J. Soc. Cosmet. Chem., 29, 727-732 (November 1978) Regulatory considerations concerning mutagenesis ALBERT C. KOLBYE, JR. Bureau of Foods, Food and Drug Administration, U.S. Department of Health, Education, and Welfare, Washington, D.C. 20204. Received June 8, 1977. Presented at Annual Scientific Seminar, Society of Cosmetic Chemists, May 1977, Montreal, Canada. As the study of mutagenesis has developed over the past decade, our understanding of mutagenesis and related phenomena has undergone a rapid evolution that may have profound implications for public health. This rapid evolution is also compelling us to reconsider our methods of testing substances for their potential adverse effect on human health and their implications for regulatory decisions. Thus a series of questions have arisen for which society must find answers. I shall attempt to formulate several of the questions and to explain some of the scientific and regulatory considerations that determine the utility of presently available test systems for regulatory decision-making. Society's concern about preventing cancer and genetic damage is readily understood in terms of public health, but in my judgment this concern may not always be properly directed. The chief question is not primarily whether mutagenesis assay systems are useful for identifying compounds that have the potential for causing genetic damage in future generations, although that is a major question in itself. Instead, the question is whether the results of mutagenesis testing alone can enable us to decide if a substance poses an appreciable risk for inducing cancer in humans. The scope of that question covers many related issues. For example, are the biological phenomena associated with positive results in mutagenesis assay systems relevant to cancer induction in humans? My answer is both "yes" and "no, not necessarily." I answer "yes" because induction of mutations in the DNA of somatic cells appears to be associated with an important pathway in the cancer induction pattern in experimental animals by many chemical "carcinogens." Furthermore, certain human cancers appear to be associated with exposure (primarily occupational) to chemicals all of which have the properties associated with the quality of carcinogenicity and, by definition from the results of mutagenesis tests, of mutagenicity as well. Hence we could label some of these substances "carcinomutagens" or "mutacarcinogens." The current leaders of this school of thought include Drs. Ames and McCann. They have accomplished a great deal by expanding their system so that many compounds can be tested rapidly, at moderate cost, and with results that are reproducible by independent laboratories. Their test system (1) has many attractive features, including the capability of assessing mutagenic potency quantitatively over a very wide range. Furthermore, their system is 727

728 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS sensitive enough to detect biological phenomena associated with many (but not all) so- called chemical carcinogens. I have a favorable attitude toward the Ames system be- cause I have been directly and indirectly associated with the development of mutagenic assay systems ever since the 1960's, even before I came to the Food and Dru. g Administration. The outstanding accomplishments of Drs. Ames and McCann represent considerable progress and improvement on bacterial systems which have also been the subject of research by others, including FDA workers, since the 1960's. FDA's activities in this field have been substantial and have included tests of many compounds, including some on the so-called GRAS list, by various systems as those systems have evolved over the past decade. As we entered the phase of testing in 1970, we encountered problems in interpreting the results from some of the systems. It was not always evident whether or not we were dealing with true positives. We have always encouraged and recommended mutagenesis testing, for we believe that it is helpful to know whether or not "positive findings" are associated with particular compounds, both for developers of products and for those primarily involved in public health. The question then arises, why doesn't FDA require mutagenesis testing now ? This leads to another question: Which test should be required for what effect, and will the results be used any differently than they are now used on a voluntary basis ? The second ques- tion involves many issues. In detecting genetic hazards, it would be helpful to measure the potency of a sample of mutagenic compounds in intact experimental mammals. This would provide a better correlation of in vivo effects inherited by succeeding generations. We could thus improve our assessment of genetic risk and our extrapolations from in vitro mutagenic test systems with high sensitivity such as the Ames system. However, the whole issue of forecasting the risk to future generations from mutagenic substances on the basis of our present testing capabilities is the subject of a "white paper" by a select committee functioning under the auspices of the DHEW Toxicology Coordinating Committee. Their report (2) may well serve as a foundation for the evolution of additional test re- quirements in the future. In deciding whether or not a test such as the Ames test should be required by a Federal regulatory agency for controlling carcinogens, it must be recognized that results of a re- quired test, both positive and negative, will have immediate regulatory significance in decision-making, either alone or in conjunction with other tests. If decisions are to be made on the basis of such tests, those tests must be able to stand up under legal scrutiny by the courts. In court, all relevant evidence will be considered and all relevant testimony by experts qualified by training and experience to assess toxicologic risk to human health is admissable. The following questions are fairly typical of those directed at chemists when they testify in court as to their findings. 1. Is the test sensitive in the sense of being capable of detecting the parameters of concern? 2. Is the test specific for detecting the parameters of concern (or are other qualities also detected as positive responses, necessitating another independent test to confirm the identity of the parameter being detected)? 3. Does the test quantitatively measure the parameters of concern and what is the range of error? 4. Is the test as performed acceptable by the current standards of good scientific practice?