Human toxicity from drugs is predicted in over 70% of cases by pre-clinical studies. Some toxicity is easier to predict than others. It is likely that prediction could be improved by concentrating on developing topics and problem areas, e.g. skin toxicity and hypersensitivity reactions are predicted poorly. Toxicokinetic data, particularly for highly metabolised drugs, might identify the most appropriate species to study. At present it appears that two species—one non-rodent—are necessary to maximise the predictive value of pre-clinical studies.
Was this article helpful?