Clinical and economic burden of cytokine release syndrome in a nationally representative cohort of patients receiving CAR-t therapy

Abstract Introduction: Chimeric antigen receptor (CAR) T-cell therapy has emerged as an effective option for several hematologic malignancies, though associated toxicities like Cytokine Release Syndrome (CRS) can complicate clinical management. The implementation of a specific ICD-10-CM code for CRS (D89.83) in late 2020 enables large-scale real world studies. This study leverages national data to assess the incidence of any-grade CRS and associated complications, quantify its economic burden, and identify independent predictors of in-hospital mortality.Methods: We performed a retrospective study using the 2020 and 2021 annual datasets from the National Inpatient Sample (NIS). Adult CAR-T hospitalizations were identified using a comprehensive list of product-specific ICD-10-PCS codes. We determined the national incidence of any-grade CRS (code D89.83*) and related complications. The economic impact of CRS was assessed using an inverse probability of treatment weighted (IPTW) model comparing total hospital charges. Results: We identified a national cohort of 992 CAR-T therapy hospitalizations. The median age was 62 years, 61.6% were male, and the primary payer was Private Insurance (48.3%) or Medicare (38.3%). The overall in-hospital mortality was 3.3%, and the median length of stay was 14 days. Any-grade CRS was documented in 41.1% of patients. Among all CAR-T recipients, common complications included neutropenia (37.3%), hypotension (14.8%), and acute kidney injury (13.7%). Tocilizumab was administered to 12.0% of patients with any-grade CRS, serving as a proxy for more severe cases. Mortality rates varied by ethnicity: 4.2% among Hispanic patients (n=72), 2.6% among Black patients (n=115), and 3.3% among White/Other patients (n=805). In the IPTW-adjusted analysis, CRS was significantly associated with 35% higher total hospital charges (adjusted Rate Ratio 1.35; 95% CI 1.11-1.63; p=0.003).Conclusions: In this national analysis, any-grade CRS is documented in 41.1% of real-world CAR-T hospitalizations, a rate lower than in clinical trials, likely reflecting under-coding of milder grades. While CRS itself was not an independent predictor of mortality, it remains a significant driver of economic burden, increasing hospital charges by 35%. While CRS was not a direct driver of mortality, our observation of higher mortality rates among Hispanic patients suggests a potential racial and ethnic disparity in outcomes that requires further investigation to understand and address. Key limitations of this administrative data study include the inability to adjust for tumor volume or stage of disease.