De-Risking Schizophrenia Trials: Strategies for Improving Signal Detection

Schizophrenia clinical trials face a daunting reality: rising placebo response rates are eroding drug-placebo separation, threatening years of development work and billions in investment.  

In a recent expert panel discussion, Dr. Jeffrey Lieberman, Dr. Alan Kott, and Dr. Chris Murphy explored the operational and scientific strategies that can help sponsors navigate what's become known as the "valley of death" in drug development. 

Key Theme 1: Understanding Why Promising Drugs Fail in Phase 3 

One of the most painful scenarios in drug development occurs when a compound shows robust Phase 2 results but fails to separate from placebo in Phase 3.  

"The question isn't whether your RTSM can handle complex randomization. The question is: Are you ready to derisk the transition from Phase 2 to Phase 3?" explained Dr. Lieberman.  

The Dose-Response Dilemma 

A critical but often overlooked factor is inadequate dose-response mapping. Dr. Lieberman pointed to pomaglumetad as a cautionary tale: "Lilly used 80mg daily in Phase 3, but their own data showed you needed at least 160mg to achieve the target mechanism. They literally missed the signal by using the wrong dose." 

Invest adequate time in Phase 2 to map the complete dose-response curve, including maximum tolerated dose and minimum effective dose. Don't rush into Phase 3 without confidence in your dosing strategy. 

Key Theme 2: Data Quality Issues Drive Placebo Response 

Dr. Kott's presentation revealed a sobering statistic: approximately 27% of data in schizophrenia trials is affected by clinically meaningful data quality concerns.  

More critically, these concerns aren't randomly distributed—they cluster at specific sites and directly impact signal detection. 

High-impact data quality issues include: 

1. Lack of diagnostic confidence: Not just misdiagnosis, but inappropriate acuity levels for acute trials
2. Aberrant variability: Erratic ratings that introduce noise
3. Baseline inflation: Artificially elevated entry scores leading to pseudo-improvement
4. Within and between-scale discordances: Inconsistent severity assessments
5. Data fabrication: Rare but devastating when present 

Data quality concerns identified during screening predict post-randomization problems. Implementing early monitoring and intervention during the screening phase offers a critical opportunity to protect trial integrity. 

Key Theme 3: The Anxiety Marker as an Acuity Indicator 

Dr. Kott presented compelling evidence that anxiety symptoms at baseline may serve as a marker of true acute exacerbation. In Karuna's CAR-XT trials: 

Phase 2 findings: 

• Patients WITH anxiety at baseline: Strong drug-placebo separation
• Patients WITHOUT anxiety at baseline: Placebo outperformed drug

Phase 3 confirmation: 

• Patients WITH anxiety: Maintained strong separation
• Patients WITHOUT anxiety: No separation (both arms showed diminished response) 

"The absence of anxiety may question the actual acute exacerbation," Dr. Kott explained, suggesting that subjects entering trials without expected anxiety symptoms may not represent the target population. 

External Factors Impact Trial Quality 

Data from Ukraine demonstrated how external circumstances affect trial quality. Following the 2022 geopolitical conflict, anxiety and depression scores measured by PANSS sharply increased in Ukrainian sites compared to stable US trajectories—a stark reminder that trial quality is influenced by real-world events. 

Key Theme 4: Predict, Restrict, and Intervene Strategies 

Dr. Murphy outlined a three-pronged approach to risk mitigation: 

1. Predict and Restrict: Screen Out Inappropriate Subjects
2. Intervene: Manage Expectation Bias

"When sites and participants are excited about a promising new compound, enthusiasm can intrude upon recruitment and assessment," Dr. Murphy explained. 

3. Monitor and Remediate: Maintain Rating Reliability

Beyond Dopamine: Emerging Opportunities 

While the webinar focused heavily on operational de-risking, Dr. Lieberman highlighted exciting therapeutic frontiers: 

Unmined Therapeutic Territories: 

• Negative symptoms (no drug clearly superior to placebo yet)
• Cognitive symptoms (untapped therapeutic area)
• Preventing disease progression
• Treatment-resistant symptoms beyond clozapine
• Side-effect-free formulations 

"The pathophysiologic dimensions of schizophrenia—psychosis, negative symptoms, cognitive symptoms—are underpinned by different mechanisms," Dr. Lieberman said.  

The Bottom Line: Operational Excellence Matters as Much as Mechanism 

Schizophrenia drug development success rates underscore the challenge: only 7.3% of Phase 1 compounds reach approval, 26% from Phase 2, and 56% from Phase 3. With the financial stakes measured in billions and the clinical need urgent, sponsors cannot afford to lose viable compounds to preventable operational failures. 

"There's great opportunity, albeit with a given degree of risk," Dr. Lieberman concluded. "The whole point is to help you understand the key things that can derisk your efforts." 

The message is clear: In an era of rising placebo response and commoditized trial infrastructure, competitive advantage comes from operational rigor—comprehensive screening, proactive monitoring, and data-driven intervention strategies that protect drug signal from the noise of poor-quality data. 

Expert Panel 

Dr. Jeffrey Lieberman - Pioneer in schizophrenia pathophysiology and pharmacology; key role in clozapine characterization and CATIE trials 

Dr. Alan Kott - International authority on blinded data patterns, error variance, and placebo response; with Signant Health since 2005 

Dr. Chris Murphy - Decades of experience in rater training, ratings quality assessment, and optimal patient selection for clinical trials 

Dr. David Daniel - Scientific and clinical leader for CNS solutions; 30+ years psychiatric trial expertise; extensive publications and treatment patents in epilepsy, anxiety, and psychotic disorders