8 - Azathioprine/biological therapy does prevent surgery but not reoperation in smokers with Crohn's disease

Abstract

Background and Aim: 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 smoking might improve the disease course in ulcerative colitis (UC). Our aim was to assess the combined effect of smoking and immunomodulator (azathioprine, AZA/biological) treatment on the risk of intestinal resection and re-operation in CD and colectomy in UC. Patients and Methods: 504 IBD patients were analyzed (CD: 277, 46.2% males, age at diagnosis: 28.7 (SD 13.2) years, mean duration: 9.5(7.8) years; UC: 227, 48.5% males, age at diagnosis: 34.3 (SD 14.5) years, mean duration: 11.3 (10.4) years). Patients’ medical records have been analyzed retrospectively and patients were asked for smoking status at diagnosis and during follow-up by standardized questionnaires. Results: Smoking was present in 47.1% in CD and 13.2% in UC. 133 CD patients (48%) underwent at least one bowel resection, while at least one reoperation was necessary in 44 (15.9%). In univariate analysis disease location (p = 0.004), behavior (p < 0.0001), AZA or AZA/biological use prior to surgery (OR: 0.19 and 0.227, p < 0.0001 for both) and smoking (OR: 1.79, p = 0.018) were associated with the risk for surgery. Perianal disease (OR: 3.83, p = 0.001) and frequent relapses (OR: 5.85, p < 0.0001) but not smoking status or AZA or AZA/biological use after first surgery were predictive for reoperation. Smoking (OR: 1.91), AZA or AZA/IFX use prior to surgery (OR: 0.19) and disease behavior (OR: 3.13) were independently associated with risk for surgery in a logistic regression analysis. The deleterious effect of smoking was most striking in stenosing disease (pLogRank: 0.028) and females (p = 0.006) in a Kaplan Meier analysis. AZA use decreased the risk for first surgery (p < 0.0001 for both) but not re-operation in patients with and without smoking in the same analysis using LogRank/Breslow tests. In UC, 12 (5.3%) patients had colectomy. Disease location (p = 0.001) but not smoking status was associated with risk for colectomy. Of note, none of the patients with colectomy smoked compared to 14% of patients without colectomy (p =NS). Conclusion: Our data suggest that AZA/biological therapy reduces the risk for first operation but not reoperation in CD in both smokers and none-smokers. The deleterious effect of smoking was most pronounced in females and in patients with stenosing disease.