Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
CCase Studies of the EPA's Application of the Delaney Clause in the Tolerance-Sett~ng Process RICHARD WILES Case studies of nine pesticide active ingredientsfosetyl Al, benomyl, captan, chlorobenzilate, dicamba, the ethylenebisdithiocarbamates (EBDCs), metalaxyl, permethrin, and thiodicar~are presented. Descrip- tions of each chemical are included along with regulatory status, special review criteria that have been triggered, oncogenic findings and risk estima- tion, alternative pesticides, and a discussion of issues relevant to tolerances and the Delaney Clause (section 409~. FOSETYL AL Description of the Chemical Common name: Aluminum tris (ethyl phosphonate) or fosetyl Al Trade name: Aliette Pesticide type: Systemic fungicide Chemical family: Organophosphate Year registered: 1983 Major producer: Rhone-Poulenc, under patent Volume of use: Small Tolerances: Fosetyl Al has section 408 tolerances of 0.1 parts per million (ppm) for pineapples, pineapple fodder, and pineapple forage. A recent petition for sections 408 and 409 tolerances on hops was denied. An application for a section 408 tolerance on citrus is pending. 196
EPA'S APPLICATION OF THE DELANEY CLAUSE 197 Regulatory Status General-use pesticide Special Review Criteria Triggered Evidence of weak oncogenicity was dealt with in the registration process. Summary of Oncogenic Findings and Risk Estimation The incidence of kidney tumors was statistically significant in the high dose of one chronic feeding study. However, the high-dose feeding level was approximately 35,000 ppm, a rate equivalent to about 4 percent of the total diet of the test animal. Tumors termed by the EPA as of "question- able significance" also appeared at the mid-dose level of 8,000 ppm, about 1 percent of the diet of the test animal. Alone, the mid-dose findings would not support a finding of oncogenicity, but because of the high-dose tumors, the pesticide was classified as oncogenic. According to informal discussions with EPA staff, the data base for fosetyl A1 is complete and of high caliber. In fact, the data are of such high quality that they may actually be working against further registration of the chemical. Fosetyl A1 has an extremely low acute toxicity. Thus, in complying strictly with EPA guidelines instructing registrants to study the effect of chronic feeding at the maximum tolerated dose (MTD), the registrant fed extremely high doses of the chemical to test animals. The only oncogenic effect observed was that mentioned above. The oncogenic risk from dietary exposure to fosetyl A1 is calculated by the EPA at about 1 x 10-8, or 1 in 100 million. Tolerance and Delaney Clause Issues A recent request for a section 3 registration on hops was denied because it was determined that the use of fosetyl A1 on hops required a section 409 tolerance. This tolerance could not be granted under current law (the Delaney Clause) because residue studies showed that during drying, fosetyl A1 concentrates to levels above the proposed 408 toler- ances for hops. The proposed tolerance level for fosetvl A1 was 10 oom on green hops and 15 ppm on dried hops. ,7 ~ The registrant anticipated that fosetyl A1 would make significant in- roads into the U.S. fungicide market and quickly become a major product. Fosetyl A1 is designed to control downy mildew in vines, as well as numerous fungi in fruits and vegetables. Its use in these areas has
198 APPENDIX C expanded since its introduction in Europe in 1978. It is possible that fosetyl Al could eventually command a significant share of the U.S. market currently occupied by more toxic or less-studied fungicides such as the ethylenebisdithiocarbamates (EBDCs), captan, or benomyl. For example, currently the most widely used fungicides on hops are the EBDCs. The estimated dietary oncogenic risk from residues of EBDCs and its metabolite ethylenethiourea (ETU) in and on hops is 1 x 10-4 to 1 X 10-5. The risk from the same use of fosetyl Al is 1 x 10-8 or less. Were fosetyl Al to have acquired any share of the EBDC market it may have lowered exposure to the more oncogenic EBDC and ETU residues. BENOMYL Description of the Chemical Common name: Benomyl Trade name: Benlate Pesticide type: Systemic fungicide Chemical family: Benzimidazole Year registered: 1972 Major producer: Du Pont, under patent Volume of use: Benomyl accounts for 55 percent of the $320 million worldwide benzimidazole fungicide market. In 1984, U.S. sales amounted to approximately $60 million. In 1979 about 3 million pounds were used in the United States on 43 food crops and 41 ornamen- tals. Tolerances: Benomyl has numerous section 408 tolerances and several section 409 tolerances. New sections 408 and 409 feed additive tolerances were issued for Benomyl in wheat, barley, and other small grains on November 7, 1984. Regulatory Status A notice of Rebuttable Presumption Against Registration (RPAR, now known as a special review) of Benomyl was initiated in 1977 because Benomyl exceeded the risk criteria cited below; oncogenicity was not an initial risk consideration. The Position Document (PD) 1 was published December 6, 1977. Findings of oncogenicity were made subsequent to the EPA's proposed decision in the PD 2/3. A PD 4, or Notice of Determination, was published on October 12, 1982. The notice allows continued registration of all uses with protective
EPA'S APPLICATION OF THE DELANEY CLAUSE 199 clothing requirements and registrant submission of field studies to identify residues that may enter aquatic sites after use on rice. A registration standard for benomyl was completed in 1986. Special Review Criteria Triggered Reduction in nontarget species Mutagenicity Teratogenicity Reproductive effects Hazard to wildlife Summary of Oncogenic Findings and Risk Estimation In tests with benomyl, hepatocellular carcinomas or combined hepato- cellular neoplasms in both male and female mice were observed at all doses (the low dose was 500 ppm). Similar tests with a metabolite of benomyl, methyl-2-benzimidazole carbamate (MBC), revealed combined hepatocellular neoplasms in male mice and hepatocellular adenomas, carcinomas, and combined hepatocellular neoplasms in female mice. These data were received subsequent to the EPA's proposed regulatory decision (PD 2/3) but prior to the final Notice of Determination (PD 4~. Based on findings of oncogencity for benomyl and MBC, an oncogenic potency factor (Q*, or extra incidence of tumors/unit dose) of 2.065 x 10-3 was determined. Multiplying projected human exposure by Q* will estimate the 95 percent upper bound on cancer risks to humans from lifetime exposures. Using the multistage model and residues at the tolerance level, the upper limit of oncogenic risk to the general public via dietary exposure to benomyl was estimated as 6.8 x 10-5. On the basis of residue analyses, the lifetime oncogenic risk from dietary exposure to benomyl at average expected residue levels was calculated at 7.2 x 10-6. Tolerance and Delaney Clause Issues The case of benomyl illustrates the relationship between the reregistration process, tolerance reassessments, and the Delaney Clause. Benomyl currently has section 409 feed additive tolerances on apple pomace, grape pomace, citrus pulp, rice hulls, and tomato products. Benomyl was granted these tolerances prior to knowledge of its oncogencity. There is also evidence (contained in data submitted by Du Pont to the EPA) that benomyl concentrates in orange juice, dried apricots, plums, and grape juice. On the basis of current studies indicating
200 APPENDIX C oncogencity, section 409 tolerance applications for these uses would probably be denied under the Delaney Clause.2 Benomyl is the first pesticide registered before 1972 for which the EPA will have residue data sufficient to support tolerance actions pursuant to the Delaney Clause at the time of a major regulatory action (registration standard). If the Delaney Clause is strictly applied, benomyl could lose section 409 and possibly section 408 tolerances for apples, grapes, citrus, rice, and tomatoes. These uses account for around 1.1 million pounds of benomyl applications, or approximately one-third of all benomyl sales. Reduction in estimated dietary oncogenic risk from revocation of these tolerances would largely be a function of the oncogenic risk associated with benomyl's replacements. (See Chapter 5 for further discussion of this issue.) Pest Resistance A distinct feature of benomyl is that it acts systemically. Because of this, benomyl has many more uses and does not have to be applied as often, in as high rates, or prophylactically, as do nonsystemic fungicides such as captan and the EBDCs. However, this characteristic has led to the development of resistance in target fungi. According to Dr. George Georghiou of the University of California at Riverside, of the 70 species of fungi reported as resistant to fungicides by 1979, 69 species (84 percent) were resistant to one material benomyl. Du Pont has recommended lower doses per application and more precisely timed use of benomyl in order to control the exacerbation of this problem. To retain the advantages of benomyl use and to retard the spread of resistance, growers often curtail use or apply benomyl in combination with captan and/or the EBDCs. For example, Pacific Northwest apple and pear growers use benomyl only for post-harvest disease control to reduce the possibility of tolerant fungi strains. Alternatives For several pests and diseases, there are no registered substitutes for benomyl. For example, benomyl is the only pesticide registered to control rice blast and stem rot, which cause approximately a 12 to 15 percent loss in rice production annually. And according to the EPA, neither cultural practices, crop rotations, nor water management are effective in control- ling these diseases. The principal replacements for benomyl in fruit and vegetable produc- tion are captan, the EBDCs, captafol, or newer systemic fungicides such as metalaxyl or fosetyl Al. However, both captan and the EBDCs are
EPA'S APPLICATION OF THE DELANEY CLAUSE 201 under special review for oncogenic, mutagenic, and teratogenic effects, pending the receipt of data. Although a weak oncogen, fosetyl Al has been denied tolerances due to the Delaney Clause. CAPTAN Description of the Chemical Common name: Captan Trade names: Merpan, Orthocide, Vondcaptan, Vancide-89, and SR-46 Pesticide type: Nonsystemic fungicide Chemical family: Dicarboximides Year registered: 1951 Major producers: Stauffer Chemical and Chevron Chemical produce the technical material. There are over 600 registered products containing captan, with registrations held by 139 formulators and producers. Volume of use: Approximately 9 to 10 million pounds are applied annually. Tolerances: Captan has more than 70 section 408 tolerances ranging from 0.25 to 100 ppm. One section 409 food additive tolerance is established for raisins and one feed additive tolerance is in place for corn seed used as animal feed. Regulatory Status An RPAR and PD 1 were issued on August 18, 1980. At that time the EPA sought information on the oncogenic, mutagenic, teratogenic, and other reproductive effects of captan. PD 2/3 was issued June 21, 1985, in which the EPA proposed to cancel all uses of captan on food crops unless "data are submitted that demonstrate that actual residues are sufficiently lower than current tolerances or that modification to application practices will sufficiently reduce dietary risk." Special Review Criteria Triggered Oncogenicity Mutagenicity
202 APPENDIX C Summary of Oncogenic Findings At the time of PD 1, the strongest evidence of captan's oncogenicity was in studies by the National Cancer Institute (NCI) and Innes et al.3 These studies showed that captan can induce adenocarcinomas, adeno- matous polyps, and mucosal hyperplasia in both sexes of mice. Two subsequent studies for Chevron, one a high-dose (6,000 to 16,000 ppm) and one a low-dose study (0 to 6,000 ppm), replicated the positive finding of adenocarcinomas of the digestive tract in both sexes in mice. A concurrent rat study, cosponsored by Stauffer and Chevron, found statistically significant increases in combined malignant and benign kid- ney tumors. The determination of oncogenicity has been contested by captan's registrants. In support of its finding of oncogenicity for captan, the EPA cites the rarity and replication of intestinal tumors in mice, and the fact that captan is structurally similar to captafol and folpet, both of which have demonstrated oncogenic effects in laboratory animals. Of particular significance is the occurrence of rare intestinal tumors, includ- ing adenocarcinomas, in chronic feeding studies of both captan and folpet. On the basis of this information, the EPA has assigned captan to category B2 in their modification of the International Agency for Research on Cancer (IARC) classification "probable human carcino- gen." The EPA calculated the Q* (potency) factor for captan as 2.3 x 10-3. The EPA is also requesting chronic data on tetrahydrophthalimide (THPI), a metabolite of captan. There is some concern within the agency that THPI may also cause tumors in laboratory animals. Estimate of Dietary Oncogenic Risk Using the multistage model for risk assessment, the EPA has calculated two estimates of dietary oncogenic risk. One is based on residues of captan at tolerance levels; the other is based on data from market basket surveys conducted by Chevron, Stauffer, the FDA, and the Canadian government.4 Using Food Factor consumption estimates assuming that 100 percent of all crops with tolerances for captan are treated and that residues are at the tolerance level, the agency estimates a dietary upper-bound oncogenic risk of 10-3 to 10-4 When market basket survey residue figures are used, the risk is calculated at 10-6 to 10-7. Although calculations using tolerances probably overstate exposure and risk, the use of market basket data may
EPA'S APPLICATION OF THE DELANEY CLAUSE 203 underestimate exposure and risk because the frequency of treatment was not stated and because both treated and untreated foods were examined. Tolerance and Delaney Clause Issues Captan illustrates the difficulty of conducting risk assessments using raw agricultural commodity (section 408) tolerances based on little or no data. These tolerances are generally high and in many cases not supported with valid data. They are often estimates set to accommodate the greatest conceivable residue of the chemical. Thus, they tend to inflate risk estimates. For example, the tolerance for apples- the major use of captan is 25 ppm, whereas the highest residue detected in the studies cited above5 was 0.08 ppm. The difference in these exposure estimates could alter risk estimates for consumption of cap/an-treated apples by three or four orders of magnitude. Further, in many cases, pesticides registered in the 1950s and 1960s (such as captan) have been subjected to few if any of the studies necessary to determine whether residues concentrate in processed foods or animal feeds. Although the residue data cited above6 indicate that captan residues generally decline with the processing of foods, it is possible that residues could concentrate in animal feed portions of many crops, thus necessitating section 409 feed additive tolerances for these crops. If tolerances were denied or revoked because of the concentration of residues in animal feeds, significant adjustments would be required of growers highly dependent on captan. Other Chronic Health Risks Captan illustrates the limitations of the Delaney Clause in reducing a non-oncogenic dietary risk in this case a reproductive risk. The No Observable Effect Level (NOEL) for toxic effects in reproductive studies using captan is 12.5 mg/kg body weight/day. Using a safety factor of 100, the allowable daily intake of captan would be 0.125 mg/kg body weight/day. For a 60-kg person this translates into a maximum allowable intake of 7.5 mg/day. However, using the EPA theoretical maximum residue contributions (TMRCs) based on dietary exposure to captan residues at the tolerance level, a person would consume 12.2 mg/kg body weight/day of captan residues, or 63 percent more than the estimated safe daily dose. In the absence of concurrent oncogenicity, these risks could not be reduced by the Delaney Clause.
204 APPENDIX C CHLOROBENZILATE Description of the Chemical Common name: Chlorobenzilate, or ethyl-4,4'-dichlorobenzilate Trade name: Chlorobenzilate Pesticide type: Acaricide Chemical family: Organochlorine Year registered: 1956 Major producer: Ciba-Geigy Volume of use: Approximately 1.5 million pounds per year are applied on citrus. Tolerances: Chlorobenzilate has a section 408 tolerance of 5.0 ppm in citrus. There are no section 409 tolerances. Regulatory Status An REAR against the registration of pesticide products containing Chlorobenzilate was issued May 26, 1976, on the basis of findings of . . . . Oncogen~c~ty In mice. A PD 4, Notice of Intent to Cancel Registrations of pesticide products containing chlorobenzilate, was issued February 13, 1979. This notice canceled all uses except on citrus in Florida, Texas, California, and Arizona. These remaining uses are classified as restricted and require protective clothing during application. A registration standard was com- pleted in 1984. Special Review Criteria Triggered Oncogenicity Summary of Oncogenic Findings and Risk Estimation Evidence of oncogenicity was found in an 18-month feeding study in which male mice exhibited a statistically significant increase in liver tumors when fed chlorobenzilate. An NCI study also found statistically significant increases in total tumors and hepatocellular carcinomas in mice. The EPA's Cancer Assessment Group classified Chlorobenzilate as a class C "possible human carcinogen." For most foods with tolerances, tests revealed no detectable Chlorobenzilate residues. In these cases, a residue level of 0.1 ppm (the level of detection) was used in calculating dietary exposure. Exceptions were made for apples and pears because the whole fruit is consumed; residues in these cases were set at 5.0 ppm. Because residues of 0.01-0.02
EPA'S APPLICATION OF THE DELANEY CLAUSE 205 ppm were detected in milk and beef, a 0.04-ppm residue level was used. Using the one-hit model and these residue levels, the oncogenic risk from dietary exposure to chlorobenzilate was calculated at 0.4 x 10-6 to 2.1 x 10-6 throughout the U.S. population. Tolerance and Delaney Clause Issues A potential conflict with the Delaney Clause arose during the review of residue chemistry data for the preparation of the registration standard. Residue studies reveal that chlorobenzilate concentrates by a factor of 5 in citrus oil. Because of the Delaney Clause, agency findings of oncogen- icity would normally block the issuance of a section 409 tolerance and, moreover, would draw into question the section 408 tolerance for use in citrus. Were chlorobenzilate a new product, it would most likely have been denied both section 408 and section 409 tolerances on citrus (see Chapter 31. However, in this case the section 409 tolerance for citrus oil was not issued, and the section 408 tolerance for use on citrus remains in effect. The EPA has taken the position in this case that the oncogenic potential of chlorobenzilate is so weak, and the consumption of citrus oil so small, that a quantitative assessment of the oncogenic risk from consumption of citrus oil cannot be supported by the available data. Benefits and Alternatives At the time of the REAR, the EPA estimated the increased total cost to citrus growers from the cancellation of chlorobenzilate at $57 million. Other benefits of retaining chlorobenzilate use are its application in integrated pest management (IPM) programs and its effectiveness for control of mites. However, some experts have argued that the EPA's analysis exaggerated the value of chlorobenzilate in citrus production. A 1980 study by the National Research Council7 (NRC) used chlorobenzilate as a case study of the RPAR process. The NRC analysis concluded that "the evidence indicates that the yield and quality of citrus crops will not be diminished appreciably, if at all, if farmers are required to replace chlorobenzilate treatments with some alternative" (p. 1961. Further, the NRC calculated the added cost of these alternatives to be in the $~$3 million range, rather than the $57 million cited by the EPA. The principal alternative pesticides are ethion, carbophenthion, sulfur, and dicofol. Dicofol is currently under special review and the remaining are in the registration standard process with no evidence of oncogenic effects. Ethion and carbophenthion are potent organophosphate insecti- cides which present other types of risks to applicators and the environment.
206 APPENDIX C DICAMBA Description of the Chemical Common name: Dicamba Trade name: Banvel Pesticide type: Broadleaf herbicide Chemical family: Benzoic acid Major producer: Sandoz Year registered: 1967 Volume of use: 3 million pounds annually Tolerances: Dicamba has numerous section 408 tolerances. The establishment of a section 409 tolerance for dicamba residues in sugarcane molasses involved the application of the FDA "constituents policy" as discussed below. Regulatory Status General-use pesticide. Registration standard was completed in 1983. Oncogencity data are due in October 1987. Special Review Criteria Triggered None Oncogenic Contaminants and the Delaney Clause Animal studies submitted to the EPA do not show dicamba to be oncogenic. However, these experiments were conducted at Industrial Biotest Laboratories (IBT) and are considered invalid. Replacement tests are scheduled for submission to the EPA in October 1987. Studies with a contaminant of dicambadimethylnitrosamine (DMNA) have shown it to be an animal oncogen. The presence of an oncogenic contaminant in a non-oncogenic food or feed additive (pesti- cide residue) could in theory trigger the Delaney Clause. However, because dicamba as a whole is considered non-oncogenic in spite of the presence of an oncogenic contaminant, the EPA employed the "constit- uents policy" articulated by the FDA in D&C Green No. 68 to issue a section 409 tolerance. The FDA's constituents argument states that the safety of undesired (oncogenic) nonfunctional constituents (in non-oncogenic substances) should be judged under the general safety provisions of the FDC Act (not
EPA'S APPLICATION OF THE DELANEY CLAUSE 207 the Delaney Clause), using risk assessment as one of the decision-making tools. In this case the additional oncogenic risk from exposure to DMNA residues in sugarcane molasses, with expected residue levels of =16 parts per trillion (ppt), was calculated as 2.9 x 10-~. This level of risk was deemed acceptable in lieu of losing the benefits, and the section 409 tolerance was granted. Discussion There are at least three basic issues involved in the issuance of a section 409 tolerance for dicamba. One centers on the regulation of oncogens under the general safety clause of the FDC Act, the second involves the FDA's so-called "constituents policy," and the third entails the consid- eration of comparative risks and benefits under the FDC Act. GENERAL SAFETY CLAUSE Several comments on the rule establishing a section 409 tolerance for dicamba argued against the use of the general safety clause of the FDC Act on the grounds that it is generally established in the scientific and regulatory literature that there is no safe level of exposure to a carcino- genic substance. In other words, there is no threshold level below which tumors are known to not be induced. This argument contends that because residues in the diet at tolerance levels must be shown to be safe (in section 409 of the FDC Act and in the FDA's constituents policy), and because there is presently no known safe level of exposure to an oncogen, then section 409 tolerances for dicamba cannot legally be issued under the general safety clause. On the other hand, the FDA argues that, using a set of conservative assumptions, exposures that create an additional oncogenic risk of less than 1 in 1 million (1 x 10-6) throughout the lifetime of the population shall be considered safe. The oncogenic risk from exposure to DMNA in dicamba residues is estimated to fall in the 1 in 100 million (1 x 10-~) range. THE CONSTITUENTS POLICY Regarding the definition of a constituent and the future use of the constituents policy for the issuance of section 409 tolerances for pesti- cides, the EPA stated its policy as follows: EPA does not regard deliberately added active or inert ingredients, or metabolites thereof, as potential candidates for clearance under the constituents policy
208 APPENDIX C rationale.... Rather, EPA will only consider applying this rationale to impurities arising from the manufacture of the pesticide (residual reactants, intermediates, and products of side reactions and chemical degradates). Furthermore the Agency will consider using this rationale in issuing a food additive regulation only where the potential risk from the impurity is extremely low.9 (How low was not specified in the Federal Register notice; however, 1 x 10-6 has been the criteria used by the FDA.) The EPA has not subsequently invoked the constituents policy to grant any food or feed tolerance. BENEFITS OF D~cAMsA Dicamba is one of several products available to replace the suspended and eventually voluntarily withdrawn herbicides 2,4,5-T and silvex. Other alternatives in sugar production include paraquat, glyphosate, and 2,4-D. It is likely that the EPA's desire to find a replacement for the suspended herbicide 2,4,5-T played a part in the use of the constituents policy in this case. Although the comparative risks of these two compounds were not discussed in the Federal Register notice establishing this tolerance, the EPA's fact sheet and the registration standard document both note that "The performance of dicamba containing herbicides is such that they are viable alternatives to the suspended uses of silvex and 2,4,5-T." One result of this use of the constituents policy was to provide an alternative to 2,4,5-T and silvex which, although not risk free, clearly presented less risk. In this way the constituents policy provided a mechanism to move toward the use of safer pesticides. On the other hand, the constituents policy does allow small amounts of theoretical risk (generally less than 1 x 10-6) from residues of oncogenic pesticide contaminants in food if the product when tested as a whole is non- oncogen~c. ETHYLENEBISDITHIOCARBAMATES (EBDCs) Description of the Chemical Trade names: Major products include maneb, zineb, mancozeb, and metiram Pesticide type: Nonsystemic fungicide Chemical family: Dithiocarbamate Years registered: Introduced from the 1930s through the 1960s Major producers: There are over 40 manufacturers worldwide. In the United States, EBDCs are produced by Rohm and Haas, Du Pont, FMC, Stauffer Chemical, and seven
EPA'S APPLICATION OF THE DELANEY CLAUSE 209 other corporations. Major foreign producers are Montedison (Italy) and Rhone-Poulenc (France). Volume of use: EBDCs are the most widely used group of fungi- cides in the world. The global market was estimated at $525 million in 1984. In the United States, more than 30 million pounds are used annually to control a wide variety of fungal diseases on fruits, vegeta- bles, field crops, seeds, and ornamentals. Approx- imately one-third of all fruits and vegetables in the United States are treated with EBDCs. They are also used as industrial slimicides. No other com- mercial fungicides have as broad a spectrum of activity on as many crops as the EBDCs. Tolerances: There are more than 150 section 408 tolerances for EBDC fungicides. Mancozeb has been granted sec- tion 409 tolerances for raisins, cereal grain, and bran, as well as animal feed tolerances for barley, oats, rye, and wheat. No meat, milk, or egg tolerances have been established, although EBDC fungicides are ap- plied to numerous commodities used as animal feed. No feed additive tolerances have been established for the many EBDC-treated vegetable and fruit by- products that are used as animal feed. No tolerances have been established for ETUa contaminant con- version product and metabolite of the EBDCs. Regulatory Status An RPAR against the EBDCs was initiated in August 1977, on the basis of oncogenicity, teratogenicity, and acute toxicity to aquatic organisms. The PD 4 was completed on October 14, 1982. None of the presump- tions of risk were rebutted, yet all registrations were continued contingent on label modifications to include requirements for protective clothing for mixers and loaders, an aquatic toxicity warning statement, and the completion of specified chronic toxicology, metabolism, and dermal absorption studies. The EBDCs have extremely low acute toxicities. The EPA was sued by the Natural Resources Defense Council (NRDC) on this and 12 other RPAR decisions. NRDC charged that the RPAR process was not sufficiently open and that industry had an unfair influence on the resolution of these RPAR proceedings. In a consent decree agreement with the NRDC, the EPA agreed to review its decision on the EBDCs during the registration standard for the chemicals originally
210 APPENDIX C scheduled for FY 1986.~° The EPA recently proposed to postpone the completion of the registration standard to 1990. Special Review Criteria Triggered EBDCs Oncogenicity Teratogenicity Acute aquatic toxicity Thyroid toxicity ETU Oncogenicity Teratogenicity Acute aquatic toxicity Mutagenicity Thyroid toxicity Summary of Oncogenic Findings An RPAR was initiated against the EBDCs largely on the basis of studies indicating the oncogenic potential of EBDCs and their contami- nant and conversion product ETU. These studies found · Increased lung adenomas in three short-term (6- to 11-week) single-dose EBDC feeding studies with mice; · Increased liver and lung tumors and lymphomas in mice fed a single dose of ETU for 80 weeks; · Thyroid carcinomas in both doses of an 18-month ETU feeding study of rats; and · Dose-related thyroid carcinomas, thyroid adenomas, and thyroid hyperplasia during a 2-year ETU feeding study of rats. The agency agreed with rebutters that the EBDC feeding studies could not be used for quantitative risk assessment purposes. Several registrants argued that the liver and thyroid tumors found in these studies could have been indirectly induced. The EPA recognized this possibility and requested more data on the subject. The agency rejected epidemiological evidence provided by registrants claiming to show that EBDCs and ETU are not oncogenic. The EPA has requested new chronic feeding studies from the registrants.
EPA'S APPLICATION OF THE DELANEY CLAUSE 21 ~ Dietary Exposure Because ETU is a metabolite of the EBDCs and because the chronic data on ETU are of higher quality and generally indicate that ETU is a more potent oncogen, teratogen, and thyrotoxin than the EBDCs, the EPA conducted its risk assessment on the basis of exposure to ETU resulting from EBDC usage. ETU is also a contaminant and degradation product of the EBDC fungicides and is thus present on raw agricultural commodities prior to processing. In estimating dietary exposure to ETU, the agency made several assumptions. First, it factored the percentage of each crop treated with EBDCs in estimating a high and a low dietary exposure level. Although 11 to 48 percent of EBDCs have been shown to convert to ETU during cooking, the agency assumed a one-to-one conversion rate for the purpose of an upper-bound dietary exposure calculation. This was justified by the general uncertainty in estimating overall exposure, as well as the exclusion of major exposure factors from sources such as drinking water, and meat and dairy products, for which there are no tolerances, but where residues have been detected in the past. (The EPA expects to find ETU residues in meat and dairy products. These ETU residues are produced by animals during metabolism of feeds such as apple and tomato pomace and citrus pulp, known and suspected to have EBDC and ETU residues.) The worst-case ETU dietary exposure for a 60-kg person when estimating EBDC residues at the tolerance levels, a one-to-one EBDC- to-ETU conversion rate, commonly used food factors, the percentage of food treated, and residues at the level of detection (0.02 ppm) for milk and meat is 3.65 x 10-3 mg/kg body weight/day. The lowest-case estimate for ETU dietary exposure calculated using the same food factors, the level-of-detection residues for milk and meat and all other foods except spinach and tomatoes (where residue survey data were utilized), the percentages of crops treated, and a mean averaging of positive residue samples for raw and processed foods is 2.4 X 10-4 mg/kg body weight/day for a 60-kg person. When exposure to ETU residues through metabolism of EBDC residues calculated at survey levels or the level of detection were added to the lowest-case estimate, the dietary exposure level was calculated as 3.4 x 10-4, or an additional 0.00010 mg/kg body weight/day. Risk Assessment Using the one-hit model, oncogenic risks from dietary exposure were calculated for both a worst-case and a lowest-case estimate. The worst
2 ~ 2 APPENDIX C case assumed residues at the tolerance level and a 38 percent conversion of EBDC to ETU (not one-to-one as with the worst-case exposure assessment), whereas the lowest case used the 38 percent conversion and residues based on survey data or the level of detection. Both assumed exposure over 70 years. The lifetime worst-case estimate of oncogenic risk through dietary exposure was calculated as 4.9 x 10-4, whereas the lowest-case estimate was calculated as 4.8 x 10-5. The agency has stated that it may have underestimated the upper limit on risk because of inadequate data on animal metabolism, ETU in processed foods, residues on raw agricultural commodities, and residue in drinking water, meat, milk, eggs, and animal feeds. Tolerance and Delaney Clause Issues Specific issues include · Prior sanctioned tolerances; · Readjustment of section 408 tolerances in 1972; · Conversion of EBDCs to ETU, and thus ETU concentration during cooking, canning, and other processing; · Absence of sections 408 and 409 tolerances for ETU; · Absence of section 409 feed additive tolerances for many vegetable and fruit by-products; and · Absence of tolerances for milk, meat, and eggs, even though EBDCs are applied to numerous commodities used as animal feed, and ETU has been detected in those foods. HISTORY In 1955, section 408 tolerances for Zineb (the first EBDC to receive tolerances) were set at 7 ppm for fruits and vegetables on "very little data showing residues from actual commercial use."' In 1957, section 408 tolerances for Zineb in spinach, lettuce, and six other related crops were increased from 7 to 25 ppm. For some of these uses, there may be approved residues in processed foods sanctioned prior to the Food Additive Amendments of 1958 which includes the Delaney Clause. Other EBDC tolerances were established throughout the 1960s, ranging from 7 to 15 ppm. However, in 1970, final action on a petition for a 1-ppm section 408 tolerance for Zineb in potatoes was never taken because the FDA notified the EPA that an FDA rat feeding study confirmed findings of ETU carcinogenicity. In October 1971, the EPA and the FDA tolerance-setting staffs recommended the revocation of all
EPA'S APPLICATION OF THE DELANEY CLA USE 2 ~ 3 EBDC tolerances except those where no residues were detected. According to EPA staff, no tolerances were revoked, however, because of insufficient data on ETU residues on crops. Representatives of Du Pont, FMC, and Rohm and Haas met with the EPA in early 1972 and agreed to lower tolerances on major-use crops. Because many of these tolerances had been previously raised, these current (lowered) toler- ances remain in the 5- to 15-ppm range. Further, the residue chemistry and toxicological data to support EBDC tolerances are generally not complete. }2 CONVERSION TO ETU In addition to the presence of ETU in EBDC fungicides as applied to raw agricultural commodities, EDBC residues are known to degrade readily to ETU during the commercial processing or home cooking of various foods in particular, during the cooking or canning of spinach, carrots, potatoes, snap beans, and apples. A study of processed foods by Du Pont found ETU in 23 percent of the samples.~3 A 1978 FDA study found ETU in 100 percent of both raw and canned spinach samples. 14 Although ETU's presence as a contaminant and a degradation product could necessitate tolerances for ETU, to date no section 408 or 409 tolerances have been established. For enforcement purposes, ETU is considered to be covered by EBDC tolerances and to be present at levels equivalent to a 100 percent conversion of EBDC residues. There are problems, however, with this enforcement system in relation to the Delaney Clause. Although the accepted average rate of EBDC-to- ETU conversion of 38 percent indicates that ETU residues per se concentrate during cooking and processing, it is unlikely that ETU residues in a processed food will exceed the EBDC residues in the raw agricultural commodity. However, where conversion takes place, the ETU residues in processed foods will be greater than the ETU levels in the raw agricultural commodity. In other words, ETU is an oncogenic by-product of an oncogenic pesticide, concentrating to levels in processed foods that are not likely to exceed the relatively high section 408 tolerances for the EBDCs that for enforcement purposes are applied to ETU; but ETU residues are potentially higher in processed foods than they are in raw agricultural commodities. Available metabolism studies, although not complete, show that EBDCs are metabolized to ETU in animals. An adequate understanding of this problem is further complicated by the dearth of residue studies and the absence of EBDC or ETU tolerances for meat, milk, and eggs,
214 APPENDIX C even though EBDCs are applied to numerous commodities used as animal feed. Both EBDCs and ETU have been detected in milk and butter. Alternatives EBDC fungicides control nearly all foliar pathogens of vegetables found in the United States. The EBDCs are desirable to growers because they are inexpensive, have few phytotoxic effects, have no problem with pest resistance, can be used in some integrated programs, control a wide spectrum of diseases, and are compatible for tank mixing with other pesticides. Effective alternatives are registered for nearly all registered uses of EBDCs, but they are generally more expensive. In some cases the increase in cost to achieve equivalent control would be quite significant, particularly in humid areas such as the Southeast. For example, with the currently available fungicides, Florida growers use an average of 20.2 EBDC applications to control early and late blight in celery, whereas in California the average number of applications is around 7. According to a USDA/State/EPA assessment team, the cost per acre is the only real difference in controlling early- and late-season blight with the following fungicides: $3.00 for EBDCs, $3.80 for captafol, and $5.80 for chlorothalonil. If the EBDC registrations were canceled, south- eastern vegetable growers would suffer the greatest economic losses. Indeed, it is the use of these fungicides that has permitted the expansion of the production of certain vegetables into the humid areas of the Southeast. It is noteworthy, however, that many of these crops have no processed form, and thus remain beyond the scope of the tolerance- setting limitations of the Delaney Clause. METALAXYL Description of the Chemical Common name: Metalaxyl Trade name: Ridomil Pesticide type: Systemic fungicide Chemical family: Benzenoid Year registered: 1979, conditional registration Major producer: Ciba-Geigy, under patent Volume of use: Approximately 400,000 pounds were used on to- bacco in 1982. Other uses include vines, potatoes, and vegetable crops.
EPA'S APPLICATION OF THE DErANEY CLAUSE 215 Metalaxyl has numerous permanent section 408 tolerances on vegetables as well as section 409 tolerances on potato and tomato products, and feed additive tolerances for tomato and potato by- products. Regulatory Status General-use pesticide with a section 3 registration. Evidence of poten- tial oncogenicity was reviewed extensively for several years. The final EPA decision was that metalaxyl is not an oncogen. Special Review Criteria Triggered Issues of oncogenicity were dealt with in the registration process. Summary of Oncogenic Findings Chronic rat feeding studies to support the registration of metalaxyl were initially accepted, and the determination was made that the fungi- cide was not an oncogen. However, questions later arose regarding the possibility that the EPA staff had "cut and pasted" Ciba-Geigy's analysis onto EPA letterhead to expedite their review of the chemical. Subse- quently, a reevaluation of the data and a lab audit were ordered. During the data reevaluation, concerns arose surrounding the appear- ance of statistically significant parafollicular adenomas of the thyroid in female rats at the low and middle dose, but not at the high dose, of a two-year rat feeding study. Concurrently, the lab audit team could not validate that the chronic feeding studies were in fact done using metalaxyl. At this point, December 1983, all actions on metalaxyl were halted, including tolerance approvals and emergency exemptions. Under instructions from the registrant, the test samples were unsealed and results showed that the studies were in fact conducted with metalaxyl. However, upon further investigation, the EPA staff found evidence of pheochromocytomas of the adrenal gland medulla in male rats. Questions also appeared regarding whether the maximum tolerated dose had been administered during the teratology and chronic feeding studies. The toxicology branch turned the oncogenic evaluation over to the EPA's Cancer Assessment Group which, in conjunction with other agency staff, decided in early 1986 that metalaxyl should not be classified as an oncogen.
216 APPENDIX C Estimation of Risk There was a difference of opinion among the EPA staff as to whether metalaxyl is an oncogen. One statistical analysis of the data submitted by Ciba-Geigy concluded that when the upper-dose finding of thyroid adenomas in females rats was excluded, a significant dose response relationship emerges. Using a risk estimation derived from this statistical analysis, the EPA staff have calculated eight upper limits of oncogenic risk from dietary exposure ranging from 2.41 per 10,000 (2.41 x 10-4) to 2.27 per 1,000 (2.27 x 10-31. However, the validity of this interpretation is disputed within the agency. Several staff have argued, in agreement with Ciba-Geigy, that when the upper-dose finding is included in a calculation of oncogenic potential, there is no dose response relationship and the incidence of this tumor in this species in comparison to the control group is not statistically significant. Tolerance and Delaney Clause Issues This example illustrates the importance of data interpretation and the far-reaching consequences of a borderline decision on oncogenicity. The EPA staff interpretation of data on metalaxyl ranged from classifying it as non-oncogenic, to characterizing the dietary risk at 2.27 x 10-3. If metalaxyl had been declared an oncogen, the Delaney Clause would have been invoked, and metalaxyl would have been denied permanent section 408 and section 409 tolerances in cases where residues concen- trate in processed foods or feeds. However, because it has been declared non-oncogenic, it will retain current tolerances and will presumably be granted permanent tolerances for pending petitions. Potential Uses and Alternatives In 1979, metalaxyl was granted a conditional registration based on its potential economic benefits to tobacco farmers in controlling blue mold and downy mildew. Temporary tolerances exist for many other uses, but the economic benefits of its use are not well quantified. However, metalaxyl is generally more expensive than currently used compounds. Pest resistance has also been a problem in isolated areas. Currently registered alternatives to metalaxyl include the EBDCs, captan, benomyl, captafol, and chlorothalonil. Although all of these have chronic toxicity data indicating adverse effects usually more severe than those associated with metalaxyl, the agency has yet to apply Delaney to section 409 tolerances for old chemicals found to be oncogenic. Theoret-
EPA'S APPLICATION OF THE DELANEY CLAUSE 2~7 ically, metalaxyl could replace some percentage of uses of the currently used fungicides. In the absence of use cancellations or tolerance revocations for these compounds, however, metalaxyl is expected to complement rather than replace older fungicides for most crops. PERMETHRIN Description of the Chemical Common name: Permethrin Trade names: Pounce, Ambush Pesticide type: Pyrethroid insecticide Chemical family: Synthetic pyrethroid Major producers: Imperial Chemicals Industries and FMC Year registered: Conditional registration in 1978 Volume of use: Used on cotton, corn, soybeans, fruits, and vege- tables. As a family, synthethic pyrethroids are the fastest growing sector of the insecticide market, with projected annual sales growth rate of 30 percent. Permethrin, however, has experienced a recent decline in use, partially due to pest resis- tance. Tolerances: Numerous section 408 tolerances are established. No section 409 tolerances have been issued be- cause of oncogenicity and the Delaney Clause. Regulatory Status General-use pesticide Special Review Criteria Triggered Positive findings of oncogenicity in mice were dealt with in the registration process. Summary of Oncogenic Findings and Risk Estimation Among six long-term mouse and rat oncogenicity studies, an increase in malignant tumors was evident only in the lungs of female mice from one test. For total tumors, dose response relationships were established in two mice studies. No evidence of mutagenicity was observed in a battery of tests including a test for DNA damage. All other oncogenicity and mutagenicity tests were negative. After an evaluation of the weight of toxicological evidence, the EPA concluded that at doses above 250
2 ~ ~ APPENDIX C mg/kg body weight/day, permethrin exhibits low oncogenic potential in mice. The EPA concluded that although permethrin is a possible human oncogen, the potential for oncogenic effects in humans at expected exposure levels is "extremely low." Delaney Clause Issues As a result of these findings, permethrin is being regulated as an oncogen. Most section 408 tolerances were finalized in a rule published October 13, 1982, in the Federal Register. ]5 However, the same Federal Register notice identified tomatoes, corn, soybeans, and apples as com- modities in need of section 409 tolerances that would be acted upon separately "because the results of the mouse oncogenicity studies raise questions under the Delaney Clause." Because residues of permethrin concentrate during processes associated with these commodities (that is, because section 409 tolerances are required), final section 408 tolerances for corn, soybeans, and tomatoes were delayed. No section 409 toler- ances have been issued for these crops. All 409 tolerances have also been denied for apples. For tomatoes, corn, and soybeans, three different methods were used to grant section 408 tolerances. Each case involved eliminating the need to promulgate the associated section 409 tolerances which could not be set because of the Delaney Clause. TOMATOES During processing of tomatoes, permethrin residues concentrate about 230-fold, clearly necessitating section 409 tolerances for processed to- mato products. Because of positive findings of oncogenicity in mice, the Delaney Clause prohibits the granting of section 409 tolerances. Prior to the issuance of section 408 tolerances for tomatoes, no section 408 tolerance had been granted for any oncogenic pesticide in a commod- ity where any portion of that commodity would be processed and need a section 409 tolerance. For enforcement purposes, it was deemed impos- sible to determine whether any portion of the treated raw agricultural commodity would be present in any processed food or animal feed. Permethrin was granted a section 408 tolerance, however, for use only on "Tomatoes Grown in Florida for Final Marketing as Fresh Tomatoes." By prohibiting the use of permethrin on tomatoes for processing, the Delaney Clause was not invoked. Three factors were cited by the EPA to support this decision: 1. Approximately 98 percent of all tomatoes grown in Florida are for the fresh market.
EPA'S APPLICATION OF THE DELANEY CrAUSE 219 2. All Florida canneries (a total of four that process tomatoes) have signed agreements that no cannery waste from the canning of whole (not processed) tomatoes will be used as animal feed. 3. Shipping to canneries in adjoining states is economically unfeasible. Because no waste will be fed to animals, no section 408 tolerances for meat, milk, or eggs were deemed necessary. CORN Tolerances were initially proposed for residues of permethrin and its metabolites in or on the following raw agricultural commodities: corn fodder at 5 ppm, corn forage at 12 ppm, and corn grain at 0.05 ppm. The petition for corn fodder was subsequently raised to 12 ppm, whereupon applications for both forage and fodder tolerances were dropped after section 408 forage and fodder tolerances were granted on sweet corn. These section 408 tolerances for sweet-corn forage and fodder carry over to field-corn forage and fodder. The initial tolerance petition for residues of permethrin in corn pro- posed a use pattern for permethrin that would have allowed application of permethrin after ears had formed and included a 30-day preharvest interval (PHI) and a prohibition on cutting for silage within seven days of the last application. However, residues resulting from this practice necessitated section 409 tolerances for corn oil and soap stock. Because the Delaney Clause will not allow section 409 tolerances for permethrin, section 408 tolerances for permethrin in corn were denied. In order to avoid the application of the Delaney Clause to these uses, label restrictions were developed, supported by residue data that show that if permethrin were not applied after ear formation, no detectable residues (at 0.02-ppm level of detection) would remain at harvest. Lowering residues below the level of detection (or theoretically eliminating these residues) eliminated the need for section 409 tolerances; thus a section 408 tolerance of 0.02 ppm for permethrin in or on the raw agricultural commodity, corn grain, was granted. Subsequently, data were submitted to support a label change allowing application after ear formation but prior to brown silking. This policy also prevented the need for section 408 tolerances to cover residues in eggs, milk, fat, and meat by-products of cattle, goats, hogs, horses, poultry, and sheep. SOYBEANS Section 408 tolerances for permethrin on soybeans were also denied pending resolution of similar Delaney issues. The need for section 409
220 APPENDIX C tolerances were obviated and section 408 tolerances were ultimately granted, when residues were lowered below the level of detection through the application of a 60-day PHI. Discussion One major criticism of the tolerance-setting system is that the process is generally devoid of incentives to drive tolerances to the lowest levels necessary for efficacious use of the product. Usually, it is only when the Theoretical Maximum Residue Contribution (TMRC) approaches the Acceptable Daily Intake that registrants seek to lower existing tolerances to allow for new uses. Presumably, a tolerance lowered in this fashion could have been lower at the time it was granted. In the case of permethrin, the Delaney Clause provided this incentive and clearly forced a reduction (in theory an elimination) of residues on two major food crops corn and soybeans while at the same time allowing these uses, and providing agriculture with new insecticides that are generally less toxic and provide significant benefits when compared with their major alternatives. However, where the elimination of residues cannot be achieved, the Delaney Clause does not allow the use of permethrin when processing will necessitate a section 409 tolerance for that crop. THIODICARB Description of the Chemical Common name: Thiodicarb Trade name: Larvin Pesticide type: Insecticide Chemical family: Carbamate Year registered: 1979 Major producer: Union Carbide, under patent Volume of use: Not available Tolerances: A section 408 tolerance of 2.0 ppm for thiodicarb residues in sweet corn was established in 1984. Sections 408 and 409 tolerances for thiodicarb and its metabolite methomyl in or on cotton, cotton- seed, soybeans, and soybean hulls were initially denied under the Delaney Clause. Using the FDA's sensitivity-of-the-method approach, these toler- ances were finalized on October 10, 1985.
EPA 'S APPLICATION OF THE DELANEY CLA USE 22 ~ Regulatory Status General-use pesticide Special Review Criteria Triggered Positive findings of oncogenicity for acetamide, a metabolite of thiodicarb, were dealt with in the registration process. Summary of Oncogenic Findings and Risk Estimation Thiodicarb has a complete data base of acceptable grade. All thiodicarb oncogenicity studies have been submitted, and all are negative. However, animal metabolism studies show that acetamide, an animal oncogen, is a metabolite of thiodicarb. Four tests performed from 1955 to 1980 show that at doses ranging from 12,500 to 80,000 ppm, acetamide is oncogenic in test animals. Although none of these studies meet current standards for oncogenicity testing, it is the conclusion of the agency that "the studies collectively demonstrate that, at least under certain conditions, long-term dietary administration of acetamide at high doses is associated with the occurrence of liver tumors in rats." Further, "the agency believes it is prudent to assume for present purposes that acetamide is a possible human carcinogen." Using the positive results in male rats from the most recent study, the EPA calculated a level of risk from acetamide in the human diet as a result of thiodicarb residues. This exercise employed a set of conser- vative principles in which, among other things, the agency assumed that · The metabolic pathway of thiodicarb is the same as that found in test animals, and the highest value of risk obtainable from the animal data is applicable to humans; · All consumed residues of thiodicarb are converted to acetamide (which the EPA states is unlikely as suggested by the available data); and · All cotton and soybeans grown in the United States will be treated with thiodicarb. On the basis of this body of evidence and these presumptions, the EPA calculated an upper-bound estimate of total dietary oncogenic risk from acetamide in the diet as a result of thiodicarb residues on cotton and soybeans, of approximately 1 x 10-6. Yet, it is clear that the agency does not believe that the dietary risks are this high. In conclud- ing comments discussing these finds in the Federal Register, the EPA states that because of the extremely conservative methodology em-
222 APPENDIX C played in the risk estimation, it "believes that the actual risk is less than 10-6. " 17 For purposes of the committee's work, it is noteworthy that the risks involved here were insufficient to trigger a special review of thiodicarb. As stated by the EPA in the proposed final rule: There are no regulatory actions pending against the registration of thiodicarb. On the basis of the available studies on acetamide and the chronic oncogenicity studies for thiodicarb, the Agency has concluded that the human risks posed by the use of thiodicarb on cotton and soybeans does [sic] not raise prudent concerns of unreasonable adverse effects and that a special review under 40 CFR 162.11 is not warranted. Put another way, this statement means that in the opinion of the agency, the regulatory actions surrounding thiodicarb arise entirely from the Delaney Clause, and concern the issuance of tolerances rather than the granting of product registration, in a case where the risk involved would not otherwise warrant review or a delay in the issuance of such tolerances. TH~oD~cARs AND METHOMYL Thiodicarb breaks down to methomyl soon after application. In fact, the tolerances at issue here are for "thiodicarb and its metabolite methomyl." Methomyl itself is a registered pesticide with two valid studies showing no oncogenicity. Even though it is very likely that acetamide is also a metabolite of methomyl, it has not yet been detected in animal metabolism or residue studies accepted by the EPA in support of methomyl registrations. To date, methomyl and its metabolites have not been regulated as oncogens, nor has the Delaney Clause been invoked against any tolerances for methomyl. The available data do not show methomyl to concentrate during the processing of food or animal feeds. Therefore, although section 409 tolerances have been a major issue for thiodicarb, section 409 tolerances have not been required for methomyl. Until concentrating oncogenic residues of methomyl and/or its metabolites are detected, the Delaney Clause will not apply to methomyl, regardless of its chemical similarity to thiodicarb. Tolerance and Delaney Clause Issues Because thiodicarb is known to concentrate in cotton seed and soybean hulls, its use on soybeans and cotton requires section 409 feed additive tolerances. These tolerances were initially denied because of acetamide
EPA'S APPLICATION OF THE DELANEY CLA USE 223 oncogenicity and the Delaney Clause. As stated in the Federal Register, the additional cancer risk from the proposed uses of thiodicarb is less than 1 X 10-6 19 According to agency sources, thiodicarb is, in a sense, a victim of the high quality of its supporting studies. In particular, the animal metabolism studies that detected acetamide pursued thiodicarb metabolites to an exceptional level of detail. Had these studies not traced the metabolism of thiodicarb so thoroughly, they might have met the EPA requirements but not have detected acetamide as a thiodicarb metabolite. Data supporting methomyl the major metabolite of thiodicarb and itself a registered pesticide active ingredient have been reviewed and accepted by the EPA, yet acetamide was not detected as a metabolite. Were acetamide not detected as a metabolite, thiodicarb would have received section 409 tolerances and section 3 registrations on the basis of its negative oncogenicity. To summarize, thiodicarb needs a section 409 tolerance because it concentrates in cottonseed and soybean hulls used as animal feed. Thiodicarb is non-oncogenic and regardless of its concentration in feed, in the absence of an oncogenic metabolite, the Delaney Clause would not apply. However, thiodicarb is metabolized by livestock into acetamide, an oncogen, which is present in meat, milk, and eggs. Thus, the EPA interprets the Delaney Clause to prohibit the use of thiodicarb on crops fed to animals that produce these foods. Sensitivity-of-the-Method Procedure Because acetamide is an animal metabolite of thiodicarb, and not present in foods derived from soybeans and cotton treated with thiodicarb, the setting of animal feed additive tolerances under section 409 of the FDC Act is the focal point of this exercise. Within section 409(c)~31(A) of the FDC Act is the so-called "DES proviso" which states that the Delaney Clause shall not apply with respect to the use of a substance as an ingredient of feed for animals which are raised for food production, if the [Administrator] finds . . . (ii) that no residue of the additive will be found (by methods of examination prescribed and approved by the [Administrator] by regulation) in any edible portion of the animal after slaughter or in any food yielded by or derived from the living animal. The FDA has extensively analyzed the meaning of this exception in a document published in the March 20, 1979, Federal Register.20 Therein, the FDA concludes that the proviso should be implemented by requiring that residues of an oncogenic compound should not be
224 APPENDIX C allowed in the total diet of humans unless it can be verified ana- lytically that they occur at levels less than those that, as deter- mined by prescribed methods of extrapolation based on animal bioassay data and a series of conservative assumptions, yield an insignificant excess cancer risk (which the FDA sets at 1 in 1 million or 1 x 10-61. Although this analysis of the DES proviso has not yet been formally adopted through a final rule, the EPA employed this rationale in issuing section 409 tolerances for thiodicarb: "For the purposes of this action, EPA adopts the reasoning and methodology of the FDA document." Several comments on the issuance of these tolerances criticize the EPA's use of a procedure that has not been finalized through a formal rule making. Using the FDA's methodology as explained in the July 3, 1985, proposed rulemaking for thiodicarb tolerances, the EPA calculated that meat and poultry could contain 90 ppb acetamide residues, and the excess lifetime cancer risk would not exceed 1 x 10-6. The agency further estimates that at the proposed tolerance levels for thiodicarb, maximum concentrations of acetamide residues in beef and poultry liver, which on average contain 17 and 6 times the residues found in muscle tissue, would be 1.8 and 0.6 ppb, respectively. For enforcement purposes (the liver will be monitored to detect violative residues in meat) one should multiply 17 times 90 ppb to get 1,530 ppb, the maximum level of acetamide residues allowed in liver. The lowest levels of reliable measurement for acetamide in beef and poultry liver, using the analytical method submitted by Union Carbide, are 700 and 400 ppb, respectively. In the EPA's judgment, this method is sufficient to detect violative residues in beef and poultry. Clearly, both the level of detection and the allowable level of residues of acetamide are far above levels expected to result from residues of thiodicarb at the tolerances, 1.8 and 0.6 ppb. In the cases of milk and eggs, allowable levels of acetamide of 30 and 90 ppb, respectively, were determined. In contrast, the maximum ex- pected acetamide levels in milk and eggs resulting from thiodicarb residues on cottonseed and soybean hulls are 0.3 and 0.07 ppb, respec- tively. Union Carbide requested a waiver from the requirement for an analytical method of detection because milk and egg samples purchased at grocery stores in 11 states contained levels of 275-500 ppb acetamide for milk, and 75-350 ppb acetamide for eggs far above anticipated maxi- mum residues from use of thiodicarb as well as those equivalent to a risk of 1 x 10-6 (30 and 90 ppb). Because EPA tests also found acetamide residues to be ubiquitous, the requirement for an analytical method was waived. i'
EPA'S APPLICATION OF THE DELANEY CLAUSE 225 NOTES 1. U.S. Environmental Protection Agency. 1982. Benomyl Fact Sheet. Washington, D.C. 2. Holder, J. A. 1980. Memorandum to C. Chaisson, U.S. Environmental Protection Agency, regarding correction of worst-case dietary exposure to benomyl in the United States for percent crop tracked and direct sampling of selected crops in a Du Pont market booklet survey. U.S. Environmental Protection Agency, Washington, D.C. 3. Environmental Protection Agency Rebuttable Presumption Against Continued Regis- tration of Products Containing Captan. 1980. Federal Register 45(No. 161, August 18): 54938-54985. 4. U.S. Environmental Protection Agency. 1985. Captan Special Review Position Docu- ment 2/3. Washington, D.C.: U.S. Environmental Protection Agency. Sec. 2, pp. 91-92. 5. Ibid. 6. Ibid. 7. National Research Council. 1980. Regulating Pesticides. Washington, D.C.: National Academy Press. Federal Register 47(April 2, 1982): 14136. 9. Federal Register 49(December 5, 1984):47482. 10. Environmental Forum. December 1984. PesticidesChanging the Way EPA Does Business, pp. 17-18. 11. U.S. Environmental Protection Agency. 1982. Ethylenebisdithiocarbamates. Decision Document, Office of Pesticides and Toxic Substances. Washington, D.C., p. I-11. 12. Ibid. 13. Ibid., p. I-38. 14. Ibid., p. I-39. 15. Federal Register 47(0ctober 13, 1982):45008. 16. Federal Register 50(No. 128):27453. 17. Federal Register 50(No. 197):41342. 18. Federal Register 50(July 3, 1985):27464. 19. Federal Register 50(No. 197):41342. 20. FederalRegister44(March20, 1979):17020. 21. Ibid.