Real-world effectiveness of aripiprazole once-monthly REACT study: Pooled analysis of two noninterventional studies

Abstract

Background: Noninterventional naturalistic studies are an important complement to randomized controlled trials. Aripiprazole once-monthly (AOM) is an atypical antipsychotic in a long-acting injectable formulation.

Methods: A pooled analysis of two noninterventional studies was undertaken to validate previous results on AOM effectiveness and safety in a larger population and improve statistical power for preplanned subgroup analyses. We analyzed data from 409 patients with schizophrenia who were treated with AOM and were enrolled in noninterventional studies in Germany (via noninterventional studies registry 15,960 N) and Canada (NCT02131415). Data collected at baseline, 3 and 6 months were analyzed. Among the endpoints were psychopathology (brief psychiatric rating scale [BPRS]) and disease severity (clinical global impression [CGI]).

Results: Mean patient age was 38.9 (SD 14.8) years, and 59.9% were male. BPRS decreased from 48.1 (SD 15.6) at baseline to 36.5 (SD 13.7) at month 6 (p < 0.001). CGI decreased from 4.47 (SD 0.90) at baseline to 3.64 (SD 1.16) at month 6 (p < 0.001). A total of 54.4% were responders (at least 20% reduction) on the BPRS, and 56.5% had a CGI-S-score that was at least 1 level better than baseline. A total of 43.4% were considered responders on both the BPRS and CGI scales. A total of 45.2% were considered in remission. Adverse events were rare and corresponded to the previously known safety profile of AOM.

Conclusions: Treatment with AOM for patients with schizophrenia appeared effective and safe under real-life conditions.

Keywords: Aripiprazole once-monthly; long-acting injectable; noninterventional study; real-world evidence; schizophrenia.

Figures

Figure 1.

BPRS total score. (A) Change of BPRS total score in the total population (n = 395). (B) Change of BPRS total score in patients ≤35 years (n = 194) and >35 years (n = 201). Error bars represent standard deviations. Missing data were imputed using the last observation carried forward method.

Figure 2.

Proportion of responders. Patients were considered responders if they showed an improvement of at least 20% of the BPRS total score.

Figure 3.

Proportion of remitters at month 6. Patients were considered remitters if they had a score of 3 or less on the following BPRS items for at least 3 months, that is, two consecutive visits: grandiosity, suspiciousness, unusual thought content, hallucinatory behavior, conceptual disorder, mannerisms, and blunted affect.

Figure 4.

CGI. (A) Change of CGI in the total population (n = 395). (B) Change of CGI in patients ≤35 years (n = 193) and >35 years (n = 202). Error bars represent standard deviations. Missing data were imputed using the last observation carried forward method.

Figure 5.

Proportion of patients with defined changes of CGI-S scores between baseline and month 6.

Figure 6.

Correlation of BPRS total score and CGI-S score. (A) Correlation at baseline. Pearson correlation coefficient is 0.65. (B) Correlation at month 6. Pearson correlation coefficient is 0.67. Blue solid line shows regression, blue shaded area shows 95% confidence limits.

Figure 7.

Proportion of responders on both the BPRS and CGI scales. Patients were considered responders if they showed an improvement of at least 20% of the BPRS total score and an improvement in CGI of at least one level.