Evaluating toxicity from definitive radiation therapy for prostate cancer in men with inflammatory bowel disease: Patient selection and dosimetric parameters with modern treatment techniques.

ABSTRACT

PURPOSE: Inflammatory bowel disease (IBD) is considered a contraindication to abdominopelvic radiation therapy (RT). We examined our experience in men with IBD who were treated with definitive RT for prostate cancer. METHODS AND MATERIALS We queried our institutional database for patients with a diagnosis of ulcerative colitis, Crohn disease, or IBD not otherwise specified. Endpoints were acute and late ≥grade 2 (G2) GI toxicity and IBD flare after RT. Outcomes were compared with controls using propensity scoring matched 3 to 1. We matched controls to the IBD cohort according to RT technique, RT dose, risk group, hormone use, treatment year, and age. We determined predictors of acute outcomes using the Fisher exact test and time to outcomes using the log-rank test. RESULTS: Between 1990 and 2010, 84 men were included. Sixty-three men served as matched controls and 21 with IBD 13 ulcerative colitis, 7 Crohn disease, and 1 IBD not otherwise specified. For men with IBD, median age was 69 years, and median follow-up was 49 months. Median flare-free interval before RT was 10 years. Seven were taking IBD medications during RT. There was no difference in acute or late gastrointestinal (GI) toxicity in the IBD group versus controls. Among IBD patients, IBD medication use was the only predictor of acute ≥G2 GI toxicity 57.1% with medication versus7.7% without (49.4% absolute difference, 95% confidence interval [CI] 10.0%-88.9%, P = .03). The 5-year risk of late GI toxicity in men with IBD versus controls was not statistically significant (hazard ratio = 1.19, 95%CI 0.28-5.01, P = .83). The crude incidence of late ≥G2 GI toxicity was 10%. CONCLUSIONS: Acute GI toxicity appears to be exacerbated in patients on concomitant medical therapy for IBD. Overall, late GI toxicity was relatively low and not significantly different between patients with IBD versus no IBD. However, the small sample size limits the interpretation of our estimates and the wide confidence intervals indicate these patients warrant careful selection.