A double-blind randomized placebo-controlled study of relapse prevention in remitted first-episode psychosis patients following one year of maintenance therapy

Abstract

Background: Currently, there is no consensus regarding how long antipsychotics medication should be continued following a first/single psychotic episode. Long-term maintenance therapy carries side-effects and may not be necessary for some patients. Clinically, patients often request discontinuation after a period of remission. This is one of the first double-blind randomized-controlled studies designed to address the issue.

Methods: Patients with DSM-IV schizophrenia and related psychoses (excluding substance induced psychosis) who remitted well following a first/single-episode and had remained well on maintenance medication for 1 year were randomized to receive either maintenance therapy with quetiapine (400 mg/day), or placebo for 12 months. Relapse was defined by the presence of (i) an increase in at least one of the following Positive and Negative Syndrome Scale (PANSS) psychotic symptom items to a threshold score (delusion, hallucinatory behaviour, conceptual disorganization, unusual thought content, suspiciousness); (ii) Clinical Global Impression (CGI) Severity of Illness 3 or above and (iii) CGI Improvement 5 or above.

Results: 178 patients were randomized. 144 patients completed the study (80.9%). The relapse rate was 35% (31/89) for the maintenance group and 66% (59/89) for the placebo group (log-rank test, χ2 = 15.858, p < 0.001). Relapse was not related to age or gender. Other significant predictors of relapse include medication status, pre-morbid schizotypal traits, verbal memory and soft neurological signs.

Conclusions: This study provides new data for relapse risks in remitted first-episode psychosis patients following 1 year of maintenance therapy. There is a substantial risk of relapse if medication is discontinued. Maintenance medication halves relapse rate. On the other hand, 34% of patients discontinued medication and remained well. Explorations of potential predictors of relapse may help further to estimate risks so that an informed decision about medication discontinuation can be made.

Acknowledgement: This study was supported by an investigator-initiated study award from AstraZeneca and the Research Grants Council, Hong Kong (Project number: 765505).