induce cancer through mechanisms that do not occur at lower doses, thereby generating false-positive inferences of hazard and risk for humans who are exposed at lower doses. The committee was given pharmacokinetic and other mechanistic arguments, such as induced cell proliferation, that support this hypothesis.
Even in cases where effects might occur as a result of low dose exposure, the results of an MTD test might have little utility in defining the dose-response relationship. Some agents could have nonlinear dose-response relationships that reflect pharmacokinetics, induced cell proliferation, or other mechanisms. The result of the nonlinearity could be overestimation (or, in some cases, underestimation) of low dose risks. Overestimation could occur where the dose-response curve has a shallow slope at low doses and becomes markedly steeper at higher doses. Underestimation could occur where the dose-response curve flattens out or curves downward at high doses.
Statistical analysis of bioassay results for many agents has shown strong correlations between estimates of carcinogenic potency and measures of toxicity (including the MTD) that suggest that carcinogenicity is inherently related in some way to other toxic effects produced by a chemical. However, some investigators have concluded that those correlations, and possibly estimates of carcinogenic potency, are determined in some way by the bioassay design or the mathematical and statistical methods used to estimate potency and investigate the correlations, rather than by inherent biologic properties of the agents.
SCOPE OF REPORT
The above points are all addressed in various degrees in this report. Particular attention is focused (in Chapter 2) on the fourth point—questions concerning the observed correlations between measures of carcinogenic potency and the MTD. The report explores the extent to which the correlations appear to reflect some underlying biologic reality, as opposed to being determined solely by experimental design or statistical methods. It further considers the relationship of the correlations to possible biologic mechanisms of carcinogenesis and the implications of the correlations for risk assessment.
The report discusses both what bioassays conducted at the MTD can
tell us and what they cannot tell us, qualitatively and quantitatively, regarding carcinogenic hazard in humans (Chapter 3). Several proposals are discussed (Chapter 4) for modifying the design of the bioassay, for modifying the process of selecting chemicals for testing, and for augmenting the results of the bioassay with additional testing to improve risk assessments.
The committee's conclusions are presented in Chapter 5 with the recommendations of the majority of the committee concerning the better use of bioassays, specific results from bioassays, and other types of data to assess carcinogenic hazards in human populations. The dissenting recommendations of a minority of the committee are also described.
An active discussion is in progress in the scientific community concerning the extent to which high doses produce increased mitogenesis (cell division) and how much the increase contributes to the incidence of cancer at the MTD and lower doses. That was the subject of a presentation given at the MTD workshop conducted by the committee (see the summary of the workshop at the end of the report). This report reviews and evaluates the recent research related to the issue.
Although the MTD concept is used in other contexts (e.g., tests for reproductive toxicity and teratogenicity), the committee essentially limited its investigation to the use of the MTD in bioassays for carcinogenicity associated with exposure to chemicals.