Drug-induced parkinsonism is generally known as a side effect of antipsychotic drugs. Recently, the induction of parkinsonism has been reported with the use of various peripherally acting drugs such as cal cium channel blockers, antiarrhythmic agents and so on. Therefore, it is eagerly required to quantitatively assess the extent of drug-induced parkinsonism prior to the dispensing of drugs. We have developed a strategy to quantitatively predict the intensity of drug-induced parkinsonism based

more »

... n the receptor-occupan cies of the dopamine DI, D2 and muscarinic acetylcholine (mACh) receptors in vivo and in vitro. First, we as sessed the in vivo dopamine D1, D2 and mACh receptor occupancies in the striatum in mice after drug ad ministration for twenty five drugs. Simultaneously, the intensity of catalepsy induced by these drugs were as sessed in the same animals. The intensity of catalepsy can be predicted by in vivo occupancies of the dopa mine D1, D2 and mACh receptors using a pharmacodynamic model. Second, we applied this pharmacodynam ic model to predict quantitatively the risk of drug-induced parkinsonism in humans. The in vivo dopamine DI, D2 and mACh receptor occupancies of each drug in humans were estimated from the pharmacokinetic data in humans and the in vitro affinities for the receptors in mice. The estimated risk of each drug to induce parkin sonism well coincided with the incidence of parkinsonism reported during the clinical trial period. The phar macodynamic model was applicable not only to an individual drug but also to a prescription. The clinical risk of every prescription to induce parkinsonism can be quantitatively predicted by this strategy from the set of pharmacokinetic data in humans and the affinities for the receptors of each drug in the prescription. Finally, we incorporated this model into a prescription-ordering system to develop a computer-aided drug information system for rational prescription. Developed system can display the risk of parkinsonism for an individual prescription and provide essential drug informations to the physician to prescribe. The developed system may be also applicable to quantitatively predict other receptor-mediated adverse reactions of drugs.