Prior benzodiazepine use and buspirone response in the treatment of generalized anxiety disorder

Background: An earlier preliminary report suggested that prior treatment with benzodiazepines might predict a reduced response to buspirone in patients diagnosed with generalized anxiety disorder (GAD). To confirm or refute this hypothesis, the present data analysis was conducted. Method: One large data set (N = 735) of GAD patients (DSM-III) treated with buspirone, a benzodiazepine, and a placebo was analyzed by dividing all patients into 3 prior benzodiazepine (BZ) treatment groups: no prior BZ treatment, recent (< 1 month) BZ treatment, and remote (greater than or equal to 1 month) BZ treatment. Using an intent-to-treat last-observation-carried-forward (LOCF) data set, acute 4-week treatment response was assessed in terms of clinical improvement, attrition, and adverse events as a function of these 3 prior benzodiazepine treatment groups. Results: Patient attrition was significantly higher (p < .05) in the recent BZ treatment group than in the remote and no prior BZ treatment groups with lack of efficacy given as the primary reason by patients receiving buspirone but not benzodiazepine or placebo. In the buspirone group, adverse events occurred more frequently in the recent BZ treatment group than in the remote BZ treatment and no prior BZ treatment groups. Finally, clinical improvement with buspirone was similar to benzodiazepine improvement in the no prior BZ treatment and remote BZ treatment groups, but less than benzodiazepine improvement in the recent BZ treatment group, leading to the smallest buspirone/placebo differences in improvement in the recent BZ treatment group. Conclusion: These data suggest that the initiation of buspirone therapy in GAD patients who have only recently terminated benzodiazepine treatment should be undertaken cautiously and combined with appropriate patient education.