Growing placebo response rate: the problem in recent therapeutic trials?

Recently the development of several promising new compounds for anxiety and depression was discontinued because of difficulty demonstrating therapeutic effects. This article explores alternatives to "increasing placebo response rate" as explanations. We reanalyzed a study of 81 panic patients treated with placebo, alprazolam 2 mg or 6 mg, or imipramine 225 mg daily to investigate the effect of baseline pathology and selective effects of treatment on biological and cognitive components of panic disorder. The regression of endpoint on baseline number of spontaneous panic attacks differed among treatment groups, with lower slopes for the more active compounds. Only patients with many panic attacks at baseline benefited from the active compounds. Also, treatment effects declined progressively on the more cognitive aspects of the disorder (situational panic attacks and phobia ratings) for alprazolam and were entirely absent for imipramine. Implications for the etiology of panic disorder, its treatment, and therapeutic research are discussed.