Predictors and markers of clozapine response

Rationale: With other atypical antipsychotics now available, having predictors of clozapine response would be of considerable value, offering clinicians guidance in their decision as to when, and if, a trial of clozapine is warranted. Objective: The aim was to review existing evidence regarding identified predictors and markers of clozapine response. Methods: Relevant studies were identified through PUBMED searches (1975-June 2004) and cross-referencing of reviews and included studies. The data were summarized under two main categories: clinical (general, neurological, cognitive/neuropsychological, clozapine levels) and biological (biochemical, endocrine, genetic, metabolic, morphological, dopamine D-2 receptor occupancy). 'Reliable' predictors/markers were defined a priori as those with support of at least two independent reports that addressed overall response, with no contradictory findings to date. 'Potential' predictors/markers had the support of a single report that addressed overall response and at least one other evaluating treatment outcome but not directly addressing response status. Results and Conclusions: Higher baseline clinical symptoms and functioning in the previous years and low cerebrospinal homovanillic acid/5-hydroxyindoleacetic acid levels were identified as reliable. Three potential measures were identified: reduction of frontal cortex metabolic activity, reduction of caudate volume, and improvement in P50 sensory gating.