First dose of potential new medicines to humans: how animals help

Abstract

The aim of a first study of a new drug in humans is to explore in a safe and ethical manner the dose and exposure range that is well tolerated, and, if possible, to identify any dose-limiting adverse events. Achievement of these aims represents a major leap from the laboratory bench to humans. It requires a substantial body of information characterizing the test substance, which can only be derived from animal studies. Despite these activites being regulated by many national and international guidelines, the approach to preclinical studies remains largely empirical. There is a paucity of evidence about the performance of the widely employed preclinical tests in prediction of toxicity in humans. There is a need for much better performance data in this area. The available limited, retrospective evidence indicates that the conventional approach using experimental pharmacology alongside toxicity studies of one month's duration reasonably predicts adverse events in the first human studies. The conventional methods identify more than 90% of toxicities that can be detected in animals. If toxicity studies are shorter than one month, there is a risk of certain organ toxicities being overlooked. However, single studies seem to have the capacity to detect many of the most important potential adverse effects. Data obtained from dog studies are frequently better predictors than data from rodent experiments. Although uncommon, serious idiosyncratic drug reactions involving skin, liver and haemopoiesis — which conventional animal studies usually fail to predict — are major problems in drug development.