Surgical resection rate of Crohn's disease over time: the Sydney Inflammatory Bowel Disease Cohort (1942-2012)

Abstract

Background: Medical therapy for Crohn’s disease (CD) has changed significantly over the last two decades. Studies from Europe and Northern America have shown a trend towards reduced surgical rates in the recent times. However data from the Southern Hemisphere is lacking. Aims: The study aims to determine the changes in intestinal resection rates over longitudinal follow-up in a large specialist-based metropolitan cohort of CD patients. Methods: The Sydney Inflammatory Bowel Disease Cohort is an ongoing epidemiological study recruiting metropolitan patients of Sydney Local Health District. CD patients were recruited and that demographics, disease characteristics, drug treatments and surgical procedures recorded. Kaplan-Meier analysis was used to estimate the cumulative probability of undergoing intestinal resection over 3 time periods and statistical analysis according to the log rank test. Factors that may affect the likelihood of intestinal surgery over time were investigated using the Cox proportional hazard multivariate regression model. Results: In total, 751 CD patients (56.6% females) with complete follow-up data were recruited. The median length of follow-up was 12 years (IQR: 5–21) with an overall 10,383 patient years of follow-up. Inflammatory behaviour was noted in 74.3% patients at diagnosis and ileal, colonic and ileo-colonic locations of disease in 24.6%, 35.6% and 39.8% respectively. Perianal involvement was found in 18.0% at diagnosis. There were total of 305 (40.6%) patients who had intestinal surgery. Ileocaecal resections accounted for 54.8% of surgeries. A total of 32.1% of resections occurred in the first year of diagnosis (IQR: 0–7; Figure 1). There was a significant increase in time to first surgery between the three time periods (P = 0.0001). The proportion of CD patients having surgery in each subsequent time period significantly decreased with 68.1% having surgery in pre-1980 group compared to only 22.0% in the post-2000 group (P = 0.0001).Early introduction of immunomodulators prior to surgical resection was significantly associated with the reduced surgical rate in CD patients (Hazard Ratio [HR]: 0.02; 95% confidence interval [CI]: 0.001–0.438). Decade of diagnosis, disease location, behaviour, presence of perianal involvement at diagnosis and smoking status were all associated with time to intestinal resection (all P < 0.05).Conclusion There has been a significant decrease in the surgical rate of Crohn’s disease over the last 7 decade. This reduction is most likely associated with earlier introduction and greater use of immunomodulators.