Modeling Multiple Agent Interactions in Behavioral Pharmacology

Abstract

Drugs of abuse frequently are self-administered in combinations, thus the study of the behavioral pharmacology of drug interactions has important public health implications. The effect-addition and the dose-addition models are predictive models frequently used to interpret such investigations. The aims of the present study were to empirically validate these predictive models using simple drug combinations. The effects of combinations of pentobarbital (PB) with itself or PB with secobarbital (SB) were studied using an inverted screen test in mice and a milk consumption test in rats. When evaluated using the effect-addition model the interactions between PB and itself were generally greater than effect-additive, suggesting that greater effects were obtained when giving the drug by two separate injections. In contrast, when evaluated using the dose-addition model, the interactions between PB and PB were dose-additive as would be expected for a drug interacting with itself. Similar results were obtained for the interaction of PB and SB using the mouse inverted screen test. These results suggest that the dose-addition model is a better model for evaluating drug interactions. The interactions observed for PB and SB using the rat milk drinking test generally were greater than effect-additive when analyzed using the effect-addition model. In contrast, using the dose-addition model, the interactions between PB and SB were generally infra-additive, suggesting that these two drugs have different mechanisms in decreasing milk consumption.