Amisulpride for very late-onset schizophrenia-like psychosis: The ATLAS three-arm RCT

Background: Very late-onset (aged ≥ 60 years) schizophrenia-like psychosis (VLOSLP) occurs frequently but no placebo-controlled, randomised trials have assessed the efficacy or risks of antipsychotic treatment. Most patients are not prescribed treatment. Objectives: The study investigated whether or not low-dose amisulpride is superior to placebo in reducing psychosis symptoms over 12 weeks and if any benefit is maintained by continuing treatment thereafter. Treatment safety and cost-effectiveness were also investigated. Design: Three-arm, parallel-group, placebo-controlled, double-blind, randomised controlled trial. Participants who received at least one dose of study treatment were included in the intention-to-treat analyses. Setting: Secondary care specialist old age psychiatry services in 25 NHS mental health trusts in England and Scotland. Participants: Patients meeting diagnostic criteria for VLOSLP and scoring > 30 points on the Brief Psychiatric Rating Scale (BPRS). Intervention: Participants were randomly assigned to three arms in a two-stage trial: (1) 100 mg of amisulpride in both stages, (2) amisulpride then placebo and (3) placebo then amisulpride. Treatment duration was 12 weeks in stage 1 and 24 weeks (later reduced to 12) in stage 2. Participants, investigators and outcome assessors were blind to treatment allocation. Main outcome measures: Primary outcomes were psychosis symptoms assessed by the BPRS and trial treatment discontinuation for non-efficacy. Secondary outcomes were extrapyramidal symptoms measured with the Simpson–Angus Scale, quality of life measured with the World Health Organization’s quality-of-life scale, and cost-effectiveness measured with NHS, social care and carer work loss costs and EuroQol-5 Dimensions. Results: A total of 101 participants were randomised. Ninety-two (91%) participants took the trial medication, 59 (64%) completed stage 1 and 33 (56%) completed stage 2 treatment. Despite suboptimal compliance, improvements in BPRS scores at 12 weeks were 7.7 points (95% CI 3.8 to 11.5 points) greater with amisulpride than with placebo (11.9 vs. 4.2 points; p = 0.0002). In stage 2, BPRS scores improved by 1.1 point in those who continued with amisulpride but deteriorated by 5.2 points in those who switched from amisulpride to placebo, a difference of 6.3 points (95% CI 0.9 to 11.7 points; p = 0.024). Fewer participants allocated to the amisulpride group stopped treatment because of non-efficacy in stages 1 (p = 0.01) and 2 (p = 0.031). The number of patients stopping because of extrapyramidal symptoms and other side effects did not differ significantly between groups. Amisulpride treatment in the base-case analyses was associated with non-significant reductions in combined NHS, social care and unpaid carer costs and non-significant reductions in quality-adjusted life-years (QALYs) in both stages. Including patients who were intensive users of inpatient services in sensitivity analyses did not change the QALY result but resulted in placebo dominance in stage 1 and significant reductions in NHS/social care (95% CI –£8923 to –£122) and societal costs (95% CI –£8985 to –£153) for those continuing with amisulpride. Limitations: The original recruitment target of 300 participants was not achieved and compliance with trial medication was highly variable. Conclusions: Low-dose amisulpride is effective and well tolerated as a treatment for VLOSLP, with benefits maintained by prolonging treatment. Potential adverse events include clinically significant extrapyramidal symptoms and falls. Future work: Trials should examine the longer-term effectiveness and safety of antipsychotic treatment in this patient group, and assess interventions to improve their appreciation of potential benefits of antipsychotic treatment and compliance with prescribed medication. Trial registration: Current Controlled Trials ISRCTN45593573 and EudraCT2010-022184-35. Funding: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 22, No. 67. See the NIHR Journals Library website for further project information.