Bringing the Bayesian Method to Clinical Trials: Q&A with Dr. Stacy Lindborg

Key Takeaways

• FDA draft guidance authorizes Bayesian designs in drug trials
• Bayesian methods enable adaptive trials and smaller sample sizes
• Imunon evaluated borrowing historical data for IMNN‑001 Phase 3
• Simulations showed modest power gain, not justifying extra effort
• Regulatory acceptance may accelerate oncology drug development

Summary

The FDA issued a January 2026 draft guidance formally recognizing Bayesian methods for drug and biologic trials, offering sponsors a clear regulatory pathway to incorporate prior data and adaptive designs. Imunon CEO Dr. Stacy Lindborg explains how the guidance could boost innovation, citing its advanced ovarian‑cancer therapy IMNN‑001 as a case study. Bayesian approaches enable flexible interim analyses and potential sample‑size reductions, but Imunon’s simulations showed only modest power gains, leading the company to retain a conventional size while using Bayesian monitoring. The agency stresses rigorous pre‑specification and transparency, positioning Bayesian tools as complementary rather than a replacement for frequentist methods.

Pulse Analysis

The FDA’s January 2026 draft guidance marks a watershed moment for statistical methodology in drug development. By formally recognizing Bayesian approaches, the agency offers sponsors a clear regulatory pathway to incorporate prior information, real‑world evidence, and adaptive decision rules into pivotal trials. This shift reflects years of experience with Bayesian designs in medical devices and rare‑disease studies, and it leverages modern computing power that now makes extensive simulation feasible. Regulators emphasize that Bayesian methods complement, rather than replace, traditional frequentist analyses, demanding rigorous pre‑specification and transparent operating‑characteristic evaluation.

For biotech firms like Imunon, the guidance translates into tangible design choices. The company’s ovarian‑cancer immunotherapy IMNN‑001 is being positioned for a Phase 3 registration trial where Bayesian borrowing from its 112‑patient Phase 2 dataset was considered. Extensive simulations showed only a modest two‑percentage‑point increase in statistical power, insufficient to offset the added regulatory and operational complexity. Instead, Imunon opted for a conventional sample size while retaining Bayesian interim analyses to monitor predictive probabilities of success and safety, thereby preserving flexibility without compromising rigor.

The broader impact of the FDA’s Bayesian endorsement could reshape oncology development pipelines. Smaller patient populations, high unmet need, and costly biologics make adaptive, learning‑oriented trials especially attractive, potentially shortening timelines and reducing expenditures. However, successful implementation hinges on robust prior selection, transparent justification, and extensive simulation—areas where many sponsors still lack expertise. As more companies adopt these tools, we can expect a gradual shift toward probabilistic decision‑making, greater use of real‑world data, and ultimately faster access to innovative therapies for patients.