Predictive toxicology in drug development.

Abstract

A critical issue in drug development remains the failure to identify toxicological problems in new chemical entities sufficiently early in development to avoid the expense of terminating drugs in late-stage clinical development. The Society of Chemical Industry's BioActive Sciences Group arranged a one-day meeting in London, United Kingdom, on February 27, 2003, devoted to progress in the development of in silico modeling and expert systems. The meeting provided an assessment of the progress that has been made in the development of models that can be used to evaluate chemical libraries at the design stage, or to predict potential toxic effects while chemical projects are still in the lead optimization stage. Although success in developing such models might reduce the number of new chemical entities progressing to clinical development, its potential for reducing the high attrition rate (approximately 50%) in late-stage (phase II and later) clinical development could produce a substantial cost saving and might also shorten the time frame of drug development.