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.
Executive Summary In February 1985, the Environmental Protection Agency (EPA) asked the Board on Agriculture of the National Research Council to study the EPA's methods for setting tolerances for pesticide residues in food. Specifically, the EPA asked the board to examine the impact of the Delaney Clause on the tolerance-setting process. Although the Delaney Clause appears on its face to be a minor feature of the complex statutory scheme governing the regulation of pesticides and pesticide residues in food, its potential impact on the EPA's future decision making is great. The EPA establishes tolerances for pesticide residues on raw commod- ities under section 408 of the Federal Food, Drug and Cosmetic (FDC) Act. Enacted in 1954, this law stipulates that tolerances are to be set at levels deemed necessary to protect the public health, while taking into account the need for "an adequate, wholesome, and economical food supply." Section 408 thus explicitly recognizes that pesticides confer benefits and risks and that both should be taken into account in setting raw commodity tolerances. Pesticide residues that concentrate in processed food above the level authorized to be present in or on their parent raw commodities are governed by the FDC Act's section 409, the law governing food additives. Under section 409, such residues must be proven safe, which is defined as a "reasonable certainty" that "no harm" to consumers will result when the additive is put to its intended use. Consideration of benefits is not authorized. Moreover, section 409 contains the Delaney Clause. This clause prohibits the approval of a food additive that has been found to
2 REGULATING PESTICIDES IN FOOD "induce cancer" (or, under the EPA's interpretation, to induce either benign or malignant tumors i.e., is oncogenic) in humans or animals. The dichotomous statutory standards applicable to tolerance setting inspired this study. A pesticide regulated on a risk/benefit basis at the time of registration and in the setting of tolerances for residues in or on raw agricultural commodities becomes, solely because it concentrates in processed food, subject to the Delaney Clause's ostensible zero-risk standard. This shift in criteria has potentially far-reaching effects. If any portion of a crop to which an oncogenic pesticide has been applied is processed in a way that will concentrate residues, the EPA's current policy is to deny not only a section 409 tolerance for the processed food but also a section 408 tolerance for residues of the pesticide in or on the raw commodity. Further, if required section 408 tolerances cannot be granted for a food-use pesticide, the EPA must also deny registration of the pesticide under the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA). The EPA's policy concerning application of the Delaney Clause to new pesticides and new uses is fairly clear. The EPA's policies are currently less settled regarding the large number of already-registered pesticides, however, and are expected to come under intense pressure in coming years. The sources of this pressure are the following: · The EPA has instituted programs to expand substantially the data on the toxicological properties of pesticides and the tendency of individual pesticides to concentrate in processed foods. · Once tested in accordance with contemporary standards, many older pesticides are likely to be identified as oncogenic or potentially oncogenic, inviting regulatory action by the EPA. · As more data on the tendency of pesticides to concentrate in processed food become available, many more currently registered pesti- cides will be shown to need section 409 tolerances and thus will become subject to the Delaney Clause. It is unlikely that the EPA will be able to avoid applying the Delaney Clause to registered pesticides. Thus, as the agency proceeds through the reregistration process, it must determine how to apply the zero-risk standard of the Delaney Clause to a significant number of currently registered, commercially important pesticides. Because of the potential magnitude and complexity of this task, the EPA asked the Board on Agriculture to undertake this study. The committee-undertook three principal tasks. First, it examined the statutory basis of tolerance setting for pesticide residues in food and the operation of the tolerance-setting process at the EPA. Second, it devel- oped a computerized data base to estimate the potential dietary oncogenic
EXECUTIVE SUMMARY 3 risk associated with pesticides determined or suspected by the EPA to be oncogenic. Third, using the data base it had developed, the committee analyzed the impact of alternative approaches to tolerance setting on estimated cancer risks and pesticide use and development. THE COMMITTEE'S ESTIMATES OF ONCOGENIC RISK . The analytical methods involved in estimating oncogenic risks from pesticide residues in food contain many areas of scientific and technical uncertainty. The assumptions made to account for these uncertainties can have profound implications for the resulting risk estimates. For example, the calculation of exposure to pesticide residues in a given foodstuff can yield risk estimates that vary by an order of magnitude. The committee assumes all residues are at the tolerance level, although actual residues may be different. Likewise, assumptions regarding how and when to aggregate risks from a pesticide used on a variety of crops can signifi- cantly alter risk estimates. The quantification of a pesticide's oncogenic- ity potency, called a Q star or Q*, can also vary by orders of magnitude, depending on such factors as whether a surface area or body weight correction is made in extrapolating risks from rodents to humans, the choice of extrapolation model, or whether malignant and benign tumors are combined when calculating the response to a given dose. The Q~'s used by the committee were calculated by EPA scientists and have not been formally peer reviewed. The EPA generally follows a conservative policy in estimating risk. Whenever assumptions must be made, the agency attempts to make them in a way that minimizes the chance of underestimating risks. The cumulative result of these assumptions is an upper-bound estimate of additional oncogenic risk above the background cancer risk of 1 in 4 or 0.25 (2.5 x 10-~. For perspective, it is worth noting that an additional dietary oncogenic risk of 1 in 1 million or 1 x 10-6 would raise this background risk of 0.25 to 0.250001. In developing the risk estimates contained in this report, the committee adopted what it understood to be the EPA's current methodology for quantitative risk assessment, recog- nizing that many key elements of the agency's risk assessment procedures are under review. This report only notes the importance of these assumptions and underlying issues; it does not offer resolutions. In arriving at a regulatory position on an oncogenic pesticide, the EPA considers the relative significance of all the evidence of oncogenicity for a compound. This "weight-of-the-evidence" approach involves consid- ering many qualitative factors such as the tumor type, results of mutage- nicity bioassays for the compound, and negative oncogenicity test results. In calculating the distribution of dietary oncogenic risk and the conse-
4 REGULATING PESTICIDES IN FOOD quences of regulatory scenarios examined in this report, the committee did not use weight-of-the-evidence techniques. Instead, it relied entirely on the quantitative risk assessment methods that the EPA uses. Risk estimates are generally shown with the EPA's oncogenic classification for the compound, however. This classification system is designed to char- acterize a pesticide's oncogenicity in humans (see Chapter 31. The reader should understand that a wide margin of uncertainty surrounds nearly all of these numbers. With this in mind, the reader should focus on general patterns of risk distribution and how key parameters change when policy alternatives are assessed, not on specific point estimates of risk. The committee further emphasizes that all risk estimates in this report are limited to oncogenic risk from pesticide residues in food. This does not imply, however, that other risks presented by pesticides or other routes of exposure are less important. Indeed, the regulation of pesticides involves a consideration of many health and environmental risks, only one of which involves residues of oncogenic pesticides in food. ESTIMATED ONCOGENIC RISK AND ITS DISTRIBUTION IN THE FOOD SUPPLY To characterize the universe of oncogenic pesticides, the committee adopted the list of 53 suspected oncogenic compounds that the EPA transmitted to Congressman Henry Waxman (D-Calif.) in October 1985. Of these compounds, the committee limited its analysis to three types of pesticides herbicides, fungicides, and insecticides. The risk estimates discussed in this report are derived from only 28 of these 53 active ingredients (see Chapter 3~. To roughly characterize the benefits associ- ated with the use of these oncogenic compounds, the committee assem- bled crop use and farm-level expenditure estimates averaged over three years on all 53 compounds. Findings The committee's analysis indicates that the Delaney Clause will be central to the EPA's decision making in future years. First, the EPA considers a substantial fraction of all herbicides, fungicides, and insecticides to be oncogenic or potentially oncogenic in animal studies. On the basis of pounds of pesticide applied, 60 percent of all herbicides are oncogenic or potentially oncogenic. (This number includes two compounds not on the "Waxman list" that have since been found to be suspect oncogens, raising the percentage from around 40
EXECUTIVE SUMMARY 5 percent to just over 60 percent.) By volume, 90 percent of all fungicides fall into this category, as do about 30 percent of all insecticides. Second, for the 53 oncogenic compounds, the committee has identified 31 processed foods with approved section 409 tolerances. All of these appear to conflict with the Delaney Clause. Moreover, the committee has identified an additional 778 processed foods with no 409 tolerances for which oncogenic pesticides are registered. Residues of these pesticides are expected to concentrate in many of these processed foods. Hence, over the next few years, the EPA will face bringing several hundred additional pesticide uses into compliance with section 409 of the FDC Act and the Delaney Clause. Third, over the next five years, the EPA is scheduled to complete regulatory actions that will force decisions on compliance with the Delaney Clause for 10 currently registered oncogenic pesticides. These pesticides account for between 80 and 90 percent of the total estimated dietary oncogenic risk from residues of the 28 compounds that comprise the committee's risk estimate (see Chapter 31. Fourth, the committee explored the distribution of dietary oncogenic risk from residues of these 28 pesticides. Fungicides account for about 60 percent of all currently estimated dietary oncogenic risk from these 28 compounds. Of the remaining risk, 27 percent stems from crop uses of herbicides and 13 percent from insecticides. Fungicides, however, ac- count for only about 7 percent of all pesticide sales and less than 10 percent of all pounds applied. Further insights derived from the committee's analyses of these 28 pesticides include the following: · About 55 percent of the total estimated dietary oncogenic risk stems from residues on crops that have raw and processed food forms. · About 20 percent is associated with consumption of the processed forms of these crops. Approximately 35 percent is from consumption of the raw form of the same crops. · About 45 percent of estimated dietary oncogenic risk is from foods that the EPA considers to have no processed form. These foods include many fruits and vegetables and all meat, milk, and poultry products. These figures lead to several observations: · At most, the Delaney Clause could apply to processed-food residues responsible for only one-fifth of the estimated dietary oncogenic risk from pesticides. However, its implementation could eliminate another 35 percent of the estimated risk from residues on the raw forms of these processed foods because it is the EPA's policy to deny section 408 raw food tolerances when section 409 tolerances cannot be established for the processed forms of the same crop.
6 REGULATING PESTICIDES IN FOOD · Foods accounting for nearly one-half of the total estimated dietary risk (all meat, milk, poultry, and pork products and many fruits and vegetables) are ostensibly beyond the scope of the Delaney Clause, because under current EPA guidelines they have no processed form. Fifth, dietary oncogenic risk appears to be concentrated in a relatively small number of pesticides and crops. Nearly 80 percent of the estimated dietary oncogenic risk (from all 178 food uses of the 28 compounds that comprise the committee's risk estimate) is from residues of 10 pesticides on only 15 different foods. IMPACTS OF FOUR ALTERNATIVE WAYS TO REGULATE ONCOGENIC RESIDUES The committee studied the implications of four theoretical policy scenarios, or frameworks, for regulating residues of oncogenic pesticides in food. The committee emphasizes that these scenarios are artificial constructs chosen not because they reflect any current regulatory ap- proach but rather because they represent a plausible range of approaches. Further, the committee does not offer a legal opinion on the comparability of any of these scenarios with current law or interpretation. These tasks were not in the committee's charge. Scenario 1 applies a zero-risk standard for oncogenic risk to all pesticide residues on both raw and processed foods. If the EPA determined that a pesticide was oncogenic, all food tolerances for that pesticide would be revoked. Scenario 2 applies a zero-risk standard for oncogenic risk to all pesticide residues in processed foods: any residue of an oncogenic pesticide in a processed food would be disallowed. This scenario further assumes that when residues of an oncogenic pesticide are present in the processed portion of a crop, tolerances for both raw and processed forms would be revoked. Scenario 3 would revoke all tolerances for a pesticide on a crop when the combined estimated cancer risk from the residues of that pesticide on both the raw and processed forms of a crop exceeds 1 in 1 million or 1 x 10-6. Scenario 4 would revoke all tolerances for a pesticide on a crop when the total risk from residues of a pesticide on all processed forms of a crop exceeds 1 x 10-6. As under scenario 2, when residues on the processed form of the crop trigger revocation, both raw and processed food tolerances would be revoked. Results of the Scenarios Scenario 1 would revoke all tolerances for all oncogenic pesticides and eliminate all estimated dietary oncogenic risk.
EXECUTIVE SUMMARY 7 Scenario 2 would revoke tolerances that would reduce dietary oncogenic risk (from the 28 pesticides that constitute the committee's estimate of dietary oncogenic risk) by 55 percent, ignoring about 45 percent of total estimated dietary risk from foods with no processed form. Scenario 3 would reduce total estimated risk from these 28 compounds by 98 percent, while revoking only 32 percent of all tolerances for the 28 oncogenic pesticides. Scenario 4 would eliminate just 35 percent of the estimated dietary oncogenic risk, while revoking the smallest percentage of all tolerances. The most significant finding from the committee's crop level analyses is that under a "negligible risk" standard applicable to residues on both raw and processed food illustrated by scenario 3 a high percentage of total risk is eliminated while a low percentage of "benefits" (as measured here by acre treatments and expenditures) are lost. For certain crops, scenario 4 would also eliminate a high percentage of total risk while affecting a low percentage of benefits. Scenario 4 would be less effective at reducing overall risk, however, by not addressing tolerances on three of eight crops. Scenarios 1 and 2 have the same effect on all crop-pesticide combina- tions examined except peanut fungicides; all oncogenic risk from residues on these crops would be eliminated. In most cases, tolerances associated with a significant percentage of current pesticide expenditures and acre treatments on these crops would be lost. For certain crops in certain regions, the loss of all oncogenic compoundsparticularly fungicides- would cause severe short-term adjustments in pest control practices because of the lack of economically viable alternatives. Findings The four scenario analyses suggest that progress toward risk reduction could be the greatest and most uniform when raw and processed foods are subject to a consistent risk standard. The potential advantages of a negligible risk standard, with no consideration of benefits, are also highlighted. Such a standard, consistently applied, could eliminate most existing dietary oncogenic risk while allowing continued use of and benefits fromcertain low-risk compounds. AN ANALYSIS OF THE FUNGICIDES: A SPECIAL CASE Fungicides present a unique problem. The EPA needs to establish hundreds of new processed-food tolerances for these compounds, but about 90 percent by weight of all fungicides now applied are considered potential oncogens. All but 1 of the 14 oncogenic fungicides were
~ REGULATING PESTICIDES IN FOOD registered more than 15 years ago. These older pesticides are relatively inexpensive, remain effective in various applications, and provide signif- icant benefits to agricultural producers and consumers. These fungicides are important in the production of many high-value fruit and vegetable crops, particularly in the eastern and southeastern regions of the country. The eleven fungicides in the committee's total risk estimate represent less than 10 percent of all acre treatments with pesticides, but are responsible for about 60 percent of the total estimated dietary oncogenic risk. The mode of action of fungicides makes it difficult to develop com- pounds that are nontoxic to genetic material. As a result, few elective non-oncogenic new fungicides are being developed. Only four fungicides registered since 1972 have gained greater than 5 percent of the market share for any food crop. Data on several of these compounds indicate oncogenicity and other chronic effects, however. The combination of the above factors makes the regulation of oncogenic risk from fungicide residues extremely complex. It is the committee's view that literal application of the Delaney Clause will significantly complicate the EPA's efforts to reduce dietary oncogenic risks from fungicides. The committee performed several simple analyses (see Chapter 5) showing that sequential tolerance revocations or denials for one active ingredient at a time could in some cases actually increase human dietary oncogenic risk and in many other cases only lower it slightly. These results would occur when the use of a fungicide presenting an equal or greater risk increased after tolerances for a less-hazardous compound were revoked. In response to this dilemma, the committee examined the effect of cropwide tolerance reductions as a means for reducing estimated dietary risk from fungicides. Even a cursory analysis suggests that this and other new regulatory strategies warrant detailed study in terms of their potential to bring about significant reduction in dietary oncogenic risk while preserving beneficial fungicide uses. Cropwide tolerance reductions could reduce the total estimated dietary oncogenic risk from fungicides by up to 50 percent with only modest enforcement effort and minor adjustments in the agricultural sector. PESTICIDE INNOVATION AND ALTERNATIVE PEST-CONTROL METHODS Despite the development of pest resistance to chemical pesticides, environmental damage, applicator risk, and fears of cancer associated with pesticide use, most farmers believe that pesticides are a critical part of a reliable and cost-effective pest control strategy. If a large number of tolerances for oncogenic pesticides are revoked over the next five years,
EXECUTI VE 5 UMMAR Y 9 adequate replacement pesticides, particularly fungicides, probably would not be available for several major fruit and vegetable crops. For herbi- cides, however, the prospects are more optimistic. Considering all alternatives being developed, it appears that the loss of several major oncogenic herbicides would not pose a serious threat to agriculture. The outlook for insecticides lies somewhere between that for herbicides and fungicides. The simultaneous loss of several oncogenic fungicides as described above could present serious disease control problems in certain crops in major production regions. Further, the rate of successful new product development within the three major categories of pesticides is almost inversely proportional to dietary risk; innovation has not been occurring to address the problem of oncogenicity. Alternative Technologies Advances in classical plant breeding, innovation in biological and cultural pest control systems, and progress in genetic engineering offer some promise for nonchemical pest control in the future. Nonchemical approaches will be encouraged by tolerance revocations if more profitable chemical controls are not available. For many crops, especially fruits and vegetables, there are few equally effective technologic alternatives to chemical pest control. More important, as with synthetic chemical pesticides, R&D efforts in alternative technologies do not appear directed toward the pest problems most likely to be affected by Delaney Clause-driven tolerance revocations; breeding for disease resistance is an important exception. Because of the time needed to further develop plant genetic engineering techniques, new technologies involving or derived from biotechnology will not be available to reduce the impact of the next five years of regulatory actions. Although plant breeding offers the promise of more pest-resistant crop varieties, these new varieties will probably lower rather than eliminate the need for pesticides. The objective of biocontrol techniques is to establish a more stable pest control situation in the long run, but these methods are often complex and do not always provide the certainty that chemical pesticides provide. The problems with these alternative technologies will delay the adoption of nonchemical tech- niques in the absence of incentives. The Effect of the Delaney Clause on R&D In the long term, a rigorous application of the Delaney Clause to both existing and new pesticides is likely to shift the focus of public and private
10 REGULATING PESTICIDES IN FOOD R&D away from oncogenic compounds and reduce the total investment in chemical pesticides. This shift and reduction of investments will depend on factors that the committee was not able to examine quantitatively, including the structure of the industry (corporate mergers), regulatory developments, and scientific breakthroughs. If gross sales of agricultural chemical companies are reduced or if net returns become more variable as a result of tolerance revocations, however, revenues available for these R&D activities will decline and overall innovation is likely to fall. These declines in R&D investments are more likely if revocations lead to increased use of nonchemical controlsdevelopments that provide no profit for pesticide companies. The Minor Use Issue All fruit and vegetable crops grown in the United States are considered "minor" crops in terms of pesticide use. Most minor crops have no recognized processed forms under EPA regulations. Most minor crops also present relatively small oncogenic risks. The important issue with these crops is whether pesticides currently vital to their production will remain available when the tolerances for oncogenic pesticides on other crops are revoked pursuant to the Delaney Clause. Certain minor crops with processed forms do present potentially significant risk, however. Apples, tomatoes, grapes, potatoes, and citrus are directly vulnerable to tolerance revocations under any version of the Delaney Clause. In contrast to most minor crops, these crops represent intermediate markets for pesticide producers and therefore provide an incentive for product development. In this fashion, the continued avail- ability of pesticides for small minor crops is linked to the continued availability of pesticides on "minor" crops representing larger markets. Liability for crop failures or crop injury is another potential obstacle to future pesticide availability for small minor crops. This results from the limited acreage of these minor crops and hence the limited pesticide market they provide, relative to potential liability for control failures. Liability is high because these crops have high per acre value and must meet high-quality standards. The Delaney Clause and Pest Resistance A distinct value of many widely applied, suspect oncogenic fungicides is that even after years of use, pests show little if any resistance to them. In contrast, a number of pests have developed some degree of resistance to many of the new, non-oncogenic or more weakly oncogenic fungicides and insecticides. The viability of many of these newer compounds is often
EXECUTI VE 5 UMMAR Y ~ ~ linked to their simultaneous use with the older oncogenic compounds. Given the large number of tolerances for older fungicides that may be revoked under the Delaney Clause, there is a critical need to develop non-oncogenic fungicides as well as resistance management programs for these fungicides in the next five years. MAJOR CONCLUSIONS AND RECOMMENDATIONS The committee's analysis and findings support four basic conclusions. The first deals with the legal and institutional base for regulating oncogenic pesticide residues in food. The second addresses the possible roles of other than zero oncogenic risk criteria in targeting regulation on pesticide uses that pose potentially significant human health hazards. The third involves the need for and structure of an overall strategy for the EPA as it moves ahead with the task of bringing all existing pesticide registrations and tolerances into compliance with the law, including the Delaney Clause. The fourth describes the adoption of an analytical framework to facilitate a more systematic examination of the risks and benefits associated with pesticides. Legal Basis for Regulation Pesticide residues in food, whether marketed in raw or processed form or governed by old or new tolerances, should be regulated on the basis of consistent standards. Current law and regulations governing residues in raw and processed food are inconsistent with this goal. First, the committee can discover no scientific reason for the law's different treatment of raw and processed food tolerances. Because the committee could find no scientific reason for this disparity, it recommends the consistency and simplification of treating them alike. Second, residues stemming from the use of older pesticides should be subject to the same scrutiny and standards as those applied to residues from new pesticides. Although concentrating on new pesticides that might present new dangers was a reasonable policy when the regulation of pesticides began, new analyses, data, and criteria provide compelling reasons for uniform treatment of old and new compounds. Neither the FDC Act nor FIFRA provides a clear basis for treating residues of new pesticides differently from those of old pesticides. Subjecting old toler- ances to contemporary safety criteria is essential. The committee's analysis demonstrates that about 90 percent of esti- mated dietary oncogenic risks from pesticides stems from uses sanctioned by tolerances granted before 1978. The only way for the EPA to reduce
12 REGULATING PESTICIDES IN FOOD dietary risk substantially is to subject old pesticides to the same regula- tory scrutiny it applies to new agents. Doing so in the context of reregistration will provide the EPA with an opportunity to reduce public exposure to at least some of the oncogenic pesticides that now are routinely present in food. Third, the two agencies charged with implementing and interpreting the Delaney Clausethe EPA and the Food and Drug Administration (FDA) should continue to work to achieve consistent legal interpreta- tions of their common statute and compatible regulatory policies based on the best contemporary science. As long as the Delaney Clause is a part of the FDC Act, better cooperation and communication between EPA and FDA scientists and policy-level officials will be increasingly important as the two agencies work to develop policies for implementing this provi- sion. The effective functioning of both agencies depends on the develop- ment of consistent policies for achieving the widely supported goal of eliminating added carcinogens from the food supply whenever feasible and prudent. A Nonzero Standard for Oncogenic Risk A negligible risk standardfor carcinogens in food, applied consistently to all pesticides and to all forms offood, could dramatically reduce total dietary exposure to oncogenic pesticides with modest reduction of benefits. The committee believes that the elimination of oncogenic pesticide residues from human food is an appropriate aspiration of regulation. The committee recognizes, however, that residues of several dozen oncogenic pesticides may be found in hundreds of different foods. Many such residues pose little risk to humans, whereas some clearly warrant attention and, quite probably, regulatory action. The problem of imple- menting action against many pesticides with limited personnel and resources should be minimized. Moreover, the challenge for regulators grows increasingly complex as science and technology advance. Improve- ments in analytical chemistry and residue detection capabilities, new toxicological data, changing pesticide use practices, and the development of new pesticides and foods establish an urgency and the feasibility to devise a strategy for attaining a safer food supply. One option for regulators is to adopt a negligible risk standard in setting and revising tolerances for all oncogenic pesticides found in food. The committee sees merit in such a standard if its adoption can speed up progress toward risk reduction and help the EPA focus its limited resources on pesticides and crops that pose significant oncogenic risks.
EXECUTIVE SUMMARY 13 The committee notes with concern that current EPA policy has allowed continued use of pesticides that pose estimated dietary oncogenic risks as high as 1 in 10,000 (1 x 10-44. The adoption of a negligible-risk standard would provide added justification for the agency to reduce relatively high risks while deferring actions on relatively low or perhaps even zero risks. The committee would not endorse the adoption of such a standard if it were not consistently applied to all pesticides and all forms of human food. The committee is aware that a zero-risk standard applied to pesticide residues in all foods would eliminate that source of oncogenic risk in the diet. The committee believes, however, that in certain regions and on certain crops, the implementation of such a policy would cause severe adjustments in agricultural practices, particularly in control of plant diseases. This policy could impede agency discretion necessary to achieve significant risk reduction over the next 5 to 10 years while maintaining viable disease control alternatives. The committee's analysis highlights several advantages of a consis- tently applied negligible-risk standard over even strict adherence to the traditional zero-risk interpretation of the Delaney Clause, which applies a zero-risk standard only to processed foods and their parent raw commod- ities: · The committee found that, if consistently applied, a negligible-risk standard applied to raw and processed foods (assuming no consideration of benefits) could lead to the elimination of 98 percent of existing dietary risk from exposure to the 28 pesticides comprising the committee's estimate of dietary oncogenic risk. In contrast, a zero-risk standard applied only to oncogenic residues in processed foods and their parent raw commodities would reduce estimated risk by just 55 percent. In reality, however, benefits must be considered, and not all residues will concentrate. Also, the risk reduction achieved under both scenarios will probably be less than suggested. Nonetheless, the committee believes that the relative effectiveness of these two scenarios will remain constant and that any plausible negligible-risk standard that treats section 408 and 409 tolerances consistently will lead to greater risk reduction than a zero-risk standard applied only to section 409 tolerances. · A uniform negligible-risk standard could give the EPA the flexibility needed to reduce dietary oncogenic risks over time. One important option would be the ability to grant tolerances for new chemicals that might pose a slight oncogenic risk (currently prohibited by the Delaney Clause), if use of such pesticides would displace more hazardous products now routinely used. In the past, the EPA has applied the Delaney Clause to deny tolerances for weak oncogens on certain crops. To date, the agency has not invoked the clause to revoke any existing tolerances (in most cases
14 REGULATING PESTICIDES IN FOOD citing lack of acceptable data on oncogenicity and residue concentration). This application of the Delaney Clause has had the effect of preserving the market share for, and continuous dietary exposure to, pesticides that present relatively greater dietary oncogenic risks. · A negligible-risk standard for tolerance setting would aid the EPA in focusing regulatory resources on the crop and pesticide combinations presenting the greatest oncogenic risk. On the other hand, a zero-risk standard does not encourage and may not allow the EPA to discriminate between relatively significant and relatively insignificant oncogenic risks. Indeed, the Delaney Clause has focused considerable agency resources on protracted scientific assessments designed to determine whether a pesticide is an "oncogen" per se and whether the risk associated with a particular use of that pesticide is zero, or some very small level that is well beyond the predictive power of currently available toxicological tests and risk assessment methods. The advantages of a consistently applied negligible-risk standard will depend on how fast the EPA will reregister old chemicals, the sequence that will be followed, and perhaps most important, how the benefits will be taken into account. Risk reductions of the magnitude that the committee's analyses suggest are not guaranteed because the analyses did not incorporate benefit considerations. The benefits of some pesticide uses may justify greater risks. The level of risk reduction achieved through the adoption of such a negligible-risk standard will depend on the importance the EPA places on the benefits of specific pesticide uses. Targeting High-Risk Pesticide and Crop Uses The committee's analysis (described on pages 50~6J suggests that about 80 percent of oncogenic risk from the 28 pesticides that constitute the committee's risk estimates is associated with the residues of 10 com- pounds in 15 foods. Logic argues that the EPA should focus its energies on reducing risk from the most worrisome pesticides on the most- consumed crops, and compelling reasons support such a strategy. First, if the EPA developed a regulatory position on all oncogenic pesticides used on a given crop, the agency would have a realistic chance of dealing with the special problems that arise when there is more than one oncogenic pesticide used on a particular crop. As a class, the fungicides present special difficulties because nine oncogenic compounds account for about 90 percent of all fungicide sales. Further, these nine compounds present comparable risks and generally are viable substitutes for one another. In this situation the agency's historical approach- regulating one pesticide at a time across all of its uses is not well suited
EXECUTIVE SUMMARY 15 to ensuring real risk reductions. This approach could even increase oncogenic risk when revocation of one agent allows a more potent oncogen to gain wider use. The recommended strategy would also help preserve the benefits of pesticide uses that pose very low, but possibly not zero, risks. For example, most suspect oncogenic pesticides used on corn, soybeans, and cotton present very low dietary oncogenic risks. Further, the committee's analysis indicates that when estimated dietary oncogenic risks from herbicides or insecticides on these crops are deemed high, they can generally be substantially reduced through actions affecting one or two compounds. Most pesticides used on these crops would probably be passed over if a policy were in effect that targeted pesticide uses presenting relatively high dietary oncogenic risk. Such pesticide uses could come under scrutiny for other reasons, however, such as other health effects, ecological problems, or groundwater contamination. The Adoption of an Analytical Framework The EPA should develop improved tools and methods to more systemat- ically estimate the overall impact of prospective regulatory actions on health, the environment, and food production. Rapid advances in com- puter technology, as well as the EPA's successful efforts to computerize major data sets like the Tolerance Assessment System (TASK make such progress readily attainable. The EPA's current approach to pesticide regulation focuses on the risks and benefits of one active ingredient at a time across all its uses. Much of the committee's analysis rested on new analytical manipulations of EPA data sets. Insights gained on the distribution and relative magnitude of risks and benefits associated with pesticide use are the foundation of the committee's recommendations. The analytical framework and data base that the committee developed on a prototype basis are described in detail in Chapters 3-5 and Appendix B. The framework the EPA might develop and utilize could be more thorough and precise. The committee's prelim- inary work provided intriguing new insights, however. The data base that the committee developed is extremely valuable in comparing the effec- tiveness of different regulatory policies that reduce dietary oncogenic risks from pesticides. Use of new analytical tools and data bases could help the EPA get ahead of its growing work load. The refinement of such a system would allow the EPA to project with increased confidence a wide range of impacts associated with its regulatory actions. For example, the com- mittee's rudimentary analysis demonstrates that certain strategies for
16 REGULATING PESTICIDES IN FOOD implementing the Delaney Clause could increase dietary risk, and vigor- ous application of the Delaney Clause to tolerances for residues in processed foods may not be the most effective strategy for minimizing dietary exposure to oncogenic pesticides. It is important to note that this study's preoccupation with oncogenic hazards simplifies the challenges that the EPA actually faces. In devel- oping regulatory decisions, the EPA must take into account trade-offs between oncogenic risks and other sorts of health and environmental hazards among all pesticides registered for a particular use. Nonetheless, this more complicated task would be aided by the analytical tools discussed in this report.
