Review of: Long Y et al, BMC Medicine 2023;21:263. 

Study Type: Randomized open-label clinical trial.

We know that many patients with schizophrenia do not show meaningful clinical improvement after a full trial of antipsychotic monotherapy. If a patient is not responding, how long should we wait before switching to a different medication? Established guidelines vary widely, suggesting a range between 2 and 8 weeks, and few studies have examined how baseline severity of psychosis impacts early prediction of nonresponse. Interestingly, studies in depression have shown that early nonresponse within the first 2–3 weeks often predicts later nonresponse at 6 weeks, supporting the value of early monitoring across psychiatric conditions (Kudlow PA et al, CNS Drugs 2014;28(7):601–609).

This multicenter, 8-week, open-label randomized trial was conducted across 19 psychiatric centers in China. The study included 964 patients with schizophrenia, who were randomized to receive monotherapy with one of four atypical antipsychotics: olanzapine (average dose 17.2 mg/day), risperidone (4.6 mg/day), amisulpride (634.9 mg/day), or aripiprazole (20.4 mg/day). All participants were diagnosed with schizophrenia within the previous five years. Those with serious physical illness or substance abuse were excluded. Participants were stratified based on illness severity (mild, moderate, or severe) and were evaluated using the Positive and Negative Syndrome Scale (PANSS) at baseline, week 2, week 4, and week 8.

The primary outcome was to determine the predictive value of early nonresponse—defined as a less than 20% reduction in PANSS total scores from baseline—for later nonresponse. Various cut-off points for PANSS reduction at weeks 2 and 4 were analyzed to find the optimal predictive thresholds.

The study found that early non-response at week 2 and week 4 can indeed predict later nonresponse at week 8. Specifically:

At week 2, a reduction of less than 10% in PANSS scores was the best predictor for nonresponse in patients with moderate schizophrenia (accuracy 84%) and for those treated with olanzapine (79.2%) and aripiprazole (77.4%).

For patients with severe schizophrenia or for those treated with risperidone or amisulpride, a reduction of less than 5% at week 2 was the most accurate predictor (accuracy 75% for severe schizophrenia, 82.4% for risperidone, and 78.2% for amisulpride).

By week 4, a reduction of less than 20% in PANSS scores universally served as the best predictor across all severities and antipsychotic treatments (accuracy 89.8%–92.1%).

Overall, the results indicate that the response to treatment within the first 2–4 weeks is critical for predicting longer-term outcomes. The study also highlighted that the optimal predictive cut-off values for early nonresponse vary based on the type of antipsychotic and the baseline severity of the illness.

Carlat Take
The sooner we can predict antipsychotic nonresponse, the faster we can hope to provide relief for our patients. In clinical practice, we’re not usually measuring symptom reduction in percentage points. However, if your patient with schizophrenia shows no or minimal response to antipsychotic monotherapy after 2 weeks, and certainly by 4 weeks, it may be best to consider switching to another drug or making a change in treatment.