A Single, Interpretable Vocal Biomarker for Enriching Antipsychotic Clinical Trials

Area of Science:

• Psychiatry
• Clinical Neuroscience
• Biomarkers

Background:

• Speech latency, an objective measure of verbal response time, is sensitive to cognitive, social, and motivational factors.
• Previous research indicates speech latency can enrich participants in antidepressant trials, nearly doubling the drug-placebo effect.
• This study investigates speech latency as an enrichment tool in a schizophrenia clinical trial.

Purpose of the Study:

• To evaluate if speech latency enrichment improves outcomes in a schizophrenia clinical trial.
• To assess the utility of speech latency as a predictive biomarker for treatment response.

Main Methods:

• Analysis of 2,320 audio recordings from psychiatric interviews of 406 participants in a Phase 3 schizophrenia trial.
• Participants were from three countries, spoke eight languages, and received brilaroxazine monotherapy.
• Speech latency was measured and analyzed for internal consistency, stability, and correlation with negative symptoms.

Main Results:

• Speech latency demonstrated excellent internal consistency and temporal stability.
• Patients with moderate-to-severe negative symptoms exhibited significantly longer speech latencies (Cohen's d=0.83 to 1.47).
• A vocal biomarker (VBM-pos) derived from speech latency identified participants with longer latencies, in whom brilaroxazine showed statistically significant and large treatment effects versus placebo.

Conclusions:

• Speech latency is a valid, objective biomarker from standard clinical interviews.
• Utilizing speech latency as an enrichment tool reduces sample size requirements and enhances clinical trial outcomes.
• This approach has significant implications for lowering the cost and burden of clinical trials.