Abstract
Background Chimeric antigen receptor (CAR) T-cell therapy has revolutionized survival outcomes for relapsed or refractory B-cell malignancies, yet its toxicities are still poorly defined in routine care. To inform surveillance and prevention strategies, we aimed to quantify post–CAR T cardiotoxicity incidence, spectrum, and its associated hospitalization burden in routine U.S. care.
Methods We performed a retrospective cohort study using the 2017-2022 Merative MarketScan® Commercial database. Adults (≥18 years) receiving CD19-directed CAR T were identified by CPT procedure codes. Continuous enrollment ≥6 months before index (CAR T administration date) and a diagnosis of diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL), mantle-cell lymphoma (MCL), acute lymphoblastic leukemia (ALL) or primary mediastinal large B-cell lymphoma (PMBCL) were required. Patients with baseline cardiotoxicity diagnoses during the 6-month pre-index washout period were excluded.
Incident cardiotoxicities comprised ICD-10 codes for acute myocardial infarction, pulmonary embolism, heart failure, cardiac arrest, serious arrhythmias, myocarditis/pericarditis/endocarditis, severe valve disease, and cardiomyopathy. Inpatient and outpatient claims were scanned from index to censoring (insurance disenrollment or study end on 31 Dec 2022) with the cardiotoxicities list. Cardiotoxicity-associated hospitalizations were defined as admissions “related to cardiotoxicities” when the principal diagnosis matched the cardiotoxicities list.
Results Among approximately 10,046 patients with B-cell malignancies in the database, 204 patients met all eligibility criteria for inclusion in the final analysis cohort. Median age at CAR T infusion was 56 years (IQR 49–60), and the majority (69%) were male. Malignancy distribution was DLBCL 77%, MCL 7%, FL 6%, and ALL 9%.
Cardiotoxicity Incidence: 65/204 patients (32%) experienced one or more cardiotoxicity diagnoses in either inpatient or outpatient settings.
Ambulatory spectrum: The most observed cardiotoxicities included pericardial effusion (12%), atrial or other specified arrhythmias (11%), drug-induced or toxic cardiomyopathy (9%), unspecified pulmonary embolism (9%), and tricuspid valve insufficiency (5%). No cases of acute myocarditis or myocardial infarction were identified.
Hospitalizations Related to Cardiotoxicity: Five patients (2% of the entire cohort) had a hospitalization with a principal-diagnosis cardiotoxicity event, including 1 ventricular arrhythmia, 3 pulmonary embolism events, and 1 supraventricular tachyarrhythmia or cardiac arrest. The mean length of stay was 4 days (IQR 1–8).
Conclusions In this real-world cohort of adult CAR T recipients, nearly one-third developed a serious cardiotoxic event post-infusion, with a small proportion hospitalized specifically for cardiotoxicity. The broad spectrum of cardiotoxic presentations highlights the need for systematic cardiovascular monitoring and individualized risk assessment in CAR T recipients. Given the heterogeneous coding practices, potential unmeasured confounders, and inherent database limitations, complementary prospective studies are warranted to more accurately estimate the cardiovascular risk in this population.
