ACCELERATED INPATIENT INFLIXIMAB INDUCTION FOR PATIENTS HOSPITALIZED WITH ULCERATIVE COLITIS REDUCES RISK OF COLECTOMY AT 2 YEARS COMPARED TO STANDARD INDUCTION

Abstract BACKGROUND Real-world data have shown accelerated infliximab induction for patients hospitalized with acute ulcerative colitis (UC) reduces the risk of short-term but not longer-term colectomy. METHODS We retrospectively reviewed patients hospitalized with acute UC treated with infliximab induction as rescue at 2 health systems (one tertiary, one safety-net) from 2012 to 2022. Patients treated for Clostridium difficile or cytomegalovirus were excluded. Accelerated induction (AI) received ≥10mg/kg within 10 days and standard induction (SI) received one 5mg/kg dose. Two years of follow-up data were analyzed. Wilcoxon rank-sum testing was used to evaluate baseline differences. Subgroup analyses were conducted via multivariate logistic regression and Cox proportional hazards modeling, with variable selection via LASSO (least absolute shrinkage and selection operator) regularization. RESULTS Seventy patients were identified, of whom 33 received AI and 37 received SI. Patients that received AI were more often treated at the tertiary site; severe disease by Truelove and Witts, endoscopic Mayo score of 3, CRP and prior anti-TNF use were numerically higher in the AI group. Other baseline characteristics were similar between groups. 18 patients underwent colectomy (25.7%), of whom 11 had received SI (29.7%) and 7 had received AI (21.2%). All colectomies in patients who received AI occurred within the first 2 months (median 9 days, IQR [5, 15]), while SI colectomies tended to occur later (median 87 days, IQR [13, 194]). Logistic regression showed significantly lower adjusted odds of colectomy at 2 years in the AI group (aOR 0.153, 95% CI [0.026, 0.896], p = 0.037). The lower odds of colectomy in the AI group did not meet significance at 120 days (aOR 0.464, 95% CI [0.091, 2.359], p = 0.354) or 1 year (aOR 0.320, 95% CI [0.064, 1.608], p = 0.167). Having an infliximab level checked at any time was also associated with lower two-year colectomy odds. Survival-time analysis supported the findings of the logistic regression, showing significantly lower hazard of colectomy over two years (aHR 0.316, 95% CI [0.102, 0.975], p = 0.045). CONCLUSIONS AI significantly reduced odds of undergoing colectomy within 2 years in this cohort of patients hospitalized with acute UC. The lower odds of short-term colectomy in the AI group did not meet significance, potentially reflecting disease severity biasing dosing selection. For patients who did not undergo early colectomy, therapeutic drug monitoring, more commonly done in the AI arm, was associated with lower odds of colectomy. This suggests that outpatient care patterns influence the durability of inpatient response. Additional real-world data that includes patterns of outpatient care following inpatient infliximab rescue, alongside data from the PREDICT-UC trial, will further inform dosing practices.