Adaptive prior variance calibration in the Bayesian continual reassessment method.

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Pharmacometrics

Background:

• Model-based designs, like the continual reassessment method (CRM), are increasingly used for Phase I clinical trials to better identify the maximum tolerated dose (MTD) compared to traditional methods.
• A key challenge with CRM is its sensitivity to the prior distribution's variance, particularly with small sample sizes.
• Existing methods for variance calibration are often limited to the trial's initiation, lacking real-time adaptation.

Purpose of the Study:

• To propose and evaluate novel methods for adaptively calibrating the prior variance throughout Phase I clinical trials using the CRM.
• To compare the performance of these new adaptive methods against existing methods that calibrate variance only at the trial's start.

Main Methods:

• Development of three distinct systematic approaches for adaptive prior variance calibration within the CRM framework.
• Simulation studies to assess the performance of the proposed adaptive methods.
• Comparison of adaptive calibration methods with conventional, non-adaptive variance calibration techniques.

Main Results:

• The proposed adaptive methods demonstrate a potential for improved performance in identifying the MTD.
• Simulation results indicate that adaptive calibration can mitigate the sensitivity of CRM to initial variance choices, especially in early-phase trials.
• Comparative analysis highlights the advantages of dynamic variance adjustment over static, pre-trial calibration.

Conclusions:

• Adaptive calibration of prior variance offers a promising advancement for CRM-based Phase I clinical trial design.
• The developed methods provide a more robust approach to MTD identification, enhancing trial efficiency and patient safety.
• Further research and validation are warranted to implement these adaptive strategies in clinical practice.