Background/aims: Smoking may alter the natural course of Crohn's disease (CD). Smokers are more likely to develop complications, relapses and have a greater risk for surgery. In contrast, in ulcerative colitis (UC), smoking might improve the disease course. Our aim was to assess the combined effect of disease phenotype, smoking, and immunomodulator [azathioprine (AZA), AZA/biological] treatment on the risk of intestinal resection/reoperation in CD and colectomy in UC.
Patients/methods: Six hundred and eighty-one inflammatory bowel disease patients were analyzed (CD: 340, male/female: 155/185, duration: 9.4+/-7.5 years; UC: 341, male/female: 174/164, duration: 11.5+/-9.7 years). Patients were interviewed on their smoking habits at the time of diagnosis and during the regular follow-up visits. Medical records were retrospectively analyzed.
Results: Smoking was present in 45.5% in CD and 15.8% in UC. CD patients who underwent at least one bowel resection comprised 46.5%. In an univariate analysis, disease location, behavior, AZA, or AZA/biological use before surgery [odds ratio (OR): 0.26 and 0.22, P<0.001] and smoking (OR: 1.61, P = 0.03) were associated with risk for first surgery. Smoking, AZA, or AZA/biological (P<0.001) use before first surgery and disease behavior were independently associated with risk for surgery in a proportional Cox-regression analysis. Perianal disease (OR: 3.2, P = 0.001) and frequent relapses (OR: 4.8, P<0.001) but not smoking, AZA, or AZA/biological use after first surgery were predictive for reoperation. In UC, the rate of colectomy was 5.6%. Disease location (P = 0.001) and smoking status (P = 0.02) were independently associated with risk for colectomy in a proportional Cox-regression analysis.
Conclusion: Our data suggest that early AZA/biological therapy reduces the risk for first operation but not reoperation in CD, in both smokers and nonsmokers. In contrast, smoking was associated with a decreased need for colectomy in UC.