# Immune Checkpoint-Induced Colitis: A Single-Center Retrospective Cohort Study

## TL;DR

This study examines the features and treatment outcomes of colitis caused by immune checkpoint inhibitors in cancer patients.

## Contribution

The study provides real-world insights into the management and outcomes of ICI-induced colitis using corticosteroids and biologicals.

## Key findings

- Combination ICI therapy was linked to higher disease severity and need for biological treatments.
- Endoscopically normal patients often showed histological signs of colitis.
- Biologicals achieved 100% clinical response rates compared to 79.6% with corticosteroids.

## Abstract

Background: Immune checkpoint inhibitors (ICIs) have transformed cancer therapy but are often complicated by immune-related adverse events, particularly colitis. With increasing ICI use, understanding the clinical course and management of ICI-induced colitis is essential. Objectives: To characterize the clinical, endoscopic, and histological features of ICI-induced colitis and evaluate treatment outcomes, focusing on the use of corticosteroids and second-line biologicals (infliximab and vedolizumab) in a real-world setting. Methods: A retrospective cohort study was conducted at Ghent University Hospital, including 77 adult patients diagnosed with ICI-induced colitis in between 2012 and 2023. Clinical, biochemical, endoscopic, and histological data were analyzed, along with treatment response and safety outcomes. Results: Patients with ICI-induced colitis received anti-PD-1/PD-L1 (64.9%), anti-CTLA-4 (9.1%), or combination of both (26.0%). In patients with normal endoscopic findings, histological signs of colitis were observed in 88.0%. Combination ICI therapy was associated with higher Mayo scores (p = 0.029) and increased need for biologicals (p = 0.011) compared to anti-PD-1/PD-L1 monotherapy. Clinical response rates were 79.6% with corticosteroids and 100.0% with biologicals. Rechallenge with ICIs lead to a 17.4% relapse rate. No colitis-related deaths were observed. Conclusions: In this retrospective study, we demonstrate that random colon biopsies reveal microscopic ICI-induced colitis in most patients with absence of endoscopic disease. Combination ICI therapy predicts a corticosteroid-refractory course, supporting the need for early escalation to biologicals. ICI rechallenge appears feasible, as relapse rates were relatively low and colitis morbidity remained manageable. Prospective studies are needed to refine therapeutic strategies and improve patient outcomes.