults. But I think to presume in calculating the risks that a child under eight years of age will consume 500 grams of meat every day, respectfully Sir, I think it is somewhat of an exaggeration.
WT/DS26/R/USA
Page 268
The other point that I would bring to your attention and again I apologize for responding on the fly, is I think the point that Dr. Arnold has already elaborated on and that is that the numbers that you see before you are presented as, how shall I put this, as if they were rigid when in fact hormone levels, either present as residues or normally circulating endogenous levels, you would normally be accustomed to seeing them as a range. These do not have rigid values. If you look, for example, at the, let us take the most extreme example of the table before you, if you take a look at fat, the 23-fold excess factor is of course assuming that the five-year old or the four-year old or the two-year old has consumed 500 grams of meat. I think if that calculation were redone with a more realistic figure the 23-fold excess might in fact be reduced to a two-fold excess. But in addition to that, I would suggest to you Mr. Chairman that the test value of 41.4 microgram is a value from a range if you like and that the range is very broadly distributed. Similarly the 1.82 in the untreated control is extracted from a range. Depending on where in the range these two values fall, respectively, may impart the impression that it is high or low, when in fact what we should be doing when talking about ranges is just that, we should be expressing a range rather than a rigid point estimate.
Thank you. Many of the numbers are similar to the numbers that I put together in preparation for the meeting - certainly the range of oestrogen production per day in young boys and girls is about what is reported here from that IARC publication, the range of 0.4 to 0.2 micrograms per day. This is about 140 times, from my estimates, higher than the amount obtained from eating if one of these young boys or girls ate 500 grams of meat containing the maximum level permissible. In addition, if you looked at a pre-menopausal woman in terms of oestrogen that would be about 200 times that is what is normally seen in a young boy or girl, so that ratio instead of one over 140 becomes one in 28,000. So eating that amount of meat per day for a pre-menopausal woman would add one molecule for every 28,000 she would normally have in her body, and I take the point that this is a point estimate and of course that is an average estimate for the women, not encompassing the range that Dr. Ritter just described which is in fact true. But at least a point estimate it would be one molecule per every 28,000 normally present in a woman's body. In addition, probably only about ten per cent of that is going to be absorbed because it is degraded in the stomach, so you can add another factor of ten on that so it becomes one over 1,400 or one over 280,000 for the young boys or girls and the pre-menopausal woman, respectively. But some of these numbers are very consistent with the numbers that I have had - just used differently.
Mr. Chairman, if I could make two comments following on from my colleague. The other thing is that depending on the mode of cooking we can lose up to 80 per cent of these steroid hormones in the cooking process and not terribly many people eat raw meat.
The main point I would like to make was to address the matter of misplaced implants and violations of administration. In my country and in the United States, but particularly in my country, we have targeted in residue surveys trenbolone and zeranol, very specifically looking for violations, and in fact to all intents and purposes violations do not occur. Now this means that even if there is misuse, and that is difficult to prove, we are still not getting residues that exceed the MRL. I would suggest that a similar situation exists with the naturally-occurring hormones, where it is not possible to determine whether a violation has occurred or not because the levels that you find in the carcass sit within the normal range. But I think that you can take some comfort in the residue surveys of zeranol and trenbolone which show to all intents and purposes violative residues not present.
Are there any further comments from the Panel? May I ask Professor Epstein to comment briefly on what was said?
WT/DS26/R/USA
Page 269
Unfortunately American (you may gather from my accent that I am not American) but one of the things that has impressed about American boys is their appetite for "Whoppers" and for hamburgers. It is not at all uncommon for a child of seven or eight to have a hamburger at lunch and a "Whopper" at lunch and another one in the evening, and that is roughly 500 grams. That is the first point.
The second point is we are really dealing with a situation in which we have provided the Panel with data which are entirely inconsistent with data provided to them previously by JECFA. One can try to explain this away in various ways and means but the data are fundamentally different and the pre-pubertal levels in pre-pubertal boys are clearly the margin of safety is less than one order of magnitude. Again, let me also emphasize that influence has been made that misplaced implants are the exception rather than the rule. I repeat, in 1986 a survey of USDA feedlots showed that some 32 major feedlots or feedlots by the USDA, showed misplaced implants in half of them.
My final point is that in the absence of routine monitoring chemical analysis we do not have the slightest idea, we have no information at all, as to whether we have 100 per cent misplaced implants, ten per cent misplaced implants or any number at all.
I meant to ask a specific question, this question of potential misuse, for the scientists' reply, but if they want to come in earlier that is fine with us as well. Because in our view the question about the pre-pubertal boys has nothing to do with misuse. We are talking about proper use of these implants and in that case the JECFA Report said, and the calculation was based on the 6 micrograms per 24 hours and they said the residue that would result is 1,000-fold whereas in this case we see it is eight-fold less than is the physiological level. This issue has to be disconnected from the potential increased risks from misuse on which I will come to later. We are talking about proper use of these implants and what is the sensitive part of the population which may be affected by this. Thank you.
As I understood there were two lines of argument but the scientific data was for regular use. I give you the floor. I am sorry I have to give the floor first to Doctor Randell, I think he was first.
Thank you Mr. Chairman. I would just like to dispel the impression given that the data in the table just presented to us, which is entitled Symtex NADA 9-576 1983 Synovex Steers, was not available to JECFA. Those data are in fact abbreviated data from Table 6 of the JECFA evaluation of 1987 published in 1988. It is one line in one table of the data which JECFA considered and I would point out that the JECFA also examined the variations in the data, that is the JECFA data contains the ranges of each of these data points as well. So the data which have been presented here are taken in isolation from a larger set of data which JECFA considered. Thank you.
Mr. Chairman, in the meantime I have found the tables I have produced in preparation of today's meeting. I used a brand new text book of endocrinology and metabolism published in 1995. What you can easily see from these tables that in fact there is a tremendous range of values. So first of all, whether it is two or six, it is the same order of magnitude. Then one needs to know whether it is an average, a mean or a medium; there can be a tremendous difference because the range and I have all the figures from foetal life, all stages of puberty and adult life, the range sometimes is a 250-fold
WT/DS26/R/USA
Page 270
or even greater and therefore you can fully explain the difference if two people take the same data and one scientist takes the mean and the other takes the medium and therefore what I did - I put together the mean, the minimum and the maximum and here you can see that for some of these hormones there is just a 500-fold range. This has been put together from all sources available in a most recent 2,500 page text book on this subject. If you want the full reference is given and I can leave it for you. It was my personal preparation but if it is of value to the Panel, you may have it. It is not included in my statement. I did it after.
Could it be given to the Panel and we would hand it also to the parties.
Mr. Chairman, we would be grateful of course to have a look into those. At least I can explain our sources. Our sources, and I think there might have been again a misunderstanding, it is on page 19 of the JECFA Report, where it is stated and I quote "even in pre-pubertal boys the amount of oestradiol 17 beta produced daily (6.5 micrograms) is a 1,000 times the amount derived from ingestion of 5 grams of treated meat." We did not look into the tables referred to by Dr. Randell, we looked into the text where this is a citation. Then we tried to find out what the citation was and it turned out that it was an article published in 1974.
Our data, which you see from this paper we circulated, came from the IARC volume and they are based on a more recent article by Brown and Tolle, 1987. So this is our basis of calculation and we would of course appreciate that there are other calculations made that would show a different calculation. But we also take the points that may be arranged and there may be variations but we all based ourselves on what it is in the publication of the International Agency for Research of Cancer.
Mr. Chairman, I can now proceed with posing the other questions. I would like to request because some of my scientists would like to have a precise, if possible, reference on what Dr. McLean is saying, that upon cooking about half of these are destroyed and we do not know of any such reference in any academic reviews. I think it will be useful if we can get the reference so that we can control this but we do not have this type of information and I would appreciate a reference to that.
My next question is linked to the misuse because this is a slightly different issue. Dr. Epstein has made the reference to the misuse that this has been reported in the United States. We have also made the reference to this misuse in our submissions. I will remind only with a number of figures here and our exports and then I will raise the question.
For example, we have examined the national plans, the annual control plans in the United States and it would appear that between 1972 and 1994 there was no control on trenbolone. There was never any control on progesterone. There was never any control on testosterone. On oestradiol it was checked only for the years 1987 to 1990. MGA was checked only for the years 1978 to 1983. then it was discontinued, and started again in 1987 to 1990 and then discontinued and then in 1993 only.
For zeranol it was tested for the years 1973 to 1974. Again then in 1977 and in 1985 to 1989. This has been referred to the Panel it is paragraph 57 of our second written submission of the United States Panel. We have also calculated on the basis of the samples that are examined annually in the United States that, for example, in the year 1993 there were checked only 39,128 samples. On making the calculations what does this number represent in terms of total livestock production the United States tells us this is only 0.005 per cent. I link then these factual statements and I also note that in the year 1993 there were only 22 samples of MGA for MGA tested, 22 samples only and there was one violation. If we extrapolate this for the further amount of livestock that enter the food chain, that can amount to 4.5 per cent of tests positive.
WT/DS26/R/USA
Page 271
The question is because Dr. McLean also said there is no incentive for the farmers to misuse because there will be penalties if they are caught. In light of this, and I can cite similar numbers for Canada, they are in paragraph 11 of our second written submission, if you would like I can make similar references. In light of this 0.005 per cent - check - do you really think it is? - the threat of penalties it is a real dissuasive instrument so that farmers will not misuse those implants, because those implants are very cheap apparently and they may speed the growth rate and we have heard this morning that they may be an additional injection of an implant before the withdrawal period has expired. So how credible really is this disincentive on farmers to misuse in view of the potential advantages of either misusing or adding more implants. Thank you.
Would you like to comment this?
As far as I can Mr. Chairman. The residue surveys that are carried out in the United States, I cannot comment precisely on the numbers, but the violation rate of five per cent is significantly less than the violation rate that was being quoted before of 50 per cent, so there is a variation and I think that the United States delegation could throw light on the levels of violation.
The second thing is that the residue misuse that we were talking about was in 1986. I would suggest now that with the advent of pellets and pelleting administration devices then it is my impression that the implant administration to the ear have improved quite dramatically.
In answer to the question about cooking, I was fairly careful to say that steroids in general there could be destruction of up to 80 per cent. My figures come from a submission made on another steroid for registration and as I said I was very careful to say it was up to 80 per cent and I would suspect that the destruction of steroids by cooking would be similar across the range.
Thank you very much. Any other comments on this question? Doctor Arnold?
Mr. Chairman, I cannot speak on the details of American residue surveillance plans. What could perhaps help, maybe other people have that information, what is the result of European residue control on imports. I have seen that, for example, France does a lot of work in this field, but as far as I know, I saw more than 1,000 samples in 1995 but no positive.
Thank you. I would like to stay with this residue controls and ask what is the situation in the European Community operating under the ban? You need to have controls as much as when you allow it, but how is it done, do you have figures on this? Does any other delegation have figures on this?
Mr. Chairman, it is in our submission brought for the United States and Canada and this has not been disputed as far as I know and I should know because I am always in all meetings present. There are samples tested every year, there are about 200,000 samples of animals tested in the European Community. This is what is the level of testing in the European Community.
WT/DS26/R/USA
Page 272
We also have provided the reply to the question by Canada concerning the natural hormones and we said we test annually 59,000 of blood to verify the levels for the natural hormones. These are the figures we have already given to the Panel and I have so far no comments on this. But this is our official figures.
Mr. Chairman, since I am also from the European Community I may slightly correct this figure. This is the total number of samples for all residues and not for hormones. For hormones the number with all food animals and all substances altogether is in the order of 20,000 and almost half of that figure comes from one country.
Mr. Chairman, for the natural hormones we have verified each individual check done in response to the question by Canada for that we are sure. For the natural hormones the blood samples tested are 59,000. For the total number of checks done it is indeed 200,000 samples of animals tested. We are not going to verify how much of this, our understanding is very much higher than 20,000 mentioned by Dr. Arnold. I will clarify this for the total hormones, but for the blood residues, for the natural hormones checked last year, it was 59,000.
Do you have any indications of percentage in herds? What that would mean?
Mr. Chairman, we need to understand also that the natural hormones are not allowed to be used in the EC for growth promotion. The checks are done because only one of those natural hormone, oestradiol, is allowed to be used for therapeutic or zootechnical reasons. So the information is there is no excess but I do not have the exact figure if there was any violation or if there is any physiological level for the natural hormones. We do not have any figure of violation in the Community because they are not used and they cannot detect high violations in those 59,000 we checked last year.
Thank you Mr. Chairman. I offer just a brief commentary. I of course will not endeavour to speak for the United States or for Canada with regard to how many samples they have sampled but I think it is important to recognize that the number of samples that one takes and measures can be somewhat misleading. 20,000 may appear to be larger than 5,000 for example, but the intent of any monitoring programme is the development of a programme that has the statistical confidence necessary to detect violations to the extent that the compound is used, if in fact residue levels which are out of compliance are present. I would be surprised that either the Canadian or the US scheme has not considered that statistical design in the monitoring programme.
I think the issue of how many samples were measured can really be somewhat misleading. I think the more relevant question is - is the monitoring programme statistically confident and is it able to detect in either country, or in the European Community for that matter, is it able to do what it is intended to do? And generally speaking most monitoring programmes are intended to detect a five per cent violation rate 95 per cent of the time. The number of samples that are required to do that will vary from country to country and from commodity to commodity because it is a function of the use practice. I make this point only because I think it would be relatively fruitless to get into a debate as to whether or not 20,000 is larger than 5,000. It clearly is. But that quite frankly, I do not think, is the relevant issue.
WT/DS26/R/USA
Page 273
I would just add one other commentary perhaps on that. The Council for the European Community has just made reference to the fact that there maybe somewhat less than a rigorous programme for some of the anabolics because the compounds are not used and hence of course, one would not expect to detect them. But the international experience with clenbuterol, for example, which is not permitted for use as an anabolic, is that it is widely used and abused particularly in Europe, with very real cases of adverse effects in humans, in some cases I think death, resulting from the use of a anabolic, clenbuterol, in this case, which most certainly is prohibited for use. So the fact that it is not permitted, I think as you referenced yourself a moment ago Mr. Chairman, certainly is not evidence that it is not used. To draw some confidence from the fact that it may not be approved for that use and therefore should not be present I think is euphoria.
I do not wish to go on in these disputes of a number of samples because it is a regulatory problem, it is not a real scientific one. But I just would like to invite you to observe what is done with the sample, because you can manage a lot of samples for one or two hormones and you can have samples for 30 or more hormones, and in these cases the efficiency of the control is very different.
To my knowledge, in France as well as in most countries in Europe now, we developed very nice multi-residue control programmes. It means that for one sample we check sometimes for 30 anabolics, different anabolic steroids, and in this case it shows that we discover sometimes black-market compounds. It is not that everybody knows it.
But my question is to know when our colleagues say that in their country they do not observe misuse of hormones, my question would be what hormones are you checking for? If it is only trenbolone zeranol they can be sure that there is no misuse of these hormones - these two or three hormones - in that they have no higher residue levels than allowed as MRL. But to my knowledge I do not think that all these hormones are assayed in these countries as well as we do up to now, so they cannot say that they have no misuse of other xenobiotics.
Thank you. May I just follow up one question. I know this is not a strictly scientific question but you have extensive experience in this. Do you consider the control to be more efficient, more feasible, under a system which allows and restricts or under a system which at the outset prohibits and tries to find those who step over the law. Just from your assessment, from your background and experience: I know you have not extensively discussed this in your papers because it is not a scientific question, but I think from the legal point of view it has some relevance. Would you like to speak on that?
I am a little bit reluctant because this is certainly subject to speculation. I would assume, and I expressed this in my paper, I would assume a certain competition between legal products and illegal substances. I would assume that a significant number of farmers would use legal products because they are efficient and not too expensive, but to say this would result in no misuse I think would be pure speculation.
If I may just add one more sentence to the sampling and residue control. It is in fact very important to consider the importance of multi-residue methods and also whether the programme is based on random sampling, statistically-based sampling or on suspicion. In the EC, for example, we have at least these two aspects: random sampling and sampling of suspected animals, plus control of imports.
WT/DS26/R/USA
Page 274
If we look at the random sampling, then the situation is not so bad except for certain substances in certain countries. But if we look at the suspect sampling, then we find all substances including zeranol and trenbolone from our domestic production. The only area where I feel that even under conditions of random sampling we have some problems is with veal calves and the natural hormones. If we look at the plasma levels, there we have sometimes in some countries some problems with a high incidence of violating levels. But here as a scientist, I must say, distribution of physiological levels and of levels resulting from misuse are overlapping to such an extent that this may just be a course because the decision limits have been inappropriately defined to separate the two populations.
May I add something to Dieter Arnold's presentation? We have also the opportunity, more than random samples and suspected samples, to go on farm and to check for misuse during breeding of animals and not only at slaughter house and this is sometimes very efficient tool.
As I said this morning, in France we, concerning the operation, we had four years, or five years, I do not remember, during when these hormones were allowed and after this long period during which these hormones were banned and really I was in charge of control during this two different periods I can confirm that even if in my country hormones were allowed, the misuse of other hormones, the black market was also present based on the idea that when you allow something you have an official label and people have an official label and can have treated animals, but then try always to give more and more and to give different compounds to have more results and we never observed differences in the black market when the hormones were allowed or after. It was so important.
If I can offer my speculation on this. My feeling would be that the violations probably would occur more frequently in cases where it was banned if you are comparing it to a controlled use in which there was well-defined educational programmes, very good communication efforts and appropriately stiff penalties for misuse. If all of those things were in place with the control programme, my guess is that you would have fewer violations in that case - if it was a very systematic approach for controlling them that adequately addressed those and also very stiff penalties for violations.
Can we go on with your questions and comments?
Mr. Chairman, I think one of our experts would like to make a very short statement on the issue of cooking - the percentage of the natural hormones destroyed during cooking - I think he might have a couple of words to say, then I will proceed to the next question. Thank you.
First of all I assume that we are talking about cooking and I do not have to add anything to that. I only have questions. I very recently did a literature research on behaviour of veterinary drugs in general during the preparation in the kitchen - so that is what I call cooking: frying, braising, etc. I could not find any data on steroids but what I found on other types of veterinary drugs in general, ranging from very old work with diethylstilboestrol to more recent work.
The general pattern is that most of these residues are quite stable. Some are lost, not because they are destroyed, but they are cooked out of the tissue when they are fat soluble and so on, so if you do not use the drippings you will lose something. That is the general pattern. So I am very much
WT/DS26/R/USA
Page 275
surprised about the statement of Dr. McLean (sic) that up to 80 per cent or 80 per cent of these residues will be destroyed. I would be very interested in having a reference or the data of that. That is my remark, so it is more of a question.
Thank you very much. Would the Panel like to comment on the effects of cooking?
Mr. Chairman, I made the comment and it was in relation to a submission that was made about a steroid that is commercial in confidence and before I left I approached the manufacturer or the sponsor and at this stage they are still trying to get clearance for me to release the information. I can say no more and very carefully I must say that the cooking methods vary - they involve braising, boiling and the figures varied quite considerably. I can say no more.
Are there any other data on cooking available or not now?
Mr. Chairman, since we have such a wide range of expertise present here, I think, Dr. Liehr has done some research himself and would like to say a few words on this particular point on cooking.
Not on cooking itself, but the process of heating. I have done chemistry, chemical work, chemical preparations of oestrogens which involved heating these oestrogens to 120-140 degrees and we found no evidence for decomposition in various organic solvents. Oestrogens are quite stable products and I cannot imagine an 80 per cent destruction of these compounds under cooking procedures. Thank you Mr. Chairman.
Thank you Mr. Chairman. The discussion puzzles me because there is nothing in the JECFA or Codex evaluation that takes into account possible reduction factors as a result of cooking. In fact the JECFA Codex figures would be applicable to those people, who for some reason of their own, ate raw meat, raw kidneys, raw liver and raw eggs.
So you are saying that basically you would assume that 100 per cent is still in the food which is being prepared and eaten by people.
It was me that introduced it and I think that the important thing to realize is that the evaluations that normally occur through, say for example, the JECFA process, do look at the worst case and the worst case is that you get the full load and it is a benefit if processing removes some of it.
WT/DS26/R/USA
Page 276
I think that what Dr. Liehr has said is absolutely correct. The steroid hormones are so stable you find them in petrol, in mineral oil, they survive even the conditions under which these products have been formed. Mr. McLean was not speaking about steroids, he was meaning another substance.
Yes, I think that is absolutely correct. There is no reason to think that they would be extensively destroyed. The only way you would get rid of them, I think was mentioned earlier, is in getting rid of the drippings which would contain many of the steroids just by extracting them out because they are so fat soluble.
When the artificial hormones are used and added to what is naturally there on stock are they going to stay in that way in the environmental situation given their stability?
Yes, they should stay in the same form. Through the cooking process you mean?
No, just when they are released and they are not destroyed by cooking or other ways and you said they would even be found in petrol. My question is whether the amount of hormones on the globe are just increasing by adding artificial hormones.
Most of the steroids themselves once in the biota and subjectedto living organisms on the planet, those organisms would tend to degrade most of the steroids. They have some of the same abilities to deactivate them as people have so the biota itself is capable of degrading them, so you would not get a gradual accumulation of these in the environment as you would for say some of the more persistent organochlorine compounds such as the PCBs and DDEs and these kind of things which are much more resistant to this kind of degradation. You would not get that same kind of bio-accumulation with the steroids.
If I could add to it cholesterol, which is the basic nucleus upon which the steroids are based, is extensively oxidized on cooking, and in fact it is thought that it is oxidized cholesterol which is the contributing factor to heart disease rather than cholesterol per se. I would suggest that to compare an environment where a steroid is an inorganic solvent, which we know protects the non-extraction compared with the cooking environment which contains water, oxygen and a number of other salts and other compounds would be indeed the ideal environment for destroying compounds. So I do not want to explore it any further - I think the jury is still out, but I am not sure that the steroid hormones in the cooking environment are quite as stable as people would believe.
Mr. Chairman, I do not know if you wish any further comment on this table that was recently circulated but I have just taken a moment now. I am a little slower perhaps than most and I need a moment to absorb the information and try to come to terms with it, but as I look at these numbers
WT/DS26/R/USA
Page 277
now and under No. 2 - this Syntex data - I am not entirely sure what this is intended to represent. I am left with the impression that these numbers are derived from use data which has no withdrawal period whatsoever, because when I look at the food standards programme - the residue levels - which I think we all referenced in our submissions, and I look on page 9 of that, I note that, for example, with withdrawal periods of as little as 24 hours, following implantation with 24 milligrams of oestradiol 17 beta control-released implants and withdrawal periods as little as 24 hours, residue levels were in fact identical to the control - they were not higher at all. If one increases that withdrawal period, in fact the scenario continues - withdrawal periods in this particular study were measured up to 72 hours, which is still a very short withdrawal period and certainly much shorter than what is recommended in the actual use practice, the residue levels for the E2 are identical for all tissues to their respective control. They are not 23-fold higher, they are precisely the same number. So I am not sure I know what the basis for these numbers are at all, or what the experimental conditions are, which are represented in this table.
Would I be correct in assuming that this table does not represent or reflect any withdrawal period whatsoever?
Can I give the floor to Doctor Randell and then Professor Epstein.
If you take the volume of Nutrition Paper 41 which is the residue studies or the evaluation of the residue studies done by JECFA at its 32nd Session, you will find on page 11 in Table 6 the numbers that are in this piece of paper just handed to us. They are the numbers associated with the study done at 15-day withdrawal period. The JECFA table, as I said earlier, puts these numbers in context by giving ranges of the data that were presented. I might point out that of all of the studies, and of all of the withdrawal periods looked at by JECFA and reported in this particular volume, I think this is about the highest set of numbers that you could select out of this group of data. However, they have to be seen within the overall context, Mr. Chairman, of all of the studies that JECFA looked at when it did its evaluation and these data presented to us today are not new data; they were data that was part of the package evaluated by JECFA.
I am glad to have the opportunity of explaining this table further. If you refer to Table 2 under the column marked "test" it says 15D - this is for 15 days. As you know when it comes to implants in the US, there are implants relied on two occasions: one on entry into the feedlot and the other half-way through the feedlot period, but this is a sample taken at 15 days and I should simply say that I simply picked this out of a wide range of NADAs which I have and this is by no means atypical. For instance I have another NADA for Synovex H referring to testosterone where levels in fact were 30-fold times. So if you really would like to have a detailed analysis of all the NADAs of which I have a fair sample, I am sure you could get them all from Syntex, Upjohn and all the other companies and do comparable simple calculations. This is by no means a selected sample, but just one I happened to pick out and I can assure you that there are many others that show similar high ranges and the one on Synovex H is another example: 30-fold increase in fat.
Mr. Chairman, forgive me, I do not really want to take up a great deal more time with this residue issue but it is perhaps an important one. If I refer to the same page, page 11, of the same document, the residue document that Dr. Randell referred to a moment ago, we should perhaps make this specific table available to anyone so that we are all speaking from the same piece of paper.
WT/DS26/R/USA
Page 278
Table 5 specifically refers to residue results in muscle, liver, kidney and fat from both treated and controlled animals following a 63-day withdrawal period after the administration of a 24 milligram implant of oestradiol 17 beta in 12 bulls weighing approximately 850 lbs. In that particular study and reported here in Table 5, the fat values by way of example are barely twice what they are in the control - not 23 times as is reported in Table 2 that Dr. Epstein has presented.
I think the point that Dr. Randell was attempting to make is a very important one. That is that these values represented in Table 2, while they are of course taken from data that has been presented in some context, they are, in my view, at one extreme. That is I refer to Table 5, for example, on page 11, these values are barely twice, and in fact when one looks at the statistical range of the values reported here, one might even argue that there is some significant degree of overlap between the control and the treated values.
I have of course made the point because it refers specifically to my submission with regard to Table 3 in the footnote - Truhaut et al 1985 states that unlawful use can result in residue levels from 300-fold in excess of established tolerance limits. I have of course made the point, Mr. Chairman, but I think this misrepresents somewhat the context of the remark made by these authors.
Are you saying Dr. Ritter that taking a test after 15 days is unusual? That would never be the normal ordinary slaughter date from which it would come into the food chain?
The withdrawal period is more typically in the range of 60 days and what I am saying is that with a 63-day withdrawal period reported here in Table 5 the residue levels, utilizing withdrawal periods which is more consistent with typicaluse practice, the comparison between the treated and the controlled is very much narrower than a 23-fold range as is reported here. It is barely twice.
I think the situation has been for some time the FDA allows a separate implant half-way through the feedlot period. So 63 days in fact would represent the time when the initial pellet for the initial implant is tailing off. So 15 days is a much clearer indication of levels in the animal than 63 days at which time levels have tailed off and furthermore, as I said before, you have second implants. So at 63 days now you would have the initial residue from the initial implant when the animal enters the feedlot, plus a second round of extra hormone added at 50 days. So in fact your 63-day period would be equivalent to 13 days after the second implant.
Mr. Chairman, the table that has been circulated and discussed contains two elements: one is the one we are discussing now, the other element was the level in pre-pubertal boys and since we are now all looking at the JECFA Report in page 15 for pre-pubertal boys the value 6 micrograms per 24 hours a day is given. That is what also what we contest strongly because this is based on data of 1974 and our calculations are on the basis of the Institute for Research of Cancer which are more recent, but this value is incorrect. There are a number of things on this page as well which I hope will be clarified in this case.
WT/DS26/R/USA
Page 279
Are you saying that the numbers that you believe are correct are 0.4 to two micrograms per day as opposed to the six. I believe that the 0.4 to two probably is more in line with reality. I believe you are correct on that.
Could you repeat that again and just elaborate on the impact?
The impact is if you use a number - say one microgram per day that the young boys or girls produce - and if you compare that number to the amount that is ingested from 500 grams of beef you will have a ratio of about 140 above that. In other words, the boys are making 140 times more than what they are ingesting, just from their natural bodies. If you use a number of six, then that will be six times higher than that, and they would be making 840 times what they are ingesting in that meat. So it means that the safety factor would change from 140 to 840. Is that correct?
If there is no argument I can continue ..
Can I just raise the point, Mr. Chairman, I think it might be important - I am not absolutely sure that the figures in Table 6 for the controls are actually correct. I suspect that they are out by one order of magnitude all the way through because we are talking, for example, of muscle at .84 whereas most of the other control muscles in all of the other tables are somewhere around the six, seven, eight mark. Liver is .91 whereas most of the others are 8, 10, 16. I would draw it to your attention, without being able to go any further, I suspect that the control values in that first bit of the table for oestradiol are in fact out by a factor of ten. They differ very significantly from all of the other controlled data and that would not be beyond the realms of possibility.
I think we should go on with remarks and questions.
First this information to Dr. Arnold. It seems that the number of samples tested is very close to 2,000,000 samples per year for the hormones. But this is our own figures. For example in Germany 205,000 checks were made for inhibitor substances for the year 1995. The number is really very high. For the year 1995 we checked, as I said, 69,000 for blood for the natural hormones. That shows that the total number of the checks made for the other hormones is much higher and it is confirmed that it is this figure that was cited at 200,000 samples for hormones is correct.
The next question relates to since now has been clarified what is the level of what we would consider to be normal physiological levels, especially in the area of the sensitive population that is pre-pubertal boys. Then we would intervene in the area of threshold and whether it is really appropriate in that case and indeed it seems in the JECFA Report, that the 1,000 fold as it said higher than what it is in the normal so-called, but is now contested. In that case by fixing the DAI and by suggesting and recommending the maximum residue limit, in this case, they did not recommend because they thought the JECFA Report may fall within the physiological level, but that is not correct; the second
WT/DS26/R/USA
Page 280
reason was that they cannot control and check. In that case, and also this is linked with the argument made about the existing physiological levels are already dangerous; that was suggested by one of the five experts here. Dr. Liehr would like to intervene in that case with respect to threshold and genotoxicity and he will have three questions to ask in that respect.
I would like to start out with genotoxicity since this was already earlier brought up by several of the experts because, as you remember Mr. Chairman, it was said that the genotoxicity was tested largely only in vitro at too high doses and that the data were irrelevant. When I started out the studies in the early 80s, the general consensus was that oestrogens, which had been known at that time already to be carcinogenic in several animal models, that oestrogens were carcinogens by a mechanisminvolving solely receptor mediator pathways. In other words that oestrogens acted only the way hormones act as we understand it.
I started out with these genotoxicity tests at higher doses because indeed, as pointed out by Dr. Ritter, the aim was to establish whether any genotoxicity could be found in the meantime many other laboratories and many other researchers participated and contributed in this and there are several or multiple classes of genotoxicity established for oestradiol - the natural hormone. And just to ... [tape ends]
Single strand breaks, Dr. Arnold, has been determined by us in vivo and not in vitro as you asserted this morning. It has been determined by Dr. Nater and Abul Haj in Minnesota in cells in culture, in breast cancer cells in culture, 8-hydroxy radical DNA damage to guanine bases(?) has been established by us in vitro and in vivo chromosomal abberations have been established by Dr. Karl Barret in vitro and also in vivo, by Dr. Jonathan Li whom you cited this morning Dr. Arnold. Furthermore, DNA adducts have been determined and established in vitro by Dr. Cavalieri, who can certainly elaborate on this later on than, in vivo. We have also determined in vivo DNA adducts by lipid pro-oxidation products that are generated by oestrogen administration. So there is a whole variety of different types of DNA damage of genotoxicity established for oestrogen and I would submit to you Mr. Chairman that this whole variety of DNA damage constitutes what Dr. Lucier this morning called compelling evidence for genotoxicity of oestrogens. Now the JECFA report in 1988 considered for genotoxicity tests mainly bacterial assays. This is an assay test called the Ames Test and it is well known and well established that oestrogens are not mutagenic in the Ames test. This is a bacterial test system where many carcinogens have been tested for mutagenic activity. Now when Dr. Ames introduced this test in the 70s, a number of carcinogens were tested and a high correlation between mutagenicity and carcinogenicity was established. And so when oestrogens were tested and were found not to be mutagenic it was generally assumed there is no mutagenicity and there is no genotoxicity. However, as we have been testing in the Ames test thousands and thousands of compounds, carcinogens, it has been determined that the Ames test is actually a very poor predictor for carcinogenic activity. The mutagenicity test that I reported to you for oestrogen by myself and by other laboratories have all been done in mammalian systems and I submit to you, Mr. Chairman, that these mammalian genotoxicity tests are much more important than the Ames test that has been done in bacterial test systems, we are quite different from bacteria. Now there is another point that I would like to bring up that is very important. This DNA damage by itself might or might not be highly significant. These tests were done at elevated doses and the impression was given to you this morning that testing genotoxicity at elevated doses was highly unusual and was a skewing of the scientific evidence and Mr. Chairman this is not the case. Because carcinogens are routinely tested, initially at elevated doses and when carcinogenicity is found then people test at lower and lower doses to find out whether this is still the case. The same is true for genotoxicity initially as it is tested at higher doses and then later on lower and lower doses are tested. And these tests have only begun, we have determined genotoxicity at elevated doses, I agree. They need to be tested at much lower doses to determine if at the levels that are found in beef, genotoxicity exists and I submit to you that the parties that discuss this case might want to fund such studies because it is difficult to obtain funding for this type of work.
WT/DS26/R/USA
Page 281
One additional point I would like to make, we have demonstrated earlier on or postulated earlier on, that oestrogens are carcinogens by a combined pathway - both their hormonal action which is stimulating cell division and genotoxicity. Now this is an important fact that oestrogens can have both these qualities or both these characteristics, because a cell which has been damaged genetically can then at the same time be stimulated to divide and this represents a fixation of the mutation mechanism. And these things need to be explored in the future further, but the preliminary evidence exists that such possibilities exist. At the moment all we have is the genotoxicity data at elevated doses, yet how can we establish that at "physiological levels" these carcinogens present no risk when oestrogens in line with many other carcinogens are genotoxic at elevated doses and we have not determined how low a level is safe. So I would like to end with a quote. Dr. Arnold has quoted Dr. Jonathan Li and I would like to end with a quote from Dr. Karl Barrett, a colleague of Dr. Lucier at the National Institute of Health in North Carolina and that quote comes from the Journal of Environmental Health Perspective' s review article of Dr. Barrett's in volume 100, pages 9-20 1993. Dr. Barrett states that "It can be concluded that there is significant evidence that certain oestrogens can also cause genetic alterations by a mechanism not involving the classical oestrogen receptor, [in other words, the physiological hormonal pathway] and that hormonal carcinogenesis is most likely a result of the interplay of both genetic and epigenetic factors", exactly what I have been arguing. So based on this, I would like to conclude that it is difficult to establish that low doses are not genotoxic when these experiments have not been done.
Thank you very much. May I ask the Panel to comment on what was said and to tell us whether they believe this is somewhat far-fetched or whether there is reason to believe that there is a potential that such effects could occur within the ADI levels which have been defined. Dr. Ritter?
Thank you Mr. Chairman. I will attempt to begin perhaps. I think Dr. Liehr makes the point that when referring to the studies at the elevated concentrations and he quite correctly points out that these are high concentration compared to what one might expect in terms of exposure to food residues. I think he is quite correctly in order that the relevance of these high dose effects to actual human risk remains to be evaluated.
But he is not talking about high doses now. He submits that future trials have to take them down and that there is reason to believe that the risk exists and now I would like to know what your reaction is on the potential of having the risk within the accepted levels because this seems to be a new series of trials.
I would offer only two very simplistic views, Mr. Chairman, in that regard. The first is to refer to the arguments that were advanced this morning by Dr. Lucier, and that is that at the very low levels that we are talking about in terms of food residue exposure, the contribution of that residue to the already existing biological load that we normally carry in our bodies, I think that the example that Dr. Lucier used was as if we are adding one additional molecule, in one instance he referred to every 28,000 that are present. I do not think anybody can reasonably assure you that the presence of these residues albeit at low levels constitutes zero risk, in fact I am not even sure that is a fair question to ask, scientifically it would be impossible to ever test to a certainty. But I think the point that Dr. Lucier attempted to make is that if these low levels do constitute a risk, they constitute a risk of a magnitude which approaches zero. This is the order of the magnitude of risk that we are talking about for these very low levels and the example that he referred to as I say is that it may constitute one additional
WT/DS26/R/USA
Page 282
putative molecule in the presence of 28,000 others. Additionally though, I would remind the panel and in fact I think the argument is made in one of the Canadian submissions, that particularly in the case of oestradiol, we have what I believe are some very relevant human experience, never mind trying to model from bacterial or mammalian genotoxic systems or from animal studies for that matter. We have a population literally of hundreds of millions of women who have taken these compounds, as I mentioned this morning, in many cases for periods of 25 to 35 years. In spite of that exposure study if you like which has now gone on for a period in excess of a generation and populations which would certainly be large enough to detect an increase in cancer risk if it were associated with these compounds, I think that it is fair to say that while there has been reports in literature reporting or purporting to report such an increase from time to time, that the weight of evidence with regard to increased cancer risk in association with the use of steroid oral contraceptives is that they do not constitute an increased risk. The best testimony of that conclusion which I have just drawn is the fact that their use continues largely unabated, in spite of the fact that all of this work has gone on now for, I'll say 35 years with regard to the risk of the use of oral contraceptives in women throughout the entire world. There has certainly not been any restriction on prescribing this medication for women, there certainly has not been a reduction in use, in fact if anything, and I don't have the numbers in front of me, but I think use in general has somewhat increased over a period of years. And this is an experiment conducted in the ultimate species. Thank you
Thank you, Dr. Arnold.
Mr. Chairman, this morning and obviously this afternoon I will not make any attempt to challenge the excellent quality of Dr. Liehr's work, this is not my problem, he is doing excellent work and he should continue doing this work because these are interesting possible pathways he is trying to elucidate. What my problem was this morning and still is that I feel that there are still too many missing links. For example, if you would say catechol oestrogens or the 16-hydroxyestrone whatever they bind in vivo to DNA, OK that would be strong evidence. But in your very excellent review article in Annual Review of Pharmacology and Toxicology of last year, you write, yourself, you write that you were unable to confirm these data in vivo. You say you that it could not be demonstrate in vivo using the post- labelling techniques. So for the moment at least, I think we exclude this. I am also not at all opposing the idea to further investigate the probable pathway of creating reactive oxygen species, but since you were referring to the work of Nater, I have seen there that they for example say in one of their publications in Chemical Research and Toxicology, while the reactive oxygen theory in oestrogen induced DNA damage is a reasonable one, there is no direct evidence utilizing oestrogen quinones and the physiological conditions that support this hypothesis and so on. So I see there are still some missing links or gaps in the evidence. Also if I have correctly understood Dr. Cavalieri's work, but we will later hear from himself, he has synthesized these quinones and he has synthesized with more or less chemical methods the possible adducts and this is quite an interesting piece of information because now we have reference substances for further research. And I see in your paper which you have for example published in Chemical Research and Toxicology last year, that you were saying in conclusion the adducts described here provide insight into the type of DNA damage possible when catechol oestrogen quinones are generated in vitro and in vivo and will be used in studies designed to elucidate the structure of oestrogen adducts in biological systems. So again I have the feeling that it is quite attractive and if I look at these substances with the eyes of the chemist, I would immediately agree with chemical methods you can produce all these things. The question to me is does it occur in living cells, at what concentrations, what are the enzymes involved, what is the compartmentalization of these enzymes, etc. etc. In some of your papers, Dr. Liehr, you were talking about indirect DNA adducts, I am not sure whether I have correctly understood these papers, but might it be that these were adducts of malondialdehyderesulting from stimulation of lipid peroxidation? What do you think of the mechanism of stimulation of lipid peroxidation by oestrogens? So this is just a small number of questions I have
WT/DS26/R/USA
Page 283
and I reiterate it is not the purpose of my statement to criticize in any way the quality of Dr. Liehr's excellent work, but I feel there are many gaps in the evidence. I can't exclude as a scientist that at the end somebody you will show that oestrogens can act directly on the gene, nobody can exclude. For the moment I don't see that the evidence, the convincing evidence. Thank you, Mr. Chairman.
Dr. Lucier.
Yes just a few comments on Dr. Liehr's comments. One is on the ability of non-genotoxic materials to cause cancer. We have analyzed at a national toxicology programme 500 bioassays and approximately one third of them that turn out to be positive are negative in the salmonella test for mutagenicity. So that statement is correct. There are many carcinogens in animals systems that do not cause mutations, so this statement is correct, and that is about one-third of the ones that we have tested, the 500. The other points that I want to make, one relates to the receptor mediated response versus the genotoxic response. The receptor mediated response being the one in which an oestrogen binds the oestrogen receptor and stimulates the growth, the division of cells that already contain a mutation. There is a great deal of evidence in the scientific literature that supports that as a critical event in the ability of hormones to cause cancer. So I think that, and Dr. Liehr says that that is clearly an important aspect of hormonal carcinogenesis. In addition he is proposing that genotoxic events may also be facilitating the carcinogenic process of hormones. I would say the evidence, although not weak is not as strong in that case as it is to the receptor mediated pathway. That does not mean that it is not occurring and not playing a role, but is not as compelling at this point in time as the evidence is for the receptor mediated pathways. But if say if it is an important event, I think Dr. Ritter gave my example, 28,000 molecules of oestrogen being naturally produced to the body for every one that was introduced by meat containing the growth promoting agents. Once the body sees those molecules, it doesn't differentiate, it isn't that there is the one from the growth promoted animals, it treats all 28,001 as identical so the chance of that one moleculebeing converted to a genotoxic molecule is equal to that of any other molecule of oestrogen that the body produced. So it just gets thrown into the mix and I think that may be an important point to consider. So even though it would be unusual for an event to occur in that one molecule, it is possible, and we said earlier we can't rule out that not occurring, even though it wouldn't contribute much because the number of molecules are so small relative to what is naturally occurring. The other point with breast cancer and oral contraceptives that was made, there has been a number of studies looking at breast cancer, I don't know how many maybe 15 or so and some of them show an increase, some of them don't show an increase. If you add them all up you don't know what to make of it. If you add them all up you get about a 15 to 20 per cent increase in breast cancer in women who use oral contraceptives but this not statistically significant. So that may just be occurring by chance. Epidemiology studies are very insensitive in picking up the changes in tumour incidence especially when the tumour incidence occurs in very high frequency. Since one out of ten women get breast cancer, a ten per cent increase or five per cent increase or two per cent increase is significant in terms of the number of cases, but this could never be picked up by epidemiology studies, they are simply not sensitive enough. It is not like the case with diethylstilboestrol, we had a very rare tumour so it was very easy to pick it up. The vaginal adenocarcinoma in female offspring who had been exposed at that age and during their gestations, so that generally occurred in low frequency and was very easy to pick up. Changes and incidences of breast cancer are very very hard to pick up from epidemiology studies.
WT/DS26/R/USA
Page 284
Thank you very much. We would propose maybe we close the debate at this point and we would briefly have a break maybe for 10, 15 minutes. I was informed that the coffee bar closes at 5:30, so ... you make it and then we will continue perhaps until 7:00 which is the usual time. ........ Well, during the Uruguay Round it was even later! But how shall we proceed, would you like to comment on this question?
Somehow I think Dr. Liehr would like to make a comment and then Dr. Cavalieri and Dr. Epstein would like to intervene on this particular point.
OK, so we have three interventions briefly. Thank you.
First of all Mr. Chairman, considering the comments of Dr. Ritter, you said that you quoted this case of one additional molecule over 28,000 and that this constitutes a very low risk and that the risk is approaching zero at these low levels. I would like to submit to you Mr. Chairman this is pure speculation, we have no genotoxicity data at low doses. This is exactly the point that I am making. The experts make statements that they cannot back up with data, and until we have data I would like to first see this carcinogen oestrogen tested fully before large populations are being exposed to this. Secondly, I agree concerning Dr. Arnold, I agree with you that at the moment this is an interesting hypothesis and I have never labelled it more than a hypothesis and I also agree with you that many pieces are missing but this is exactly our point. I think that in face of a substance which where there are many troubling aspects, we should proceed carefully and we should proceed in a very measured way before we storm in and permit the exposure of wide populations until these troubling questions have been clarified. Concerning the indirect DNA adducts and the stimulation of lipid peroxidation, Mr. Chairman is one aspect, a non-genotoxic aspect of action of free radicals in cells induced by oestrogen metabolites. The stimulation of lipid peroxidation nevertheless is very important because it re-routes or alters metabolism, including metabolism of oestrogen, and permits the formation of reactive metabolites at much much higher levels. So they are processes that are set into motion and then feed on themselves which have really insufficiently been examined and we have really only started to scratch at the surface. I agree with you but nevertheless they are troubling aspects that really need to be examined in more detail. Concerning Dr. Lucier's comments, that in the debate of receptor mediated responses versus genotoxic responses, thegenotoxic hypothesis may not be as compelling at the moment, again in the absence of data I don't know whether we can make such a judgement or statement at the moment. It is possible that genotoxicity is elicited only in the relatively rare instances, but let's please not forget at this point we are dealing also with a hormone and so when a hormone stimulates cell proliferation and a mutation gets fixed, Mr. Chairman, I agree with Dr. Lucier that because we breath oxygen and because we are exposed to a wide range of genotoxic substances this is a process that is occurring on a daily basis within all of us. However, and Dr. Lucier made an argument that these oestrogens would provide an incremental increase that is so small that it is negligible however, since oestrogens are also hormones, the hormonal stimulation of cell division then may well fix the mutation and induce tumours. So at the moment in the absence of more data I don't see how we can arrive at a safe passage for oestrogens.
I would like to comment in answering Dr. Arnold that, I thank him for having acknowledged just the tip of the iceberg of all of the problem. So practically what he has recognized is that these
WT/DS26/R/USA
Page 285
metabolitesof the oestrogen can react with part of the nucleic acid and this has been very vital in creating the standard compounds that now we have demonstrated in vitro and in vivo, as you can see in my report that I have submitted to you. The basic principle, Dr. Arnold, on which that it is very difficult to demonstrate the genotoxicity of oestrogens has been there because we didn't know what a chemical carcinogen is. That has been the big problem and we know now what a chemical carcinogen is since two years and this is the basic paper that now has been published in 1995 in the Proceedings of the National Academy of Science. A chemical carcinogen is a compound, and we have demonstrated that for the polycyclical aromatic hydrocarbon that is huge class of carcinogen and can be extended to compounds like aflatoxin etc., is a compound that produces in excess what we call apurinicides. It means that it takes out of the DNA two bases. One is called adenine and the other one is called guanine and this process is so rich in the sense that it is so high compared to the normal depurination that we have everyday. You know that a carcinogen becomes carcinogenic, it initiates cancer, when these processes of repair of these apurinicides is not any more longer possible. By adding this basic knowledge on aromatic hydra-carbon, we have found out the same formation of apurinicide by a specific metabolite of the oestrogen that is called 4-catechol oestrogen quinone. These they bind to the DNA forming an amount of apurinicide. These we have demonstrated that in vitro and you can see from the report and in vivo by injecting the catechol on the mammary gland of the rat and finding out this kind of atom. We think as Dr. Lucier has mentioned in the morning that the adducts that they form apurinicide inside they are the most mutagenic adducts, that they are consistent with the definition of a chemical carcinogen. This is the initiating event is cancer and I think and I not only think, I am sure because of uniqueness of these phenomena of forming these depurinating carbons that oestrogen they are the origin of breast cancer as Dr. Liehr and other they already have very strong compelling evidence for prostrate cancer and for other human cancers. Therefore, I really think the major problem that we have in our lifetime is that we have the oestrogen. When the oestrogens are placed correctly, you know in the normal human being they are placed correctly, they are very well controlled, they are all beneficial, but if something goes wrong in the control because we have all protective mechanisms in order not to go on the wrong place. But if this protective mechanism for some or for some reasons or for complex they don't work, these oestrogens can become chemical carcinogens and initiating the series of events that lead to cancer. What in hormonal carcinogen is being studied and what the data that here are based the same data as you call it, ADI etc., is based only on the hormonal effect of the receptor mediated process. The initiating defect can occur in a non-protected person, in a person very likely that when he is 30 has breast cancer because this protective mechanism they don't work, and this can come with a very minute amount of oestrogen. You know because the protective mechanism they are not there to impede that the wrong pathway occurs. This wrong pathway is the metabolism of oestradiol to catechol oestrogens. These two exist, two kinds of catechol oestrogens. The most abundant in general is called 2-catechol oestrogen is not dangerous, it doesn't produce cancer. The other one that is called 4-catechol oestrogen if oxidized further to quinone can induce cancer and that is the thing that we have to take into account. Unfortunately, or fortunately, this research is on the making now but is going to be any moment out and the reason that we worked on a different paradigm of other people creates the problem in communicating here today and communicating in general with the scientific community. But you know, I think that all human endeavours work on different paradigm and at a certain moment the one that they count they have to be considered. Thank you for the possibility.
Thank you very much, Professor Epstein please.
Apart from the fascinating technical details of the genotoxicity, I think that mammalian genotoxicity data which we have been talking about for the last half hour, I think that the importance is, really relates to the fundamental fact that in the FAO/WHO and JECFA documents the basis for classification of oestradiol as epigenetic was largely derived on thenegative Ames test, negative bacterial tester and because of this ADIs, the concept of ADI and thresholds crept into the whole of the
WT/DS26/R/USA
Page 286
documentation because you can have thresholds for epigenetic carcinogens. Now I would submit that the evidence for that is to say the least shaky and it's high time that the FAO/WHO documents reflected the growing body of data on mammalian genotoxicity. Even at the time of the JECFA document, there was a body of data on this which was in no way reflected, I think that's the first point I would make. Secondly, let me mention three of the most potent known human carcinogens: benzene, asbestos and arsenic, for a long time these were classified as epigenetic because they were negative in the Ames test, but subsequent work over the last decade or so has shown the mammalian genotoxins. In fact in one of them, a Dr. Legator, a colleague of Dr. Liehr has titrated the levels with mammalian genotoxicity tests and shown genotoxic effects at the lowest level possible, 40 parts per B, which isn't very much higher than when you go to a garage and fill your car with gasoline and stand next to it and sniff it, you'll get it. So in fact, in the one or two instances where genotoxicity has been titrated down, we find no effect levels similarly to what we know from our vast knowledge of radiation biology on linear dose responses. Over and above that, in view of the fact that the negative Ames test was used to establish the concept of thresholds, I should point out that there are various other issues that have to be taken into account when you want to consider setting thresholds. One is the question in this context of the interrelationship of the fact that you give two different anabolics together, two different hormones together, like oestrogen and progesterone which we know have additive or synergistic effects. It ignores the synergism between the hormones and between pesticide residues, which some of which are carcinogenic and some of which oestrogenic. It also ignores the high fact of the sensitivity of children. Nowhere in the JECFA document is there any reference to the high sensitivity of infants and children. Furthermore, to say that a compound has a threshold it puts the onus on those that say it to produce bioassay data to confirm that. In the absence of bioassay data you can't say there is a threshold if you haven't titrated it down and found a no effect level. Let me move very briefly on to two other issues, namely, the differentiation between the hormonal and the proliferate effects of oestrogens and the carcinogenic effect. A great deal has been made of the fact that the carcinogenic effects of the oestradiol is directly related to the hormonal things. I would suggest that we should re- thinkthat. For instance, oestrogens increase the incidenceof salivary cancer, thyroid cancer and melanin cells cancer or melanotic cells in humans and in fact other oestrogen receptors, I don't know. Are these tissues reproductive? Certainly not. Are they proliferative? No. So the whole concept of hormonal, of carcinogenesis being due to hormonal proliferative effect needs very very careful re-examination. Coming back to the oral contraceptive, I would agree with Dr. Lucier that it is difficult to do epidemiological studies when you are dealing with a relatively common cancer. However, I should point out to him, that in the October issue of Contraception, there was an analysis of every single study done on the carcinogenic effects of contraceptives, oral contraceptives, in relation to breast cancer. Over 60 studies were analyzed and metoanalysis and detail every single possible permutation and combination was studied. And with due respect and contrary to Dr. Lucier, what was found was as follows: that when you focus on adolescents or women starting to take oral contraceptives at 15, 16, 17 which is not uncommon now, and taking it for prolonged periods of times, there was a highly significant increase in rates of breast cancer, a very highly significant increase, this is the largest compilation of studies ever analyzed. Let me end up by making one comment on reproductive cancers. Dr. McLean this morning made some comment to the effect that there hasn't been any increase in reproductive cancers. With due respect Sir, this is quite contrary to the fact. Let's take first what testicular cancer. Testicular cancer in the United States based on National Cancer Institute and SEER data show somewhere in the region of about 150 per cent excess than age standardized data from 1950 to now, and when you concentrate on an age group of 28 to 35, the figure comes 280 per cent, nearly a threefold increase in testicular cancer in that period of time, number one. Number two, breast cancer. To suggest that the increase in breast cancer rate is due to improved detection mammography is largely incorrect and the evidence for this is as follows: in the United Kingdom, large scale mammography has only relatively recently started. We've seen the same rate of increase in the United Kingdom as we see in the United States. In the United States large scale mammograph started in about 1981 and
82. If you look at the data from 1950 to 1981, you find that the rate of increase in breast cancer is perhaps the same or a little bit greater then that from 1982 onwards. As far as prostate cancer, so in other words, to suggest that there hasn't been any major increase in reproductive cancers is entirely contrary to an established database from the National Cancer Institute and Surveillance Epidemiology
WT/DS26/R/USA
Page 287
and End Result Programme. As far as prostrate is concerned, there is no question that there has been a significant increase in prostrate cancer. However, I would agree that a certain amount of this is due to over-diagnosis and particularly with the PSA test and other things, so one cannot exclude that there is an element in the increased incidence of prostrate cancer due to over-detection. However, there is also an increase in mortality rate and mortality rate can't be an expression of improved early diagnosis. So the points that have been made, that there are no data whatsoever relating oestrogens and human cancer, exogenous oestrogens and human cancer, is contrary to a very substantial body of epidemiological data, which in the time available to me and I've had to summarize terribly briefly and terribly hardly. Thank you Sir.
Thank you very much. Well I think we will briefly break here so you can still make it and we would inform the cafeteria that there will be a queue. Could we resume at 5:45, 6:00 please, the meeting is adjourned.
Over to Dr. Lucier to comment briefly on what was said before the break. I'd just like to indicate the intention of the Panel. We intend to conclude with the part of the European Communities as a response to the United States for this session, because tomorrow we cannot have a meeting in the morning, because the experts are not available due to other commitments. But I would like to assure that we start at 2:00 with the statements and questions by the Canadian delegation so that we have enough time for this round and then the final comments by the EC and then the experts. So I would like to urge the experts on the EC delegation to have in mind that we should finish by tonight with their presentations which is now being scheduled in this first block here and I hope this can be done by 7:00-7:15 so please keep this in mind. Okay thank you. Dr. Lucier, can I give you the floor in response to the statements by the EC experts before the break.
Thank you, Mr. Chairman. I just wanted to make a few comments regarding Dr. Epstein's presentation. One regarding the presence of the oestrogen receptor in various tissues of the body. He had made the point that there were oestrogen caused tumours in numerous sites. It is also true that oestrogen receptors are found in every tissue of the body, so there is no reason to think that oestrogen action isn't occurring through a normal receptor mediated pathway. That doesn't mean genotoxic events aren't also occurring, but it's clear that classical receptor mediated events would be occurring as well. The other point relates to the increased incidences of some tumours occurring in various countries in the world. Regarding testicular cancer, I think that it's probably true that there is a real increase in testicular cancer and this increase appears to be predominant in young men which is especially disturbing. Why this is occurring, no one really knows, but as a public health person it is very disturbing. There is no reason however to think that this has to be associated with oestrogen, there are a lot of other kinds of things that could be causing that, so there is not a direct link to it, although that increase is in fact a real one. Regarding breast cancer, that is also likely increasing. However, one can't necessarily attribute that to exposure to exogenous external oestrogens. There are several studies in the literature which suggest that or demonstrate that exposure to genotoxic agents during the time that a woman is a teenager causes a dramatic increase in breast cancer. This was shown in atomic bomb survivors from Japan. Those women who survived the atomic bomb, who were teenagers at the time, had very elevated risk of breast cancer later in life, women who survived who were in their 20s or 30s had no increase in risk and there is very plausible biology for this which I won't go into. It's also been shown now in three publications that women who start smoking as teenagers have a higher increase in breast cancer later in life, whereas if they start smoking later, they do not. So there are a lot of reasons why exposure to genotoxic agents could be accounting for the rise in breast cancer rates that are seen throughout the world. So we need to be cognisant of other factors and not
WT/DS26/R/USA
Page 288
just blame the exogenous oestrogens for everything. Regarding the oral contraceptive issue, I think Dr. Epstein is probably right that there would be an increase in breast cancer risk for women who start taking the pill very young because that extends the period of time in which they are exposed to high levels of oestrogen and this is a known risk factor for breast cancer. The same exposures a little bit later probably wouldn't have an increase, so when you average everything out together you don't get a statistically significant increase in breast cancer from oral contraceptives. I don't think that anyone is challenging the fact that oestrogens are carcinogenic, IARC is classified conjugated oestrogens as a known human carcinogen, they have classified oral contraceptives as an ... [tape ends]
... carcinogen based on an increase of liver cancer not on the increase in breast cancer. They have also classified tamoxifen as a known human carcinogen because of its oestrogenic activity in the uterus. I don't think that anyone is disputing that fact that oestrogens do cause cancer.
Thank you.
Just by way of explanation, I did not mean to mislead and I can see how that appeared. I was talking in the context of tumours relating to the use of hormones as growth promotants, and I just make one observation that the increase in tumours occurs in those countries where the use of growth promotant is not widespread and the increase began before the hormonal growth promotants received widespread use. But it was in the context of the hormonal growth promotants rather than an increase in cancer per se that I meant to convey, and if I misled then I apologise.
Thank you. Mr. Christoforou can I come back to you. Oh Dr. Arnold, you wanted to speak as well, sorry.
Just a short remark, Mr. Chairman. Dr. Liehr and Dr. Cavalieri have proposed interesting mechanisms how oestrogens could induce cancer by genotoxic mechanisms. The Panel should know that other well-known people working with other models came to different views. For example, I mentioned Dr. Schulte-Hermann who is a well-known expert in rodent liver carcinogenesis. He says for the hormones to exert their action, there must be some pre-neoplastic lesions already present in the animals, and although his evidence is also not complete, he has quite a lot of data supporting his ideas. He for example has shown that the incidence of tumours after oestrogen treatment depends on the age of the animals at the time they start the treatment. And he has developed experimental techniques to reduce the number of pre-neoplastic lesions by restricting the diet, and he could show that under these conditions the incidence of tumour formation was lower when he started with old animals he kept on the restricted diet and then administered oestrogens. So there is some evidence in other models that perhaps there is no need that oestrogens are complete carcinogens but the first step could be an unknown initiation, the formation of pre-neoplastic lesions and then in addition the hormone promotes the further development. Thank you.
Thank you very much. Can we go to your next comment and question?
WT/DS26/R/USA
Page 289
We have been addressing until now, the potential at least genotoxic effects of the natural hormones. But three of the hormones in dispute in this case also are synthetic. They don't occur naturally by the other animals. So we would like also to address the potential genotoxic and carcinogenic effects of the synthetic hormones. I would give the floor with your permission to Dr. Metzler who has produced a paper and who will after this short presentation conclude by framing a question to the scientists. Thank you.
Thank you Mr. Chairman. Although the synthetic oestrogens or I would rather call them xenobiotic oestrogens and androgens, trenbolone and zeranol have been far less thoroughly studied in terms of their genotoxicity, they have most been subject to routine toxicological testing which is difficult to pick up the genotoxicity of hormonal carcinogens. Nevertheless, there is some evidence that these xenobiotic hormones also are genotoxic and I'd just briefly summarize the evidence. There are data from three different laboratories showing that there is DNA binding at a low but significant amount or degree with a covalent index of about 10. There is induction of chromosomal damage in mammalian cells by trenbolone and other by the non-hormonally active isomer diethyl trenbolone and there is evidence for cell transformation in vitro in different laboratories. For zeranol there is also evidence for DNA binding with a similar covalent binding index and for closely related compounds, zeralenone there is recent evidence of DNA adducts in vivo in mice in different organs. So although there are not many studies available, the few studies indicate that there is also genotoxic potential for these xenobiotic hormones and I would like to finish this brief presentation by addressing a question to Dr. Lucier. Dr. Lucier has indicated that for the natural hormones it does not make much sense or he called it irrelevant to discriminate between genotoxic and non-genotoxic compounds. How would you view this issue for the xenobiotic compounds?
Before I answer, you have to answer one of mine before I answer your question Manfred. What is, at the risk of getting too technical, what is the evidence for the DNA adduct formation for zeralenone or zeranol? Is this simple binding of radioactivity to a DNA fraction or has altered nucleoside isolated and characterized?
For zeralenone or the microtoxin, or the micro-oestrogen I should probably say, there has been a recent study with the post-labelling assay indicating in vivo in mice both in the kidney and in the liver a level of adducts, of various adducts as shown by the post-labelling assay at the level of about 1,000 nucleitides, modified nucleitides per 10(?) nucleitides. So it's such a spining of discrete adducts that shown for zeralenone. No such study has been done for zeranol to my knowledge, for the growth promotant.
One other quick question. Is it possible that P-30 post-labelling which doesn't characterize an individual adduct, is it possible that that is just arising because a change in self-proliferation rates because when a cell divides, of course, it loses its DNA adduct, or if it doesn't divide it keeps its adducts. So a measurement of the endogenously occurring adducts could lead to an alteration in the apparent number but not be related to the zeranol itself. In other words, so that adduct hasn't been characterized it's just been seen as a spot on a TLC plate?
WT/DS26/R/USA
Page 290
Well these adducts have been seen as spots but there were controls on untreated animals of course and there were also species differences like in rats. None of these adducts were seen with zeralenone, they were only seen in mice. So there is a species difference which also is in favour of real adduct.
Sorry Mr. Chairman for that digression, but I needed that information to answer Dr. Metzler's question. The adducts that may arise from synthetic materials are likely different and of more concern than the ones, in terms of a risk assessment, than the ones that arise from the naturally-occurring ones, because we already know that there is a given body burden of the naturally-occurring oestrogens. So whether DNA adduct formation is occurring or whether its cell replication you are just adding a very small amount to an existing burden, whether it be genotoxic or whether it be non-genotoxic so that mechanism is irrelevant as Manfred said for the naturally-occurring oestrogens. That is not necessarily true for the synthetic materials and in fact an adduct or DNA damage has been characterized and identified this may be starting out from ground zero. We have no kind of damage like that is currently in the body, so there the issue of a threshold versus linear models becomes much more important in terms of a hazard identification. So that in fact an adduct has been characterized and I would probably say at this point the data is suggestive of that. That a DNA adduct is occurring that's related to the zeranol itself or zeralenone, but not convincing because that adduct has not been identified in terms of the altered nucleoside. In other words it could be just be altering the occurrence of naturally-occurring adducts is a possibility. Although the date the Dr. Metzler has briefly presented is suggestive that such an adduct independentof the naturally-occurring adducts is in fact being formed.
Thank you very much.
Mr. Chairman, I think we have reviewed the potential to toxic and carcinogenic effects of now all the substances in question. The next question is related to the response the distinguished experts gave to the question whether the JECFA report of 1987 and 1988 is taken into account this type of evidence which you have been hearing this afternoon. As you already noticed, this evidence has started appearing quite late, I would rather put it roughly, mid 1980s and it is growing very fast in this area. The scientists have replied that all these aspects of genotoxicity, synergistic effect, long-term exposure to these substances have indeed been taken into account in the JECFA report. But there were very strong differences in my views. For example Dr. Ritter would say that they have not addressed mutagenicity for all hormones, they have not addressed definitive studies for genotoxicity or carcinogenicity of combinations although this is the preferred method for use. The other scientist, Dr. Arnold for example, would say that genotoxicity and carcinogenicity have been examined and Dr. André would say that synergistic effects have not been taken into account. I would appreciate if all the scientists could summarize once again whether this synergistic effect, the long-term exposure and the potential effects or increased effect from the use of combinations have indeed been taken into account in the JECFA report. And if they have not been taken into effect what is the consequences of that? Do we have to review that report now in view of that evidence? Dr. Ritter has said there is no compelling evidence that we need to review it today. In his written reply he said the weight of evidence would not appear to suggest that there is no, that they are genotoxic and carcinogenic. So there are some variations and slight nuances in the expression. So I would appreciate if we can come back on this issue by hearing the five experts on this.
WT/DS26/R/USA
Page 291
Did you get the question and could you, who would like to start? Dr. Ritter.
Thank you Mr. Chairman. In the written comments which I offered with regard to the xenobiotic hormones, I think that it will be evident to all that MGA for example, has not been subjected to a JECFA evaluation. As I indicated earlier on in the day, my assessments were drawn primarily, let me go further, they were drawn entirely from information which was published in the open literature and obviously dominated by JECFA reports and other ones. It is clear that MGA has not been subjected to such an evaluation and consequently, I have no information other than of a proprietary nature which I did not use, from these evaluations to offer any public opinion on MGA. In so far as the carcinogenicity testing of the combinations are concerned it will also be evident to any reader of the JECFA reports, that the assessments and conclusions offered therein are based on the single compound that were evaluated by the Committee and it was in that context that I offered those views. When I concluded that there is no compelling evidence to suggest that these compounds should be immediately re-evaluated, it was on the basis again of my understanding not only of the historical reviews which have been carried out by JECFA and other organizations, but indeed how much more recent reviews, including those published as a result of the European Conference on Growth Promotion in late 1995, early 1996, and indeed on the basis of the statements presented by our colleagues here today, Dr. Liehr, Dr. Metzler and others. I think as Dr. Liehr indicated, the work which he has carried out recently the work which others made out recently, the work which others have presented recently, suggest that there are circumstances under which adverse effects may be demonstrated in association with the multitude of these compounds. But I think the scientists in my view have quite correctly noted that their relevance to the doses at which this debate focuses is unknown, and I'm simply suggesting that that is not in my view necessarily a reason to argue that it should therefore be immediately redone. In other words, let me put it another way, if Mr. Chairman, an organization like JECFA or indeed a national regulatory authority, were to immediately conclude that every evaluation that it had carried out should always be subject to an immediate re-evaluation at any time that a report indicating that there are circumstances under which an adverse condition can be demonstrated with a particular compound, then I would respectfully submit we will always be re-evaluating everything that we had ever done. Because work goes on, not only on these hormones but indeed on every substance that we can think of. But as scientists, I think we are compelled to look at the totality of the evidence as it's available. I think that totality of evidence, certainly recognizing the information that has been presented here today as well as the historical information, suggest to me that the assessments that were provided then continue to assure a reasonable degree of safety to consumers of these commodities. As I say this is not only my opinion but in fact it was the conclusion of a conference held very recently and sponsored by the European Commission on exactly this issue. That these compound when used in accordance with accepted practice, I should perhaps use a phrase that we have all been using, good veterinary practice, good practice in the use of veterinary drugs, whichever acronym you would like to use. I think this has been the consensus view of this Conference if you like. I use the word consensus because clearly in any scientific endeavour of this sort one can quite properly and realistically anticipate that there will be a divergence of opinion. The nature of scientific interpretation is that legitimate bona fide knowledgeable scientists may reach different conclusions from the same set of data. But I think the consensus opinion of that Conference Mr. Chairman, was that the weight of evidence used then continues to prevail now, and that the assessments and conclusions drawn then are still consistent with the available information now. I think as Dr. Arnold has indicated, this in no way is intended to cast any doubt or validity on the work which others such as Dr. Liehr have performed. And certainly this sort of work should continue, and I don't think anybody can assure you that in some time down the road, that it will not require or dictate that there be a re-evaluation. But as we sit here in this room today, I continue to be of a view that those assessments served the community well then, the international community, and they continue to do so now. And that there has not been compelling evidence that has been provided to suggest that those assessments have been in serious error.
WT/DS26/R/USA
Page 292
Thank you very much. Professor McLean.
Thank Mr. Chairman. I'll be brief. I'd like to highlight the fact that in my submission I made no comment about melengestrol acetate because there hasn't been a large amount of data package available for evaluation and so I restricted my comments to the other five hormones. The only other comment I would make that if any national body or group believe that the time for re-evaluation of some of the hormones previously evaluated by JECFA was upon us, then I would invite them to make an approach to JECFA as has been done on a number of occasions with the number of compound of varying chemical classes. It is interesting that if this information that has concerned some has been so compelling, to ask why an approach to JECFA has not been made.
Thank you very much. Dr. Arnold.
Thank you Chairman. Concerning the table I have presented in my written contribution, I just wanted to facilitate the work of this Panel so that you don't need to read all these documents and so I summarize the kind of information, I was saying information pertaining to the effects have been considered by the study, it's on page 9. I have not stated that I felt that all this evidence was complete at that time, I explicitly mentioned that some data had been completed only at the 34th meeting of the Committee. This was only to facilitate the discussion. What concerns the six substances, I made no statement on melengestrol acetate because I haven't seen the data. The only statement I made was on the availability of analytical methods because this is something I know, but otherwise I have limited my statements on the five endogenous hormones. I did however say that from my point of view, the significant new evidence which has been produced since that Committee did not invalidate the basic conclusions and therefore I still am feeling very comfortable with the conclusions although I admit that a lot of new evidence has been produced by the scientific community. Coming to the question of combinations the synergistic and antagonistic effects of whatever, I had somedifficulties to understand this question because every time if we administer fixed combination to animals or humans, it's quite clear and in all countries participating in this dispute there are mechanismsthat these fixed combinations are evaluated. But I didn't understand the significance of these questions when we are talking about the residue levels because at those concentrations, all these substances occur simultaneously in human beings of all sexes all ages and it's quite obvious that these substances have both synergistic or antagonistic effects in living animals including human beings. This is their normal physiological function and they can synergize in one tissue in one cell with respect to one effect and they can antagonize each other. Therefore I had difficulties if these substances enter the internal pools of the hormones they are immediately confronted with all kinds of hormones, so I didn't understand what the question meant at the residue level. At pharmacological levels of course they are tested in all states if they are administered as fixed combinations.
Could you Professor Epstein precise the question?
I am sorry that my point that I raised was so obscure. I would have thought that if you add two carcinogens to an animal or to a human, you stand the risk of interactions of one kind or another.
WT/DS26/R/USA
Page 293
There are data showing that are positive interactive effects whether they be additive or synergistic such as those between oestrogen and progesterone. With oestrogen and progesterone together, produce a far greater increase in mammary cancers in rodents than do either separately. So that is clear what one means by synergism. I'm unaware of any evidence of antagonistic, any experimental evidence of antagonistic effects of any anabolics and would be happy if you could refer me to the appropriate citations.
I agree with Dr. Epstein that there are examples of synergistic effects but I could list a number of situations where these hormones counteract. I'm not prepared at the moment to give you such references but I'm sure I can do it tomorrow.
Dr. Lucier.
Probably the best example is the one that you just cited; oestrogen and progesterone in which oestrogen unopposed by progesterone causes an increase in uterine cancer. If progesterone is given at the same time then that effect is blocked. So where as oestrogen and progesterone seem to synergize each other in terms of breast cancer risk, they actually antagonize each other for uterine cancer risk. So it just points to the fact that's very difficult to fully appreciate the complexity of trying to estimate synergistic responses, additive responses or antagonistic responses. Even the same two hormones in different tissues will do exactly the opposite, one synergize, the other antagonize.
Thank you very much. Can I give you the floor?
We will speak about the six hormones and not of MGA, it is clear we are just speaking about the other ones. I think that concerning this question when JECFA has considered these hormones, they have been considered as drugs, as classical drugs and for this reason, all the data concerning toxicological assays has been asked to companies and evaluated by very well scientific expert groups. But I think they have not been evaluated as for the reviews for growth promoters, I mean with a very large scale use. I think interference between this large scale use of hormones and the other drugs used also in animal breeding has never been studied, and its clear that the use of hormones has been demonstrated to longer some withdrawal period and elimination rates of other drugs. The probability to have interference between two drugs when they are used for therapeutic use is very small because you have no opportunity to have at the same time two different illnesses and two drugs at the same time. But when you are using such compounds in large scale as hormones, then you have a higher probability to see on the same animal at the same time the use of these hormones and use of another drug. And I think in this particular case of growth promoting substances, the assay has to be done. And concerning the new scientific evidence we are speaking about very recent data from this morning altogether and I would just make two remarks that many of these data concerning mode of mechanism of action carcinogenicity and other are recent data from '90 to now and most of them have been published during 1995 and 1996, the two last years. So this could also explain in some ways that many of these data, recent data have not been taken into by the EC Conference in '95 because the bibliography and the investigation has been done for this Conference up to the beginning of '95 and not more recently.
WT/DS26/R/USA
Page 294
Can I give you the floor?
Mr. Chairman, Dr. Liehr would like to make a comment on the synergistic effects and then Dr. Bridges would like to intervene on the so-called 1995 EC Scientific Conference what was ....... during that Conference. Then if you allow me I will summarize once again what was said I may be making probably slightly legal but directly related question to what is the JECFA standards in this case and what is the legal context in which we are litigating in this case. I think that will be important for our scientists. Thank you.
Mr. Chairman, concerning the synergism of hormones, I agree with Dr. Lucier who stated that the combination of oestrogen plus progestin inhibits uterine tumour formation whereas it appears to enhance mammary tumour formation. It certainly does so in animals and the epidemiological evidence in humans is also quite strong. In this context, I think it is very important for all of us to realise that when several hormones are combined and given as a combination that the balance of hormones within the body, the regulation by various hormones appears to be altered or influenced in ways that we have not sufficiently taken into account. If for instance progestin inhibits uterine tumour formation then and enhances mammary tumour formation, then it is obvious that circulating within all of us are hormones in a quite well-defined balance and as humans proceed through life, through stages, the endocrine regulation is quite well balanced. Once combinations of hormones are added this balance is being destroyed and I think an obvious example of how this, effects, synergistic effects can arise, is the fact that the metabolism of oestrogen is clearly or has clearly been shown to be inhibited by progestins. So combinations of hormones they easily alter the balance between these, between endogenous hormone levels and for the many synthetic progestins, I don't know to what extent this is known. A number of synthetic progestins have been examined but certainly not all of them. Thank you Mr. Chairman.
Thank you very much.
Mr. Chairman before giving the floor to Dr. Bridges, I would like to clarify one point and Dr. Stephany will intervene just afterwards. Dr. Ritter has said, JECFA evaluated only single compounds, single substances. From what we know in the market, the implants, there is only one implant that is one substance only marketed. All the others are implant of compounds of similar substances and I think Dr. Stephany would like to say a couple of words about this. So in fact, what it is administered to the animals in this case, they are no single substances except one. The majority of all the rest of the implants are several compounds together. That I think Dr. Stephany would like to say a couple of words on this.
Yes? Dr. Ritter please.
WT/DS26/R/USA
Page 295
I don't want to pre-empt the comments, but in addressing the issue of combinations, I'm not sure if you are dealing with the issue of animal safety, who presumably would be at greatest risk from the combination, if the combination was to produce synergistic effects? Because I'm left with the impression, perhaps these comments will be addressed, that in terms of human safety, the issue is the exposure to the levels which are present as food residues, which indeed are present in combination with hundreds of other substances, not only these two alone. So I'm not sure if the comments are intended to be more relevant to the issue of safety in the cow, or to the issue of safety to humans who consume commodities from which the residues may indeed be present in combination but at very much reduced levels and again in present in combination not only for the two steroids in question, but indeed present in combination with hundreds of other residues of both natural and exogenous origin.
Thank you, thank you very much indeed. Because there is one aspect which is not practically discussed in this room but has been argued in the submissions of the parties, especially in submissions of the European Community. This case concerns both the facts on human beings of course, but also on animals, and we did not elaborate because the effects on the animals are even more obvious than what the effects are of this residue for human beings. But Dr. Stephany will also given you some information on animal health also in this case, that is the prostate of animals that are affected. Our question is not only the potential effects of the animals but also the potential effects of those combinations for human beings. And the link that was made, has been explained, is since the implants which are used on the market, apart from one the others are compounds of several substances, this is linked to our question, what are the possible synergistic effects of this combination? Has this been studied by the JECFA Report? That was the aim or the point that we were trying to make. But I give the floor first to Dr. Stephany then will come back to Dr. Bridges. Thank you.
Thank you Mr. Chairman. First of all I have to make one very small correction, of the implants that are on the market, there are two having a single active component but only one having oestradiol and the other one is zeranol. So there are two, one is oestradiol the brand name is Compudose and the other one contains zeranol, xenobiotic and the name is Ralgro. All others as far as I am aware are combinations of having at least one xenobiotic hormonal active compound in it. What surprises me a little is that we always talk about, that the implants contain compounds that are identical to the natural ones. That's simply not true because the implants contain either the natural ones in chemical different form like an ester like oestradiol benzoate and or the testosterone as a testosterone propionate. And at least testosterone propionate what is xenobiotic circulates freely in the body as is known from clinical chemical studies, and I'm not aware of any residue studies of testosterone propionate as such and it is certainly not true that all testosterone propionate or all oestradiol benzoate only will hydrolyse and will yield the natural identical hormones. So I think that's one thing and coming or responding to the remarks of Dr. Ritter, if you have two hormonal active components in your implants you will have two at least two hormonal active residues. So the consumer will in his food have, in between a wealth of other things, I fully agree, the probability of being served with two hormonal active hormones or residues at the same time, at least two different ones. Well what's next? Oh the remark on animal health, it's only a small remark. Some recent research in the Netherlands with, picking up old research, we know that if you treat bull-calves with oestrogenic compounds then at least a few organs will change like the prostate and with some new technology based on histo-chemical its a new thing, we found that at least this effect is much more extended than we have found previously. Whether that will effect the health of the animal as such, I can hardly say especially for calves only having a full life of half a year. But at least it's another signal that a lot of things change in the body of the animal, besides that more meat is produced. And also that's an item that is touched in the European Conference, ranging from animal health to animal welfare, but that's completely another thing. I think that's it for the
WT/DS26/R/USA
Page 296
moment, and once again I think the most important thing is that the implants do not contain in most cases natural identical compounds.
Thank you could. I'm sorry I missed before Dr. Randell he wanted to intervene. May I give you the floor?
Thank you Mr. Chairman. I would not like the Panel to go away with the idea that JECFA did not evaluate the residues present in animals from these mixed implants. The monographs which were prepared by JECFA at the 32nd session clearly indicate that trials were done on mixed implants as well as on single substances implants and exactly the sort of mixtures that we have just heard, that is, oestradiol together with testosterone; oestradiol together with progesterone; oestradiol benzoate together with testosterone propionate and also oestradiol with trenbolone acetate under the oestradiol evaluations. These were known to JECFA at the time and the pharmacokinetics of these substances as they were gradually metabolized and excreted by the animals were studied by the experts at that time. Thank you.
Sorry, Dr. Arnold.
Thank you Mr. Chairman. Dr. Stephany is absolutely right that the esters are administered. However, this is also the case in the European Union. If these substances are legally used for therapeutic purposes because, and I think here the rules are entirely right, it is recognized that the esters are readily hydrolyzed to yield the substances. Therefore it doesn't make a great difference if we have talked about oestradiol, testosterone and progesterone because these esters are in fact readily hydrolyzed. The reason why the esters are used is that the free substances are more slowly released because we discussed this earlier today. The plasma half-life is so short. However, in order to give you an idea how ineffective these substances still are, when testosterone propionate was used in earlier times as androgen substitution therapy, it was necessary to inject 25 mg. every day intra-muscularly, so although he is from an academic viewpoint entirely right, I think we can continue discussing the effects of the free substances. Thank you.
Thank you very much. I give the floor back to the Community.
Chairman, I'd just like to clarify the nature of the December 1995 Conference because several speakers have referred to it as if it was an up-to-date comprehensive risk assessment on steroids. It wasn't really of that nature at all, it had nearly 400 participants, which is a pretty difficult way of going about risk assessment. It was covering three areas, development of new growth promoters generally, risk assessment across a range of growth promoters and then monitoring and surveillance. So there wasn't and the way the papers were identified was that we had a small organizing committee who chose the speakers. Now in their wisdom or otherwise, they chose not to have any body who was involved in genotoxicity and so as a consequence, and I was a plenary speaker for risk assessment so I remember the brief well, as a consequence this issue of genotoxicity was not addressed at all by the conference and therefore the conclusions of the Conference did not take it into account. So it isn't and was never
WT/DS26/R/USA
Page 297
intended to be a comprehensive risk assessment and it didn' t really consider any other mechanism because of the way the speakers were selected other than just to look at the hormonal mechanism and how that may be related to cancer.
Can I ask you, the fact that you did not invite speakers, was that an indication that this was not part of the mainstream thinking at the time, or?
Chairman, I wasn't a member of the organizing committee but they were clearly very wise because they chose me as a plenary speaker so I couldn't argue with them. I think really the thinking was that they wanted to follow up to the so-called Lamming Report, and the Lamming Report had concentrated very much on hormones acting as hormones or hormonal mechanism. So I think the natural assumption was that they would get an update on those sort of issues rather than look at broader issues and I've taken the opportunity to talk to members of the organizing committee recently to confirm that that was their view.
But could it be said that if this would be organized today they would probably take a different approach? Is that, I mean these issues we're discussing have they moved into the debate?
I'm sure Chairman that if we were organizing the same conference today, genotoxicity issues would loom very large in the discussion and I'm sure I would have had to refer to it particularly in my paper.
Thank you very much. Dr. Arnold.
Thank you Mr. Chairman. I totally agree with you that this Conference has not produced the slightest element of risk assessment, I totally agree with you. But I have recognized that some important information has been disseminated including your own very important key lecture, but I wanted to underline this Conference has not produced the slightest element of risk assessment.
Thank you very much. Dr. Ritter please.
Thank you Mr. Chairman. I think it may be useful, a number of us have referred to the report, I have the report with me which I brought at the risk of great personal injury to transport it ... [tape ends]
It is addressed, particularly for Dr. Ritter, Dr. McLean, Dr. Arnold and of course, Dr. Randell.
WT/DS26/R/USA
Page 298
Thank you very much. Could I start with you Dr. Randell because the procedural questions are addressed to JECFA.
Thank you, Mr. Chairman.
First I'd like to say that the JECFA Secretariat, of which I had the honour to be a part for a period of my career, would feel very disturbed if they were aware that data were available to the scientific community that were not available to JECFA, or were not being proposed to be available to JECFA, in a manner that would render JECFA's advice to FAO and WHO member countries, either obsolete or inappropriate.
Therefore, my comment earlier, was really to maintain, what I believe to be the standing excellence of JECFA's evaluations in all of the fields that JECFA deals with, in order that, we could perhaps encourage those scientists that have these data to assemble them and bring them to JECFA for review. It was meant as a positive, not a negative, statement.
A few minor corrections to the history which we've just heard. The vote which took place, in which the hormones were failed to be adopted by the Codex Commission, was in 1991, not 1992.
The four principles concerning the role of science in Codex decision-making, and the extent to which other factors are taken into account, was never discussed, as far as I am aware, by JECFA. It was not processed through the JECFA system. It is entirely a Codex matter and, therefore, entirely operated by the member governments of Codex and not by the experts in JECFA.
Let me put the JECFA material into perspective. JECFA provides advice to FAO and WHO and through them to the members of those two organizations and also to the Codex Alimentarius Commission.
The Codex Alimentarius Commission has, as a matter of procedure, taken the JECFA recommendations in a variety of areas and plugged them into the eight-step elaboration system and the result of this, in most cases, is a consensus adoption of these points of view which are then available for countries to apply, if they feel that they wish to. They are not obliged to under the Codex rules.
There have been any number of emergency re-evaluations in JECFA which have been initiated in Codex and which have found themselves reflected in Codex opinion.
As to the direct question if there is one that has occurred in the nine-month period since the entry into force of the Uruguay Round Agreements, I'd have to check that in but I can't think of one. The most recent one which has similar import would be the decision of JECFA to withdraw previous evaluation in regard to potassium bromate, as a flour-treatment agent, and the subsequent withdrawal by Codex of all approved uses of potassium bromate in flour, primarily because potassium bromate residues remain in the flour at the time of ingestion by the human consumer, and potassium bromate is a carcinogen.
The maximum residue limits are not exclusively concerned with trade. They are concerned with the control of good agricultural or good veterinary practice and they are applied in trade. To be quite practical about it, it's definitely beneficial to trade if countries have the same numerical values for maximum residue limits, or at least if you are exporting a product, the level at which you are exporting is lower than the limit applied by your importer.
WT/DS26/R/USA
Page 299
However, the maximum residue level should reflect the good agricultural, or good veterinary practice, on your territory and not be linked exclusively to trade considerations.
I would take umbrage at the point that the MRL is a trade-derived figure. It is, in fact, a figure derived from good veterinary or good agricultural practice.
Can I ask you to go into the question which was put forward?
Very quickly, I think the principle that was being elaborated, the philosophy is more popularly referred to as the precautionary principle.
As to what is the most appropriate action in the face of the kind of information, for example, that has been presented today, I take that really to be the work of the Panel, and so I will make no attempt to answer that question.
I think it goes well beyond the jurisdiction that I've been provided here to comment.
I will, very quickly, comment on, if you like, the age of the assessment and I've already made comments on this a number of times, but it perhaps bears repeating.
I think it is incorrect and I would respectfully submit that it is misleading to suggest that 1988/1989 is the last time that this issue on the safety of the hormones, in question, has been examined.
I've already mentioned that certainly in my view, this issue, without going into a lengthy debate as to whether or not the strict definition of risk assessment was made by the European Conference, I think it goes without question that the issue of the safety of the hormones, in the general sense, was most certainly re-evaluated in December 1995, which is a little more than one year ago.
Indeed, a number of people who are here today were active participants at that conference.
I must say that I disagree with the view that 1989 is the last time that this issue was formally evaluated on an international level. I think that's incorrect and I think it's misleading.
More specifically, on the issue of risk assessment, whatever that means. We could spend the rest of the evening debating whether or not risk assessment have, or have not, formally been discussed by the Conference.
I would refer the Panel to a specific section entitled, Risk Assessment, on page 3 of the Conference Report, and it falls under the general section of the Report entitled Report and Conclusions of the Steering Committee, specific section entitled, Risk Assessment.
Although the Conference was concerned with manytechnicalmatters, it is inevitablethat general interest should have centred on the risk assessment of different classes of both promoting substances, the subject of discussions of Working Group 2.
Then I would refer, Mr. Chairman, the Panel to a number of specific papers which I would respectfully submit, dealt with very much with the issue of risk assessment. I refer to Workshop No. 2 on pages 245-401 and I would refer specifically to a number of papers, including the one presented by Professor Bridges, but also including papers presented by Dr. Miller from the United States, Dr. Hoffmann from Germany, and so on and so forth.
WT/DS26/R/USA
Page 300
If I refer, just by way of example, not to give particular preference to any one of these papers, to the paper presented by Dr. Miller from the USFDA at that Conference, she concluded, at that time, "hormonal growth promoting agents include a variety of compound which very markedly in pharmacology and biochemistry" so on and so forth.
"When these animal drug products are used according to label directions, the edible tissues from treated animals, are safe for all consumers."
I would respectfully submit, Mr. Chairman, that that is very much an assessment of risk, regardless of whether or not we can come to a formal definition of what constitutes a risk assessment. It's certainly my view that this is most definitely the element of risk assessment. It strikes the heart of what a risk assessment is, because after all the product of a risk assessment, is the statement of safety to the target population. That's why we carry out risk assessments. It's not merely a mathematical exercise.
I would suggest to the Panel that there are a number of papers here that deal very much with the issue of risk to human populations, resulting from the use of these compounds, and that the consensus of this Workshop was that the use of these compounds, as recently reviewed as December of 1995, does not constitute any risk.
May I add some comments for the Panel? It can also see in the same paper, page 378, conference.
I will read what it says:
"New quantitative risk assessment models are needed for the safety evaluation of chemicals with toxic actions, like general carcinogen."
It's in the same paper so it's not possible to extract something when you don't extract all the information.
I didn't intend to mislead. I mean obviously I'm not going to read into the record the entire Conference Report it is available.
There is no question that a number of deficiencies, in terms of our state of knowledge, as identifiedthroughout the report, but those efficiencies, those deficiencies were weighed in the conclusion which the Panel has reached. They weren't weighed by me. But the conclusions of the Working Group was that the present state of knowledge is sufficient to demonstrate the safety, in their view, of the use of these chemicals in contemporary times.
I think it would be silly for any scientist to presume that the day will ever come, on this issue or on any other, where we could say we know enough and that there is no need to do any further work. I have never been associated with a scientific issue where I have ever heard a scientist say, there is no need to do any further work on this topic, we know all that we need to know.
I agree that there is further work that is indicated. I agree that statements made by scientists, such as Dr. Liehr will continue to contribute to our understanding, but I also agree that the totality of evidence re-evaluated, as recently as December 1995, suggests that the way in which these substances are used and the residues which they produce, do not constitute a risk to human health.
WT/DS26/R/USA
Page 301
I don't find those statements to be contradictory at all.
Dr. Lucier.
My question may show a certain amount of naivety but did JECFA document or the 1995 meeting address the issue of carcinogen risk assessment, in any kind of way?
It seemed to me like the risk was based on the hormonal activity of the agents, in primates.
We discussed that briefly before. Would you like to give your views on this.
I think only to say that unless we were to spend a great deal of time strictly defining what we mean by the phrase "risk assessment" we could continue to go around this circle, I think, all night. The point that I was trying to make that, in the opinion of the scientists that were gathered at the Conference, it was their collective wisdom, for whatever reasons they considered. I think, as Professor Bridges has already quite correctly indicated that it did not include extensive examination of some of the evidence which was presented today. But for all the totality of reasons that they considered, at that time, they came to conclusions which have already been made clear. One may raise the question if they were redoing it tomorrow, would they come to the same conclusion? I think this is speculation. One could raise the question, if they were doing it again a year from now, would they come to a different conclusion? I don't know.
Were the acceptable levels, or the acceptable daily intakes, were they derived, primarily, from the tests of hormonal activity in non-human primates? Is that correct?
No, I don't think that's correct. I think there were extensive carcinogenicity data that were available to the JECFA Committee. These data have been described, the assessments provided, at least in the case of two of the xenobiotic hormones. To the best of my recollection, monographs on the three natural hormones are not available, but they are available for the two synthetics, and certainly they were based on chronic toxicity carcinogenicity studies that had been conducted in a very classical, NTP kind of methodology.
No, the naturals were evaluated, what I indicated, to the best of my knowledge, monographs are not available on the naturals. But data was evaluated by the Committee, on which they drew their conclusions, I have not been privy to that data. But then again, I have to be candid, I have not re- evaluated the carcinogenicity data for any of these compounds, in order to prepare this session. This would have been neither practical or arguably even necessary. I would never pretend to suggest that I have greater wisdom in this area than the collective wisdom of the JECFA Panel.
Dr. Randell, please.
WT/DS26/R/USA
Page 302
Mr. Chairman, just perhaps to clarify that point.
In the case of zeranol, for example, the JECFA Committee did evaluate carcinogenicity studies, in rats and in mice, and so forth, but the conclusion that JECFA came to in this substance and in the other substances, that the carcinogenicity was linked with the hormonal effect, and therefore the establishment of the "no hormonal effect" level was the point at which the evaluation hinged, it hinged around that point.
This is why the new data, which we're hearing about today, is of interest because it would supplement that evaluation and help us considerably. However, the information which was available to JECFA definitely did include carcinogenicity studies for all of the substances concerned.
I'm looking at my watch. It's about 7.30. Could you come to a close for today?
Mr. Chairman, I will come to a close on this particular point, if you allow me.
I will read from what the response of Dr. Ritter was to Question No. 5, because that is what we are discussing now.
I quote page 12 of his reply to Question 5, at the end of the first big paragraph:
For progesterone:
"The report was silent on the results of mutagenicity studies which were available to the Committee for review."
For testosterone:
"Results from mutagenicity studies, although available to the Committee, were not described in the report."
For the synergy:
"The issue of potential synergy has been addressed, at least in part, through the context of biochemicalstudies, directed at the effect of excretion of hormone combinations when compared to single hormones alone."
This is on page 14.
"It is, however, clear, that definitive studies relating to genotoxicity or carcinogenicity of hormone combinations have not been carried out even though this is frequently the preferred method of use."
I'm slightly puzzled by the positive terms in which Dr. Ritter is now phrasing his replies for the review of the, in the JECFA report. All issues we have been discussing here now about the genotoxicity, not only due to the hormonal effects, as has been rightly pointed by Dr. Lucier but all the other possible sources of risks we have identified.
WT/DS26/R/USA
Page 303
At least the way I understand his replies is that these studies have not been carried out. They might have been available but they have not been reported in the report. We don't know if they were real studies.
He was not part of the JECFA committee, in that particular case. So I don't know where he has got this information. Whereas Dr. McLean was part of this report and also he might have an interest to explain to us whether this has been done. From the reports we have in front of us, Dr. Arnold has provided a table, where there are some black holes where things which have not been examined. I see there is, at least some, contradiction between what our scientists are saying on this particular issue. If one reads the entire report of JECFA, it is clear that only the hormonal effects of those substances have been studied for the potential carcinogenic effect. So what is the exact scientific basis on which Dr. Ritter is making these comments?
One more comment.
I think Dr. André, because he was part of the Organizing Committee of the 1995 Conference, he has, in my view, direct inside information on how this conference was organized. The Community only funded the Conference and left the scientists entirely to address some issues.
In the report of Dr. Ritter, it is frequently referred to as the EC Conference. It is not the EC Conference. It is a conference funded, but no more than that. It was entirely left to the Steering Committee who to invite and what to discuss in this conference.
I would propose that we have a final round on this question and then we will close the meeting for today, and I would give the opportunity to respond to the scientific experts.
Maybe Dr. Ritter would speak at the end.
I'll start with Dr. McLean.
I thoroughly concur with the comments of Dr. Randell. We were trying to put a positive view on the availability of JECFA, if people wanted to take that opportunity, rather than forcing people to do it.
In relation to the mutagenicity data for the naturally-occurring substances, the mutagenicity data that was reviewed is referenced in the report, in each case, and so therefore, not so much that is was silent but it referred to mutagenicity data that was in the open literature. So, in the case of progesterone, testosterone and oestradiol 17 beta, there were a number of published documents that the Committee did look at, and they are referenced in the report.
Dr. Arnold.
Mr. Chairman, I have two points.
WT/DS26/R/USA
Page 304
The first point is I have difficulty understanding how it could happen, but the legally competent authorities of the EEC, on the advice of the competent advisory scientific body, have established no MRL for oestradiol and this happened two years ago.
The public summary report states, inter alia, "The conclusion of the FAO/WHO Expert Committee on Food Additives, JECFA, that no ADI and MRL for oestradiol need to be established, as adopted." And then all the other things I mentioned this morning, this is what I have difficulty in understanding, this happened in 1994. And this Regulation has been implemented, it has passed all the EC institutions and the result has been published in the Official Journal. I refer to this in my answers to the question of the Panel.
My second point is I was rapporteur of the Codex for five consecutive years. That means I had jointly, with the secretaries of the two agencies of the United Nations and with the US Secretariat of the Committee, to prepare the report. I was in a really difficult situation because it's true that the EC objected to move these MRLs to the next step in the procedure, but without raising any health issues. So, I was really in trouble and you can find the result in the report. I finally got a written statement and the arguments you will find there are, inter alia, the EC is opposing because they have specific legislation prohibiting the use and the EC consumer doesn't wish to receive meat from animals treated.
These were the arguments, so I'm a little disappointed that these questions have not been raised earlier, first of all, in earlier years during the Codex discussions. It could have happened, for example, to refer the whole matter back to JECFA with new health arguments for re-evaluation, at that time, but this proposal was not made.
And that secondly, two years ago it happened that we came to the same conclusions by the competent authorities of the EC regarding oestradiol. Progesterone has also been finished but not yet published in the Official Journal, and testosterone is still under evaluation.
I think we will reserve this for tomorrow.
I'd like to conclude and if anybody of you would like to speak?
Just to say a little more about this Conference because I have been asked to do so.
It's clear that the EC has just asked four experts to organize this Conference: Sir John Maddocks is the Head Director of the Nature Review, a scientific paper, an Irish colleague, Bergen colleagues and myself.
I can witness that we were very, very free of our to our choices, in terms of selecting scientific people. And we organized this with a first group of two people for the three Work Shops, and they proposed that we invite scientific people and we agreed, on the basis of the publication and the famous of these people.
But, I think, and I personally regret now that we didn't take enough care with the scientific publication. I discovered this morning a lot of publications on this topic and some of them were available and the scientific were also available, and maybe we have not done our best. We have done our best but maybe it was not good enough, and that will explain also the missing of some information in this Conference.
WT/DS26/R/USA
Page 305
On the last point, I would like to say that the conclusions are a reflection of a common discussion, and it is not a complete reflection of the majority of people. It is not a total opinion of the members of the Committee and it cannot be taken into account as something coming from an official body. Our objective was just to put some light, some knowledge, on these hormones in order to inform politicians preparing new regulations and no more. It is not the same thing as to establish an official MRL. The responsibility was not exactly the same order.
Dr. Lucier.
I'm just surprised in reading through this report that, given the fact that oestradiol is a known human carcinogen, that an ADI was considered unnecessary for it. There is a tolerance but not an ADI, if I'm reading the report correctly.
What page?
471. 19.
The JECFA Report. This was the 1988 JECFA Report. Is that correct?
In the middle of page 19 it said the Committee considered an ADI unnecessary for a hormone that is produced endogenously in human beings and shows great variation and level according to age and sex.
The difficulty that JECFA had of setting any ADI was the problem of the increase that one saw upon treatment, against a background of very large levels that occurred naturally. And so therefore it meant that the setting of the ADI really wasn't possible, because the treatment altered the levels by such a small amount when compared with the endogenous levels in cattle to make any ADI meaningless, and that was the difficulty we had.
I would suggest, today, that is not possible to regulate the use of the three naturally-occurring hormones by the setting of an ADI because you can't tell the difference between treated animals and untreated animals, by a method that is suitable for regulatory purposes.
It's a practical consideration, a real and practical consideration. I don't quarrel with what you say, but in practise there's no way of regulating it.
WT/DS26/R/USA
Page 306
But if one knew, which we do know, what are given body burden in the cattle would be produced by the use of growth promoters, we can certainly estimate various eating habits of people, what they would take in their body from those agents that were exogenously administered.
One could come up with reasonable estimates on that and establish an ADI. In my mind, that could be done. From my comments this morning, I don't argue with your point that it's a very small number of molecules, one in 28,000 I came up with in my calculations, not taking into account the fact that it is poorly absorbed and that would make it one in 280,000.
But nevertheless, some sort of an assessment should have been made, in my mind, of what that risk was, as low as it is. I'm not saying that the risk is high but I believe the risk is extraordinarily low or zero. That's not my point. I am just surprised that that exercise was not done.
If more points on this are to be discussed, I would like to discuss them tomorrow because I think we are a little tired.
Would you like to make statements tonight?
I apologise for the length of the meeting but it has been, in my view, very informative and I would like to thank the experts and also the delegations for their support.
As agreed, we will meet tomorrow at 2 o'clock, hopefully sharp, and we will open discussions with the statements and questions by the Canadian Delegation.
Thank you very much and good night.
The meeting is closed.
As announced yesterday, we shall proceed in the following manner. We will go first to Canada for their comments and questions to the experts. We will then turn again to the European Communities and I would urge the Community to limit their interventions to the utmost necessary in order to save time. We will see how much time can be allocated when the Canadians have made their statements and questions and then there will be final questions from the Panel to the experts and then the final statements by the experts in the end. I have to be able to conclude this meeting by around 6 o'clock. I am not able to say whether we will be able to do so. It depends on you and I am in your hands to a large extent.
We would really do our best to respect what you have indicated, but our delegation has a serious problem in the sense that the planes of two of our scientists are leaving at 6 o'clock and we consider their testimony and their presentation very important. It is just a request and we hope we have the understanding of the Panel in this room. If it is possible by 4 o'clock, Canada does not finish, I do not know how many questions they have, if we can stop for a while they can make their presentations
WT/DS26/R/USA
Page 307
and if there are any questions or answers and then they could probably be able to take their taxis and get to the airport in time. That is the request.
Thank you very much. I think we will see where we are with Canada at 4 o'clock.
Mr. Chairman, a procedural question. Last night we had left that the Panel was going to deliberate on our question about the follow-up submission. That was requested by the Europeans and I wondered if the Panel had come to a conclusion with respect to the matter we had raised last night. It is important, in part, because as we understand, the procedure that was just described, there would be no more opportunity for comments by either Canada or the United States after the Canadians have finished their questions today (and obviously we finished our questions yesterday), and there are a number of points that have been made that I think it would be appropriate to give us an opportunity to respond to.
Thank you very much. I intended to take this up at the end of the meeting but since you raise it I can give the answer of the Panel. We have looked into the question and based upon the text which was communicated on 3 February by the Secretariat to the Parties, which I will read in a second, we reviewed the decision I made yesterday and we will not accept any ex-post submissions from the Parties. The United States has submitted some materials, you have submitted some materials yesterday which we took on the file, but we will not take on any more written materials after this day. The text reads "Please find attached the responses of four of the five scientific experts responses to Panel questions. Professor Arnold requested a few more days to submit these responses. [They came later on.] The Parties are informed that their comments to these questions at the 17-18 February meeting with scientific experts should be made available, as is customary with the Parties' oral statements at panel meetings, in the written form and if possible on diskette to the Secretariat". So this means that we expected these comments by today and I think we will conclude the proceedings based on this letter. So if you have additional materials, please hand them in by the end of this meeting.
Mr. Chairman, the practice though is that we are allowed to submit a document at least the next day after the oral presentation. This has always been the case. We are adjusting the text while we are speaking and I have been doing that for so many years and it is always the next day we are allowed to submit in writing the oral presentation. It is very restricted this time-limit of today. I would appreciate it if tomorrow there is also a chance for us to submit a document. Thank you.
I think the problem is that then we need to have another go on rebuttals. And I think we stick to this here for the moment because this has been a hearing of experts by a panel and we do record on tape and will transcribe everything which is being said at this meeting so I do not think that we would need another written statement on the side of your part.
Mr. Chairman, if you apply this rule then the entire text which is in the text that the United States has circulated has not been read out orally. Then I would have unfortunately to request you to exclude
WT/DS26/R/USA
Page 308
all those parts of the documents which have not been read out. I really regret that in this part of the document there are so many things which have not been said yesterday.
I think it was not required that this be read out because it says that the comments on these questions should be made available to the Secretariat. It is not exactly the same proceeding in my view as in the ordinary first and second substantive meetings here. I do hope that you are able to provide some of the materials as it was done yesterday, today - the statements by the experts and so on.
Thank you Mr. Chairman. If I am going to get started before my 4 o'clock deadline then I had best get quicker with the finger on the button.
It had been Canada's original intention to comment quite briefly on the experts' answers to the Panel's questions. Primarily by noting in point form some seven areas where the experts were in general consensus and then to put what we hoped were a few focused questions. We still intend to do that, however some of the events of yesterday require some broader comments on Canada's understanding of the purpose of this meeting and the rulings that you made on how we were to proceed.
We came here prepared to comment on and explore the Panel of Experts' answers to your questions and abide by your ruling that no new evidence would beintroduced after 8 February. I believe, Mr. Chairman, that you will recall the purpose of that ruling was intended to provide all participants with an equal opportunity to review relevant material in advance so that informed commentary could be made.
Yesterday we all had the privilege to hear some of the world's leading experts comment and debate the properties and characteristics of the substances in issue. Some highly technical and complex issues were discussed with clarity and precision. Unfortunately and unfairly, in Canada's submission, we were sometimes deflected from that very high level of debate by the introduction of new material, some of which was apparently presented out of context and in a selective way. The Panel's experts were asked to make extemporaneous replies on matters they had not had an opportunity to consider in advance. This sometimes led to prolonged discussion clarifying facts on matters of marginal importance to the issues this Panel must decide.
An example, no better or worse than many I could choose, is Dr. Epstein's table on oestradiol residues in eight-year olds, which ultimately led to the somewhat surprising assertion that two quarter-pounders or 110 gram whoppers were the equivalent of 500 grams of meat. If I could be permitted a pun, that was a real whopper! With respect, it is important to refocus the issues these experts are here to help elucidate.
The complaint before this Panel concerns a measure of the EC which bans the importation of beef or any of the six hormones, alone or in combination, that have been used for growth promotion purposes in the face of properly established international standards of safe use.
One goal of the SPS Agreement is to further the use of harmonized sanitary measures between Members so as to protect health but not unnecessarily interfere with trade. The Agreement permits a Member to depart from these standards for certain prescribed circumstances. The relevant one for the issue concerning the Panel's experts, and the only one in my submission, is a scientific justification. Dr. Randell has spoken eloquently of the high standards of JECFA and Codex and the subject matter that was taken into account in establishing the standards of both the natural and synthetic hormones and the very conservative assumptions adopted to ensure safety through the use of ADIs and MRLs
WT/DS26/R/USA
Page 309
where required. It is clear that these organizations are to be commended for their work and nurtured rather than criticized.
We have learned that studies have been conducted for more than 12 years to try and prove a hypothesis that oestrogen is genotoxic. Many of these studies were known to and considered by JECFA. Dr. Liehr, in one of his studies, has injected comparatively massive doses of oestrogen in a male Syrian hamster in an experiment designed to produce tumours. The quantities are about twice the dose given to cattle. Dr. Lucier, as I understood him, agreed that in light of this evidence there was a risk, but he puts that risk at a level somewhere between zero, a concept which cannot be achieved in absolute terms, and a factor so small that it cannot be measured.
The question for this Panel is this in my submission: is that a risk that justifies departing from an international standard in light of its impact on trade and the objects of the Agreement. One way of testing that, Mr. Chairman, is by comparing other EC sanitary measures where there are similar or greater risks. Canada will pose some questions on the issues of consistency or lack of it in the EC's handling of these and other veterinary drugs.
Before turning to those questions I would like to review what Canada submits are the points of consensus between the Panel's experts: (1) good animal husbandry practice is a broad term encompassing good herd health, management practices and includes good practice in the use of veterinary drugs; (2) in the event that good practice in the use of veterinary drugs is not followed, higher residue levels may result.
However, given the fact that the MRL was derived from an ADI where a large safety margin has been applied, such an event is unlikely to cause an adverse health effect; (3) consumers are not able to distinguish the meat of an animal treated with growth-promoting hormones and the meat of an untreated animal; (4) observing withdrawal times for the hormones does not guarantee zero residues in meat. No residues detected does not imply that there are no residues present in meat as they may be present at levels that are below detection; (5) following treatment with the natural hormones, for either therapeutic, zootechnical or growth-promoting reasons, hormone levels will fall to those normally associated with untreated animals. There are no differences between the residues of natural hormones that are endogenous to the animal and the residues of the same hormones administered exogenously to the animal. Extensive data relating to genotoxicity and carcinogenicity were available to the JECFA and were considered by the Committee. The Committee concluded that the five compounds evaluated are not genotoxic carcinogens; and (6) growth hormones have been used in animal husbandry for many years; in some cases up to 40 years and there are no human epidemiological data that suggests a hazard. Adverse health effects have not been observed in those countries using growth hormones.
I have just one other point I wanted to raise briefly, not for purpose of resolution but rather to alert the Panel to an additional concern that Canada has. It has become apparent that the European Communities intend to present a slide demonstration at some time this afternoon. I have not seen that demonstration and I do not know the contents of it, but I am concerned that it may well be fresh evidence which would be introduced after the February deadline the Panel has set and I am also advised by some people who claim to have seen these slides that it is of little scientific assistance and may well be inflammatory. If and when the EC chooses to present that slideshow I would hopefully have an opportunity to discuss whether it is appropriate or not.
If I may I would like to turn to my questions.
Dr. Arnold, on pages 10 to 12 of your written answers you described the evaluations conducted by the EC for oestradiol and progesterone pursuant to Regulation 2377/90. As a result of the evaluation oestradiol has been put into Annex 2 of that Regulation and progesterone is pending I understand. Could you describe the significance of putting these substances in Annex 2 and could you contrast
WT/DS26/R/USA
Page 310
it with the significance of putting a substance in the accompanying Annex 4 and perhaps describe some of the substances that are in Annex 4?
I should perhaps spend the first minute describing the working relationships between JECFA Codex and the competent bodies of the Community because I think then you would better understand how this happened.
We have set up in the Community a working party obliged to propose maximum residue limits for residues of veterinary drugs in 1984. We started at the beginning without having a true legal basis so it was a big advantage when Regulation 2377/90 became effective on 1 January 1992. If you look, for example, at the Official Journal, or maybe at Volume 6 of the Rules Governing Medicinal Products in the European Community, you will see a whole volume devoted to this issue. You will see the regulation, you will see guidelines, what the requirements are for the evaluation and you will see an annex to that regulation with a list of studies required in order to scientifically evaluate the substances. This list is absolutely identical with the list of JECFA requirements because it has practically been copied from JECFA.
I was the Chairman of this Safety Group at this critical time for more than three years and was helping drafting these rules and regulations at that time. This was also an interesting time because JECFA had just started doing reviews and we quickly realized that this could be of great help to us in the European Community.
If I look back at the time between 1984 and 1988-90 we had not achieved to set any MRL because it is really a tedious procedure, and as soon as we were able to cooperate with JECFA and Codex and harmonized our procedures with a system, we also harmonized with the Food and Drug Administration we had regular meetings the EC Commission and various experts with the Food and Drug Administration, there was an exponential growth. If you check the list of MRLs we have so far adopted it is quite a considerable list. But you will find the exposure limits, the ADI, are almost every time identical, those proposed by JECFA with those used in the EC. There are slight differences in the MRLs and this can be readily explained because veterinary practices are not the same in all countries.
Coming to the hormones, when JECFA has proposed MRLs and ADIs for hormones it was impossible to discuss this issue in this Safety Group and in the Committee of Veterinary Medicinal Products because it is well known that we have had already at that time the specific agricultural policies and the specific rules. Therefore it was surprising to me, and I told you yesterday, that six years after JECFA had discussed these substances, a Committee of the EC came to the same conclusions that the JECFA position is adopted and they also felt it was unnecessary to set an amount in ADI for oestradiol and was unnecessary to set an MRL and this has been signed by the competent authorities of DGIII, of DGVI, it has been published in the Official Journal. And this demonstrates that, scientificallyspoken, what JECFA produces, what the JECFA Codex system produces, even if the results have not been officially accepted in the Codex procedure they are used everyday and to the benefit of the consumer protection in Europe.
I wanted to say this at length because otherwise you could have the impression that we feel very uncomfortable with JECFA. In fact the opposite is true.
Coming back more specifically to your question Annex 2 is meant for substances which no MRL is needed. I give you maybe two examples. If you use sometimes a local anaesthetic on a horse it might be necessary to set a withdrawal time but this is a ridiculous problem in view of setting MRLs for consumer safety in a community like the EC. This could be a candidate for such a substance, or if the substance absolutely harmless after the evaluation had been done. The usual procedure is that
WT/DS26/R/USA
Page 311
the company has to apply for inclusion in Annex 2 and they have to submit scientific evidence. Harmless substances, mainly harmless substances, for which there is no need to set an MRL, for example, elements which endogenously occur in the body. It is a rather long list and they are put into Annex 2.
In Annex 4 it is a little bit more difficult. Originally this Annex had been developed to include substances for which it is impossible to develop any conditions of safe use. This means for real hazardous substances, this was the original meaning. But what then occurred that later we had substances for which there were indications that they could be hazardous and gaps in our knowledge missing information. Since these were all out of patent substances and since no sponsor applied to prolong the presence of these substances on the market in order to be able to develop the missing data, some of these substances have also been included in Annex 4. That means we find there substances for which it is quite sure that they are hazardous under all circumstances and other substances for which there is a strong indication that they might be hazardous and no sponsor could be identified who is willing to produce the missing data. Maybe this is enough for the moment from my point of view.
My second question is also drawn from some of the material in Dr. Arnold's answers. By way of prelude I can let you know it is a compound question and there will be some follow-ups that come with it.
In your answer to question 2.2 on page 5 of your written answers there is a table that describes the uses of nature identical hormones and their esters in the European Communities. My first series of compound questions are: How are these substances used? Why are they used for oestrus-synchronization and when used for synchronizing oestrus, are the animals treated because they are sick or for a zootechnical purpose? This question may also venture into the expertise of Professor McLean.
I cannot satisfactorily answer all your questions and sub-questions. I wanted to show only a few examples of substances for which I absolutely know that they are used. The list might be longer because a list of substances which are in compliance with our rules in the EC has been set up early on the advice by the Committee for Veterinary Medicinal Products, there are more substances on the market. I am not a veterinarian so I cannot really tell you whether this is justified to use these substances for this purpose. What I wanted to show you was how similar some of these substances are which are used on the one side for growth promotion and on the other side for therapeutic and zootechnical purposes, but there are also differences in the esters, and based on this table I developed my answer so these were just examples and I can as a non-veterinarian not justify their use but I am sure they are used in accordance with the directives in the EC.
I wonder if Professor McLean or one of the other experts can assist me with that information?
Thank you Mr. Chairman. Oestrus-synchronization is a commonly-used animal husbandry practice. You synchronize the reproductive activity of your herd for a variety of purposes; it may relate to the availability of feed, it might relate to the availability of markets, etc. It is a relatively common animal husbandry practice.
WT/DS26/R/USA
Page 312
After use would residues of these administered substances be present in meat and milk of the treated animal?
I would suggest that all of the experts may take the floor if they wish to do so. Who would like to go first?
Yes, the residues would be in meat or milk if the animal was slaughtered. It is interesting that oestrus-synchronization that is often used in dairy cattle in a variety of production circumstances and the residues would be in milk.
If I could continue with that table and some of the information on it. I wonder if any of the experts who feel confident could help me with the use of these various esters in the EC formulations. What are the chemical differences between the esters and why are different esters used and is there any comparison of persistence in the body of these various esters?
The esters are generally used to alter the rate of uptake of the drug from either the injection site or if it is in some sort of intravaginal device for example, although generally not there, it alters the rate of uptake. But what generally happens is that the oestradiol ester is actually metabolized either at the site of injection in the blood or in the tissue and the bond between the active substance, the oestradiol, and the side chain to which it is bound is readily broken and the active constituent oestradiol 17 beta, or testosterone in its various forms, is the active constituent.
There can be under some circumstances small quantities of the ester appearing in the blood and that is picked up by analysis. But those levels are small and no consideration when it comes to looking at the toxicity because, for example, if there were esters in the meat and someone ingested that meat then the enzymes of the gastrointestinal tract of humans that digests normal food readily cleave the ester bond.
Mr. Thompson, just for the benefit of the Panel, would you anticipate to draw some conclusions from these statements? I am not in a position to do so myself.
I thought that was the purpose of tomorrow' s meeting when I would be presenting oral argument. But the general purpose of these lines of questions and a few others that I have is to try and demonstrate that the European Communities use some of these hormones either alone or in combination for zootechnical purposes such as in increasing the rate of production in sheep and cattle. Notwithstanding that other alternative means may be available they continue to use these substances for these purposes. What I would be arguing tomorrow in a like manner is that North America and other places use growth hormones.
WT/DS26/R/USA
Page 313
Mr. Chairman, I think there is a very big difference between the use of such drugs for therapeutical or zootechnical purposes. In very precisely defined animals, as Professor McLean said, and the use of the same hormones or similar hormones in large scale for growth promotion. No scientist has said that it is a bad thing to use these hormones for therapeutical use. The problem is for growth promotion on large scales on all the animals, it is not the same thing as is being discussed now.
Is it true that for therapeutical reasons this is done under prescription by assistance of the veterinarian, while the growth promotion is not done so? Can I ask this just for my clarification to one of the experts?
In Europe it is done under veterinarian control. Directly by veterinarians in some countries or under veterinarian control by prescription in other countries. For growth promotion we have no experience.
Perhaps I could continue with some of the other parts of my question and it may become clearer as to the reasons I am asking this. My understanding is that it is not so much therapeutic uses as zootechnical uses in order to increase the production of sheep and cattle so that their gestation periods are reduced and happen more frequently.
Can the experts help me with whether there are any synthetic hormones used in the European Communities for synchronization of oestrus such as medroxy progesterone acetate or allyltrenbolone, or whether oestradiol is combined with progesterone in some of these proceedings?
I would like to ask my colleagues to assist me. What concerns allyltrenbolone there is an exemption in the directives. This substance can be used because, as I said yesterday, it is not suited for growth promotion. Although the name is suggestive, it does not have the properties of trenbolone. Allyltrenbolone is totally different from trenbolone with respect to the biological facts. So this is the first part of my question.
The next part I am not so sure. I know that the European Medicines Evaluation Agency and the CVMP has just finished an evaluation of medroxy progesterone acetate under my recommendation and this does not fit into my picture of what is legal in the EC. Maybe you, François, you know?
No, I have no information of this regulatory point but in any case this compound is used for zootechnical purposes and not for growth promotion. It will never be evaluated by EC Organization as growth promoters. It is a very different thing. We are speaking about things that are not of concern with growth promotion here.
My problem is that it is synthetic and it doesn't yield the natural hormone upon hydrolysis. That is my point. It is certainly not suitable for growth promotion.
WT/DS26/R/USA
Page 314
Medroxy progesterone acetate is sometimes included into an intravaginal device along the lines of the second line on table 2.2 of Dr. Arnold's submission. It is used in a number of countries.
If I might just add for the sake of clarification, whilst these substances for zootechnical purposes are used in individual animals, it is not unusual in some practices for a significant portion of the herd to be treated over the life of a reproductive season, if I could put it that way.
In other words, whilst individual animals are treated you might get 10 or 15 or 20, I am not fully familiar with the production procedures of EU, of the herd treated over one part of the year that is associated with the breeding season.
Mr. Chairman, may be you are not very familiar with the reproduction control in farm animals. In this case, may I explain to you that the medroxy progesterone acetate is included in a sponge and the sponge is put into the vagina of the sheep with a small cord. During a period between 10 days, 15 days, two weeks or maybe a little more but it is not a problem. Then the sponge is pulled out with a small cord and then after this treatment the animal comes into oestrus and is then inseminated and becoming pregnant in most cases. Only one per cent fail to be pregnant and this maybe could be a problem for these animals, but usually people try again and it is not really a problem. It is not a problem of residue of synthetic hormones because these animals are bred to have sheep and they will stay a long time in the farm. They will not be slaughtered after such treatment. It is not the objective. It is very different to treating animals for growth promotion and to slaughter them at the end. It should be very clear in your mind.
Is this a step of preventive medicine or is this a treatment for animals who have difficulties to conceive?
It is just a so-called zootechnical practice. That means it is more easy for the farmer to have all the females at the same date on oestrus and to inseminate them at the same time and then to have products at the same time. It is more convenient for management of animals. No more.
But it is a standard procedure which is applied to healthy animals?
It is familiar procedure. It is applied in sheep mainly. More in sheep I think than in cows.
Can I understand you to say, Professor André, that farmers did it to their whole herds at a single time in order to assist conception occurring at the same time?
WT/DS26/R/USA
Page 315
If your question is to know if they can buy this product freely, the answer is no. This product is always under veterinary control.
I had not asked that question. May I proceed with my next question Mr. Chairman?
Again I have a somewhat compound question. In answer to question 30 of the Panel concerning the potential adverse effects on human health from residues of carbadox, monesin, olaquindox, avoparcin, benzylpenicillin and carazolol, ivermectin and organophosphorus compounds, someof the experts noted that when used correctly, the residues from approved veterinary drugs should not produce adverse effects in the human population.
In question 17 of the Panel the experts were asked to consider the implications for human health of residues from the misplaced implants or improper administration of the six hormones in dispute. And I wonder if the experts could extend their analysis and comment in respect to the residues of the veterinary drugs I read out and listed in question 30, as to what are some of the implications for human health when they are misused in a similar way described in question 17? For example, could the experts explain what adverse health effects may result if a hypersensitive person ingests meat or milk from an animal that was treated with benzylpenicillin, when the required withdrawal period has not been observed and the residues from benzylpenicillin in such meat or milk exceeds prescribed MRLs, as an example.
Maybe I can partly cover this question. I am a little bit reluctant to compare the hormones with carbadox because both substances have been regulated, or proposals have been made by JECFA concerning these substances, that in both cases if good practises are observed there is no appreciable risk. So I could say they are equally safe if good veterinary practices are observed. Because although carbadox itself is a genotoxic carcinogen, it is so quickly metabolizedthat if good practices are observed neither carbadox nor its main carcinogenic metabolite is present as a residue. Only an innocent metabolite quinoxaline carboxylic acid, which has been extensively tested including carcinogenicity, and this is the compound on which the MRL has been proposed.
If somebody does not respect the withdrawal time there might be an increase in this innocent metabolite, but the closer you come to the administration there is an increasing risk that the carcinogenic compound itself or its metabolite is present. For example, at zero time of withdrawal it is obvious that it would be present and that now you have a different quality of risk and how do you compare this? If I am a little bit reluctant and I must insist that it is the aim of our procedures, to check whether conditions of use can be proposed which if followed guarantee that there is no appreciable risk. And if the possible risk is qualitative in nature how do you prepare this? This is the first part of my lengthy answer.
Concerning benzylpenicillin, I had looked carefully at benzylpenicillin, I was consultant to WHO at the time benzylpenicillin had been evaluated. What is interesting to see is that about 15 per cent of the world population, some figures are higher, other figures are lower, are sensitized against benzylpenicillin. But we have carefully looked at all the reported cases. Only a very small number of cases one can discuss between five and ten but in no way more than ten, have been observed over the many years of use. Billions of doses administered of benzylpenicillin to human individuals causing sensitization but few cases where residues were the cause of an allergic reaction. The levels proposed by JECFA of benzyl penicillin are so low that in the whole literature you find only three cases where a lower level has caused an effect in a human being. The dairy industry needs these low levels too
WT/DS26/R/USA
Page 316
because the most sensitive population in this case are the starter cultures used for yogurt production, for example. They are much more sensitive than human beings. So the human consumer benefits from this that the mix cannot be used for food processing purposes if it contains high levels of penicillin.
Thank you Dr. Lucier.
I have a couple of questions, before I attempt to answer the question. So carbadox does produce tumours in the long term bioassays in rodents, is that correct? And are there multi-sites more than one site of cancer in the animals?
I don't know if I'm qualified to answer that question. I understand that it's 11 out of 12 sites; perhaps Man Sen Yong can answer the question for us?
Well, actually I think the JECFA has done the concrete evaluation of the toxicology of carbadox, including the genotoxicity and carcinogenicity. So maybe I would have the JECFA persons to answer for that.
Let me ask one other question while you are looking that up. How are, in relation to Dr. Arnold's answer, carbadox residues ever found in products consumed by people? And if so are the MRLs exceeded from time to time?
An MRL was not set for carbadox and nor was ADI specified. And it's permitted for use providing with a sensitive regulatory method you cannot detect residues. It's a metabolite desoxycarbadox that's carcinogenic. It's a short-term intermediate metabolite between carbadox and quinoxaline compound. And also quinoxaline-2-carboxylic acid, the end-metabolite, there were carcinogenicity studies on that as well as the intermediate desoxycarbadox. So you have the parent compound desoxycarbadox and the final metabolite. It' s a compound that' s used widely in some countries including the EU and the recommendation of JECFA was that they could not on the evidence presented state an ADI or an MRL for genotoxic carcinogenic.
So residues are never detected I guess of carbadox?
If the withdrawal time is observed we will not even find it quinoxalinic acid with routine methods. However, with more sophisticated methods you can show that up to 70 days after treatment this metabolite persists.
WT/DS26/R/USA
Page 317
Can I briefly just for my clarification now ask the Panel whether they consider it possible to compare the use of the hormones we are talking about and the substances we were just discussing, you were reluctant to do so. Could I have the views of the others whether this is coming into a category where we could really start comparing the use.
I will try and shed some light on it Mr. Chairman. When an ADI set for a compound, providing that ADI is not exceeded, then the risk associated with consumption of produce containing the compound in question up to the ADI is essentially zero. Now I use the word essentially to putting emphasis on the fact that zero risk is not an obtainable figure. However, if you exceed the ADI, then the nature of the risk varies from compound to compound. For example, if you seriously exceeded the intake of certain compounds, then depending on this toxicological profile, you may cause some quite serious effects. However, many of the ADIs are set on a relatively minor change such as a body weight change and change in an organ weight, some minor variations in some constituents of blood. So therefore to exceed the ADI in that case would not be as serious. And so what we do is we determine a level where there is no observed effect in the test animal systems, then we apply a safety factor and by doing that you essentially reduce the risk of consuming an amount equal to the ADI on a daily basis for lifetime to zero. However, once you begin to exceed that ADI you introduce risk. The rate at which that risk comes in, if I can put it that way, and the intensity of that risk varies from compound to compound, depending on that toxicological end point which made the no effect level upon which you derived the ADI. You've got a further question I guess?
Dr. Lucier, on this comparison question.
That's where I was trying to get some information so that I could determine whether or not such a comparison could be made, in terms of at least carcinogenic risk. I think that what Dr. McLean was talking about was not essentially a carcinogenic risk that was a hormonal activity risk. So they are two different things, two different methods to establish risk, but I was just trying to determine with the information available if it would be possible to compare the carcinogenic risk at residues that might be normally encountered versus the residues that might be normally encountered from the six hormones in question.
But the issue Mr. Chairman, I think as a number of people have already attempted to explain, is that at permitted residue levels, none of these compounds constitute a risk. That's the nature of the process of establishing an ADI and subsequently a MRL. It presumes that there can be lifetime dietary exposure, every day of your life, and that continuous dietary consumption will pose, for lack of a better term I'll say essentially no risk at all. The question that you are asking in fact is what if that residue level is exceeded? It goes beyond what's been specified in the regulation, or internationally, and I think the attempt that's being made by way of explanation is that it would be potentially misleading to suggest to you that these risks can be compared, because the ADIs that we've referred to repeatedly are based on entirely different end points. They can range from something relatively innocuous, in which case if an ADI is exceeded the consequence would be relatively trivial. And they can on the other hand extend to such issues as carcinogenicity, in which case it would be potentially a very serious consequence. So to compare the two, simply because they both have an ADI, presumes that that ADI has been established on a similar end point. Which is almost never the case.
WT/DS26/R/USA
Page 318
Would you think that the fact that the use above the ADI levels varies greatly also induces the possibility to have them regulated differently?
I think the nature, having been involved in the process of setting ADIs both nationally in my own jurisdiction and participating in the process on an international scale, one is left with the impression that the nature of the process is such that there is a significant margin built into the process, that in fact will have the effect of compensating for that in any case. I briefly alluded to some of those yesterday. For example the assumption is, one of the assumptions we discussed yesterday, is the consumption of 500 grams of meat per day. I mean I think we can all probably agree that in first principles that's an overestimate. And so the risk has probably been exaggerated in the calculation. That you will consume that risk every day of your life and so on and so forth. So I think collectively as toxicologists involved in that kind of a process, we have the impression that MRLs that are temporarily exceeded, occasionally exceeded, infrequently exceeded, are not likely to result in a significant health consequence, because the calculation has already built in so much overcompensation, that these occasional transgressions are not likely to produce an adverse effect. That's the inherent nature of the calculation itself.
Is this a view shared by the Panel?
I'm still searching for an answer to help me. I recognize that the carcinogenic risk of these agents is very low. Within that range of zero to very low, if I could get information I could say whether or not the risks were comparable for cancer. Now this is the genotoxic carcinogen, one accepts an ADI is established not on the basis of carcinogenicity for genotoxic carcinogens its on the basis of something else.
In the case of carbadox I think Dr. Lucier, what has been explained is that there is no ADI and there is no MRL, because it was the consensus of the Panel that reviewed the material that as a genotoxic carcinogen we were unable to establish what would popularly be referred to as an acceptable intake. The acceptable intake is nothing.
We would not normally set even an end point based on carcinogenicity, because generally carcinogenicity occurs at much higher doses than we would set a no-effect level. If the no-effect level was based on carcinogenicity, we would look at the compound and alter safety factors and other things, or again we might not permit it. And so in this process, in managing the risks through the ADI, the aim is really to make sure that you don't end up with carcinogens, as a result at least of animal toxicity studies, in the food chain.
Perhaps, Dr. Lucier, you mentioned two different methods of determining risks. One would be for hormonal action and the other for carcinogenic effect. It seems to me that we have an established method for the first one and the second one is somewhat new to the field here. Could you elaborate
WT/DS26/R/USA
Page 319
a little bit on the second method and how you see the implications in that particular field here, in particular when we talk about whether these substances can be compared or be treated alike and so on.
If I saw the tumour data for carbadox, I could establish a level at which a certain proportion of the animals got tumours, and as best I could, I could extrapolate down into the very low level risks what people might be exposed to. There would have to be some residues left of carbadox, just maybe not detectable, there have to be some residues left. And I could compare that carcinogenic risk to that from oestradiol or any of these other hormones in question. Now, that's a comparative risk of carcinogenicity the comparative potency in what people ingest from 500 grams of meat a day. So that would be a direct comparison. I would have to assume what carbadox levels exist and that was why I was wondering how often they are detected in fact in meat samples, if ever, and what the limits of detection would be. Then I could take, you know, something below the limits of detection of what people are exposed to and do the calculation on a rough estimate with considerable uncertainty but at least to see if the carcinogenic potency is in the same ball park or not. That's all that I'm after.
Is such data available or not? May I ask the experts?
I have never seen positive results and I see residue monitoring data since many years. But it has not been intensively monitored on the other side, that is the other difficulty. But I could perhaps say that at least in the Community it is only used in piglets up to a maximum age of 4 months, so it's quite realistic that the withdrawal times are observed and then there is no carcinogenic residue in the tissues, just maybe a trace of this innocent metabolite.
I won't belabour this any more after this one point. If apparently there is tumour data that is summarized, if I could look at that sometime during the afternoon I'll do my little calculations. My intuition tells me that the risks are going to be similar, carcinogenic risks.
I'm advised that there may be residue data for desoxycarbadox, but perhaps if I could restate the question. I think we are not coming directly with what I was getting at, we have heard statements and, I think Dr. Ritter was drawing attention to the fact that when JECFA and Codex review and approve things and set standards it is assumed that there is no risk associated with any of these substances, any of the six hormones at issue, or carbadox, or carazolol or the benzylpenicillin. The experts had been asked to speculate if the standard administration procedures are abused in respect to hormones and proper procedures are not followed, is there a possibility of a carcinogenic effect. And we have heard that in respect of some of the hormones the answer was, while close to zero risk, yes. And I've been asking in respect to something like carbadox, which has also been described as genotoxic carcinogenic, where an MRL and an ADI cannot be set because of it, if we take the hypothetical situation where it has been abused, administration has not been properly followed, can we draw the same conclusions? Or to put the matter another way, is there a rational reason to distinguish between permitting substances which are genotoxic carcinogenic is like carbadox and permit their use, while imposing a total ban on hormones which are said to have no risk, verging on zero as well?
WT/DS26/R/USA
Page 320
Who would like the floor?
Since you are asking for data, maybe for the record I could tell you what we have seen at JECFA when pigs were treated with carbadox. The residue level of this, I call it innocent metabolite, quinoxaline carboxylic acid was 18.9 micrograms per kilogram at 30 days withdrawal; was 5.5 micrograms per kilogram at 45 days, 1.3 microgram per kilogram at 70 days withdrawal.
This goes to the previous question. Would you come back on the figures here?
I actually do have the tumour data. Carbadox is a pretty potent carcinogen in terms of multisites. So the answer to your question is yes if improperly used and there were residues it would pose a carcinogenic risk.
And just before we leave that perhaps I could ask Dr. McLean to expand on his answer to question 11 where he discusses organophosphorus compounds and the health effects there. Perhaps I could ask him to comment on carazolol which I understand is used for transporting pigs immediately prior to slaughter and the comparative withdrawal periods, and whether there are any risks associated with the use of that drug.
In relation to the organophosphorus compounds, they are quite powerful neurotoxins in humans and in animals, and are used generally to kill insects although there are other uses for them such as removing foliage from certain crops. There are two problems with them, one is delayed nerve damage and that is generally a result of direct exposure of operators and the like. However, under some circumstances residues in crops in excess of the no-effect level have caused problems, acute problems of organophosphorus poisoning. And some of them are carcinogenic and so therefore safety factors to remove the carcinogenic hazard in the appraisal, that's been taken into account although, generally, the safety limits are based on acute effects that the poisoning effects that you see. But they have been associated with poisoning associated with exceeding the MRL. In relation to carazolol, carazolol does have an effect of lowering blood pressure and heart rate, it's related to the drugs commonly used in human to control blood pressure and heart rate. It is also used to prevent a specific form of stress in pigs during slaughter. And of course pigs are sent from the farm directly to slaughter generally and if you are going to have an effective treatment for the stress associated with transport, then you have to treat them just before transport so that the drug is effective during the transport process. And there has been concern that if people were to ingest carazolol at the site of injection then they may get a dose which could have effects on heart rate and blood pressure. The sensitive human population that was taken into account in this assessment are those people already under treatment with this group of agents for the control of blood pressure and associated cardiovascular disorders, and if one was to superimpose a dose of carazolol on top of the existing medication then there could be a problem. So these two compounds in various ways, if the ADI is exceeded can cause difficulties.
WT/DS26/R/USA
Page 321
Just a quick follow up. Are the organic phosphates ever used in animal production?
They are commonly used in animal production to keep under control insects on animals, relatively widely used.
Yesterday the European Communities quoted fingers, sorry figures, on the number of tests they do for residues of hormonal substances. We understand that Professor André's expertise is in the residue testing. He noted that the EC has developed multi-residue tests. And I wonder, Professor André, could you describe what substances the European Communities test for and comment on how many of these tests come back positive and for what substances?
I can give you some comments in the field of my competence, that means on hormones, on thyrostatic compounds and on beta agonist compounds. Concerning the first part of your question I cannot give you a complete list, so today of all the xenobiotic compounds which are looked for in a real multi-residue analysis, but you can imaginethat they include progesterone, nandrolone, oestradiol and many others, usually 30 different compounds are looked systematically for. Concerning the thyrostat they are just about 6 compounds usually, small compounds, and they are all of the same groups and derivatives of uroside compounds, true - uroside compounds. And concerning beta agonists the classical research includes now something as ten different compounds from clenbuterol and its group, I mean manbuterol and others and salbuterol as well as ractopamine and many others. All these compounds have been banned in EC and we are looking for potential misuse of these compounds. The level of detection of this research is very important to consider and its always below 2 ppb. and sometimes reach more sensitive levels as some 10 or 20 ppb's. for looking at residues in meat specifically. Concerning the second part of your question, in terms of percentage of positive results I cannot answer this question as a scientific people and the numbers I can give to you would be wrong numbers for different reasons. The reason is personally I am in charge of an official national laboratory and I have more positive than usual because I confirm results of other laboratories. This question has to be asked to official bodies or maybe to our representative of official Community reference laboratories but not to me.
I understand Dr. Arnold might be aware of results of testing of residues of hormonal substances in the EC and in particular a study or testing done in Belgium. Could you help us with that Dr. Arnold?
I think François André made a good suggestion since the results of the residue testing in the EU are given to all reference laboratories. It's a good suggestion that maybe someone from these laboratories answer the question. On the other hand I also see these results on a regular basis and I, what I can say is something that I have already said yesterday, we have to discriminate between random sampling which gives you a more relevant picture and sampling of suspect animals. And I said yesterday already and I would like to confirm this, if we take the results of random sampling it's not so bad the picture. In the majority of the member States you find nothing. There are some member States, I would not name the member States, but there are some member States where in veal calf there seem to exist problems, mainly with the natural hormones. If we go to suspect sampling then
WT/DS26/R/USA
Page 322
the picture is totally different, then we have member States where we find trenbolone, zeranol, where we find protestosterone, sometimes just a few samples, sometimes a higher amount. But this higher figure is not representative for the situation because this then sometimes are follow up analysis, so I hope I made clear what my point is. Based on random sampling the situation seems to be relatively okay, with the exception of veal calves in some countries. Based on suspect sampling, we find all 30 things sometimes in some samples.
May I add just a short complement to these two communications. First of all, its clear that all the compounds I speak about are not looked for in any other countries as in EC. And when we discover some misuse of such compounds, usually and quite systematically I think, the meat is not delivered to human consumption.
Yesterday we heard quite a bit of discussion concerning hormone levels and I would like to clarify with the experts what is a physiological range and how that range is determined and what controls that range. In particular I believe Dr. Liehr cautioned us that permitting the use of growth-promoting hormones in meat might upset the balance of hormones in a human ingesting the meat. Can the experts describe what the balance of hormones in the body is; is it fixed or does it change on a daily or weekly basis? And if hormones are added directly to a human, what effect does this have on that hormonal balance? Perhaps some comparisons between the ubiquitous 500 grams of meat and a couple of eggs or a glass of milk which are also ingested on a regular basis would be a useful comparison.
Concerning the balance of hormones we need to maybe a week for discussion to explain what is now known about the inter-relations between hormones and the complex feed-backs phenomena and so on. Just to have an idea of the effect of injection of hormones to human beings, it's well known I think that body-builders or sportsmen who are using hormones as anabolizing agents, with higher doses as usually used in growth promotion it's clear, but these people have always later very many problems in terms of reproduction and sterility its clear. That is one idea to illustrate what can a hormone do in human beings.
I agree with what Dr. André said. The hormonal situation in your body is changing as we speak. It's changing in my body, you get nervous your glucocorticoids go up, that does other things to you, so they are constantly changing, in constant state of flux. How each of the hormones regulate each other is not an entirely known fact, in fact its not really known at all, except that that type of balance does occur. There' s tremendous inter-actions, cross talk as we call it, between different hormonal systems. When this is disrupted there can be a whole cascade of events which could result in changes in biology and potentially adverse effects. After having said that, the amount that one would take in from say if we come back to the example of 17 beta oestradiol in eating 500 grams of meat a day, you basically take in one molecule of oestrogen if you were a women to every 280,000 in your body. So the chances that that would cause an endocrine disruption is probably very very slight, remarkably close to zero. If you are a man your levels would be one-fifth that so it would be one molecule to every 50,000 moleculesor something like this, so the chance that that would cause in itself an endocrine disrupting event is of course highly unlikely. There's no doubt that it's causing endocrine disruption in the animals that are receiving it, that's why they grow because their hormone system is disrupted and modified and the balance is changed. It's a question of amount.
WT/DS26/R/USA
Page 323
Just by way of illustration of the variation, if people want to turn to the attachment of my submission, you will see the enormous ranges in pregnant cattle, or in cattle of progesterone, oestradiol and testosterone; the ranges and the variations e.g. in pregnant cattle in muscle for example, the levels are 10,000 nanograms per kilogram plus or minus 6,600. Enormous variations. And I think that just reflects the individual animals and so it's terribly difficult to tie this down absolutely and precisely.
I mean a very popular comparison alluded to, the egg, the amount of hormones you eat with an egg and then it's compared to what you eat when you have an intake of hormone treated meat. Is that a comparison which can be done and could you elaborate on this?
Well I'm thinking we could and in fact a number of responses contain that direct comparison. The steroid levels would be contained in a glass of milk, for example. The steroid levels that would be contained in a feeding of breast milk to a new-born infant would be thousands of times higher than what one might expect in a steak or a hamburger. But if you would like a more precise number I'm just going to.
It is just whether one can make these comparisons, whether they are sound.
I think what I'm intimating Mr. Chairman, is that one really can't because they are in different worlds...
... You can't ...
Let me put it another way. One can physically make those comparisons, the numbers are available. But there is no basis for comparison. The steroid levels associated with a glass of milk or human breast milk or any number of foods would be many many many times greater...
... Yes, but these are the same steroids?
Exactly the same. We are talking orders of magnitude difference in terms of the dimension vis-a-vis the levels of exposure from a hamburger resulting from an animal that has been treated, when compared to a glass of milk that you might have from an animal that's never been treated at all.
WT/DS26/R/USA
Page 324
My colleagues have limited their statements to the identical molecules. If we would add to those identical molecules other naturally-occurring hormones in food which have a different potency maybe less potent, but have oestrogenic action for example, then we would discover that in many kinds of natural food including plant origin we find such activity. The steroids and their derivatives, as I said yesterday, were very successful in evolution. So, even if you go down to molluscs to sea urchins, you find these substances, you can't escape eating these substances, or similar substances with related biological potential every day. So, there is no way even if you decide to castrate yourself, in addition to starve to death, to escape these hormones.
My colleague Professor McLean points out that in the second submission of the United States (28 October 1996) that sort of comparative table, I have personally not verified each one of these numbers, but that sort of comparative table is in fact is presented on page 4. The intake of oestrogen or oestrogenic equivalents ranges from tens of thousands in the case of soya beans; the case of in an egg, you referred to an egg as 17,015 nanograms, and when we are looking at an oestradiol implanted steer, levels that have been estimated are somewhere in the order of 11 nanograms. A glass of milk might represent 75 nanograms in this table, cabbage 24,000 nanograms. I emphasize that these are oestrogen equivalents because in some cases it is oestrogen directly, in other cases these will be what is sometimes referred to as phyto-oestrogen which are compounds which are not quite oestradiol but are known to have an activity that is what we sometimes refer to as an oestrogen mimic, it's an oestrogen want-to-be. Zeranol is in fact produced for exactly that purpose. It is not oestradiol directly, but it is clearly, it is produced as a result of a fungus, but it has an action which is so similar to oestradiol although the potency is somewhat different, that we call it an oestrogen.
Just to follow up on that a little bit, as is obvious to every one I'm not a scientist and simple examples help me. Someone has suggested an example for dealing with physiology is to compare it to a thermostat in a house where there is a range of temperature that is contained and within that range are normal things that occur, and the temperature may vary up and down but that's a normal occurrence. And when one drinks milk or eats eggs or consumes cabbage there will be an increase within that normal physiological range. One of the things I wanted to follow up on, and perhaps Dr. Lucier can help, as he drew the comparison with the 500 grams of treated beef of one molecule from 28,000, by comparison are you able to help us with how many molecules per 28,000 we might find with a glass of milk in a pre-pubertal boy, the glass of milk for breakfast?
Just say young boy! to save further embarrassment!
Now the numbers for milk I have are a little bit different. I have them for 17 beta oestradiol and they are a little bit higher than what you see in muscle tissue, I have no way to know, since I haven't done the measurement myself, which numbers are accurate. These things as has been stated before, are present in milk and virtually everything we eat, oestrogen. So, you get an oestrogen load in addition to your endogenous oestrogen from a number of different sources. Not only the natural 17 beta oestradiol from what you eat but also from the phyto-oestrogens and the fungal oestrogens that were already talked about. There's a considerable amount of phyto-oestrogens in soya products especially soya oil. So we are exposed to a lot of different kinds of oestrogens just from day-to-day living that are exogenous, not produced by our own body. It would be very difficult for me to total them all up and that's one of my actual research interests to try and do that to see what the body burden oestrogen that we encounter from day-to-day living from exogenous sources, to get some idea what the risk of various kinds of
WT/DS26/R/USA
Page 325
disease might be because of those exogenous sources. So I cannot give you a very good answer except to say this would only be, the amount one received from eating 500 grams of meat, would be relatively small compared to the other sources of exogenous oestrogen one receives.
Mr. Chairman may I give you another example of hormone interactions, in relation to the quantity of hormones in order just to illustrate what has been said previously. We have known hormone the name is cortisone which has various properties and anti-inflammation properties. And we have always a secretion inside of these compounds. When you have, for example, an acute inflammatory process the physician will give you a high dosage of exogenous or similar compounds as cortisone, for example. And then your own secretion will decrease dramatically in place of the exogenous compound but after this maybe a week later you will no more need the ....... [tape ends]
For a chronic skin disease you will take very, very small doses of the same compound, cortisol or cortisone or another one, and your own secretions will decrease very, very slowly. But if you take these very small compounds as drugs during months, and then when you will stop, you will never recover your own secretions. It will have stopped definitely.
That's the difference between an occasional use and a permanent use.
What will be the conclusion?
We cannot compare the food intake of drugs or hormones in meat occasionally, or the consumption of a pregnant cow, beef for example, which can give you a lot of hormones and the small amount of hormone food intake for all the population during months and years. It is two very different things.
If I could give you another example, as a male contraceptive agent, male humans are given testosterone, which by feedback mechanism, inhibits the production of sperm. This form of contraception in human males is being trialed at the moment. In other words, you are giving extra testosterone and the pituitary gland sends a message down that there is too much testosterone and shuts off the natural production, but at the same time, it shuts off some of the hormones associated with sperm production, and renders the individual temporarily sterile. And of course once you remove the testosterone being administered to the male, then it all comes back to normal. That is therapy that is undergoing trial, at the moment, in a number of countries of the world.
Could I just come in with one question which puzzles me? When we see these naturally-occurring high exposures in milk or eggs, how does that relate to the cancer research we heard yesterday? Is the risk that is being assessed today where we have seen some thesis on it, is that equally valid for this naturally-occurring intakes every day or is there something special when we talk about growth promoters? Could Dr. Lucier perhaps elaborate?
WT/DS26/R/USA
Page 326
Let me do the best I can on that. If you start out with a pre-menopausal woman and her oestrogen level which of course is very high, this results by a tumour-promoting mechanism, and I don't think I need to go into that at this point, in about one woman getting cancer in every ten. So, out of every ten women in their lifetimes, one of them will get cancer. And that's primarily the consequence of her naturally-occurring oestrogens. And that's the point, and I've made a lot of comparisons throughout this meeting.
If you look at the beef, eating 500 grams of beef, you would get one molecule for every 28,000, the pre-menopausal woman normally has, and one could estimate a cancer risk based on that.
If one says then from other sources, besides beef, there is a hundred molecules, and I'm just pulling that number out of the air, but say it's a hundred, and that's probably a conservative estimate. Then, that would contribute a hundred molecules and so the cancer risk from all other sources of exogenous oestrogens would be a hundred times that what it is in beef, from a growth promoted animal.
So the risk would be very, very small. Eating meat from a growth promoted animal would only represent a small fraction of the total cancer burden caused by exogenous oestrogens, not normally produced by the body. Again, it's impossible to give a precise number for the reasons I said, because there is a vast number of environmental oestrogens out in the environment and we only know about a few of them. We certainly don't know about all of them.
I think my conclusion would be based on the assumption that all the oestrogens are acting alike, that they all interact with the oestrogen receptor and stimulate the same battery of genes as a naturally-occurring 17 beta oestradiol. That's certainly true for the more potent oestrogens, some of the fungal oestrogens, some of the phyto-oestrogens, these kinds of things. They activate the same kind of genes so there is a reasonable scientific foundation, for my conclusion, although there is some uncertainty in it.
Mr. Thompson.
Professor André, there are a couple of questions I have concerning some of your answers to the Panel and I am hoping you can clarify them for me, in particular, your answer to Question 2. You state in part:
"On the other hand, if the physico-chemical or organoleptic criteria are concerned by this question, and the question of meat quality can be discussed, significant alteration of eating quality, Lowman et al, and significant loss of tenderness, Gerkin 1995, of meat produced with various implants have actually been reported."
When we go to the source of Lowman, the statement does not seem to be supported. I could read from page 48 of the Lowman Report. It says:
"With regard to the composition of the carcass and the meat, it was apparent that the use of the hormone implants affected a reduction in the percentage of carcass fat, an increase in the percentage lean and a marginal decrease in eating quality..."
as opposed to a significant alteration, as mentioned in your quote. It goes on:
WT/DS26/R/USA
Page 327 "... and an increase in the proportion of saleable meat for both heifers and steers."
In respect of the Gerkin Article, where you have indicated the Gerkin article identifies a significant loss of tenderness, at page 3323 of the Gerkin Article, the author states:
"Results of the present study suggest that the use of single implants containing oestradiol, TBA or the combination of oestradiol and TBA, had little appreciable effect on deposition of intramuscular fat or on beef tenderness."
Can you help me with how these articles support your statement?
It's clear that I have not invented this assertion and that they are also inside. When you see only Lowman abstract, in the last sentence. It says:
"...with marginal reduced eating quality."
That means that there is a small reduction of eating quality but there is one.
If you discuss about the difference between significant and marginally, marginally can be significant, statistically! Sorry, but statistically, it can be small, but significant!
I also had a question about Question 6A.
This is dealing with whether there is any more recent scientific evidence available with respect to the effects on human or animal health of the use of the six hormones in dispute, especially when used for growth promotion purposes, other than the evidence already taken into account by Codex. And you state as follows:
"Concerning human health, there is a discussion on the decrease of human sperm counts" and you cite Nimrod and Benson:
"An apparent increase of the incidences of hormonally mediated diseases like breast cancer and endometriosis and the decrease in the male/female ratio which are thought to be due to oestrogens in the environment. Whether the use of hormones for growth promotion is contributing to this or not, cannot be scientifically proven, at the present time."
When I look at the source, namely Nimrod and Benson, it appears that the conclusion that sperm counts have declined, is in fact controversial. And the only reported decrease in human sperm count refers to counts of men in Paris, over the past 20 years, which is within the European Community and unlikely to be directly related to eating hormone-treated beef.
Can you help me with the correlation between the decrease in sperm and the sex hormones at issue for growth promotion here? This question is addressed to Dr. André.
The sperm count issue is controversial. Some people feel that sperm counts in the industrialized world are going down, other people feel that they are not. And there is a lot of methodological issues, without going into the details, that are creating the controversy.
WT/DS26/R/USA
Page 328
The decreased sperm count has been reported in more than one publication, however, but that has not removed the controversy in this legitimate scientific debate. It is not only men in Paris. For example, men in California only have half the sperm as men in New York. I'm not sure why! The men in New York are very proud.
It is a very a controversial issue. There are data from experimental animals that show that neonatal exposure to oestrogenically active agents does cause a decrease in sperm count, later in life, in the male offspring, and there are plausible biological mechanisms for that.
But again, the question of whether or not, the amount received, I think it is highly questionable that the amount received in eating beef from hormone treated animals would be necessary to produce that.
To contribute something perhaps to the debate here, I think my colleague, Dr. Lucier puts it quite accurately. The French data that was referred to a moment ago, was published by Pierre Roget, I believe in 1992, in the British Medical Journal. It was the first of subsequent reports that were to come popularizing the issue of declining sperm counts. There were subsequently significant methodological issues identified in that report, not the least of which were that many of these subjects originated from infertility clinics or sperm banks! Consequently, I think it was widely recognized that this may have been an inappropriate sample to use to examine declining sperm counts, or for that matter, anything related to sperm.
I think Dr. Lucier puts it quite accurately, when he says that this is very much an emerging area and I think the best thing or the most accurate thing that one can say, is nothing at all, because one can find as many reports in favour of a declining sperm story, as one can find opposed to it. As a matter of fact, the New York data that Dr. Lucier referred to a moment ago, was recently reported by Harry Fish, who declared New York men to have the highest sperm counts in the United States, but no higher than New York men have been for the last 25 years.
The point that Dr. Fish was attempting to make is that it would be inappropriate to compare New York men with Californian men, but rather you have to compare New York men with New York men, California men with California men, because there can be many variables that affect sperm count within a geographic region. I think that the point that he was trying to make is that it is inappropriate to compare across wide geographic regions, the bottom line being that it is certainly an emerging area, but one about which the only certainty is complete uncertainty!
Would you like to take the floor?
I read this paper one month ago and I cannot find exactly the sentence but I am sure to find before this evening something so in this paper. I think you read it also, so we can agree at the end of this session that there is something.
You cannot say that the sperm elaboration is not under hormonal control, that's clear. And you cannot also say that some alkylphenol ethoxylates have a very dramatic effect on sperm elaboration in men and that these compounds have also some oestrogenic effects. That's clear or not?
WT/DS26/R/USA
Page 329
With respect, I'm not sure it is quite as clear as perhaps you feel it is. The nonyphenol and other ethoxylated compounds, I don't want to get into a side debate here, but there are a number of authors who have published reports indicating that when compared to E2, in terms of their oestrogenic potency, they are 10,000 times less potent. That order of magnitude. In fact, one author suggested that if we were to call nonyphenol oestrogenic, there would be few substances that would not be considered oestrogenic, on the same scale.
But I think to put it more correctly, into perspective, I think as Dr. Lucier indicated, there have been dramatic increases in breast cancer which appear to go beyond those increases that can simply be attributed to the introduction of mammography, particularly in the industrialized world. There have been increases in prostatic cancer, the origin of which seems to be at least in the first instance, related in some significant measure to the introduction of prostatic specific anogen. For those men in the room who are over 40, you'll know the experience very well.
The interest in a potential role of the environment in breast cancer, I think, stems from the fact, that when we add up all of the known risk factors for breast cancer, we include the age of first pregnancy, the number of successful pregnancies, the number of lactations, the duration of those lactations, all of those things that we know are important risk factors in breast cancer. We can't account for all of it, and consequently there has been an interest in the scientific community to look for other things that may be influencing the disease.
I think it will be some time. Certainly, I doubt very much in my lifetime, where we will be able to demonstrate a clear definitive role for a specific environmental substance. The work is justified because it is the disease that has such enormous proportions that a small contribution has an enormous impact. In the United States and Canada, to put it into perspective for you, there is a new case diagnosed approximately every 29 minutes. So you don't need to make a big dent to have a huge impact. I think the work is entirely justified, but for those who are looking for quick fixes, it's going to be relatively unrewarding. The answer isn't around the corner.
It's around 4 o'clock. How long are you planning to proceed or shall we make an inception with the Community presentation. What would you suggest?
Mr. Chairman, notwithstanding the slow start, if you'll permit me, I'll take 30 seconds to thank the scientists for their very helpful and concise answers and turn the phone over to Mr. Christoforou by 4 o'clock, as he wished.
Thank you very much for your understanding.
May I give the floor to the European Communities, Mr. Christoforou.
I would like to thank the Delegate of Canada for his understanding, in this particular case.
WT/DS26/R/USA
Page 330
Before proceeding, I would like to remind you that we have formally requested through the Panel, Dr. Ritter, the paper of Dr. Truhaut and we would appreciate if we can receive a copy of this paper for our scientists to have a look at this.
The speaker who will be leaving very shortly is Dr. Adolfo Pérez-Comas from Puerto Rico, who has been investigating the possibility or the explanation of a wide-scale premature sexual development in Puerto Rico for more than about 20-25 years.
I understood there were some problems raised by the Delegate of Canada whether these projections will be made. I can only announce there are, I think, 11 or 12 slides. Some of them are graphs showing the increase or decrease according to which diet is followed, and they are only slides which show what exactly we are talking about. I have viewed myself these slides and what they are. Most of them are documents and charts of estimations and there are only two slides with the effects on children, pre-pubertal girls, what it is about.
I would give the floor if our colleague from Canada or probably the United States, would like to make a comment.
Mr. Brinza.
As my colleague from Canada had indicated earlier, we share similar concerns with the proposed presentation. It sounded from the description I just heard, that we are talking about a presentation of new scientific evidence, not a presentation designed to fall within the purpose of this meeting, today, which is to talk about the responses to the 28 questions of the five experts, and therefore we don't see how this is appropriate, at this point in time.
It's hard to talk about something you don't know. The question is whether this comes in as new evidence or whether it supports what has been submitted by the date. As a practical matter, this material can be introduced tomorrow, at any rate. You could do this in the second substantive meeting. And from that perspective, just a very pragmatic point of view, I think it's simpler if it's done here.
Are you saying it's possible to introduce evidence of the second oral meeting tomorrow?
I'm afraid I erred. Lawyers are often known as being champions for procedures. But the very good lawyers are interested in the truth and so sometimes they subject procedures to the finding of the truth, but this does not alter the fact that, yes, we can not introduce new evidence.
Whether it's done today or tomorrow. It would really depend on whether it's within the realm of what has been submitted by 8 February.
The United States has circulated previously a document which was not submitted by the 8 February. The issue about the incidence in Puerto Rico has been argued by the European
WT/DS26/R/USA
Page 331
Communities, both in the submissions concerning the United States extensively, and in the first and second submission for Canada.
The issue has been touched. It was within the time-limits. It is not new evidence. It is an elaboration of what is already in the submissions.
I could not have the slides so I could send them before the 7th. That is what so-called the new evidence, the slides, but I have here also the documents which I would request their circulation. If you don't allow the slides, so we can circulate the documents and have a look at them, but you will miss the "nice" pictures of girls under the age of 8 with exactly what we are talking about. This is not a new document. They are well aware of the issue. We have debated this during the second substantive meeting. The United States has submitted documents about this case. They are also annexed here so we don't really take just one view. The documents submitted by the United States, are in this document. I can circulate the document if you wish and the slides are there. There's no new evidence in this case.
Both of the delegations were aware when we discussed for the first time on 7 January.
Yes, Mr. Thompson.
If I could respond, I am not, nor is the Canadian Delegation afraid of the truth. But as an experienced lawyer I know that, just as in science, there are a great many truths. And what I am concerned about, Mr. Christoforou has raised what occurred on 7 January. You will recall that I specifically asked Mr. Christoforou to identify the experts he proposed to call so that we would have an opportunity of preparing ourselves to deal with whatever evidence they had to present. He has never, until yesterday, indicated that this evidence was going to be called or who this person was.
As a matter of substantive fairness, irrespective of procedure, it is unfair to permit someone to present one side of an issue that may be controversial and inflammatory without a proper opportunity of having evidence that may go to another point of view.
Mr. Brinza.
I would also point out that whether they are slides or documents, if the point is to go back to issues or points that were raised in previous submissions and making another presentation here, that is not the purpose of today's meeting. That would make this another substantive Panel meeting. That is not what we are here for. The Community had plenty of opportunities in our Panel process to present this and they said that they already have. There is no reason to go through this again today. We are not talking about being afraid of the truth, but as lawyers, those procedures, I like to think are designed to help to make sure that the truth is brought out and having breaches in those procedures is not a way to ensure a fairness and a fair airing of the facts.
Mr. Christoforou.
WT/DS26/R/USA
Page 332
Mr. Chairman, from what I heard, they both don't deny that the issue of the incidence in Puerto Rico has been discussed extensively in the submissions. This is true. I see Question No. 7 of the Panel to the Experts saying is there any evidence and scientific evidence available which demonstrates that the potential for adverse effects on human health arise and whether there is any evidence to that effect.
As we did yesterday, I could have Dr. Pérez explain orally, as the other scientists did, since the issue has already been touched before. If you indeed insist that these slides are not projected, at least, there is a question to the experts and we are allowed to make comments.
As we did yesterday, we can elaborate five minutes and then answer the question. Is that really outside what is the understanding of today's meeting? You have been circulating documents yesterday and we can circulate those documents, as well. There is a specific question of the Panel on this issue. So as we agreed yesterday, he may make this presentation ten minutes orally and have no projections.
Mr. Chairman, I would forcefully really object to that type of procedure which is applied only unilaterally, in this case, I'm afraid. These issues and facts should be explained in the presentations and there is a question which we would like to comment on and provide a reply and ask a question.
As I said, it is difficult to talk about something you haven't seen, so logically we have to see it before we can make a ruling. But I don't want to spend too much time on these procedural questions and I would invite you to make the presentation. Then the parties would have an opportunity to comment on this, if they want to, the United States and Canada. So the fairness is preserved.
The Panel would perhaps reserve it's right to submit one or the other question to the experts. So may I ask you to go ahead with the presentation?
What should I understand by presentation. Should I make it myself?
No, no. Your expert, please.
Thank you Mr. Chairman. Good afternoon ladies and gentlemen, members of the Panel.
My name is Adolfo Pérez-Comas. I am a pediatric endocrinologist and clinical geneticist, fully trained under United States' standards and certified in endocrinology and medical genetics under United States' standards and an advisor to the Centre for Disease Control and the National Institute of Health in the United States in the area of my specialty, participating at different ongoing steering committees, both at the CDC and NIH. I am also an associate professor of pediatrics at the Ponce Medical School and previously at the Mayaguez Medical Centre and the University of Puerto Rico Medical Science campus.
We are going to present to you today accurately documented data collected in Puerto Rico that has been published in prime medical journals, such as the Lancet, the New England Journal of Medicine, the Journal of Pediatrics and the American Journal of the Diseases of Children, as well as
WT/DS26/R/USA
Page 333
in the Puerto Rican Medical Association Medical Bulletin that is a certified journal in index medicals, and all of these journals are peer-reviewed by leading scientists in the field, in the United States and in the world.
I will condense here the 50-year history of Puerto Rican exposure to hormones with 28 years in the practice of pediatric endocrinology with 14 slides presenting anomalous sexual development in my country.
I might also add that Puerto Rico is the second-largest producer, at the moment, of oral contraceptives and has been exposed for over 40 years to oral anti-conceptives because they were tested in Puerto Rico prior to its commercial disposition in the world.
Original studies were carried out in the island of Puerto Rico in the late 50s.
The aspect considered today has been presented at guest lectures in the United States, Canada, Spain, Cuba, Venezuela and the Dominican Republic. This is a summary of the patients we have seen under our care, dating back to 1969, up to the year 1985 and part of 1986, where most of the studies carried out by the CDC were done at the same time. We are not including recent data. We are just going back to the date when the data collected by the CDC was taken on.
This is a year-and-a-half old female child that presents anomalous sexual development, categorized as premature telarche, becauseshe has isolatedbreast development, that is not normal for a child eighteen months old. Breast development should appear normally in our population after eight years of age. If you have pubic hair or laxial hair, that should not appear before the age of nine years, and if you have premature menses or menarche, it should not be present in anybody before ten years of age. Anyone of those conditions that appear before those ages, is not normal. It may be secondary to these different processes or tumours in the organism or it may be secondary to exogenous hormones in different areas that can be present.
This is a ten-year old girl who presented breast tissue at seven years, who presented pubic hair at a normal age, at ten years, who presented menarche at a normal age of ten years, but who also presented polycystic ovaries, a highly elevated serum total oestrogen for her age and extremely developed breasts, that I classify as virginal hyperphacial of the breast, a condition that was not previously associated with premature telarche or with increased serum total oestrogen.
This is another aspect of the problem we are seeing in Puerto Rico. We have severe cases and we have mild cases and it affects children of any age and it affects males, it affects females and it doesn' t affect only Puerto Ricans. It affects continentals, North Americans living in the island, German people living in the island, people from Latin America living in the island. Once they get there, some of them suffer from these problems.
In the patients we have studied up to that date, we measured, in a significant number of them, the level of oestrogen. And we measured specifically that serum total oestrogen that were carried on under the same standards of Dr. Arnold presented yesterday, in the table with the normal value as specified and this procedure was carried on in different United States' research and commercial laboratories, Biosign, Intersign, Project Clinical Laboratory in mainland United States and not in Puerto Rico.
Approximately 85 per cent of those male patients who were tested and 86 per cent of the females that were tested, presented high oestrogen levels. I am speaking of 522 patients here and I am speaking here of 146 males with a high oestrogen level. I am not speaking of 30 patients, of nine patients or of 15 patients. These are a significant number of patients who presented high oestrogen levels.
WT/DS26/R/USA
Page 334
When we started studying these clinical cases in the early 70s we found an increased size of the uteri. At that moment we didn't have available sonographic studies, laparoptomies in some children we have to investigate what happened in the ovaries. The uterus was found enlarged in some of them with the present ovarian cysts. When we started having available pelvic sonogram, this study was also done in a significant number of patients who can afford to pay for it. Out of 447 pelvic sonogram studies, in females with anomalous sexual development, 276 of them presented abnormal pelvic sonograms, enlarged uteri, polycystic ovaries, the same that can be seen in oestrus, in animal husbandry.
62 per cent of those girls presented ovarian stimulation syndrome. When we studied the bone cage, we found that in 26 per cent of our patients, the bone age was accelerated. This implicates that the excess of hormones that was present in these children have been of more prolonged duration because it was accelerating the bone age of our children. This data has been published in the New England Journal of Medicine, has been published in the journals I have mentioned previously, and can be corroborated by anyone.
We were searching for the possible cause of this in our island. We looked for exogenous causes because the condition didn't fit what it should be in a normal endocrine situation.
As a pediatric endocrinologist, I should see one, two or three patients with abnormal sexual development in a year. Once upon a time I was seeing four or five a week and some days I was seeing three or four, in the same day, new cases.
We found in 1982 that the diethylstilboestrol was being sold over the counter in Puerto Rico, this was done in one local agricultural store in Puerto Rico. That Ralgro or Zeranol was also being sold over the counter at that time. DES was banned in the United States at that time and Puerto Rico has the same rules and regulations as the United States, and due to that, we started suspecting that something was in the feeding our children were getting and we started a group of our patients on a diet. We modified the diet with limited ingestion of meat, poultry, milk and eggs. In a period of four to six months, the males that were on a diet, 60 out of 103, 58 per cent of them presented a partial remission of the condition. The abnormal hormones in the blood started going down, the breast tissue that was present started to diminish, those that had pubic hair showed no difference, it did not increase, but it didn't disappear.
In the females, we observed more at the same period of time, 51 per cent reduction with the diet and a partial remission and even a total remission of the symptoms we have observed before. As I pointed out previously they have been published by that date. 11 per cent of the females and six per cent of the males on no diet, also had a remission, but the remission presented in a year and a half to two years. In an obvious period during which the diet may have changed by itself in growing children. This confirmed indirectly that we were dealing with some kind of exogenous contamination. There were not cases of true precocious puberty by endogenous origin.
This remission is presented here regarding the different ages in 1969, 1970, 1972, 1973. There was very small remission because we had not introduced the diet. When we introduced the diet modifications and after there was a big public awareness in the island when this situation appeared in one patient in the Time magazine of the United States, the remission started increasing. Not everybody remitted but a significant number of them did so with the modification in diet.
The FDA did some screening studies and this is a copy of one of those studies, where a poultry from Bonito Puerto Rico and a poultry from Royal Brothers Processing in Mississippi were found in this screening study as positive for oestrogenic substances. Posteriorly it had been reported that these were negative in the next studies that were performed by the Federal agencies.
Dr. Carmen Saenz who has been working in this area with me, who at the same time presented a publication in the same bulletin of Lancet as I did, I presented a 272 page and he presented 350 of
WT/DS26/R/USA
Page 335
the same condition. She measured percentage of zeranol through a laboratory in Paris by Dr. Morfine where five out of six patients that were so tested revealed levels of what appeared to be zeranol, by that time or a zeranol-like substance, confirming indirectly the suspect we had.
We have now a screening test positive for oestrogens and a test, possibly positive, for zeranol.
In 1986, Dr. Fred E. Titulowel(?), another collaborator from the CDC, published an article on an epidemiological study done in our patients in Puerto Rico. 50 of my patients, 60 of the patients of Dr. Saenz and approximately 20 more patients from the University Pediatric hospital in St. Juan.
Initially, we were told that this study didn't reveal any relation to what we were seeing related to exogenous hormones, but if you look in detail in the conclusions of this study, this study reflects a positive and statistical association between premature telarche and the consumption of soya-based formula, which has been widely known for a long time, various meat products and a maternal history of ovarian cysts.
Obviously the mother is eating the same food as the child and the family. With the clinical experience we have with the diet and with the presumption of the laboratory study, we must conclude that we are strongly thinking that an exogenous thing is happening in Puerto Rico and is strongly suggestive of an exogenous contamination by these substances.
Last week, on 2 February, the governor of Puerto Rico under the programme of telarche, that is the only one that exists in the world made some declaration to the San Juan Star, that is the local press. The number of children with signs of early sexual development nearly doubled in the past six years according to figures released by a survey by one of the island's top experts on this condition. The statistics adds fuel to conjecture that Puerto Rico has the highest incidence of premature sexual development in the world. There is no other place in the world with a similar situation said Carlos Bourdony who runs the Health Department's Premature Telarche and Sexual Development Programme.
A study to be published in the April issue of Pediatrics, the Official Journal of the American Association of Pediatrics suggests that girls seen in pediatric office practice are developing pubertal characteristics at younger ages than suggested in standard pediatric text books and by earlier United States' studies. Early sexual development has been monitored in Puerto Rico since 1988 when the local health department set up a register, the only one of this kind in the world, according to Dr. Bourdony. To date, 6,115 cases of children with premature sexual development have been documented by the Puerto Rican Department of Health.
Could I ask you to come to a close. You have one more minute.
This figure is nearly double. The 3,162 cases documented between 1969 and 1991 and do not include the 4,500 patients evaluated by Dr. Saenz and myself. This includes the patients evaluated by the other 11 endocrinologists. [Tape ends]
... studies ..... performed by the USDA Department of Agriculture found no significant or trace level of growth and enhancing hormone in the livestock specimen, according to Bourdony. But in the absence of conclusive research, we can't say this factor has been completely eliminated, as is stated by Dr. Carlos Bourdony, the Head of this Telarche Programme of the Department of Health in Puerto Rico. Thank you very much, Mr. Chairman.
WT/DS26/R/USA
Page 336
Thank you very much Professor for your very interesting presentation. The matter was raised, reference was made in earlier proceedings to the Costa Rica studies. We will assess to what extent new evidence was submitted here, in particular, I refer to an April study which seems to be new to us and the Panel will reserve its rights to regard or disregard the evidence submitted in the light of this test here. I would nevertheless as I indicated briefly, give the opportunities to Canada and the United States to comment on this. And I can summarize of course my impression that nothing has been said about the real causes which may have led this situation and it would take very elaborate discussions to go into this here.
Thank you Mr. Chairman. That would be my immediate reaction as well, as I am unsure how any of this material relates to the issues that the Panel must decide. If I was to be putting questions, I would be interested to know what tests methodology was used for zeranol, were the results confirmed, and what is meant by a zeranol-like substance?
Thank you very much, Mr. Brinza.
Thank you, Mr. Chairman, one moment please. I would like to have some of our experts comment on this very briefly. Dr. Miller who is here with us is familiar with this situation, as are some of our other experts as well.
Thank you very much.
I am Dr. Margaret Miller, Centre for Veterinary Medicine, FDA, and I would like to say that these types of health effects are very disturbing to the Agency and we expended a large amount of research dollars and energy, as did our colleagues at the USDA, investigating this situation. And as a result of that activity we concluded that this situation is not due to the hormones present in meat from animals treated with approved products used for growth promotion.
And this is Dr. Richard Ellis who is also familiar with this situation.
Thank you, Mr. Chairman. I had a responsibility for coordinating our sampling programme during that special study. In fact we actually conducted three studies, over the periods of which were referenced by this presentation. The third study which we conducted was by far the most extensive one and we looked at approximately 700 samples that were collected from beef animals, pig animals and poultry. We looked at both domestic production from slaughter plants, from imported product at inspection sites, and also product collected at local markets. And as Dr. Miller said, in all of our studies, we did not find the presence of any zeranol, DES or related products in our study. But in as much as we did not have methodologies for all the products that we would have liked to have looked
WT/DS26/R/USA
Page 337
at, we even submitted a portion of those samples for bioassays by FDA. And we had negative results on those bioassays of samples submitted to FDA for that purpose. If it would be worthwhile to the Chairman, when we go back to Washington I would be glad to send you copies of those documents to support my statements. Thank you Mr. Chairman.
Thank you very much. May I ask whether one of the members of the experts would like to take the floor on this presentation here? Yes, Professor Mc Lean.
I just would like to make two comments. First of all, I really do not see what this has to do with the issue of using hormonal growth promoters in accordance with good animal husbandry practice. And just a comment, the use of the word "zeranol-like" would suggest to me that there is widespread contamination of food material with Fusarium fungi, for the Fusarium fungus which produces zeranol and its related compounds, so that might give some indication as to what one might look for if one is looking at a contaminant. And I would draw attention to the Panel that zeranol was actually discovered because Fusarium contamination of pig feed, as I remember it, produced signs of oestrogen responses in pigs, as I remember it. It was from there that the research work identified that it was a product of fungal metabolism by the Fusarium species and on it went there. So I would suggest that the Puerto Ricans might like to invest some money on widespread looking of where the Fusarium is, what species it is and what it is producing. But I do not believe that this has any relevance to today's hearing.
Excuse me, if I could add a bit to that. It is known that zeralenone and zeranol-like substances produced from Fusarium do produce hyper-oestrogenism in swine and in cattle, which it fed mouldy corn that contains high levels of this, and likewise it can cause neonatal problems, but these are usually massive amounts. But then I think there have been some Indian tribes in the US who have eaten large amounts of mouldy corn in which there was oestrogenic responses which are adverse as well. So there is information in the literature as to suggest that high levels of these, can cause adverse effects, in a number of different ways. The amounts that are present in growth promoted animals, however, have been very, very much lower than this and would not produce those kinds of changes that they were reported here. Although it is possible zeralenone substances from mouldy corn and other products could do it.
Thank you very much. Can we close the discussion on this presentation. Mr. Christoforou?
Mr. Chairman, if you will allow us we would suggest that this presentation is about 90 per cent a reproduction of the document that the United States has circulated during the second oral meeting we had here on 11 November. As rightly Dr. Pérez has said, there are some uncertainties about the actual cause, but he has provided someinsights, some information on the potential possible explanations. In the article which the United States has amended to their submission during the second meeting of the Panel, that was the official explanation given by the United States and what we did is that Dr. Stephany has gone today through the article provided by the United States as an explanation of the situation in Puerto Rico. He would like to make a two-minute statement about the analytical methods used, in that article provided by the United States in this case.
WT/DS26/R/USA
Page 338
Is it absolutely necessary? Because we still have a list here of questions which are very much to the point here and we do not want to, but I do appeal to you and urge you to self restraint just to really submit issues which are pertinent. Thank you.
Thank you Mr. Chairman. Some material has been disclosed to me yesterday and tonight. I am not quite sure whether that is the document Dr. Christoforou spoke about, so have I to make that restriction, but basically it is not essential. I had the pleasure in 1983 according to my recollection, that Dr. Alice came to my laboratory and we discussed the whole case. At that time we had superior technologies available for handling these problems, unfortunately samples never showed up and for me the case was closed and only yesterday I saw some details. I will not go into any details but summarizing what I have seen, I was very much surprised about the small amount of potential substances that have been checked or could be seen by the techniques used at that time. That what especially in the field of the zeranol-like compounds, I do not think that I have any criticism basically, but some of the so-called bioassays and receptor-assays I think I have severe criticism that potentially strong oestrogens have slipped through the system, not being detected. So stating it very shortly that nothing has been seen so it could not come from xenobiotic oestrogens, maybe that is a little bit too strong. But Dr. Miller summarized that this did not come from the compounds that are in the regular implants or those implants that once have been regular, like diethylstilboestrol. So I don't want to go into any details, I only offer once again after all those years to cooperate, to clarify this very intriguing affair in Puerto Rico. But whatever the cause is, it is a very serious problem, and I learned from my colleague that the problem has not stopped, elsewhere in the world also similar problems are showing up. That is what I have to say.
Thank you very much. Mr. Christoforou you may continue.
Thank you Mr. Chairman and thank you very much indeed. Both the US and the Canadian delegation of this particular issue, which was of serious concern to the Community because we have been making this argument since the very beginning. With your permission then, our next scientist who will be leaving is Dr. Adlercreutz. He has already provided the paper within the time-limits. For this paper the United States have already made comments and I would state that Dr. Adlercreutz will touch only upon one which we think important and relevant issue in this case. I will give the floor to Dr. Adlercreutz.
I was asked to say something about the possibility of carrying over anabolic steroids to humans after eating meat. I have brought with me one of the articles, and the other because I came yesterday from the USA, I could not recover the second one. But in essence it is so that we can never be sure that the farmers really use the substances they are legally asked to use. In this article written by Demetrios Esgotas and Tom Tuten from the Department of Pathology and Laboratory Medicine Emory University School of Medicine in Atlanta. No excuse me. It was Debruyckere, Sagher and Van Peteghem and this is from the Netherlands, in Ghent, Belgium, excuse me yes [passage unclear]. I received this yesterday. In any case they state here, in this country I know that they are very careful with their testing of the meat and they tested in 1990-1991 608 injection sites. And they found that over 60 per cent contain at least one illegal anabolic agent. And I have a very long experience in the doping field and I would say that I am of the opinion that it is impossible that good veterinary practice
WT/DS26/R/USA
Page 339
is used because there is some big economical benefits of getting a high gross. So I would like to ask the American delegates and Canadians also, whether any such study has been done in the United States, testing really how much illegal anabolic steroids are used. In fact there are two articles, one which I couldn't yet get but I was present at the meeting when it was presented, and they found in people who were not sportsmen and not taking part in any such activities which would need anabolic steroids, they had found one case with nandrolone metabolized in urine which must have come from the food, because there was no other possibilities. And then they did this study here and found in the control group one subject who had clostebol which is anabolic steroid in the urine, despite that there was no possibility that this person had taken any drugs. Then they added a little bit of this drug to meat and fed it to some people and found that they could detect by doping tests these compounds. And we know from the doping field that for example, nandrolone injected into humans can be detected four to six months after they have been given. So, in my opinion it is a very serious thing if sportsmen are caught and getting doping positive samples, if they eat meat. And the risk as well is relatively high, because now it is necessary for every doping laboratory to have high resolution apparatus, so it might be that they can detect even after six months, so very small residues in meat. I mean I'm sure that if you test things in meat and you get negative results you can still find these compounds in urine, absolutely sure, because the urine methods are extremely sensitive and the meat methods are not so sensitive because it is difficult to extract these compounds. So I would like to ask whether this is taken into consideration, that the farmers don't follow the rules and that this is in fact very comparable to the sports field?
Thank you very much. According to our rules these questions may be submitted to the Members of the Panel and may I invite you to comment on what Dr. Adlercreutz. said and I also refer to the paper which was circulated, the evaluation of the 32nd report of Joint FAO/WHO Expert Committee in Food Additives and the discussion of Professors Liehr's report. Have you been able to read this? But if not, could you comment on the thesis that Dr. Adlercreutz is submitting.
Pigs and horses naturally produce 19 nortestosterone and we find that derivative of testosterone in their tissues and in urine. They can also, using the same enzymic pathway, produce 19 norsteroid derivatives of anabolic steroids, which is permitted in normal therapy in horses for certain disorders. And so it may well be possible that the human athletes have eaten pig or horse meat which could be the source of 19 nortestosterone, or even may have eaten horse meat from horses which have been treated 19 norsteroids. And so therefore it does not surprise me the sorts of results that Dr. Adlercreutz has reported are so.
It seems to confirm what I have said yesterday and today and what I have also said in my written answers. There is a significant black market in Europe and in some countries it is more important than in others. So I perfectly know that, specifically in Belgium, a long list of illegal substances are regularly found. The authorities spend a lot of time and money to analyze many samples. But I do not see what the relevance is either the legal use for growth promotion or for therapeutic purposes. I mean, on both sides of the ocean, we have strict rules and we see an example that on one side of the ocean these rules are violated.
The 1984 FAO/WHO expert consultation certainly pointed out the opinion of the experts at that time that the illegal use of veterinary drugs created a far greater potential public health problem than the regulated use for growth promotion. I think that that finding in 1984 probably still stands. I would like to point out two texts adopted by the Codex Alimentarius Commission, and I might add
WT/DS26/R/USA
Page 340
adopted on the basis of consensus, by the Codex Alimentarius Commission. The recommended International Code of Practice for the Control and use of Veterinary Drugs and the Codex Guidelines for the Establishment of a Regulatory Programme for the Control of Veterinary Drug Residues in Food. Both of these texts exist to guide governments in establishing control practices and in controlling veterinary drug residues. To go deeply into the second one of those texts, it does get into a lot of sampling statistics and unfortunately sampling statistics tend to show that unless you are willing to essentially send all of your food supply to the test laboratory, and therefore starve to death, you are not really going to find what it is you want to find at very, very low levels. So sporadic cases such as we have heard could probably exist even within the framework of a properly regulated control programme.
Before I give the floor again to Dr. Adlercreutz I would like to say something, something which probably might have escaped the point which we were trying to make. The point we are trying to make is not only from illegal use, it is also as we have heard yesterday from Dr. Lucier, that there is a probability even from the use according to the good agricultural veterinary practice. It may be an increase at what we could call, but we do not know what exactly it is the normal physiological levels. If they are allowed to be used, as it is in the United States and Canada, there is a possibility they may be, more possibility that these increases may be found. And the point Dr. Adlercreutz was trying to make is it may have repercussions not only on trade issues, these questions, but may affect legitimate sports activities because suddenly people by simply eating meat may be discovered to have increased quantities of those residues which they will disqualify them from the games without them having done anything wrong. That is the point Dr. Adlercreutz was trying to make.
This is in fact one of the points. In the other paper, I am well aware that 19 nor steroid metabolites occur in certain animals and so it would be possible naturally and theoretically that these people have been eating, I do not think horse meat as horse meat is not eaten much in Europe, it may be pig meat. But in fact there has not been shown that eating this kind of meat you can get these metabolites in human urine. But here in this paper in fact the clostebol is not found in any animal and here they found one case with clostebol in the control group so there is very little chance that it came from any other source than the diet. But what I would also like to point out is that which I also pointed out in my report, is that we know very little about the effects of this steroid in the brain. And I have personal experience also with regard to the behaviour of athletes both before and after taking anabolic steroids. And we know now because this neuro steroids in the brain occur in very very low concentrations and you need very small things to change the levels of normal steroids in the brain. They are really at very low levels active. So if you think that you consume meat containing low levels of anabolic steroids every day, they will be transferred via the blood brain barrier and theoretically could interfere in the brain. And I have been involved in police cases and in severe depression cases after taking anabolic steroids and naturally the doses have been high. But now knowing the work by Etienne Emile Beaulieu in France which shows that these are very low amounts which are needed to change, for example, an impression of animal of another animal, can change their levels very little, but significantly. So it is possible that low levels of synthetic steroids can affect our brain long term and very little is known about that and I have not been able to find any studies showing how high the concentrations of, for example, anabolic steroids, are in the brain after intake. I do not have any such material, but I am a little bit reluctant because of this physiological observations made recently in recent years.
Mr. Christoforou, how long would you intend to go? We want to reserve some time for final questions by the Panel which are important to us.
WT/DS26/R/USA
Page 341
Mr. Chairman, I think it may take me probably until quarter to six. It will depend on the replies I will be getting from the questions of course, but about quarter to six or six o'clock, but I have a number of questions and I will cut a lot of them, but it will take me really some time.
I have a feeling that a break would do good to this meeting. We would resume, if you agree, in quarter of an hour, but try to cut it down to the essentials. Thank you.
I would like to give the floor to the European Community for continued intervention.
Mr. Chairman, yesterday we left the discussion at the point where we were discussing the issue why JECFA has not fixed an acceptable daily intake for the three natural hormones. And there was an argument going between, as I understood, Dr. Lucier and Dr. McLean, whether that was first of all feasible to be done and then if it was feasible, why it was not done in that case. This is already with the record of the Panel at this point, but I would like to recall and then come back with the precise question.
We have already explained in our second written submission to the Panel that the European Community does apply maximum residue limits and tests, as I said yesterday, 59,000 in serum for the three natural hormones. They are the values given here on page 16 of our second written submission. Exactly the same, the United States has already said in the letter that was sent to the European Community in 1987, and we have quoted these values in the footnote that we have submitted in the second written submission to the United States Panel.
For example, the United States has fixed for oestradiol 120 ppt, for progesterone 3 ppb and for testosterone 640 ppt. These are publicly known and have been communicated to the European Community as being the values that should not be exceeded: the tolerances, the thresholds. But then the United States has replied that although reliable research methods are available for distinguishing these levels from background, a regulatory method is not required because under conditions of use the actual increases in the endogenous steroids are far, far below permitted concentrations. So for that reason they do not include them in their annual control plan.
Mr. Chairman, linked to that issue is the explanations were given by Dr. Lucier on the fact that probably the already existing physiological levels are carcinogenic of the three natural hormones, in this case it was oestrogens. The Community has submitted evidence, which was not discredited as far as I know, showing that the increase of the risk was probably two-fold from those residues and not two to five-fold as it is stated for oestradiol in the JECFA Report of 1988. We have also submitted evidence about the increased risks for the pre-pubertal boys - the calculation was wrong.
The two issues then linked together and my indirect question to Dr. Lucier would be, according to his own estimation that it may be one in one million the risk of additional risk from the residues of those natural hormones which, for example, if we take the total population of the European Community may lead to about 300-350 potential cancers in the European Community. And on the other side we put, this is also in the record, that the maximum economic potential benefit for the farmers of the use of these hormones ranges between minimum 30 and maximum 80 dollars - that is what it is in the literature. How would you reevaluate the response that is it really from the risk assessment, the risk management point of view, is it really worth while taking into account the potential
WT/DS26/R/USA
Page 342
benefits of the use of that hormone in view of the risks that are already there, as you said for the existing levels, and if you would take our levels that it is even higher. Would you really, as a scientist who is involved in the issue of risk assessment and probably management of that risk, how would you like to express yourself on the regulatory authority and how you would take that into account when legislating?
Risk assessment is hard enough, now you are going to ask me about risk management too? I am really not in a position to evaluate benefits versus risk. What I know is how hormones and other agents cause cancer and how to translate that information into as good a science-based risk assessment as is possible. I really have essentially no expertise in terms of quantifying benefits in terms of what the benefit might be economically or what it might be socially, politically, whatever. I have no experience in that so I am reluctant to do that. I appreciate your point and I think it is not a trivial point, but I am really not in a position to give an informed answer to the question. My estimates are, and there is a lot of variability in these kinds of estimates, but generally an upperbound estimate of the risk of breast cancer, and I narrowed it to that particular area because I should think that is where the data is strongest, would be about one cancer in close to a million of people exposed to that by eating 500 grams of meat per day over their lifetimes. So I feel relatively comfortable with that number, recognizing all the uncertainties that are associated with any kind of extrapolated risk assessment and I wish I could give you an answer in terms of how that relates to potential benefits, I don't know. One in a million is not a very high number of course - it is a very low number, but for that one person who does get cancer it is a significant risk to them, so one can never underestimate that, but as much as I would like to answer your question about the benefits part I really do not have a basis on which to do it.
I cannot offer anything with regard to benefit, but I think there is an important point of clarification here. In the estimates that Dr. Lucier presented yesterday, I think what you have done is taken the one in a million that he talked about and multiplied it by the population in Europe and from that projected an anticipated increase of 350 cancers. I think you should note that one in a million is the upperbound. The number could be zero too. The one in a million represents, in Dr. Lucier's opinion, the highest probable risk, not the likely probably risk. But in risk assessment we talk about an upperbound, that is, the highest risk that we can contemplate from a mathematically derived model is somewhere in that range. But it would be incorrect and certainly misleading, and we see this all the time in risk assessment, to use a single number because the nature of mathematically derived risk assessment is that they provide a range of values rather than a single value. I think that was the point that Doctor Lucier was trying to make; that the risk probably falls somewhere between zero and one, and it would be incorrect to cite either zero or one because the methodology simply does not have the degree of certainty that allows you to cite a number, whichever one it is, with that degree of precision
it is a range.
I wanted to make two comments on what you have just said in your recent contribution. One thing is important, I think, for the Panel to know. If you design the statistical sampling plan, it is very important what is the population to be sampled, what is the area and I know at least the American delegation can correct me if I am wrong, but I spent many hours with their experts at the Food and Drug Administration. At that year, USDA, the basis is a national residue plan. It covers all of the United States and you cannot directly compare this figure with the figure of the European Community because the plans of the European Community have to comply with certain concepts, but these are plans for the individual member States. So the minimum number required to have a certain statistical certainty, if you take a figure from the United States, and want to compare it to the EC, you have
WT/DS26/R/USA
Page 343
to multiply with the number of member States; so you cannot directly compare these figures. At least for Germany I know that your figures are wrong, but I would also be willing to provide you with the EC document showing that your figure is wrong. I have no explanation, maybe you combine all residue testing, but for the hormones definitely your figure is wrong. I have the paper with me and I am willing either to give you the document or the reference number: they are wrong. So in the EC, I did not have the time to sum up all samples, but it seems to be in the range of 20,000 to 25,000. And as I said yesterday, half of the samples are taken in one country so there are remaining maybe 15,000 or 12,000 for the rest of the member States. And if you then prepare this with the figures of the United States, it is roughly an equivalent programme. That is my conclusion.
The second point I wanted to make: I agreed that in the IARC Report the range of this production, daily production, is between 0.4 and 2. It stands in that Report. But nobody has said that the figure used by JECFA comes from this Report. I was able to trace it back to a paper given by Doctor Farber and Doctor Arcos from the Food and Drug Administration in 1983 at the OIE meeting. I could not go further back, so I am not sure from where the data comes, but I can say they definitely do not come from the IARC Report. I tried to explain how such small differences could happen if you have a very big range and one group decides to use the mean and the other decides to use the median, then you have already this discrepancy. From my point of view, it is in the same order of magnitude and we should not spend a lot of time to say who is right or who is wrong. This figure definitely does not come from the IARC Report.
Mr. Chairman, I am afraid I will need to insist on this and I will take the time from other valuable questions but we did go back, Dr. Arnold, ourselves on this issue. As you rightly to say in the Report of the International Office for Epizootics, the article by Fabre and Arkas, they cite the article by Argus Engler(?) of 1974. And if you look at footnote 39 of the JECFA Report they cite exactly the same article. So both the Fabre and the OIE Report, both the JECFA Report, relies on the same reference which is 6 micrograms per 24 hours. Whereas the IARC Report of 1987 is based on the values of Brown which are more recent and which we have discussed yesterday - they are not 6 micrograms per 24 hours. We did that research so I can show you the documents if you wish. It is wrong the conclusions you draw.
I have said the figure doesn't come from IARC and you have confirmed. Which is the better figure?
Well, in our view, we use the figures the IARC which are more recent. These figures are also discussed in the issues of Brown and they are not approved. That is the conclusion we can draw. They were coinciding with the figures of which Dr. Lucier was aware as well. But I would also, if we had the time Mr. Chairman, like to go into the comparison of the number of samples because we definitely disagree on this issue with Dr. Arnold, but that will deprive me again of valuable time which I would not like to do. We formally disagree, we have the protocols here, for serum we have checked 59,000 only for the three natural hormones of the European Community and for the total number of hormones checked in the European Community the total amount is just about 200,000 samples.
Mr. Chairman, I would like to come back again on this issue of the risk that was discussed between Dr. Lucier and version in the place of Dr. Ritter. The reason for which I pose this question is because the delegate of Canada has made the short statement starting the discussion and he said "Is this little risk enough to justify the departure from the restrictions on international trade" and of course this is related both to risk assessment and risk management and the effects of that. We have probably
WT/DS26/R/USA
Page 344
different views what the magnitude of the risk is. But these are the values we are comparing here. This risk is involuntary for the consumers. We are, of course, aware that maybe zero to one thousand, one million, but it is an involuntary risk for millions of people. The question as a scientist, if you were, you probably don't want to continue the discussion, but I was simply trying to link this to the ADI which JECFA refused to set in this case, although Dr. Lucier would say it is feasible to fix one, we do it, the United States does it, but they do not apply it, whereas we do apply it, even if the three natural hormones of the Community are only allowed for therapeutic use. The natural hormones are not allowed to be used for growth promotion, still we check for the levels of those 59,000, as I said, in the European Community. This is an involuntary risk. Why should the consumers take such a risk, probably its low, but it is there a risk.
The second question of course and I will link this to what I said to Dr. Arnold that relates to therapeutic use. He has mentioned twice the evaluation made by the Institute, the agency in London of oestradiol which is already published as a regulation and of progesterone which is on its way. But Dr. Arnold did not mention probably all the conditions which are imposed for the use of these oestradiol as therapeutic instrument. We know that they have to be administered by a veterinary if it is therapeutic use, we know the animal has to be registered and identified and we know these animals do not go into the food chain because they are not slaughtered. We have provided evidence to the Panel that this percentage is not more than one, and we have more evidence which will provide to the Panel from other member States which have not replied by the end of January which indeed demonstrates that the rate is not over one per cent. These animals might have been treated, as was said by Dr. André, once for therapeutic or zootechnical treatment. But there is in a time-lapse between the treatment and when we decide for that one per cent to slaughter it and to send it to the food chain.
May I ask you just to submit succinct questions to the experts and not to plead. I think this is not the hour to plead. You have another go on pleading tomorrow, but just please do submit questions and make the best use of this very, very little time we still have available this day. The experts are flying home and then they are gone and we only have correspondence, and I have a whole list of questions which are sustained and I will come in with these questions at 6 o'clock and that is my decision.
The precise question is, therapeutic treatment, we know what it is - the word explains it - would you think because of the need to have the animals treated for therapeutic purposes, an extrapolation can be made of that and allow the use of these hormones for growth promotion. Is this a comparable example? Is it comparable to say that because they are used for therapeutic purposes they should also be allowed to use for. Is, in other words, the potential risk to consumers the same from one and the other situation if you take into account the conditions under which the therapeutic use is allowed to the Community? This is in particular to Dr. McLean and to Dr. André.
The only reply that I can make is that you have made much of the illegal use of hormones or the abuse of the use of hormones and so, therefore, one would suggest that veterinarians are going to be no better or no worse than the rest of the population. I would remind you that it is often the case where the veterinarian leavesthematerial behind for administration by the farmer for later treatment and so that is a possibility - I can say no more.
We are all sometimes in our life ill and we may need treatment by a doctor. Are you suggesting that even that type of treatment should not be made because I sometimes have to be treated for therapeutic
WT/DS26/R/USA
Page 345
or zootechnical reasons. Either way, what we call zootechnical for the animals, if a woman cannot get pregnant she would have to go to a doctor for that treatment to be made. It is comparable, it is just once, as we said, it is done. Do you really think this is a reason enough in itself to extrapolate and allow the use of the hormones as growth promoters? That is the issue. We all get ill, but from that what is the conclusion one should draw?
It is very difficult to answer that question if you are asking me to extrapolate I would suggest to you that given the potential for misuse and the small amount of meat that might be eaten by the average person in the EU from imported meat then the risk would be just about the same.
The second question relates to the argument we have been discussing about. We get increased values of all these natural hormones from a number of sources. One of our experts has called the sea of oestrogens which are existing in our daily life and I would like to give the floor to Dr. Liehr to make a couple of comments and then probably ask one precise question.
Mr. Chairman, yesterday we heard Dr. McLean talk about and I quote him "We live in a sea of hormones and that hormones are in our diet" and he also said that it is difficult to determine where our hormone burden comes from. Also today there were references early on in discussion about the various hormones that we are exposed to.
I would like to bring to your attention an important distinction here. Oestradiol, testosterone and progesterone are hormones that are circulating within all of us and carefully controlling endocrine functions within all of us. The production of these hormones within mammalian systems, including humans, are very carefully controlled, the disposition is very carefully controlled because these hormones very carefully controlled reproductive function and a variety of other functions that I cannot go and do not want to go into here. At elevated levels such hormones may be harmful - we are talking about breast cancer risk and also Dr. McLean this morning talked about extra testosterone shutting off sperm production and lowering male fertility. Many compounds, including phyto-oestrogens for instance, in soya products, have oestrogenic activity but there is a vast difference between hormones in mammalian systems and compounds with oestrogenic activity such as for instance in phyto-oestrogens. Differences are that they are different compounds with different activities - oestrogenic activities of many of these compounds are 1,000 to 10,000-fold lower than that of the mammalian hormones. There is really a vast difference between a compound that has oestrogenic activity and the actual hormone. Some of these compounds have also been shown to be oestrogenic and others are anti-oestrogenic. As Dr. Lucier said a little bit earlier on, this is a very active area of research and I know he is involved in this, nevertheless, it is obvious also that on balance between oestrogenic and anti-oestrogenic activity it may be beneficial and much of this is not known at this point. What is clear though is that a compound with much, much lower oestrogenic activity than a natural hormone introduced into a system with oestradiol in such a system, such a phyto-oestrogen may actually be anti-oestrogenic, or act in an anti-oestrogenic manner, because it competes for the oestrogen receptor. An indication of this is the low breast cancer risk in populations where high soya bean diets are consumed. So to imply that the administration of the mammalian hormone at low levels is like applying an antibiotic or a phyto-oestrogen is really inadmissible. These are different compounds, they have different spectrums of activities and they should be weighed on their own benefits.
WT/DS26/R/USA
Page 346
Is there not a difference between the hormones in mammals and phyto-oestrogens in diet, in physical, chemical and health effects? Is there a difference between these three?
I may be able to make this short. This answer is yes. There are differences.
Could you perhaps say why?
Well, I think, as has been quite correctly pointed out, there are differences in binding capacity, there are differences in structure, there are differences in function. Clearly a phyto-oestrogen was never invented to have a physiological function in a mammal. So I agree with everything you've said except that they do not contribute much to the overall risk. That part I am less certain about than you. That they are different in their structure, that they were clearly developed through evolution to have a fundamentally different purpose, I think it is an inescapable truth. But that they do not contribute to risk, I will not engage you in a debate now, about breast cancer statistics in countries where soya bean intake for example is high, except to say that it is not quite as clear as you present it. But I think the short answer, without labouring it in the limited time available, is of course there are differences in these different categories of oestrogens.
I will quickly add to that. I agree with what you have said as well. There are different binding proteins and different receptors that are now being discovered for different agents. There is a beta-receptor that has just been discovered and so forth and this is a different binding spectra than the receptor that we've all come to know and love. The interaction issue is a very complex one. Nevertheless there is a lot of information in the published literature on cell systems and in-vivo systems that certain mixtures of exogenous hormones do produce stimulatory responses that are characteristically oestrogenic. So in total these things are producing an oestrogenic response. Exactly the magnitude of the oestrogenic response is highly uncertain.
The next question relates to what the lawyers among ourselves call the issue of consistency and in that case it was discussed by referring to a number of substances. I will touch only the substance of carbadox. The discussion was lively but we did not explain the conditions under which carbadox is allowed to be administered in the European Community. Carbadox does not act as a direct growth promoter as the normal hormones we are thinking about. These growth-promoting effects are only because it combats the bacteria and it helps the intestinal flora and in doing that it helps the growth of the animal. The hormone does not exert any other action apart helping the growth promotion of the animal. And there are no other readily available substances on the market which can do the same thing. It is administered only to piglets and there is a 28-day obligatory withdrawal period. From that discussion and because the Community allows the use of carbadox under the conditions and even stricter conditions than those recommended by Codex. I was left with the impression that probably because the Community allows the use of carbadox, should it allow also the use of hormones, because at the end the two types of substances for the same risk to human health, that is carcinogenicity. What should be the paradigm? Because we prohibit the hormones should we also prohibit carbadox or because we allow carbadox we also should allow the hormones? This is an important question. And as I said,
WT/DS26/R/USA
Page 347
we do respect exactly the conditions of JECFA as they are recommended. There is no risk because we have been doing all the samples and the checking and there is no reported case of excess. So do you think, and I address this to Doctor Lucier because at the end the question of what are we exactly comparing was critical. Do you think this comparison is really meaningful and should it be taken into account in the global risk assessment of the hormones in this case?
I think the question should be limited to the scientific distinction of the agents and not the policy options which you invoked as well, if you agree?
I hadn't heard that part of the question anyway, Mr. Chairman. I understand what you are saying that carbadox is not used as a growth promoter, it is used as an anti-microbial agent, so its purpose, its use, is much different than for the hormones in question. On the other hand, it is, and I only saw the carcinogenicity data on it today and my next-door-neighbour has gone and left with it, so I cannot refer to it again, but it was a fairly potent carcinogen - it was a carcinogen at multi-sites, at doses well below the maximum tolerated dose and it is genotoxic, so it is not an agent that you want to expose a lot of people to. That is my only point. So I think strict attention needs to be given to the avoidance of residue levels of carbadox given the availability of that cancer data.
If I am correct, the agent works assisting the flora and it is not really having hormonal effect. Would you consider this to be a critical distinction between the two agents or not?
It depends, it is a distinction, they are used for differences purposes. I do not know how to say it beyond that. The purposes are vastly different. Nevertheless, they are both present and both may have their own risk. The risk of carbadox, if residues are present, are likely greater than that of some of the hormones in question.
I think it would be useful for the Panel to know that there is another drug related to carbadox
olaquindox, and whilst JECFA has reviewed it, it has not set MRLs and ADIs because the data package is incomplete. But notwithstanding I believe it is used in the EU and it is just as effective, it has a mode of action that is similar, in other words, interfering or modifying the microbial population and so therefore it is an alternative and it is just as effective and of course there are other anti-microbial agents that are used in a number of other countries as growth promotants that are substitutes for both olaquindox and carbadox, such as the tetracyclines as an example.
But can the agent be compared with the hormonal agents or is it ...
Are you asking the question that you are asking when you ask about the difference being important between carbadox, for example, are you asking the difference in risks to human health?
WT/DS26/R/USA
Page 348
The problem is that legally the agreement requires consistent policies. One argument goes that you allow this agent but you do not allow the other one and of course you can make this argument if these are comparable agents. It is important for us to know whether from the scientific point of view you can compare these agents and put them on the same level, knowing that the one affects the hormonal household and the other one the microbiological situation in the stomach. So can we compare the two and put them on the same level and ask for the same policy under the agreement or is that a completely different story. Carbadox and the other one is olaquindox.
The hour is late and I am going to go out on a limb and even though a number of us have consistently said during the day that we feel uncomfortable with drawing comparisons, if it would assist the Panel, carbadox is a genotoxic carcinogen. I think there are few people in this room who would disagree with that conclusion. There is very much debate, some of which you have heard in the last day or two, as to whether or not the hormones are indeed genotoxic carcinogens. To put this into perspective, within the toxicological community, we would attach greater concern to a compound which was a genotoxic carcinogen than we would to one that has been referred to in the last couple of days as epigenetic. Carbadox is such a genotoxic carcinogen. I personally would be inclined to say that the evidence for oestradiol, for example, being a genotoxic carcinogen, is weak, whereas in the case of carbadox it is definitive and if one had to put on a ranking scale where you would attach greater concern or greater importance, I think most of us in the scientific community would argue that we would probably have a heightened level of concern for genotoxic carcinogens than we would for some other class. I do not know if that assists.
Yes, it does. Would you like to speak on that point?
Dr. McLean mentioned olaquindox - it is also used in the EC. It is in the same annex to the feed additives directives. It is chemically similar to carbadox but there are interesting differences. Olaquindox is genotoxic in a very large number of tests, but in the carcinogenic studies, which were available to JECFA, it did not show carcinogenicity. There was no increase in either benign or malign tumours. That is a strange finding, but it is really genotoxic and I would say in more than ten test systems. The substances used in the EC as growth promoters could replace carbadox theoretically, unfortunately the residue situation is not clear. We have no marker residue, we are not able to set an MRL, so we have no means to control the use of this substance. But theoretically it could replace carbadox and it is in the same annex to the directive.
In relation to that point and the point that Dr. Ritter made, I do not get so concerned about the distinction between genotoxic and non-genotoxic. There are genotoxic non-carcinogens and there are non-genotoxic carcinogens. What I prefer to look at is the strength of the carcinogenic response - what sites are affected, what number of sites, whether it occurs in multiple species, or in multiple organs. So I looked at the carbadox carcinogenicity data I was impressed for the number of sites that were affected where tumours appeared, the magnitude of the response, they were very high over a background and they occurred in more than one dose; suggesting this might be a sensitive effect. On top of that is genotoxic. But my concern was not so much for the genotoxicity, it was the fact that the tumour response was a very strong one for carbadox.
WT/DS26/R/USA
Page 349
One point that is germane, in my country carbadox is banned and is banned on occupational health and safety grounds - in other words, we have more concern with those that have to mix it in with the feed and feed to the pigs, than we do to the consumers of the product and so it is an occupational health hazard.
I am not a specialist of carbadox but I am happy to hear this, because I understood this morning or this afternoon, that carbadox is a carcinogenic compound, but it is not a problem to use it because it is immediately and very fast transformed in an innocent metabolite. The real risk is for farmers and for manufacturers - it is not for meat or for pigs or meat containing carbadox, meat cannot contain carbadox residue, it is okay. Just to clarify.
It is unfortunate that this is going to be the final and probably your final decision on this issue. If it is indeed final question it is regrettable because we do have many questions. Then I would only comment on that Dr. Ritter said oestrogens are weak carcinogens - that is the statement I heard - weak genotoxic, could you please restate?
What I said was that the available evidence in the case of carbadox in the example I was drawing is that carbadox was genotoxic in a variety of test systems. The data is outlined in the 1989 JECFA Report. What I said was that the genotoxicity of oestrogens is still open to some speculation. That is what I said.
I thought we heard the word "weak". But if it is so, we have already, and it is published, that the IARC, the Institute on Cancer Research, has classified oestrogens in group 1, which is genotoxic and carcinogenic.
I am going to presume that you are directing that question to me. I think we are confusing apples and oranges. That oestrogens have the potential to be human carcinogens is absolutely true. I do not think there is anybody who would argue with that who is all familiar with the discipline. I think Dr. Lucier explained that yesterday. There is no question that we know that circulating oestrogens in a woman's body have a profound impact on her risk for breast cancer. This is of no question whatsoever. What we have been discussing here for the last couple of days is whether the levels that are present as a result of the use of these products in growth promotion are sufficient to constitute a meaningful risk, and we define meaningful as being somewhere in the range of one in a million. This has been the nature of the discussion for the last day or two. Not that at some dose oestrogens are capable of causing cancer. If I have imparted the impression that somehow or another I doubted that, let me correct it. To be clear, oestrogens are human carcinogens. I never meant to impart any other impression. But I have also indicated that in my view the oestrogens present at the levels that would be present as residues as a result of their use as growth promotion constitute, in my opinion, no significant risk at all to the human population.
WT/DS26/R/USA
Page 350
This now brings us to the last stage of this hearing and I am sorry we have to advance and I would really just like to get us through these questions as rapidly as possible. We have prepared a number of so-called additional questions which were circulated with the Panel and also with delegations. What I would intend to do is to touch upon them orally and then see in the end whether it might be feasible to maybe have them done in writing in order to be fully accurate. But I would like to have the opportunity of discussing them orally since you are here.
The first question really addresses a point we have not touched upon, which is the difference in health risks of the natural and artificial hormones. Is there a difference in the risk of using natural hormones or the artificial hormones? In other words, can they legally be classified in the same group or do you see there ground for distinguishing these agents?
Just to take the floor as it comes and then we have a good debate.
Let give me a stab at that. Not that we have an answer to all the questions, but for the natural hormones we have a great wealth of information about how they work in different cells and tissues of the body. This is in the published literature. We know a lot about, as you have heard, about how they go about producing cancer as well as other adverse health effects. We know less about the synthetic materials. Now the synthetic materials are designed to mimic the natural hormones. So they will have one set of their biological properties, will be very similar to those for the naturally-occurring hormones, that is the ability to interact with the receptor system and stimulate the same kind of, say oestrogen or testosterone responsive systems that the naturally-occurring ones do. But in addition, the complicating factor is that their structure is different. So they may potentially have biological properties in addition to the hormonal properties. And that is where any level of concern would come in. Now to get any of these synthetic materials registered for use, they have to go through the significant testing protocol to determine whether not they cause cancer or produce other kinds of effects. That information is not available from the regulatory agencies because it is proprietary. So much less of it is in the peer-reviewed published literatures. You don't have as much information in peer-reviewed published literature for the synthetics as you do for the naturally-occurring ones.
Thank you very much. Any other views? Addition? I take it if there is silence that you agree? Otherwise you would just oppose or clarify.
It is a difficult question. I largely agree, but if for example, I take the molecular structure and look at things, Dr. Liehr has already described and Dr. Cavalieri and I must say this potential is not associated to such extent with zeranol. So depending on what you are looking, you could come to the idea that synthetic hormones are safer than the natural and if you look at other things you could come to another conclusion and that is the problem.
So you would agree we know less, you know about them than the normal ...
WT/DS26/R/USA
Page 351
We know less about the synthetic but they might be safer than the natural.
A short point. I think the fact that Codex did not establish MRLs for the three naturals in comparison with the others is never mind that they are safer or not. The problem is that Codex establish no MRL for these natural hormones on the basis first, that they are natural, and a priori, that if they are natural, they are less toxic, I think, may be. And they are thought so. And the second reason is that the residue level has been demonstrated to be within the physiological range. And taking account these two facts, they decided not to fix MRL but not on safety basis.
Thank you very much. Yes.
Very quickly, the evaluation process, the formal evaluation process of veterinary drugs is designed to reach a point where you effectively reduce the risk to zero. Now in the case of the naturally-occurring hormones, it was believed that the additional amount that would come from ingested food would not contribute to the overall burden in a significant way and so that risk was zero. In the case of zeranol and trenbolone, then an acceptable daily intake was set, and again, providing that wasn't exceeded, the risk was essentially zero. And the second point I would like to add that the two synthetic hormones have been subjected to a formal toxicological evaluation process designed to bring out the toxicity. And so therefore, in some ways, those processes, which are complete for zeranol and for trenbolone, are identifying the toxic hazard and bringing it out. And whilst I accept the fact the studies are not published in the open literature. The monographs that JECFA prepares are open literature and there are a hundred or so pages of material relating not to each of them, to both of them, and that is available for perusal and scrutiny. They are also included in that material, it is from the open literature, although that is not large. And so, that is the couple of additional points I would like to make, Mr. Chairman.
Thank you very much. I think we don't have to address questions 2 and 3 because, we had an extensive discussion here on carbadox and olaquindox, just a minute ago. I like to turn to question 4. Are there any qualitative chemical or other differences between the three natural hormones in dispute in the human being, depending on whether these hormones are being endogenously produced by that human being; secondly are added to the human being because of consumption of food endogenously containing these hormones, or three, of meat treated with these hormones for growth promotion purposes or four, of meat treated with these hormones for therapeutic purposes? Do any of these chemicals, or other differences, if any, pose an additional risk to human health? And if Professor André could elaborate on his thesis on the natural hormones added to human beings having different metabolites.
Very quickly. When natural hormones are administered to animals first as esters, as implants or by injection, they don't follow in the body same route as the endogenous ones. Because the endogenous ones are secreted in one gland, circulate and are metabolized in another part of the body. So they cannot be exactly compared. There are small differences in the animal. So that these animals might produce different metabolites with same hormones injected as they could do with their own hormones. For example, they could produce various proportions in conjugate forms. As sulphosteroids
WT/DS26/R/USA
Page 352
instead of glucurono steroids, for example. After this, in the meat, you may have different varieties of metabolites. When human beings are eating these metabolites, they are eating different compounds from natural hormones as their own endogenous natural hormones. Following my example, it could be that human beings absorb some sulphosteroids and it has been said this after that a new information appear now on the role of sulphosteroids in the human brain. So that you cannot exactly assimilate thenatural oestradiol, progesterone and testosterone of human beings with hormones they are consuming in food. I hope to be clear enough.
Would you like to talk on this question?
Not to precipitate a lengthy debate, but to simply present an alternative, if you like. To the best of my knowledge and again that information was published in the European Conference on Growth Promotion, as well. To the best of my knowledge, if the structure of the hormone that is if we are talking about the same hormone which is produced endogenously or exogenously, if introduced exogenously, they enter the identical pathway. And from that moment in time they are indistinguishable to the body. And I say I can provide the Panel with published references which support that view.
But is it fair to say that there are two different views in science?
I think it is fair to say that there are probably 30. As many...
As authors?
You know the story of the two-handed scientists: on the one hand and on the other. Lawyers are often looking for one-handed scientists.
There is probably some small differences in the exact pattern of metabolites or break down products resulting from a slightly different route of exposure. This might be expected because of the way they move through the body. I don't think there would be qualitative differences in the metabolites. There might be small differences in the pattern of metabolites. So I think that's what Dr. André was saying. I don't believe that this difference in pattern of metabolites would be toxicologicallysignificant.
Thank you. Would you...
WT/DS26/R/USA
Page 353
I just want to concur with that view. There could be small qualitative and quantitative differences but there would be no effect on human health due to those small differences, within the biological variations, so to speak.
On the opposite, you cannot be sure that they have not.
That's not an appropriate scientific method!
Just all I am really saying is that I accept the fact that there are qualitative and quantitative differences. I believe that demonstrate any significant effect would not be possible and it is unlikely that it would be there.
I just want to say, you don't have to agree. I am not putting a deal together. The question 5 is basically just a question of clarification about the notion of normal physiological range. If you could comment on this question, whether it does relate to the range in human beings or in animals? Then the latter part of the question, the impact on the threshold approach really goes over into question 6, which then I would like to address to Dr. Lucier.
The point that I was trying to make is, and this has been said now, I think, many times and no one seems to disagree with it, that the amount of hormones naturally circulating in the body are in fact carcinogenic. There is a wealth of information in the scientific literature, both experimental as well as human studies, to document this. Now when an additional molecule comes into the body adding to that burden, it will not be distinguished from those other hundred thousand molecules you put here in the question. Instead of one hundred thousand molecules of 17 beta oestradiol circulating there will be 100,001 and the biological systems won't distinguish that one from the other hundred thousand. So in that sense the issue of threshold is irrelevant, since we are already dealing with the carcinogenic dose. We know that if a threshold exists, it has already been exceeded. So adding that additional molecule would add a small increment to risk. A very small increment of risk, because the body won't distinguish that molecule from any of the other naturally-occurring ones. And the risk assessment that I talked about is really based on these relative proportions of molecules, recognizing that the threshold if one exists, it has already been exceeded. And that was the basis of my argument that the threshold issue was irrelevant here. Because even if it did exist, it has already been exceeded, by the woman's naturally-occurring hormones.
What is of interest to the Panel is really what does this mean for the entire concept of the threshold under which the maximum residue levels and the ADI. From your point of view, if you take this to the end, would this make this approach redundant? This is contained in question 6 and I think it is a conceptually very interesting point. You know, is this a challenge to the thesis that if we can define these levels what we do below is safe and even what we do above it is probably safe for a long time.
WT/DS26/R/USA
Page 354
Any acceptable daily intake will not guarantee zero risk.
So you would challenge that below it we can talk about a safety ...
In other words any additional incremental increase beyond we already have, it would be reasonable to assume that an additional risk is created. So an acceptable daily intake could not be established that would guarantee zero risk for the oestrogens. Because we already know, again, that we are at the carcinogenic levels, so the assumption is that an additional molecule add up very, very small incremental increase in risk to that.
But I am correct to understand that the mainstream, the traditional understanding, is that if we operate within these limits, we are safe.
Well, what that concept was developed for was looking at molecules which we normally do not have present in our bodies. So for some types of carcinogenicity you can establish a dose below which you are reasonably confident that no effect can occur. And that's what a threshold is. In cases you have in addition to an already endogenously acting mechanism and we know we have it here, the issue of threshold is really an irrelevant one. One may set an acceptable daily intake as a policy. But that policy does not guarantee zero risk.
The difficulty with the hormones in meat, the naturally-occurring ones is that they are already there in meat. And so the very act of eating meat adds more hormones. And so therefore that is a risk. And perhaps one would therefore, if you want to eliminate that risk, you wouldn't eat meat. But the argument I think is whether the extra amount produced or taken in by what is already there normally plus the small increment due to the treatment, is a significant extra risk. And with normal chemicals when we set an ADI, they are not normally in produce. And so therefore, you can say well it's not there and you can have that small increase and they are not naturally occurring. So it's easy to set an ADI. But with the case of the hormones, they are naturally there already and it's difficult to measure what the ADI is, because the levels of increase are quite small in relation to what is already there in many cases. I mean if you look at progesterone in the normal cycling cow and the amount that you add is very, very small in relation to the total. So it's not a practical consideration and I think that's where we are having the difficulty. Not between the colleagues, but the concept of managing risk.
Thank you very much, Mr. Chairman. Perhaps just to be one hundred per cent sure ... It means that looking at this problem quite likely. If there are cabbage eaters or eggs eaters, who may perhaps add some more molecules than one in case of eating the hormone meat, it means that they may theoretically through a linear increase be exposed ... more to say dangerous or threats, then for example, those who eat hormone containing meat. There are different diets, for example, you know. Such as "cabbage eaters" or "eggs eaters", that's what I had mind.
WT/DS26/R/USA
Page 355
Or those that drink milk ... You are just adding to the oestrogen burden. And whether you take it in meat with that little bit extra added from growth promotion which is difficult to detect because it sits in a rising and falling normal level and if you take a sample of meat out there, from the butcher and bring it in, you cannot really say whether that meat has been treated or not. Because the difference that you add is buried in the normal range. Therefore, when you eat it, you add that little bit. Now in case of eggs or particularly, butter and cheese, for example, that have got a lot of fat in them, then you take those hormones in any way and they are quite high when you compare them on a weight basis with meat, particularly with treated meat.
Put it a different way. For every million women alive in the United States, Canada, Europe today, about a 110,000 of those women will get breast cancer. This is obviously a tremendous public health issue. Of those 110,000 women get breast cancer, maybe several thousand of them are related to the total intake of exogenous oestrogens from every source, including eggs, meat, phyto-oestrogens, fungal oestrogens, the whole body burden of exogenous oestrogens. And by my estimates one of those 110,000 would come from eating meat containing oestrogens as a growth promoter, if used as prescribed. Does that help?
Thank you. Could I follow up on this. In the light of this, do you consider the MRLs for the synthetic hormones adequate?
From my reading of the information that I have seen the MRLs are derived by looking at hormonal responses in monkeys. And there is a series of measurements the one goes about to determine those hormonal responses. I think in any case that we are talking about carcinogenicity data has not been used except the MRL. I think that has been basically a hormone derived response. I think in the case of trenbolone, there was weak carcinogenic response. And again correct me if I am wrong, and that was essentially disregarded in establishing the ADI. In the case of zeranol, it thinks it wasn't really considered at all, there was a weak carcinogenic response that to be due to the hormonal activity of the agent. After having said that, we heard from Dr. Metzler yesterday that there is a DNA adduct which is of zeralenone, which is a structural analogue of zeranol in the metabolic chain of it. Now that DNA adduct is part of the zeralenone bound covalently to DNA, and for some chemicals this is an early step in the carcinogenic response and an early step in a genotoxic response that initial binding covalently to DNA, so when the cell replicates that adduct is fixed into the [?] or cellular mutation. If the mutation occurs in a critical target gene, it could go on to produce cancer. And that's the whole concept behind the idea that genotoxic carcinogens need to be closely regulated. Now zeranol is negative in test for genetic activity that look at the ability to produce mutations. Yet this adduct has been discovered, apparently, and there is some questions about exactly what it is an adduct of. It hasn't shown to be clearly related to the zeralenone. So I don't know, with a 100 per cent certainty of these acceptable MRLs are correct. But I see no reason to strongly dispute them at this point in terms of the zeranol. The one with trenbolone causes me a little bit more concern. Trenbolone is positive in cell transformation as says, in other words, it is capable of transforming cells. I think there are couple of studies that have shown that. And that's indicative of, it is a way that people screen for potential carcinogens, when they see this response it raises a sort of a red flag to go on and look for the carcinogenic activity of a compound that does this. So that raises an issue with the carcinogenicity studies that I guess were conducted by the company who requested that this be used. Now as I understand, the carcinogenic studies were essentially negative. That there were small increases in liver tumours. Is that correct? I think there were small increases in liver tumours that were seen. I think
WT/DS26/R/USA
Page 356
the issue of trenbolone should be looked at perhaps a little more closely for the MRL, in my mind. I am more comfortable with the one for zeranol. It's a long-winded answer without being much of an answer, but it tries to give my basis for some concern on one part and little concern on the other.
Thank you very much. Can I ask a question which actually comes into the first question, but it is related here. Do the studies considered by JECFA evaluate carcinogenic effect of the hormones at issue? Was this taken into account by the studies which JECFA considered?
Yes, they were. JECFA actually looked at trenbolone acetate on two occasions because it requested additional information. I will say thatthe cell transformation assays under some circumstances were equivocal, they weren't accepted by all as being positive. Of course the carcinogenicity was investigated thoroughly and on balance I feel that this compound, this compound is being pretty much looked at as well as any compound and indeed for some of studies we ask for electron microscopy rather than normal microscopy to determine certain end points. And of course the no effect levels are based on fairly innocuous changes in monkeys and they are very low, and then on top of that there are further safety factors for species, inter- and intra-species variation. So you have got that chain built in.
If I could add just one thing, when we were talking about the increased incidence of breast cancer of one, I think, it was in 150,000. That's I think, according to your calculations, a maximum of one. [Yes.] So it could be anywhere between essentially zero and one and I think it's important to appreciate that difference. And that comes back to the point that my colleague, Dr. Ritter made that when we are looking at risk assessment, because it's a mathematical consideration, we have a range and it lies somewhere between essentially zero and one.
Thank you very much. Would you like to talk to this point.
That's correct, between zero and one and I should also remember to say that, when I say one, I was just trying to make the point to put in comparison with the other oestrogens. I have looked at some of that data, though with the cell transformation assays and I don't find it equivocal. It's associated with increased production of oncogenes in the Syrian hamster embryo cells, the ras and myconcogenes, which are an expression of those critical target genes in carcinogenesis. And the result seems to be real one.
I was making the point that some of the studies were equivocal and some are not.
I'd like to leave question 7 to the end and move on to question 8. We have the situation of a ban or prohibition with qualifications and if I am correct that Dr. André contended that the abuse will occur in similar degrees in situations where one bans the hormones in dispute as opposed to situations where allows these hormones under certain conditions. Is there any data available in this respect which could support your thesis?
WT/DS26/R/USA
Page 357
The result, as I said yesterday I think, in France we knew the two situations and the same hormones were allowed between 84 and 89. So I don't know really if the data are available. They are in my Government and official bodies and maybe it is possible to get real data in archive from that date. But I can just give some witnesses on this. For example, during the authorization of these hormones, five years of authorization, we found some positive results, positive results is misuse, of the DES. And the DES misuse had disappeared in the early seventies. So it was a coming back of DES misuse during this period. A second fact is that it has been said. I speak for France, not for other countries, okay?, that in France, the ban has been followed by the misuse of new class of hormones, so-called beta agonist. And it's clear that we discovered the misuse of beta agonist just after the official ban. But in fact when now, with the methods we have, when we look in feed samples taken during the period where the hormones were authorized, we find also clenbuterol or beta agonists. So my idea is that the authorization, legalization, of these compounds had no effect, real effect on some black market and the misuse of hormones.
The second part of your question is concerned with the control. And to know if the control is easier under regime, with a ban or not, I would just answer that it's clearly easier technically for all the xenobiotics. Because if you have not any xenobiotics allowed each time you find one it's clearly a misuse. And it's also technically more easy because you have just to have qualitative survey, I mean, you have to check in your sample meat, urine, feeds, whatsoever, you have just to look if the compound is present. And then you say, yes it's present. So you have a qualitative analysis. When hormones are allowed, then it's the same thing for drugs, and so you are obliged to do a quantitative analysis, more complicated analysis to compare with the tolerance, with the MRL and so technically it's more easy. But it's small differences, any laboratory know how to do the two jobs.
Thank you very much. My next question related to this to all the Panel members. Is there a fundamental difference ... [I am sorry.]
Mr. Chairman, I would like to add few sentences. Sometimes a ban makes life of analytical chemists easier. But not always and there is no easy answer to that question. I give you two examples. Oestradiol 17-B, is used as benzoate illegally and according to legal authorizations for therapeutic purposes. On a routine basis, the only way to detect the substance, currently, is to detect an injection site. Although I know that many people are starting being able to detect the ester. But even if you detect the ester somewhere, not at the injection site, then it's still the question, was it illegal or has it been legally used? So it's not so easy. And most results have been obtained if people had found the injection site and if they were able then to go back to the farm and look. If somebody misuses these substances, then he never uses it on an individual animal basis. Then you trace back and you see, there are many animals treated the same way and then you are sure, it was illegal. So, it's a lot of effort and costs a lot of money to control this.
Can I follow up. As a layman, I assume that if the practice is illegal and you want to apply the drug, you can't use the earmark, because it is too obvious. So what you do is, you make massive injections in the muscle. And if we take the evidence that there has been about, was it 20,000 cattle controlled, this is a small percentage of all cattle controlled. And if you happen to be the consumer of getting that muscle, isn't there a risk that you have enormous intake? I mean, is the risk greater under a ban than when it's controlled? This is the main argument and I would like to just add this one and I have two more elements which should go with it. We should also consider the control of
WT/DS26/R/USA
Page 358
imports under the two things. And the final point is, is there a fundamental difference between controlling hormones than from other veterinary drugs and substances used. Does it allow a fundamentally different approach than what we have in other areas?
Mr. Chairman, the use for therapeutic uses is by injection. This is the only way which is permitted. So the illegal people, they do it by injection and try to hide the injection site. So you have injection in both the cases. This makes the thing so difficult.
But you can trace it, because it is used by the veterinarian on the therapeutic side.
You can ask the farmer to show his prescription, for example. So you have to do a lot of effort. Simply by analytical means it's very difficult to answer the question whether it has been used legally or illegally.
Yes, in short, concerning the control of hormones and other drugs, I think the problem is the same. Because we have also to check for banned compounds as chloramphenicol and the problem is the same to control for chloramphenicol as for trenbolone or nandrolone. It's not really different, I think. The control of imported meat, the problem is that it is imported meat. It is not a live animal and it's clear also, everybody, every scientist, analytical man can say that it's more easy to work on feed, to work on urine, as to work on meat. So it's more difficult. It's not impossible. But it's more difficult. And concerning your third short question, the occasional consumption of the meat with an injectionsite. This is highly improbable. Because, for many reasons: first of all, these injectedsolutions are mainly oil solutions of benzoate, oestradiol and other compounds. So the butcher is looking for what it is. He is cutting and it is very improbable that an injection site can be eaten. Even if it is, let me remind you that it is an occasional consumption. It will never be, you have no opportunity to eat each day an injection site. It would be an exceptional thing and maybe you will be informed by the news of that very strange thing.
The next point is relating to the labelling. You had a written question on labelling, the answers were fairly short in most cases. And we would like to back up on this one. There are two approaches. You could have voluntary labelling basically of meat which has not been treated, what we call green labels, and you can have mandatory labelling of treated meat, which has to be enforced. Reading this question, do you think labelling would be a feasible approach? If not, in what ways would this labelling procedure differ from the controls already carried out by the EC to ensure that imported meat has not been treated with hormones at all? Is the fact that one cannot distinguish between treated and untreated meat sufficient reason not to label meat? What is the difference between labelling already carried out today, even with respect to meat, such as this meat is BSE-free or American meat, or French meat, and labelling for purposes as to whether meat is or is not treated with hormones? I mean the labelling is a major policy today in making or leaving decision to consumers. Most reactions were negative except yours, Dr. Lucier. And we need to know little bit more why you think this is not a doable way. Now, but well that's an answer yes. But maybe, you have scientific reasons too which it is not doable.
WT/DS26/R/USA
Page 359
I mean, anything is possible, Mr. Chairman. It's just the cost. But if you look at the realities of labelling meat, especially when the live animal is a commodity, it's bought and sold and traded. Then what you mean by untreated, if you mean never ever treated in the lifetime, then you go to look at the calf, you got to look at as it changes hands, once or twice it moves into a feedlot. So you have got a chain of identification, you really need a passport with each animal, to which each owner attests that has never treated and their passport goes on with ...
It could be another earmark.
It's a practical problem. The other difficulty is with the naturally-occurring substances, how do you really determine whether it has been treated or not, except for the presence of the pellet in the ear. Because analytically you can't tell whether it has been treated. And there is another problem with zeranol. In some countries, Fusarium which produces the zeranol-like components contaminates pasture and feed. And there are countries in the EC, I believe, where when it comes to taking the farmer to court for example, it is impossible really to prove beyond all reasonable doubts that particular animal didn't take in a fair burden of the zeranol-like substances that have left some residue. So there are practical problems across the board. And I made the point, and I know that it's unrelated but one has to look at the cost of this against the benefit. And the benefit is largely psychological, but the cost is enormous. And when you look for example, there's the example that I gave with the outbreaks of E. coli food poisoning that is occurring all around the world, where there are hundreds of people getting ill and a number of them dying, one wonders whether one is better to spend the dollar on rectifying this problem than putting it into something that can confers a psychological advantage to the person eating the meat. A simple practical problem.
Thank you.
My written comment was probably a naive one in many respects. But I think one ought nevertheless to consider whether the primary source can in fact be labelled. Obviously there is a lot of spin-off meat products that will be developed and it will be very difficult to trace those and label each of those. But I think it needs to be looked at as to whether the primary source can be labelled such as this animal has been treated with a growth promoting agent for the purpose of growth promotion. I think it could be narrowed down to that and perhaps be not quite so costly as if full-blown labelling procedures were put into effect. I think if that's all possible and it is not going to break the bank, I think it really needs to be considered. I mean, after all if the public is concerned about these issues, the public has a right to know, if at all possible.
Thank you very much.
In trying to prepare a written response to this, Mr. Chairman, in fact I cited in a number of references who understand this topic much better than I. But I suppose the question that I ask myself
WT/DS26/R/USA
Page 360
is what would be the objective of the labelling. Because I think that helps to define the strategy. We're accustomed around the world to labelling products where an informed decision is a useful thing to offer. If you want to smoke, these are the potential consequences. Or there are many therapeutic drugs, not necessarily veterinary medicines, but human medicine, which may raise the risk of birth defects, and we label them, because they may confer enormous benefits to patients. But I was unclear when the question was raised. To turn it around, if it were the view of the Panel that the use of these agents does not constitute a risk that has to be dealt with, then I ask myself what's the objective.
Well, I am recalling a submission which was in the EC's submission to high US Government officials explaining the hormone treatment and when he was asked privately what type of meat do you prefer, he said, if I can afford it I buy non-hormone treated beef. And I think that's the sort of choice people would want to make. I think we make these choices every time, particularly now at the age of BSE. You find on the card in a restaurant, US beef, beef imported from the United States and that's another label, and that's a label in the meat sector, and so that label somehow works, I hope it does.
Well, then, I would be inclined to think, if that's the case to provide those who wish to make a choice, the opportunity to make one. And if it's not a matter of helping a consumer decide on the basis of ....[cassette ended.]
I think of some of the supermarkets particularly in the United States, South West in California where there has been a whole culture of franchises that's developed around assuring the consumer that it's pesticide free or this residue free, and my view on that is all power to them, if you can create a market niche for yourself by offering a consumer something which they want but which may not provide any direct benefit. But I understood the question to asked in the context of a regulatory imposition, and in that context I asked myself what's the objective. Because if it doesn't help the consumer make an "informed" decision, and I use the word informed to imply not the knowledge that has been treated, but rather information which can materially affect the decision, then I have to ask myself if one can justify the cost and everything else that goes with it when one could simply allow the market-place to accommodate it. I don't have that view if it's a question of risk, if we're talking about a drug that may induce a birth defect and a physician has to make a decision on behalf of his patient, than I think labelling is very appropriate. Very appropriate for tobacco, although I have other views about tobacco. I think there is a more appropriate way to deal with tobacco than labelling. So I'm not in any way trying to suggest that consumers should not be given the choice. But if the choice as I say is based on preference and not science, then my position would be that I let the market prevail. And we know from worldwide experience, I cite the US as an example but there are many others, that indeed there will be entrepreneurs who will grow up to fill that niche. And I think that if they can do that and take advantage of that sort of a situation, it is the hallmark of free enterprise.
Thank you very much. I'm trying to wrap up, there is one question I still have to address, a technical one. The experts have not addressed (this is No. 10). The experts have not addressed the hormone MGA. Has any risk assessment been carried out with respect to this hormone? Does this hormone pose different health risk than the other two synthetic hormones? Do you have any evidence on this? It's question 10 here on the paper, the last question.
WT/DS26/R/USA
Page 361
I guess all I can say Mr. Chairman is that I have not had the opportunity to look at a package of data, contemporary data, that is complete, and so therefore I'm unable to comment and I made that comment at the beginning.
Just a comment, one major risk for these hormones is the risk associated with management of an hormone, an active hormone imported directly in the feed on the farm. That's a real risk I think. In this hormone is very different from the others, but no more comment on this.
Well I'm not saddled with the same amount of information my colleagues are so it's difficult for me to make a comment because of lack of information. There is some, there is a report that it produces an equivocal increase in mammary tumours, a long-term study and that it's a very potent progestational agent. The information that I could get is that it's about 30 times more potent than progesterone and orally active so it's an extraordinarily potent progestant. So that raises some concerns toxicology just for that reason. But I haven't seen the toxicology studies that apparently are available but I would have some concern based on these two pieces of information.
Would it be possible to get this information for the Panel or is this too difficult?
No, but the issue Mr. Chairman, is that in the case of five of the six, I think the view that you've heard from all of us participating, not necessarily agreeing, but that there has been a substantial amount of information on which we have based our opinion. In a number of cases we have arrived at different opinions, but we have all agreed that there has been a substantial amount of information. In the case of MGA because it has not been subjected to an international review thus far, a JECFA review, and because information submitted in support of these sorts of petitions to national regulatory authorities is generally proprietary, I think none of us would have had the benefit of the depth of the review that would be available for the other five. So I think that the cautionary note that you are hearing, at least from Dr. Arnold, Professor McLean and myself, is that the most accurate thing we could say is that we simply don't know. I mean that not one way or the other but that I on my own, would not venture a guess one way or the other because I have not seen one per cent of the information on MGA that I've been able to examine for the others.
Thank you very much. I'd like to come to the final tour de table on the side of the experts and I'd like to link this to question seven here, I think that this will be at heart of the, also of the legal issues to be discussed in this case. Dr. Liehr has provided to us what is called the new evidence on the potential genotoxicity of the hormones. Other experts have questioned the relevance of this evidence. I would like to have your personal views on what is the relevance of this new evidence for veterinary drugs other than the hormones at issue, such as carbadox, and for the use of natural hormones, especially oestrogens, for therapeutical purposes. If this new evidence would be a reason to ban hormones added for growth promotion, would it also be a reason to ban all other uses of these hormones or to ban other veterinary drugs? And perhaps before you address this, we are not very clear on the Panel on your fundamental views to what extent one should, a scientist should or should not consider when he consider the risk for hormone residues, to what extent he should really take into account the genotoxic effects.
WT/DS26/R/USA
Page 362
There seems to be different, slightly different views. But if you could state again your position on the genotoxic effects and then perhaps follow this up. How would you advise the Panel to take into account this new evidence which has been presented particularly yesterday to us. Maybe we'll start with Dr. Ritter and then go through the Panel.
Just very quickly, I won't take much time Mr. Chairman. I think, I'm reminded that Dr. Liehr may have captured the reference point on this information that he presented better perhaps than anyone else. When he suggested himself yesterday, I think quite correctly, that in order to establish the direct relevance of his preliminary findings, he was seeking additional research funding. I think he said that the parties to the dispute would have made a better investment, if they would have funded him to examine the relevance of these findings at the low levels which are being present, than they have expended in coming to this hearing. I am paraphrasing a little bit, but I think that was the gist of his comments. I think he has put it better than anyone else. I think the findings are interesting but he himself I think has suggested that they are not in the form or of the nature where they can be directly applicable to the nature of the deliberations here. The nature of scientific investigation is that it goes forward. And if we were to meet again in a year, I dare say you would be looking at evidence which you didn't have available today. I mentioned to you yesterday Mr. Chairman, that I know of no scientist who has ever said on any issue, because we are all looking for funding, I know enough, please don't provide any more money on this issue. I know I've never done it. So I think just very quickly in the context of the relevance, that would be my view, it's interesting, but I like the way Dr. Liehr phrased it.
In terms of the relevance in a larger sphere to other veterinary drugs, again I think the only way to answer that question is to apply the same technology to other drugs. It is conceivable that if one applied the new methodology that we heard described yesterday to a range of other veterinary drugs, pesticides, other food contaminants, that it might raise the very same degree of concern that it seems to have generated in this debate over the last couple of days. I don't know the answer because it hasn't been done. But if you'd like to provide the funding Mr. Chairman ... . And finally I think on the issue of carbadox specifically, if I read this to be the third issue if you like, I think in the case of carbadox there is already general agreement that it is a genotoxic agent, so to subject that specific example to further investigation, I think would contribute little to our knowledge because we have already concluded that it is capable of this type of action. Subjected to further studies would only confirm what we have already all agreed on. So I'm not looking for more funding on that one.
Thank you very much. Professor McLean.
Yes Mr. Chairman, the way that I read it was that Dr. Liehr was putting forward an interesting, along with a number of other people, were looking at the mechanism of carcinogenesis if I could put it that way and therefore genotoxicity was important. That he had a hypothesis, he had made a series of observations and he was testing that hypothesis and that may or may not be significant, when it's finally worked its way through, to this particular group of hormones. Dr. Arnold did point out that there were other hypothesis being tested with similar sorts, at this stage, of strength, if you like. And so therefore I think in relation to the new evidence, the jury is really still out and until the hypothesis firms up, if I could put it that way, then there is little we can do. And it would be unfortunate if the hypothesis didn't stand up and other hypothesis came forward and were tested and shown to be effective if they do that we'd made a mistake there. In relation to other veterinary drugs, many of them are carcinogenic in animals. Just to name a few, dimetridazole, metronidazole which is widely used in humans is quite carcinogenic in animals, yet it is still used, the sulphonamides are quite carcinogenic in rodent carcinogenicity tests, carbadox, we know is carcinogenic and that's to name four. And to
WT/DS26/R/USA
Page 363
go that next step, we don't have the mechanism of toxicity in all of the cases. And we could develop a hypothesis and test it and then we might arrive at the precise mechanism. So I think we have a similar situation with other veterinary drugs although, with the hormones, they do have a very serious human health effect in producing for example breast cancer and so we're putting a lot of money in the area to work that through. Because it is an important human health activity and I would support continuing research with the other drugs, they don't assume that importance and so we don't put the money in.
Thank you very much. Dr. Arnold
I don't want to add very much because I largely agree with the two previous speakers. But should it happen that, for example, oestradiol would be identified as genotoxic carcinogenic and should this then cause a ban of this substances for growth promotion, I would say from the viewpoint of a consumer, I could then not understand its use for zootechnical purposes For example, then one should do research to replace this and use other substances which could be available for the same purpose. And also I would say then the substance should be restricted in veterinary therapy to a more narrow definition of therapeutic use. We have a very broad, as a consumer, I cannot agree with this broad definition of therapeutic uses. And it would have severe consequences in human medicine because then it would be very difficult to justify the use of this substance in oestrogen replacement therapy, for example, for the prevention or treatment of osteoporosis where millions of women take one microgram, in the order of one microgram of micronized and bio-available amounts of oestradiol every day. So it would have severe consequences in many areas.
Professor André.
Yes, I think that Dr. Liehr gives us new results in the mode of action of these oestrogens in their carcinogenic effects, but he is not alone to give these results. We have many different ways to understand how they act. The difference here is it a complete theory which seems very understandable and the relation between oestrogens and DNA seems to be more identified that it was earlier, for this it's clear. But I think that if we admit that these hormones are genotoxic, it's two very different things to ban them as growth promoters, as I said yesterday, in a large scale for all populations and it is very different for their use as to therapeutic use. It's the same difference as I tried to explain with my example. As therapeutic use, they are used in one animal for one time and usually for reproduction purposes, so it's a completely different thing. So they can be banned for growth promotion and they can be used for drugs. But I agree with Dieter Arnold on the fact that there is some problems in the definition of what is therapeutic use and also how to evaluate really a drug for therapeutic use, as is done classically for years now by different committees, and the use of compounds for other purposes than really therapeutic use. And in this case, to my opinion, the evaluation has to be maybe different.
Thank you very much. Dr. Lucier.
My comments will be a bit different. I think the work that Dr. Liehr is doing, I think it's important to point out, as he has, is that it's not just himself. There is a group of scientists who are
WT/DS26/R/USA
Page 364
looking at the role of oxidative damage and genotoxicity of oestrogens, himself and several other folks throughout the world. And I think this activity is very important to our understanding of how oestrogens and how hormones may cause cancer. So I'm a strong supporter that kind of work, I actually have a couple of publications in it myself. But for this narrow purpose that we are talking about today, about the influence of this on additional risk from oestrogens from eating, consuming meat containing oestrogens from growth promoted animals, it doesn't have too much consequence. If I did a calculation of risk as I've done, this one in a million thing how that has been much discussed, I would come up with the same calculation whether or not oestradiol was genotoxic. That would bear no influence on the carcinogenic risk as the assumption is we are already on a linear part of the curve in terms of amount of oestrogen, the magnitude of oestrogen and the magnitude of response, say for breast cancer. I would come up with the same answer either case; there would be no difference in the risk. So I think in that respect whether or not oestrogen is genotoxic, has less consequence then what we talked about up to this point in time.
Well thank you very much for your final statements. This brings us to the end of this expert hearing, I come to procedures, which brings us to the end of this two-day meeting. In my view, the combination both of your excellent papers, and I know you took pains to write them in a way that the normal person can read it, together with your oral presentations and the questions and comments on the parties, I think have exceedingly helped to inform the Panel on the scientific side. A side which is very difficult and very complicated. It is almost as complicated as the procedural one, but not as much. So thank you very much for your tremendous effort and also the effort you put in such a short time to assist the Panel which is operating under these very strict rules. Thank you also to the delegations for assisting in these proceedings. Thank you in particular also for the scientists who sat with the EC delegation and came here to present their views and their convictions and inform us as well.
So this brings us to the final procedural point. With a view to the meeting tomorrow, which is the second substantive meeting, the Panel has reflected and in the light of the fact that this has been a joint meeting today and we had difficulties to make decisions about what is new evidence, what is a new argument, and so on, and also in light of the fact that all the papers have been exchanged in the two proceedings, the Panel has taken a decision to come back to its right which it reserved before and we decided to invite the United States delegation to attend the meeting tomorrow. Do you have any other points on procedure?
Mr. Chairman, you anticipated correctly the question I was going to ask, thank you
Professor André
Just on behalf of my colleague experts, I would like to thank you for the excellent level of this debate, due mainly to an excellent chairmanship.
Thank you very much. So the meeting is closed, good night. I'm sorry, ... Mr. Christoforou, you wanted the floor.
WT/DS26/R/USA
Page 365
Yes Mr. Chairman. I would also like to thank the experts, and it is not usual procedure but we would appreciate if it is possible to get the typed version of what was discussed here today. I know it is not usual, but I think it would be very useful if we can get the discussions today typed and circulated to the parties.
Yes, this is underway, but it will of course take some time. But this will be done yes.
I too would like to thank again the experts for participating, I was wondering what time we are reconvening tomorrow and is it in this room.
Thank you very much, the meeting will be a 10:00 in this room here. So the meeting is closed.