Risk Factors for Late Postoperative Recurrence of Crohn's Disease: A Multicenter Retrospective Cohort Study

Introduction: Risk factors for delayed postoperative recurrence (POR) of Crohn’s disease (CD) have not been fully elucidated. This study aimed to identify risk factors for late endoscopic and surgical POR in patients who are in endoscopic remission at their index postoperative ileocolonoscopy. Methods: Retrospective study of CD patients who underwent ileocolonic resection (ICR) between 2009-2020 across three tertiary care medical centers. Patients with index postoperative ileocolonoscopy with modified Rutgeerts score of ≤ i2a disease within 18 months of surgery were included. Primary composite outcome was defined as either endoscopic (Rutgeerts score ≥ i2b) or repeat ICR after index endoscopic remission. Results: 331 adult CD patients with endoscopic remission on index postoperative ileocolonoscopy within 18 months of ICR. Of these patients, 117 (35%) experienced total composite POR (77.8% endoscopic). On univariate analysis assessing impact of active smoking at time of ICR, age at CD diagnosis, gender, prior history of ICR, history of perianal disease, and postoperative biologic prophylaxis, no variables were associated with composite POR. On multivariable logistic regression, active smoking at time of ICR was associated with a 2.5 times increased risk of composite POR (aOR 2.55 [1.10-6.03]; P=0.03) compared to never smokers. On sensitivity analysis evaluating impact on endoscopic POR, active smoking remained associated with a significantly elevated risk for endoscopic POR compared to non-smokers (aOR 2.73 [1.13-6.56]; P=0.02). In patients who had endoscopic remission on index colonoscopy, the use of postoperative biologic prophylaxis did not offer increased protection against late composite and endoscopic POR (P=0.49; P=0.15). Conclusion: Delayed postoperative recurrence in patients previously in postoperative endoscopic remission is relatively common and associated with active smoking. Multidisciplinary smoking cessation interventions should be incorporated into routine practice and may prevent POR (see Table 1). Table 1. - Multivariable Logistic Regression of Composite POR Risk Factor aOR (95% CI) P value History of ≥2 prior ICR 1.12 (0.61-2.05) 0.13 Postoperative Biologic Prophylaxis 0.69 (0.22-1.87) 0.53 History of perianal disease 0.74 (0.45-1.22) 0.49 Former Tobacco User (Reference: Never) 1.41 (0.82-2.40) 0.21 Active Tobacco User (Reference: Never) 2.56 (1.10-6.03) 0.03 Younger than 30 years of age at time of CD Diagnosis 0.90 (0.53-1.57) 0.28