Prediction of therapeutic uses of psychotherapeutic drugs from experiences with normal volunteers

One of the great paradoxes with psychotherapeutic drugs is that we are caught in a circular trap. We know the pharmacologic properties of sedatives and stimulants so well that we can identify some of these (largely those that are nontherapeutic) in normal subjects and make reasonably good predictions about drugs of this type. Antipsychotic and tricylic antidepressants have novel clinical actions unlike any we had before. Both types of drugs were discovered fortuitously in the clinic so that their pharmacologic properties were correlated retrospectively with their therapeutic effects. To make matters worse, pharmacologic effects of these drugs in most animals and human beings have little apparent relevance to the clinical disorders they are supposed to treat. Knowing these correlations, we can devise screening batteries to detect many other drugs similar to those that we now have, but none that are truly different. Until we have better human or animal models for the disorders in question or a better hypothctic basis for constructing drugs to treat these disorders, we may be stymied. It does not seem likely that one can predict these effects in any useful or innovative way fram the study of such drugs in normal, asymptomatic human subjects.