Human interindividual variability in susceptibility to toxic effects: from annoying detail to a central determinant of risk.

It is unusual to find variability issues as the central focus of a scientific conference. The discussion below first suggests why variability has often been an "annoying detail" in both basic animal toxicology and the human testing of new drugs. Then it gives some reasons why improved quantitative variability information is likely to be important. Better definition of the sources and magnitude of variability in susceptibility in the human population is a central issue for, (1) making more quantitative estimates of both cancer and non-cancer risks from occupational and environmental exposures, and (2) designing protocols for the use of drugs that maximize benefits for the risks incurred in a diverse patient population. Finally, it offers some suggestions about how better variability information is to be obtained and/or extracted from existing information.