Prognostic implications of adverse events associated with CAR-T cell therapy: a population-based global observational study

TL;DR

This study maps adverse events linked to CAR-T cell therapy using global safety data, identifying which events are most dangerous and which are less severe.

Contribution

The study provides a systematic global analysis of adverse events in CAR-T therapy, linking them to mortality and prognosis using large-scale pharmacovigilance data.

Findings

59 high-fatality adverse events were identified, including pulmonary hemorrhage and hemophagocytic lymphohistiocytosis.

31 low-fatality adverse events, such as cytokine release syndrome, were associated with better outcomes.

Cardiac and respiratory/infection-related events showed high fatality and reporting frequency, respectively.

Abstract

Chimeric antigen receptor (CAR)-T cell therapy offers a promising and transformative treatment option for patients with hematologic malignancies, with expanding potential in solid tumors and non-malignant diseases. However, it also exposes recipients to a wide spectrum of treatment-related toxicities, complicating its clinical implementation. The limited sample sizes of clinical studies hinder a comprehensive understanding of how different adverse events (AEs) may impact CAR-T treatment outcomes.

Based on two global real-world drug safety surveillance systems, this observational pharmacovigilance study identified safety signals and death-related AEs in CAR-T cell therapy. All consecutive CAR-T cell-treated cases with AEs reported to the World Health Organization’ VigiBase (as of March 2025) and U.S. Food and Drug Administration Adverse Event Reporting System (as of June 2024) were included. We evaluated disproportionate risk and death of all reported AEs by the modified reporting odds ratio method. AE fatality rates, reporting odds ratio of fatality rates among different AEs, and CAR-T cell therapy safety signals were reported.

This analysis included 12,511 CAR-T cell-treated cases in FAERS from 37 countries/regions, 2861 had a fatal outcome. Disproportionality analysis identified 266 AEs as safety signals associated with CAR-T therapy. Of these, 59 high-fatality AEs were revealed (average fatality rate 49.35% [189/383]), such as pulmonary hemorrhage, lactic acidosis, and hemophagocytic lymphohistiocytosis. Cardiac events and respiratory/infection-related complications showed notably high fatality and high reporting frequencies, respectively. We identified 31 low-fatality AEs indicating a comparatively better outcome (average fatality: 11.47% [710/6188]), such as cytokine release syndrome, bradyphrenia, and hypocalcemia, which were predominantly neurologic or immune-related. These prognostically significant AEs were also borne out as death signals for CAR-T cell administration in 5555 CAR-T cell-treated cases from VigiBase.

This work establishes a systematic mapping of prognostically significant AEs in CAR-T cell therapy, providing a data-driven resource to inform future research and toxicity management strategies. Given the inherent limitations in pharmacovigilance data, including potential reporting bias and temporal ambiguities, the identified signals require further research to validate causality.

This study was funded by the 10.13039/501100012166National Key Research and Development Program of China (2022YFC2804300), the 10.13039/501100001809National Natural Science Foundation of China (82222047, W2411079), the 10.13039/501100003399Science and Technology Commission of Shanghai Municipality (22XD1423100), and the 10.13039/100017950Shanghai Municipal Health Commission (2022XD057, 2024ZZ1008).

Figures