A quantitative review of the profile and time course of symptom change in schizophrenia treated with clozapine

Abstract

Contemporary analyses demonstrate an early response to antipsychotic treatment in non-refractory schizophrenia. The profile of response to clozapine is unknown. We used meta-analytic and statistical procedures to examine the response profile to clozapine. We identified 19 unique, randomized, double-blind controlled clinical trials with suitable time course data, representing 1745 subjects. Individual subject data were available for 419 subjects, obtained from two industry-sponsored trials. Symptom severity scores from the BPRS or the PANSS were entered into regression analyses to estimate linear and quadratic coefficients of the rate of change of symptom severity over 4 weeks. Both linear and quadratic regression coefficients for clozapine, and for comparator antipsychotics differed significantly from zero (p ≤ 0.001), indicating early response profiles. Compared with other antipsychotic arms, for clozapine the treatment response was greater (d = −0.578, p = 0.021), and the linear coefficient was steeper (d = −0.502, p = 0.042); the quadratic coefficients indicating attenuation did not differ. Analyses of 6-week data and individual subject data from non-refractory and refractory trials were consistent with the primary findings. Somewhat surprisingly, clozapine shows an early response profile, similar in pattern but somewhat larger in magnitude than other antipsychotic drugs.