Inpatient complications of immunotherapy-associated colitis in solid malignancies: Real-world data analysis.

Abstract

2657 Background: Immune checkpoint inhibitors (ICPis) and their containing regimens have been proven beneficial for the treatment of multiple malignancies, while immune-related adverse events (IrAEs) have become constant safety challenges in their clinical management through all ICPi therapies. Diarrhea resulted from immune-mediated colitis has been one of the most common IrAEs. The Aim of this study is to evaluate the real-world incidence and factors that contributed to colitis among patients hospitalized for anti-neoplastic immunotherapy using a large nation-wide database. Methods: We performed a retrospective analysis with the National Inpatient Sample (NIS) using ICD10-CM and PCS codes to identify patients with solid tumors hospitalized for immunotherapy between October 2015 and 2018. A comparison analysis was made between patients who developed diarrhea (or colitis) or not during their hospitalizations. Patient characteristics included past medical history, location, charges, length of stay (LOS) and inpatient complications variables, data were compared between the two groups. Results: The data from 5,795 admissions for immunotherapy were included, of which 615 (10.61%) inpatient events were complicated by diarrhea (colitis). Patients with diarrhea tended to be males (60%), slightly older at age 57 years, higher percentage of white patients (78%) though not statistically significant from those who did not have colitis. On the other hand, the length of inpatient stay was statistically longer (7 vs 5 days, p < 0.05), and associated with higher expenses ($214,612 vs $ 134,195, p < 0.05). In addition, inpatient admissions with colitis were more towards larger academic hospitals (80% vs 67%, p = 0.01). ICPi-treated patients with genitourinary (GU) cancers were observed more with colitis among these admissions (27% vs 15%, p < 0.05). Though no difference from mortalities observed, serious complications, such as acute kidney injury (AKI) (44% vs 20%, p < 0.05) and non-septic shock rates were higher (7% vs 2%, p = 0.01) among colitis admissions. Conclusions: This real-world data analysis of inpatient admissions from ICPi-treated cancer patients demonstrated colitis with diarrhea as a common safety concern over ICPi-therapies, a higher risk among patients of GU primaries, higher risk with serious medical complications: acute renal dysfunction and non-septic shock, led to increased length of stays (LOS) and financial burden. Special attention and further study may be placed among ICPi-treated GU cancer patients. [Table: see text]