Clinical attrition due to biased preclinical assessments of potential efficacy.

Unless it is carefully controlled, bias often distorts the results of clinical trials, usually exaggerating the magnitude of true efficacy. For that reason, procedures to limit bias have been mandated by the FDA when assessing efficacy in clinical trials. The present review shows that the effects of bias in preclinical studies are at least as large as in clinical trials, and since bias is not usually controlled in preclinical proof of concept studies, compounds that actually have little or no therapeutic potential may often be advanced into clinical trials. This possibility is supported by the fact that lack of efficacy is the single biggest reason why compounds fail in the clinic. The shift to target-based discovery during the last 10-15 years may have further increased the effects of bias on preclinical assessments of potential efficacy, and contributed to the continuing decline in clinical success rates. Procedures are available to control for bias during preclinical assessments of potential efficacy, and their use could dramatically increase clinical success rates and substantially reduce the costs of drug discovery and development.