Adaptive decision making in a lymphocyte infusion trial.

Area of Science:

• Hematology
• Clinical Trial Design
• Immunotherapy

Background:

• Patients with hematologic malignancies often relapse after allogeneic bone marrow transplant.
• Targeted immunotherapy followed by donor lymphocyte infusion is a potential treatment for relapsed disease.
• This treatment carries risks of severe liver toxicity and myelosuppression.

Purpose of the Study:

• To determine the optimal infusion time for targeted immunotherapy and donor lymphocyte infusion.
• To maximize the probability of treatment success, defined by survival and recovery without severe toxicity.
• To develop an adaptive Bayesian design for clinical trials in this patient population.

Main Methods:

• An adaptive Bayesian design incorporating a parametric model for toxicity, white blood cell recovery, and survival.
• An algorithm for between-patient immunotherapy dose de-escalation based on observed toxicity.
• Adaptive randomization to five infusion times based on posterior success probabilities.

Main Results:

• The adaptive design reliably identifies the optimal infusion time.
• The design allocates a greater proportion of patients to superior infusion times.
• Simulations demonstrate the design's effectiveness in optimizing treatment allocation.

Conclusions:

• The proposed adaptive Bayesian design is effective for optimizing immunotherapy and lymphocyte infusion timing in relapsed hematologic malignancies.
• This approach balances treatment efficacy with toxicity management.
• The design offers a more efficient and patient-centered clinical trial methodology.