Low Rates of Surgical Recurrence Following Ileocolic Resections for Crohn's Disease in the Biologic Era.

BACKGROUND

Ileocolic resections (ICRs) are the most common resections for Crohn's disease. Historical control groups have often been used for comparison when assessing postoperative recurrence, usually with temporal bias. This study aimed to (1) report contemporary rates of postoperative recurrence requiring repeat surgery (surgical recurrence at anastomosis [surgical recurrence at the ileocolic resection site (SR-ICR)] or surgical recurrence at any site) and the rates of endoscopic recurrence (ER) in the "biologic era"; and (2) determine risk factors for SR-ICR and ER.

METHODS

A retrospective multicenter study involving 12 tertiary Australian centers was performed. Patients (of any age) who had undergone an ICR for Crohn's disease between 2007 and 2023 were included. Cox proportional hazards modeling was used to evaluate clinicopathological risk factors for SR-ICR and ER (defined as Rutgeerts grade ≥i2b).

RESULTS

Overall, 875 patients were included (mean 38.7 ± 15.1 years, 51% female). Median follow-up was 63.9 months. Rates of SR-ICR were 4.5% (95% confidence interval [CI], 2.8%-6.1%) and 12.8% (95% CI, 8.8%-16.5%) at 5 and 10 years, respectively. Rates of surgical recurrence at any site were 5.6% (95% CI, 3.8%-7.5%) and 15.1% (95% CI, 11.0%-19.1%) at 5 and 10 years, respectively. Early (within 12 months) ER occurred in 24.7%. On multivariable analysis, smoking (adjusted hazard ratio, 3.49; 95% CI, 1.93-6.29) was the only factor significantly associated with SR-ICR. Smoking, positive microscopic margins, and granulomas were associated with ER, and prophylactic therapy and younger age at diagnosis (<17 years) were protective.

CONCLUSIONS

The rate of SR at the ileocolic anastomosis in this large Australian cohort was low, recorded to be 1 in 20 at 5 years. Smoking remains the strongest risk factor for both ER and SR. Histopathological factors influence ER and should be considered in future risk prediction models.