Pimozide for schizophrenia or related psychoses

Pimozide was first formulated in the late 1960s and continues to be marketed for the care of those with schizophrenia or related psychoses such as delusional disorder. Pimozide is generally well tolerated apart from extrapyramidal side effects. It has, however, been associated with cardiotoxicity and sudden unexplained deaths and electrocardiogram monitoring is now required before and during its use.

To assess the effects of pimozide for people with schizophrenia, non-affective psychotic mental illness and delusional disorder in terms of clinical, social and economic outcomes.

Electronic searches of Biological Abstracts (1982-1995), The Cochrane Schizophrenia Group's Register, EMBASE (1980-1995), Janssen-Cilag UK's register of studies (1999), MEDLINE (1966-1995), PsycLIT (1974-1995), hand-searching the references of all included studies and contacting the manufacturers of the compound.

All randomised trials relating to people with schizophrenia, or similar disorders comparing pimozide to other drug treatments were sought. Studies where randomisation was implied rather than stated were included if they did not change the results. Primary outcomes were clinically significant change in global function, mental state, relapse, hospital admission, death, adverse events and acceptability of treatment.

Studies were selected, rated and data extracted. For dichotomous data Relative Risks (RR) based on a random effects model with 95% confidence intervals (CI) were estimated. The number needed to treat statistic (NNT) was calculated where indicated. Analysis was by intention-to-treat.

This review currently includes 34 studies focusing on those with schizophrenia, none on people with delusional disorder. Few people have been randomised to pimozide versus placebo. Data from two longer term studies does suggest that the active drug prevents relapse (n=66, RR 0.45 CI 0.24-0.86, NNT 4, CI 3-22) but the confidence interval is wide. Pimozide has similar efficacy to that of typical antipsychotic drugs for the outcomes of change in global functioning, mental state, relapse and leaving the study early. People allocated to pimozide did not have a higher mortality than those taking other antipsychotics. Pimozide was more likely to cause parkinsonian tremor (RR 1.6 CI 1.1-2.3, NNH 6 CI 3-44) and lead to a requirement for antiparkinsonian medication more frequently (RR 1.8, CI 1.2-2.6, NNH 3 CI 2-5) than other drugs. It was, however, less likely to cause sedation (RR 0.38 CI 0.2-0.7, NNH 6 CI 4-16).

Although there are shortcomings in the data there is enough overall consistency, over different outcomes and time scales, to confirm that pimozide is a drug with similar efficacy to other more commonly used antipsychotics such as chlorpromazine for those with schizophrenia. There are no data to support or refute its use for those with delusional disorder.