P0855Interest of the ADMIT-ASC score for early assessment of steroid resistance in severe acute newly diagnosed Ulcerative Colitis

Abstract Background Severe acute colitis (SAC) in ulcerative colitis (UC) remains a therapeutic challenge. Intravenous corticosteroid therapy is the first-line treatment, but a substantial proportion of patients are steroid-refractory and require rescue therapy or colectomy. The ADMIT-ASC score, based on biological and endoscopic parameters, has been proposed to predict early steroid failure. The aim of our study was to evaluate the performance of the ADMIT-ASC score in predicting steroid resistance during severe acute colitis as the initial presentation of UC. Methods This was a retrospective study including patients hospitalized for an inaugural episode of SAC between January 2000 and October 2025 in our department. The diagnosis of SAC was based on the Truelove and Witts criteria. The ADMIT-ASC score was calculated retrospectively for all patients (1 point if CRP ≥100 mg/L, 1 point if serum albumin ≤25 g/L, 1 point if UCEIS ≥4, 2 points if UCEIS ≥7). Response to corticosteroids was defined according to clinical and endoscopic course. Differences between responders and non-responders were analysed using the Mann–Whitney test. ROC curve analysis was used to assess the discriminative ability of the score. Results Forty-two patients were included, 17 men and 25 women (M/F sex ratio = 0.68). The mean age was 36.6 ± 14.4 years (17–72). Steroid resistance was observed in 18 patients (42.8%). The ADMIT-ASC score was significantly higher in steroid-resistant patients (p = 0.012). The area under the ROC curve was 0.718, indicating fair discrimination. A threshold of 2.5 provided a sensitivity of 77.8% and a specificity of 58.3% for predicting steroid resistance. Conclusion The ADMIT-ASC score is a simple, reliable, sensitive and relevant tool for the early identification of patients at risk of corticosteroid failure during a first episode of severe acute colitis in UC. Its use in clinical practice is a valuable aid to quickly guide management and optimise therapeutic decision-making. Conflict of interest: Dr. Oueslati, Alaa: No conflict of interest Yacoub, Haythem: No conflict of interest Hajji, Rania: No conflict of interest Dhouha, Cherif: No conflict of interest Debbabi, Habiba: No conflict of interest Hassine, Hajer: No conflict of interest Kchir, Hela: No conflict of interest Maamouri, Nadia: No conflict of interest