Epidemiology and Predictors of 30-Day Readmission in CAR-T Cell Therapy Recipients

Abstract

The novel cellular immunotherapy using chimeric antigen receptor (CAR) T cells has transformed the management of several previously incurable hematologic malignancies. Since the first CAR-T cell product was approved by the Food and Drug Administration in 2017, five additional products have been approved for various hematologic malignancies. Although there is now more experience with outpatient administration, CAR-T therapy was initially delivered in an inpatient setting. The unique complications of cytokine release syndrome (CRS) and neurologic side effects (commonly known as immune effector cell-associated neurotoxicity syndrome [ICANS]), along with a higher risk for infection, increase the risk for hospital readmission. Given the recent approval of CAR-T therapy, large-scale epidemiologic data are lacking. The present study aimed to characterize the epidemiology of hospitalizations, readmissions, and factors associated with all-cause 30-day readmission post CAR-T therapy. This retrospective cohort study used the Nationwide Readmissions Database from 2017 to 2019 to identify hospitalizations for CAR-T therapy administration. A descriptive analysis was performed after categorizing these hospitalizations as non-Hodgkin lymphoma, multiple myeloma, or leukemia. The readmission rate was calculated, and etiologies of readmission were identified. A Cox proportional hazards model was used to elucidate factors associated with 30-day readmission. We also estimated the healthcare utilization related to readmissions, including total hospital charges and length of stay. The 2,964 CAR-T therapy-related admissions included 2,176 with a diagnosis of non-Hodgkin lymphoma, 344 with multiple myeloma, and 445 with leukemia. The median length of stay was 15 days. Most CAR-T therapy recipients were male (63.4%), admitted to a teaching hospital (99.2%); 49.3% had private insurance, and 33.2% belonged to the highest-income communities. CAR-T therapy was administered mostly in privately owned (89.5%) large-sized hospitals (74.4%) in large metropolitan regions (91.4%). Median total hospital charges were highest for non-Hodgkin lymphoma, followed by leukemia and multiple myeloma ($945,645 versus $265,034 versus $184,194; P < .001). All-cause mortality during index hospitalization was highest for leukemia at 8.6%, followed by 3.6% for non-Hodgkin lymphoma and 1.4% for multiple myeloma (P < .001). The 30-day all-cause readmission rate was 23.6%, and the median time to readmission was 7 days. The readmission rate was highest for leukemia, followed by non-Hodgkin lymphoma and multiple myeloma (34.2% versus 22.8% versus 15.7%; P < .001). Readmission incurred an additional median total hospital charge of $64,561. During readmission, median length of stay was 5 days, and in-hospital mortality was 4.9%. Top etiologies for readmission were cancer or treatment-related (22%), sepsis or infection (18%), neurologic events (15%), neutropenia or pancytopenia (11%), and fever, hypotension, or hypoxia (8%). On multivariable analysis, non-Hodgkin lymphoma and leukemia (compared with multiple myeloma), transfer to a facility at discharge, chronic renal disease, cerebrovascular disease, and noninvasive ventilation were associated with higher odds of readmission. In contrast, admission to a teaching hospital predicted lower odds of readmission. Almost a quarter of CAR-T therapy recipients are readmitted within the first 30 days resulting in additional economic burden and substantial healthcare utilization.

Keywords: CAR-T therapy; Healthcare utilization; Readmission.