[18 Fluorine(F)]DOPA Determinants and Predictors of Treatment Response in Psychosis

Schizophrenia is amongst the leading causes of global disability in adults. A major factor underlying this is that about 30% of patients show little or no response to first-line antipsychotic drugs. There is one drug, clozapine, with proven efficacy in these patients. However, currently there are no good predictors of treatment non-response and consequently patients have to undergo empirical trials with first-line drugs. This contributes to the long delays, on average 4-5 years, seen in identifying and starting patients on clozapine. Furthermore, clozapine is poorly tolerated and has potentially life-threatening side-effects, which mean that the investigators desperately need new, alternative drugs. Lack of understanding of the neurobiological basis underlying non-response has impeded the development of alternatives to clozapine in the past. However recently it has been shown that non-responders show reduced dopamine synthesis capacity relative to patients who have responded to antipsychotics. The effect size for this difference is very large, d\>1.2. This study was cross-sectional, in patients who had already received antipsychotic treatment for a number of years. The key questions now are thus: 1. is dopamine synthesis capacity different at illness onset in drug naïve patients who subsequently show non-response to antipsychotic treatment relative to drug naïve patients who respond to treatment 2. is it possible to predict who will respond to treatment To test this the investigators are going to investigate the relationship between presynaptic dopamine dysfunction and antipsychotic responsiveness in a prospective study. For this, the investigators are going to measure striatal dopamine synthesis capacity using \[18 fluorine(F)\]DOPA positron emission tomography in drug-naïve first episode psychosis and determine treatment response after 6 weeks of treatment with amisulpride. Response will be defined as a \>30% reduction in symptom ratings on the Positive and Negative Syndrome Scale. The effect size in our cross-sectional study was d=1.3. Based on this effect size a sample size of 12 per group will have \>80% power to detect a group difference with p\<0.05 2-tailed using an independent t-test. Given a non-response rate of 30% the investigators will thus require 40 patients at baseline to get 12 non-responders. To allow for 20% drop-outs we will require 50 patients at baseline.