Agonizing over antagonizing: what do benzodiazepine receptor antagonists demonstrate?

Abstract

The psychopharmacology of benzodiazepine receptor (BZR) agonists and inverse agonists is rapidly expanding, due to the ever-increasing number of BZR ligands with novel pharmacological profiles (e.g., Gardner et al. 1993) and to new and modified definitions for their clinical use (e.g., Stephens 1993; Sarter and Nutt 1994). The effects of BZR ligands in humans and animals have been frequently studied in interaction with the effects of the BZR antagonist flumazenil (Ro15-1788; FLU), although the test of the antagonist effects of FLU has by no means emerged as a standard component routinely required in studies on the effects of BZR ligands (see for example, Weerts et al. 1993; Vivian et al. 1994; Stephens and Voet 1994; Dawson et al. 1995). In fact, FLU does not consistently antagonize the effects of BZR agonists or inverse agonists. For example, FLU does not reliably antagonize the effects of BZR agonists on episodic memory and motor sedation in humans (Curran and Birch 1991; Hommer et aI. 1993; Bishop and Cmxan 1995). Despite the sporadic use and mixed efficacy of la~U as a BZR antagonist, some researchers have considered FLU-induced antagonism as an essential component of a complete experimental design for studies of BZR agonists and inverse agonists. Likewise, editors have sometimes reflexively requested the inclusion of antagonist data in research publications. From a general pharmacological perspective, inclusion of a receptor antagonism component has inherent appeal. As will be discussed, however, the significance of the ability of FLU to antagonize, or not, the effects of BZR agonists or inverse agonists is completely unsettled. To the extent that FLU fails to act in accordance with a simple model of competitive antagonism, results may be not only uninformative, but misleading. As additional antagonist manipulations typically require substantial increases in the size and/or the complexity of experimental designs, such studies need to be solidly jus-