Assessing the risks of exposures to multiple chemicals with a common mechanism of toxicity: how to cumulate?

The Food Quality Protection Act (FQPA) of 1996 requires the U.S. EPA to consider the "cumulative effects" of pesticides and other substances that have a "common mechanism of toxicity." Several different methods for combining the exposures to estimate the risk of groups of common mechanism chemicals with different potencies and exposure characteristics are critically evaluated. These are the hazard index (HI), toxicity equivalence factor (TEF), and combined margin of exposure (MOE(T)) procedures as well as the point of departure index (PODI) and cumulative risk index (CRI) methods that are the reciprocals of the HI and MOE(T) approaches, respectively. Each of these methods ideally requires, at a minimum, the availability of in vivo toxicology data for the same toxicological endpoint in the same animal species. Furthermore, all assume that the effects of the individual components in the mixture are independent in nature (i.e., are additive rather than synergistic or antagonistic) and that the dose-response functions for all compounds have a similar slope. The point of departure (POD), preferably the dose corresponding to a given effect level (e.g., the ED(10)), can be used as a measure of the relative potency of the different chemicals in the group. If appropriate exposure and toxicology data are available, and the chemicals in the group have a common uncertainty factor (UF), all the procedures yield a numerically identical result. The fact that different chemicals in the group often have different UFs raises issues for all summation procedures and, in the case of the TEF approach, the UF of the index chemical selected dictates the final result of the assessment. A major distinction between the different methods for addition is the point in the process at which uncertainty is considered. The HI and CRI approaches are problematic because they require application of policy-driven UFs (in the form of RfDs) at that stage of the process where exposure should be expressed in terms of potency. In contrast, the PODI and MOE(T) approaches require application of a single group UF(G) at the end of the risk assessment process although they will also accommodate the application of data-based adjustments earlier in the analysis. Importantly, both the PODI and the MOE(T) approaches allow policy- and data-driven UFs to be separated and thus make the process more transparent; these should be considered the methods of choice for cumulative risk assessment. Assignment of a single group UF is somewhat different from developing an UF for a single chemical and the total weight of evidence available in the group database can be used to advantage to reduce the UFs that need to be applied to the group. This larger database can also be used to refine the PODs for individual members of the group. It is important to emphasize that there remains a great deal of scientific uncertainty about how to proceed with cumulative risk assessment as described in the FQPA. The serious difficulties associated with defining "common mechanism of toxicity" and "concurrent exposure" combined with the current paucity of data and methodology required to conduct cumulative risk assessment suggest that the procedure is not yet ready for use in pesticide regulation. Copyright 2000 Academic Press.