The Lancet Psychiatry IPD analysis of 2,079 patients across six antipsychotic trials found that PANSS symptom domains improve at different speeds, with most gains early and a plateau around week 15. Hostility/excitability fell 76% (median response ~3 weeks); positive 64% (~4 weeks); anxiety/depression 57% (~4 weeks); cognitive/disorganisation 61% (~8 weeks); and negative 50% (~8 weeks). The authors recommend allowing more time before labeling patients as "non-responders" or making premature treatment changes, and suggest trial timing be reconsidered alongside regulators' focus on total-score endpoints.
Schizophrenia “Non-Responders” May Just Need More Time, Study Suggests
The paper evaluated patients from six clinical trials to look over their response times. Credit: Andrii Yalanskyi / Shutterstock.com
A new analysis of individual participant data (IPD) from six acute antipsychotic trials suggests clinicians and trial teams should allow more time before declaring patients with schizophrenia as "non-responders." Published in Lancet Psychiatry, the study maps how different symptom domains on the Positive and Negative Syndrome Scale (PANSS) change over the first six to 12 months of treatment.
Study and Methods
The authors pooled IPD for 2,079 patients treated with antipsychotics across six clinical trials to examine temporal patterns of symptom change. The analysis focused on five PANSS domains and tracked improvement trajectories week by week, identifying when most gains occurred and when symptom scores tended to plateau.
Key Findings
All five PANSS domains showed substantial improvement early in treatment, with most benefit during the first weeks and a general plateau emerging from about week 15. However, the speed of improvement varied markedly by domain:
- Hostility and excitability: 76% reduction from baseline within six months (median time to response ~3 weeks).
- Positive symptoms: 64% reduction (median time to response ~4 weeks).
- Anxiety and depression: 57% reduction (median time to response ~4 weeks).
- Cognitive and disorganisation symptoms: 61% reduction (median time to response ~8 weeks).
- Negative symptoms: 50% reduction (median time to response ~8 weeks).
Regulatory Context and Implications
Regulatory guidance from the European Medicines Agency (EMA) and the US Food and Drug Administration (FDA) traditionally emphasises total-scale endpoints such as the PANSS total score or the Brief Psychiatric Rating Scale (BPRS) total as primary short-term efficacy measures. The EMA guidance tracked by RAPS recommends short-term assessments on total-scale instruments, and the FDA's 2023 benefit–risk framework evaluates a drug's overall clinical profile without requiring domain-level temporal disaggregation in protocols.
The study's authors say these endpoint choices remain appropriate but argue that trial protocols and clinical practice should better account for differing timelines across domains. Specifically, they advise against prematurely labeling patients as "non-responders," switching treatments, or discharging trial participants when positive symptoms have improved but negative or cognitive/disorganisation symptoms have yet to show meaningful change.
Broader Context
Separately, GlobalData reported in December 2024 that the schizophrenia drug market across the seven major markets (US, France, Germany, Italy, Spain, UK and Japan) was valued at $8.4 billion in 2021 and is forecast to reach $17 billion by 2031. Their forecast expects late-stage pipeline products to capture roughly 33.4% of global sales in 2031, while atypical antipsychotics are expected to retain the largest patient share. GlobalData is the parent company of Clinical Trials Arena, which originally published the article.
Takeaway
Patients showing early improvement in positive symptoms may still require additional weeks to show meaningful gains in negative or cognitive/disorganisation domains. Allowing slightly longer observation periods could reduce premature treatment changes and better reflect true clinical response.
Help us improve.
