INTRODUCTION Although only a limited number of chemicals are recognizable as human carcinogens, [benzidene, B-naphthylamine, 4-aminodiphenyl, 4-nitrodiphenyl, bis(chloromethyl)ether, and (chloromethyl)methyl ether], animal studies are generally considered qualitatively reliable for predicting carcinogenic potential in man [1]. In this regard, animal carcinogens are considered to represent some hazard to man. Human exposure to these agents must be reduced or eliminated when possible. Complete elimination of exposure to carcinogens in many cases is impossible or impractical without producing the loss of significant beneficial effects in other areas. Therefore, it is necessary to weigh the benefits against the risks in considering these matters. Attempts to evaluate the risk to man due to carcinogens in food and the environment, and to compare the risks to the benefits from the use of some of these agents, has produced several approaches to dealing with these questions. The most promising approach would appear to be a combination of mechanistic studies to determine how xenobiotics produce their toxic responses and attempts to extrapolate this information to man. However, extrapolation of this nature is highly speculative, and is based on many unverifiable assumptions [2]. It is quite appropriate to discuss the connection between risk assessment and metabolism because the latter is part of the basis by which toxicologic mechanisms occur and only by understanding those mechanisms are we going to be able to perform risk assessment evaluations. The mechanism by which a toxic reaction, i.e., carcinogenicity occurs, defines a dose response relationship and that dose response relationship is the basis upon which risk assessment should be made. The metabolism of any material plays an integral part in its toxicity. Metabolism may involve either a detoxification or an activation process. Well-known examples of activation are epoxide formation with aflatoxin B1, and amine oxidation in B-naphthylamine. Traditionally we think of glutathione conjugation and other biological conjugations as being detoxification mechanisms. In the case of ethylene dichloride, glutathione conjugation produces a far more mutagenic material, sulfur mustard [3]. In order to better understand the relationship between metabolism and risk assessment, it is important to consider some basic assumptions in the procedures of risk assessment upon which current thinking is founded. Therefore, the following areas will be considered: (1) The proposed mechanisms of carcinogenesis; (3) The merits of several common types of models for use in risk extrapolation; The appropriateness of using these approaches in (Balance of document is held in the subcommittee files.) In response to your invitation, Monsanto Company appreciates the opportunity to provide this review of the Second Draft of the Pesticide Investigation Report, as presented to the DORFA subcommittee on December 16, 1982. Chapter 6 of the referenced Draft Report was specifically noted in your January 6 and 13, 1983 letters of invitation as containing critical issues regarding the current OPP oncogenic risk assessment policy. Our present review relates solely to these issues. As requested, our comments have been directed towards the perception of changes in Agency policies and practices which may have impacted on oncogenic risk assessment. The intent of our comments is to provide a view of the scientific aspects of carcinogenesis that scientists generally hold in common today and that have, and should, continue to serve as the basis for regulatory action. Thus, we have provided specific references from the scientific literature to support our comments. In many cases, large amounts of scientific evidence have accumulated on a given subject. In those instances, we have merely attempted to denote the "bell-weather" article or point to a treatise of information on the subject. Since Monsanto Company is a member company of the National Agricultural Chemicals Association (NACA), we also will provide support in the development of consensus industry comments on additional sections of this Draft Report. We trust that the present articulation of our perceptions will be of service to your staff and the subcommittee as they consider this important public health issue. FRJ/dlj Sincerely, Patrick R. Jhanan Frederick R. Johannsen, PhD Monsanto Monsanto Company 800 N. Lindbergh Boulevard Phone: (314) 694-1000 Specific Comments on Chapter 6 DEPARTMENT OF MEDICINE & Conclusion The Basis for Qualitative Determination of Chemical Oncogenicity Supporting Documentation I. A. THERE HAS BEEN NO DEVIATION FROM PAST AGENCY Utility of the Animal Bioassay to Detect Oncogenic Potential in Humans. cancer The chronic rodent bioassay is currently accepted within the A Animal test procedures as currently employed will continue to 2 Questions concerning the applicability of several of these procedures (e.g., use of weight-of-evidence, the issue of extrapolation based on induction of malignant versus benign tumors) have been raised in the Draft Report. Comments regarding their scientific credibility in preserving public health and safety and acceptability as natural extensions of past practices will be forthcoming below. B. Use of Statistical Methods to Determine Significance of Several times in the Draft Report it was implied that decisions regarding the significance of a toxicologic observation, be it related to tumor induction or to manifestation of other chronic effects, were based solely on statistical significance of the lesion considered. We believe this has not been, nor should it be, considered normal Agency practice. While statistical analyses of a specific data set are an aid to the toxicologist in evaluating the results of an experiment, statistical significance alone should not be considered as the ultimate determinant in arriving at a conclusion of relationship to treatment (3). Scientific judgement on the part of a qualified toxicologist experienced in the area of safety assessment is essential(4). THEREFORE, WE BELIEVE THAT THE CURRENT AGENCY PRACTICE OF THE DETERMINATION OF BIOLOGICAL SIGNIFICANCE SHOULD CONTINUE TO BE BASED ON INFORMED SCIENTIFIC REASONING, RATHER THAN BLIND MATHEMATICAL ACCEPTANCE. C. Consideration of Benign vs. Malignant Tumors in Extrapolation to Human Risk The consensus agreement between regulatory scientists worldwide on this topic has recently been restated(5). It is felt that there are many good examples of chemicals which produce both malignant and benign tumors. Thus, the development of any abnormal growth should be considered as a possible index of carcinogenicity (5). In those situations "where there is reliable evidence that a compound induces [only] benign tumors under certain experimental conditions, a great deal of suspicion becomes attached to the compound. Where justified, additional tests in different strains or species...might be indicated."(6) However, there are two significant exceptions to the above principle which take on paramount significance to the benign vs. malignant tumor issue cited in the Draft Report. Specifically, these exceptions involve the induction of either mouse liver tumors or mouse lung tumors. As noted in the Draft Report, the IARC General Principles specifically mentions that "there are certain neoplasms, including lung tumors and hepatomas in mice, which have been considered of lesser significance than neoplasms occurring at other sites for the purpose of evaluating the carcinogenic risk of chemicals to humans." [Emphasis added.] Thus, |