Statistical Methods for Chimeric Antigen Receptor Cell Therapy Outcomes: A Review and Practical Considerations

Chimeric antigen receptor T cell (CAR-T) therapy is being increasingly used as a cutting-edge treatment for blood cancers. Although the number of scientific articles focused on its outcomes is increasing, there is limited literature providing guidance on appropriate statistical analysis methods for CAR-T therapy studies. Statistical approaches applied to CAR-T therapy outcome analysis often use methods adopted from hematopoietic cell transplantation (HCT); however, there are important differences between these therapies that must be accounted for in the outcomes analysis. Additionally, much of the literature on CAR-T therapy outcomes is focused on clinical trials, which differ from real-world settings in the long-term follow-up of these patients. There are some statistical challenges in analyzing CAR-T therapy data, such as defining patients' response to therapy and quantifying the duration of response. Although some outcomes, such as overall survival and treatment-related mortality, are seen in both HCT and CAR-T therapy studies, some post-therapy complications are unique to CAR-T cell therapies, and their statistical analysis should be given careful consideration. Other challenges include establishing the appropriate time origin for each outcome, determining censoring time, and defining competing risks. Here we propose to define common CAR-T therapy outcomes, describe appropriate statistical approaches for their analyses, and address these critical statistical challenges based on the observational data collected by the Center for International Blood and Marrow Transplant Research. We used observational data to illustrate these statistical methods for analyzing CAR-T therapy outcomes. © 2026 American Society for Blood and Marrow Transplantation. Published by Elsevier Inc. All rights reserved.