P0240Predictive factors for hospitalisation duration and treatment response in immune-related adverse event colitis

Abstract Background The increasing use of immune checkpoint inhibitors (ICIs) has led to a rise in immune-related adverse event colitis (irAE colitis). Although standard therapy consists of corticosteroids and infliximab, management can be difficult in a subset of patients. Treatment resistance and prolonged hospitalization can adversely affect cancer management. The factors that affect hospitalisation duration and clinical improvement in irAE colitis remain unclear. This study aimed to identify clinical factors associated with these outcomes. Methods We retrospectively analysed patients diagnosed with immune-related colitis based on colonoscopy performed between March 2017 and June 2025 at Kyoto University Hospital. Clinical factors including demographics, underlying malignancy, causative immune checkpoint inhibitors (ICIs), endoscopic findings, medications, length of hospitalisation, and time to symptomatic improvement (defined as improvement from CTCAE Grade ≥2 to Grade 1) were evaluated. Multiple regression analyses were performed to identify factors associated with prolonged hospitalisation and delayed symptom improvement.Characteristic endoscopic findings (total 9) were defined as: edema, loss of vascular transparency, fine granular mucosa, continuous rectal-origin lesions, purulent mucus, submucosal haemorrhage, diffuse continuous erythema, ulcers, and easy bleeding. Results We enrolled forty-three patients (29 males, 14 females; median age 72 years, range 20–85) . Underlying malignancies were solid cancers in 30 cases, malignant melanoma in 11, and cancer of unknown primary in 2. Suspected ICIs were PD-1/PD-L1 inhibitor monotherapy in 30 cases and PD-1/PD-L1 plus CTLA-4 inhibitor therapy in 13. Twenty-two patients (51.1%) exhibited ≥3 characteristic endoscopic findings. Steroids were used in 31 patients (72.1%), while 12 (27.9%) improved with ICI discontinuation alone. The median hospitalization duration was 24.5 days (range 4–125), and the median time to improvement to Grade 1 was 26.0 days (range 4–381). Multiple regression analysis identified the presence of ≥ 3 endoscopic findings as a significant factor associated with prolonged hospitalization (p = 0.02). Bloody stool at diagnosis was significantly associated with longer time to symptom improvement (p = 0.03). A lower initial steroid dose (≤1.0 mg/kg vs 1.0 mg/kg) showed a tendency toward delayed improvement, although this was not statistically significant (p = 0.10). Conclusion In irAE colitis, the presence of bloody stool was the factor associated with delayed symptomatic improvement, whereas the number of characteristic endoscopic findings was the factor influencing the length of hospitalisation. Conflict of interest: Dr. Fuyuno, Takashi: No conflict of interest Onoyama, Yusuke: No conflict of interest Namikawa, Mio: No conflict of interest Kitamoto, Hiroki: No conflict of interest Okabe, Makoto: No conflict of interest Seno, Hiroshi: No conflict of interest