S218 Prevalence and Outcomes of Immune Checkpoint Inhibitor (ICI) Associated Colitis: A Population-Based Cohort Study

Abstract

Introduction: Immune checkpoint inhibitors (ICIs) can trigger an overactivation of the immune system, resulting in an autoimmune-like process in various organ systems. Involvement of the gastrointestinal (GI) tract has been frequently reported in literature. The primary aim of this study was to evaluate the prevalence and disease outcomes of ICI colitis. Methods: A retrospective cohort study was conducted using TriNetX, a multi-institutional database of more than 70 million patients from 49 healthcare organizations in the USA. ICI colitis cohort included patients on Nivolumab, Ipilimumab, Pembrolizumab, Atezolizumb, Durvalumab, Avelumab or Cemiplimab who developed colitis within 1 year of initiation of ICI. Patients with inflammatory bowel disease (IBD), GI infections and intestinal ischemia were excluded from the study. Disease outcomes that were evaluated included hospitalization, oral/IV steroid use, biologic use and mortality within 3 months of ICI colitis. Disease outcomes between patients on CTLA-4 inhibitor and PD-1 inhibitor; and CTLA-4 inhibitor and PD-L1 inhibitor were compared after 1:1 propensity matching for age, gender, race and ethnicity and reported as adjusted Odds Ratios (aOR) with 95% confidence intervals (CIs). Results: Out of 57,067 patients on ICIs, 4128 (7.23%) patients developed ICI colitis (mean age 64 ± 12, male 56%, Caucasian 77%). The prevalence of ICI colitis was 5.9% in Pembrolizumab, 15.5% in Ipilimumab, 5.5% in Nivolumab, 3% in Durvalumab, 5.8% in Atezolizumab, 3% in Cemiplimab and 2.8% in Avelumab. 34% required hospitalization, 44% required oral steroids, 6.6% required IV steroids, 6% required a biologic agent and 0.2% required colectomy (Figure). 90-day mortality was observed in 15% of patients. Patients with ICI colitis from CTLA-4 inhibitors were more likely to require oral steroids [aOR 1.62, 95% CI 1.33-1.98], IV steroids [1.54, 1.15-2.07] and infliximab [OR 2.18, 1.62-2.91] compared to those on PD-1 inhibitors. Patients with ICI colitis from CTLA-4 inhibitors were also more likely to require oral steroids [aOR 3.33, 95% CI 2.36-4.69], IV steroids [2.45, 1.49-4.02] and infliximab [OR 7.35, 3.73-14.50] compared to those on PD-L1 inhibitors (Table). Conclusion: Approximately 50% of patients with ICI colitis require either PO or IV steroids. Patients on CTLA-4 inhibitors are more likely to have a worse disease course compared to patients on PD-1 or PD-L1 inhibitors.Figure 1.: Clinical outcomes of the patients with ICI colitis Oral steroids: Prednisone or Budesonide; IV steroids: Methylprednisolone; Biologic: Infliximab or Vedolizumab; Surgical intervention: total or subtotal colectomy Table 1. - Comparison of the clinical outcomes between patients that developed Cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) inhibitor colitis vs Programmed death-ligand 1 (PD-L1) inhibitor associated colitis and CTLA-4 inhibitor colitis vs Programmed cell death protein 1 (PD-1) inhibitor colitis Clinical Outcomes CTLA-4 vs PD-L1 (OR, CI) CTLA-4 vs PD-1 (OR, CI) Oral steroids 3.333 (2.368-4.693) 1.626 (1.334-1.982) IV steroids 2.450 (1.491-4.025) 1.546 (1.154-2.071) Infliximab 7.355 (3.73, 14.502) 2.175 (1.625-2.912) Hospitalization 1.256 (0.73-1.621) 1.096 (0.916-1.31) ICU care 1.00 (0.626, 1.596) 0.972 (0.697-1.354) CTLA-4 inhibitor: Ipilimumab; PD-1 inhibitor: pembrolizumab, and nivolumab; PD-L1 inhibitor: avelumab, atezolizumab and durvalumab.