Early Phase Dose-Finding Designs for CAR-T Cell Therapies

Area of Science:

• Immunotherapy
• Oncology
• Clinical Trial Design

Background:

• Chimeric Antigen Receptor (CAR)-T cell therapy has transformed lymphoma and leukemia treatment, offering higher survival rates and lower toxicity.
• Traditional Phase I trial designs struggle to determine the optimal dose (OD) for CAR-T cells due to a lack of dose-limiting toxicity (DLT) and unclear dose-effect relationships.

Purpose of the Study:

• To propose a novel early-phase dose-finding design for CAR-T cell therapy.
• To accurately select the optimal dose (OD) by integrating both toxicity and CAR-T cell activity endpoints.

Main Methods:

• Utilized serial peripheral blood samples to measure CAR-T cell expansion over time as a sensitive indicator of activity.
• Employed a Bi-Exponential model within a Bayesian framework to analyze repeated cell count measures.
• Considered three activity criteria: cell counts at specific time points, duration of cell persistence, and area under the cell-expansion curve.

Main Results:

• The proposed design demonstrated high accuracy in selecting the optimal dose (OD), even with small sample sizes.
• Incorporating CAR-T cell expansion data provides a more sensitive measure of activity compared to traditional short-term clinical responses.
• The Bi-Exponential model effectively captures diverse CAR-T cell expansion dynamics.

Conclusions:

• The novel dose-finding design enhances the precision of optimal dose selection for CAR-T cell therapies.
• This approach addresses limitations of traditional trial designs in the context of novel immunotherapies.
• Accurate dose selection is crucial for maximizing efficacy and safety in CAR-T cell treatment.