Uncertainties in individual response to specific antipsychotics are a major limitation in improving treatment. Pharmacogenetics offers the opportunity to help the clinical decision process by allocating individuum-specific odds for response versus non-response. Pharmacogenetics in schizophrenia has up to now produced only a very limited number of valid results. A main reason might be the heterogeneity of samples under study and lack by standardization of treatment (e.g. mixture of first onset and chronic patients).

The German Competence Network of Schizophrenia has completed a long-term treatment protocol in first-episode patients with schizophrenia (double-blind randomized study with risperidone versus haloperidol). We explored the predictive power of disease-associated variants in disposition genes and modifier genes for schizophrenia. We detected predictive markers in the genes for dysbindin, neuregulin 1 and COMT. These variants were also associated with neuropsychological correlates of schizophrenia.

Thus we can be confident that time has come to improve the tools for prediction of response to antipsychotics.