On May 22, 2003, the draft Report on the toxicology and carcinogenesis studies of aspartame received public review by the National Toxicology Program’s Board of Scientific Counselor’s Technical Reports Review Subcommittee. The review meeting was held at the National Institute of Environmental Health Sciences, Research Triangle Park, NC.

Dr. J.R. Bucher, NIEHS, introduced a new report series consisting of studies performed with genetically modified mice. The toxicology and carcinogenicity studies of aspartame are the first report in that series. Dr. J.E. French, NIEHS, presented an overview of the Tg.AC hemizygous, p53 haploinsufficient, and Cdkn2a deficient mouse models used in this study. Dr. Bucher proceeded to describe the uses of aspartame, the rationale for study nomination, the design of the transgenic mouse studies, the pathology evaluation, and the body weight, survival, and neoplasm incidence results. For the new report series, the traditional Levels of Evidence of Carcinogenic Activity categories would be used to frame conclusions. The proposed conclusions were: no evidence of carcinogenic activity of aspartame in male or female p53 haploinsufficient mice.

Dr. Elwell, the first principal reviewer, agreed that the studies did not show evidence of a carcinogenic effect. He also stated that the report described well the relevance and uncertainties of positive and negative findings in these models. He inquired why positive controls were not used in the p53 haploinsufficient mouse study and suggested clarification that the uncertain findings in the literature suggesting brain tumors from this compound were limited to one study.

Dr. Storer, the second principal reviewer, questioned the statistical criteria used to score the micronucleus test, noting that in virtually every study some positive response was reported. He agreed with the proposed conclusions.

Dr. Piegorsch, the third principal reviewer, inquired about design considerations such as the number of animals per group and the spacing of doses, particularly for studies with negative findings. He felt the proposed conclusion statement was appropriate.

Dr. Bucher replied that for the p53 model, positive controls were not used because variations in response to positive control agents were viewed historically to be due to dosing irregularities rather than to a lack of responsiveness from the mouse model. He indicated that the procedure for scoring micronuclei was being changed from scoring slide smears to flow cytometry, which would enhance the power of the test. Dr. C.J. Portier, NIEHS, commented that one of the motivations for adopting the transgenic mouse models was to be able to use smaller numbers of animals, which would be justified for testing purposes if the background incidence rates for neoplasms prove to be low.

Dr. Elwell moved, and Dr. Storer seconded, that the conclusions be accepted as written. Dr. Roberts and Dr. Walker expressed concern that there was not sufficient historical information to determine whether the study design, particularly the number of animals per group, was adequate. Dr. French observed that in studies in these models, groups of 10 to 20 mice had been used, and considerations of study duration and potency of the chemical were also factors in eliciting positive responses. He suggested that in the absence of preneoplastic lesions, larger group sizes would not reveal any differences in response. Dr. Storer suggested there were two different questions of adequacy, one about the validity of the study protocol, and the other about the conduct and results of the study within that protocol. Deficiencies in the latter might fit the criteria for inadequate study, whereas questions about the overall validity of the study model might be addressed in other arenas, such as the Report on Carcinogens.

Dr. Carpenter suggested that some caveat be added to the conclusion reflecting reservations about the significance of a negative result in this type of study. Drs. Walker and Vore also endorsed adding some qualifying statement to the conclusions. The panel explored a variety of phrasings, and then agreed unanimously to add the following sentence: “Because this is a new model, there is uncertainty whether the study possessed sufficient sensitivity to detect a carcinogenic effect”. The subcommittee then unanimously approved the original conclusion, with this sentence added.

There was subsequent discussion about adding results from the Tg.AC hemizygous and Cdkn2a deficient studies to the abstract table, to accompany the results from the p53 haploinsufficient model upon which the conclusions were based. Following discussion, it was agreed that results from the other models could be summarized in the abstract table but with no reference to levels of evidence of carcinogenic activity.