Gastrointestinal adverse event in patients with and without pre-existing inflammatory bowel disease.

Abstract

e24084 Background: Immune checkpoint inhibitors (ICI) enhance immune function and may cause gastrointestinal adverse events (GIAEs), such as immune-mediated diarrhea and colitis (IMDC). Patients with pre-existing inflammatory bowel disease (IBD) are usually excluded from ICI trials. This study aims to compare GIAE in patients with and without a history of IBD, elaborating the difference in disease behavior and outcome among these two groups. Methods: This single-center, retrospective study at a tertiary care cancer center included patients who received ICI between 2015 and 2023, comprising those with and without a history of IBD who developed IMDC and those with pre-existing IBD, regardless IMDC occurrence. Results: A total of 1302 patients were included, 151 (11.6%) with IBD history, of whom 59 (39%) developed GIAE. IBD patients were more likely to receive PD-1/L1 inhibitors (86.7% vs 48.9%, p < 0.0001), and experienced more GIAE when also on tyrosine kinase inhibitors (12.1% vs 4.3%, p = 0.077). IBD patients who developed AE were more likely to be on IBD-specific therapy within 3 months of the event, including budesonide (7.9% vs 1.8%, p = 0.076), TNF-α inhibitors (7.9% vs 1.8%, p = 0.076), α4β7 integrin inhibitor (15.8% vs 4.6%, p = 0.024), and IL-12/IL-23 blocker (7.9% vs 1.8%, p = 0.076). Mortality was higher in non-AE group (32.6% vs 7.9%, p = 0.055). Comparison of GIAE between IBD (IBDG) and non-IBD (NIBDG) groups revealed a shorter median time to onset following ICI therapy in NIBDG ( 3.3 vs 4 months). Grade 2 or higher diarrhea was more common in NIBDG (80.1% vs 64%, p = 0.006), while no difference in colitis grade was observed. At GIAE onset, normal endoscopy was more common in NIBDG (38.5% vs 4%, p < 0.0001), as was positive lactoferrin (78.4% vs 53.3%, p = 0.020). Initiation of steroid (46.6% vs 90.5%, p < 0.0001) and selective immunosuppressive therapy (SIT) (17.2% vs 44.7%, p < 0.0001), or their escalation from baseline treatment, was less frequent in IBDG, but colectomy rates were higher (3.4% vs 0.8%, p = 0.039). Clinical remission occurred more often in NIBDG (93.1% vs 25.9%, p < 0.0001), but symptom duration was longer (30 vs 12 days, p < 0.0001). Hospitalization rates (27% vs 60.1%, p = 0.067) and length of stay (5 vs 6 days, p = 0.034) were lower in IBDG. No differences were observed in all-cause mortality, or rate of ICI discontinuation and resumption. Conclusions: This study is the first to compare GIAE in IBD and non-IBD patients after ICI therapy. We found a 39% incidence of GIAE in IBD patients, primary treated with PD-1/L1 inhibitors, who often received immunosuppressant as baseline IBD treatment before ICI but had lower colitis remission rate of 25.9% compared to non-IBD patients. Despite this, IBD with GIAE after ICI did not increase overall mortality compared to those without IBD.