AUTHORS
Giuseppe Maiocco, M.D., Amanda Cashen, MD2, Sai Balasubramanian, MD, JD3, Amy Schneider4, Noupama Mirihagalle, M.D.5, Sarah S. Makhani, MD6, Suzanne R. Thibodeaux, MD7, Barbara J. Bryant, MD8, Cassandra Josephson, MD9, Elizabeth Crowe, MD, PhD10, Aaron Tobian, MD, PhD11, Jonathan R Day, PharmD, MD12, Shaji Kumar, MD13, Ruchika Goel, MD14
AFFILIATIONS
1 Mayo Clinic, Rochester, Minnesota, United States
2 Washington University in St. Louis, St. Louis, Missouri, United States
3 Southern Illinois University School of Medicine, Springfield, Illinois, United States
4 Southern Illinois University School of Medicine, Springfield, Illinois, United States
5 Southern Illinois University School of Medicine, Springfield, Illinois, United States
6 New York University Grossman School of Medicine, New York City, New York, United States
7 Washington University in St. Louis, St. Louis, Missouri, United States
8 Carter BloodCare, Bedford, Texas, United States
9 Johns Hopkins All Children's Hospital, St. Petersburg, Florida, United States
10 Johns Hopkins University School of Medicine, Baltimore, Maryland, United States
11 Johns Hopkins University School of Medicine, Baltimore, Maryland, United States
12 Mayo Clinic, Rochester, United States
13 Mayo Clinic, Rochester, Minnesota, United States
14 Carter BloodCare, Bedford, Texas, United States, Southern Illinois University School of Medicine, Springfield, Illinois, United States
ABSTRACT
Background: Chimeric antigen receptor T-cell therapy (CAR-T) utilization has rapidly expanded across hematologic malignancies since first approval in 2017. Clinical experience, toxicity recognition, and inpatient delivery models have evolved concurrently. We present national data describing five-year trends in inpatient CAR-T utilization and outcomes.
Methods: Hospitalizations from 2019-2023 were evaluated using the National Inpatient Sample, the largest all-payer US inpatient database. Hospitalizations involving CAR-T infusion were identified using ICD-10-PCS procedure codes. National estimates were generated using discharge weights. Annual utilization was calculated as CAR-T hospitalizations per 100,000 total NIS hospitalizations. Outcomes included in-hospital mortality, length of stay (LOS), inpatient complications and resulting interventions including tocilizumab (ICD code effective 2021), cytokine release syndrome (CRS; ICD code effective 2021), and immune effector cell-associated neurotoxicity syndrome (ICANS; ICD code effective 2022).
Indications were identified from coded diagnoses and are not mutually exclusive. Trends were assessed using Stata 18.0 with survey-weighted regression models and calendar year modeled as a continuous variable. CRS was assessed 2021-2023 and ICANS 2022-2023 to capture full calendar years after respective ICD-10 coding implementation.
Results: Between 2019-2023, an estimated 17,785 hospitalizations received inpatient CAR-T administration nationally. Annual CAR-T associated hospitalizations nearly tripled from 2,040 in 2019 to 5,585 in 2023, an increase from 5.8 to 16.6 per 100,000 hospitalizations (p-trend=0.003). Hospitalizations predominantly involved male patients (61.7%) with median age of 63 years (interquartile range [IQR] 52-70). Most common indications were diffuse large B-cell lymphoma (DLBCL; n=8,380, 47.1%), multiple myeloma (MM; n=4,010, 22.5%), acute lymphoblastic leukemia (n=1,975, 11.1%), follicular lymphoma (n=1,355, 7.6%), and mantle cell lymphoma (n=1,100, 6.2%). MM increased from 8.8% in 2019 to 33.9% in 2023. CAR-T was predominantly autologous (96.2%), with allogeneic CAR-T coded in 3.8% of hospitalizations (effective 2021 onwards).
Overall in-hospital mortality was 2.8%, decreasing from 2.9% in 2019 to 1.9% in 2023 (p-trend=0.025). Across all 5 years, the median CAR-T day of infusion was day 1 of hospitalization (IQR day 0-5). Mean overall LOS decreased from 20.7 to 14.5 days (p-trend<0.001; 5-year mean, 16.9 days).
Coded inpatient complications decreased across several domains, including febrile neutropenia (31.4% to 22.6%; p-trend=0.018), tumor lysis syndrome (13.2% to 1.6%; p-trend<0.001), sepsis or septic shock (8.8% to 4.4%; p-trend=0.003), and pneumonia (8.6% to 4.9%; p-trend=0.023). From 2021-2023, CRS (any grade) was coded in 61.5% of hospitalizations and remained stable by year (61.2% to 62.8%; p-trend=0.49; 35.1% grade 1, 20.4% grade 2, 3.0% grade 3, 1.2% grade 4, and 1.8% grade unspecified). CRS grade 2-4 was observed in 24.6% of all CAR-T hospitalizations, decreasing numerically from 26.8% in 2021 to 22.6% in 2023 (p-trend=0.06). From 2022-2023, ICANS was coded in 26.0% of hospitalizations (7.6% grade 1, 4.6% grade 2, 6.9% grade 3, 2.0% grade 4, <0.1% grade 5, and 4.8% unspecified grade). Mechanical ventilation decreased from 6.1% to 4.0% (p-trend=0.025), while vasopressor use (2.1%), dialysis (1.7%), and red blood cell or platelet transfusion (10.0%) remained infrequent and without significant changes. From 2021-2023, tocilizumab administration increased from 8.1% to 10.7% (9.1% overall; p-trend<0.001) with mean day of administration decreasing from day 7.6 to 5.7 (6.3 overall; p-trend=0.036).
Conclusion: National inpatient CAR-T utilization nearly tripled across 5 years between 2019 and 2023, accompanied by rapid growth in MM. Concurrent decreases in all-cause mortality, LOS, and severe inpatient treatment-related adverse outcomes suggest improved safety of inpatient CAR-T delivery over time, potentially reflecting evolving CAR-T products, greater center experience, and more standardized toxicity management. As NIS captures inpatient encounters only, these trends may also reflect a concurrent shift of lower-risk administrations to the outpatient setting and likely underestimate the total national CAR-T use. These findings provide evidence for expanding national use and improving safety of CAR-T across hematologic malignancies.
DISCLOSURES
Giuseppe Maiocco, M.D.: No relevant conflicts of interest to declare.