# Clinical, Endoscopic, and Histological Characteristics of Severe Immune Checkpoint Inhibitor-Induced Colitis

## TL;DR

This study examines severe immune checkpoint inhibitor-induced colitis, showing that clinical, endoscopic, and histological features vary and often don't align, emphasizing the need for biopsies and better diagnostic tools.

## Contribution

The study provides a detailed characterization of severe immune checkpoint inhibitor-induced colitis with a focus on clinical, endoscopic, and histological features and their limited correlation.

## Key findings

- Severe IMC presents with heterogeneous clinical, endoscopic, and histological features with limited correlation between domains.
- Histological abnormalities were frequently present even in patients with normal endoscopic findings.
- Endoscopic severity correlated with bleeding and general condition but not with stool frequency or pain intensity.

## Abstract

Background: Immune checkpoint inhibitors (ICIs) have revolutionized cancer therapy. They can cause immune-mediated colitis (IMC), a potentially severe adverse event. Current data on severe IMC (grade 3–4) are limited, particularly regarding clinical, endoscopic, and histological features. Methods: We conducted a multicenter, retrospective observational study promoted by GETECCU, including adults with solid tumors who developed grade 3–4 IMC requiring hospitalization and systemic therapy. Clinical symptoms, endoscopic findings (Mayo and UCEIS indices), and histological features were systematically collected and analyzed. Results: A total of 196 patients were included. Diarrhea was universal (median 8 bowel movements/day), with 76% reporting abdominal pain and 39% rectal bleeding. Endoscopy (n = 139) revealed vascular pattern loss (80%), mucosal lesions (69%), and Mayo scores ≥2 in 69%. Histopathology (n = 141) showed abnormalities in 85%, including cryptitis (50%), lymphocytic infiltration (48%), and crypt abscesses (37%). Notably, 72% of patients with normal endoscopy had histological inflammation. Endoscopic severity correlated with bleeding and impaired general condition but not with stool frequency or pain intensity. Histological and endoscopic severity were modestly associated. Conclusions: Severe IMC presents with heterogeneous clinical, endoscopic, and histological features, with limited correlation between these domains. Endoscopic findings were often ulcerative and inflammatory, yet histological abnormalities were frequently present even in endoscopically inactive disease. These findings highlight the importance of biopsy in all suspected IMC cases and underscore the need for validated multidimensional assessment tools for accurate diagnosis and management of severe IMC.