Early Ileocecal Resection for Crohn's Disease Is Associated With Improved Long-term Outcomes Compared With Anti-Tumor Necrosis Factor Therapy: A Population-Based Cohort Study

Background & aims: Early Crohn's disease (CD) treatment involves anti-tumor necrosis factor (TNF) agents, whereas ileocecal resection (ICR) is reserved for complicated CD or treatment failure. We compared long-term outcomes of primary ICR and anti-TNF therapy for ileocecal CD. Methods: Using cross-linked nationwide registers, we identified all individuals diagnosed with ileal or ileocecal CD between 2003 and 2018 and treated with ICR or anti-TNF agents within 1 year of diagnosis. The primary outcome was a composite of >= 1 of the following: CD-related hospitalization, systemic corticosteroid exposure, CD-related surgery, and perianal CD. We conducted adjusted Cox's proportional hazards regression analyses and determined the cumulative risk of different treatments after primary ICR or anti-TNF therapy. Results: Of 16,443 individuals diagnosed with CD, 1279 individuals fulfilled the inclusion criteria. Of these, 45.4% underwent ICR and 54.6% received anti-TNF. The composite outcome occurred in 273 individuals (incidence rate, 110/1000 person-years) in the ICR group and in 318 individuals (incidence rate, 202/1000 person-years) in the anti-TNF group. The risk of the composite outcome was 33% lower with ICR compared with anti-TNF (adjusted hazard ratio, 0.67; 95% confidence interval, 0.54-0.83). ICR was associated with reduced risk of systemic corticosteroid exposure and CD-related surgery, but not other secondary outcomes. The proportion of individuals on immunomodulator, anti-TNF, who underwent subsequent resection, or were on no therapy 5 years post-ICR was 46.3%, 16.8%, 1.8%, and 49.7%, respectively. Conclusion: These data suggest that ICR may have a role as first-line therapy in CD management and challenge the current paradigm of reserving surgery for complicated CD refractory or intolerant to medications. Yet, given inherent biases associated with observational data, our findings should be interpreted and applied cautiously in clinical decision making.