Abstract Background Immune checkpoint inhibitor (ICI) colitis is a common adverse event related to ICI treatment in patients with cancer. Escalation to selective immunotherapy (SIT) after initial treatment with corticosteroids is necessary in approximately 40% of cases. The recently proposed Immune-Mediated Colitis Endoscopic Score (IMCES) aims to stratify endoscopic severity and guide early SIT initiation1. IMCES specificity for SIT use was found to be 82.2% with a cutoff score of 4. This study assessed the external validity of the IMCES in an independent two-centre cohort. Methods Patients diagnosed with histologically confirmed ICI colitis between 2014 and 2024 in two Dutch academic centres were retrospectively identified. Endoscopic images were revised by two gastroenterologists to determine the IMCES. Only patients with an endoscopy performed ≤5 days after initiating colitis treatment were included. The primary outcome was SIT use during follow-up; a secondary outcome was early SIT use (i.e. within 3 weeks after starting first colitis treatment). Discriminative performance was evaluated by receiver operating characteristic (ROC) analysis and corresponding area under the curve (AUC). Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated for IMCES cut off score of 4. Univariable logistic regression was performed to assess associations between individual components of the IMCES and early SIT use. Results Out of 279 patients diagnosed with ICI colitis, 255 were eligible for analysis. The majority had melanoma (53%) or non-small cell lung cancer (21%) and patients received aPD(L)-1 monotherapy (51%), combination therapy (43%), or aCTLA-4 monotherapy (5.9%). A total of 98 patients (38%) was treated with SIT. Median IMCES was 3 (Q1 – Q3: 2 – 4.5). For SIT use during follow up, the AUC was 0.57 (95% CI 0.50 – 0.65, sensitivity 44%, specificity 65%). Highest discrimination was achieved within a subgroup of patients treated with aCTLA-4 monotherapy or combination therapy (AUC 0.64 95% CI 0.55 – 0.74, PPV 66%). Highest NPV (96%) was observed among patients treated with aPD(L)-1 immunotherapy for early SIT use. Among individual IMCES components, presence of erosions (OR 1.83, p = 0.047) and exudate (OR 2.00, p = 0.045) were associated with early SIT use, whereas ulcerations were not (OR 1.60, p = 0.222). Conclusion In this external validation study, the IMCES demonstrated poor discriminative ability to predict overall SIT use. Performance improved when predicting early SIT use and in patients treated with aCTLA-4 based immunotherapy. Individual components of the IMCES show promise in identifying patients with a need for SIT. Reference: 1. Wang Y, Abu-Sbeih H, Tang T, et al. Novel endoscopic scoring system for immune mediated colitis: a multicenter retrospective study of 674 patients. Gastrointest Endosc. 2024;100(2):273-282.e4. doi:10.1016/j.gie.2024.01.024. Conflict of interest: Ms. Naber, Myrthe: No conflict of interest van Eijs, Mick: No conflict of interest Huitema, Jildou: No conflict of interest de Haan, Jacco: No conflict of interest Suijkerbuijk, Karijn: No conflict of interest Visschedijk, Marijn: Speakers fees from Jansen-Cilag, Abbvie, Ferring, Alfasigma, Takeda. van Schaik, Fiona: Consultancy fees from Takeda, Galapagos and Johnson&Johnson, lecture honoraria from Galapagos, Lilly and Janssen-Cilag B.V., hospitality fees from Ferring and dr. Falk Farma and an unrestricted research grant from Takeda.
Naber et al. (Thu,) studied this question.