Increased risk of colorectal neoplasia in patients with primary sclerosing cholangitis and inflammatory bowel disease: a meta-analysis of 16 observational studies.

Abstract Ulcerative colitis (UC) patients with concomitant primary sclerosing cholangitis (PSC) carry an increased risk of colorectal neoplasia (dysplasia and cancer), whereas the association between PSC and the development of colorectal neoplasia in Crohn's disease (CD) is controversial. A meta-analysis was carried out to compare the risk of this neoplasia in patients with inflammatory bowel disease (IBD) with and without PSC. A systematic research of MEDLINE, EMBASE and the Cochrane Central Register of Controlled Trials was performed to identify studies that compared the risk of colorectal neoplasia (dysplasia and cancer) in patients with IBD with and without PSC. Quality assessment was performed using the Newcastle-Ottawa Scale. Pooled odds ratio (OR) was calculated using the random-effects model by STATA 12.0. A total of 16 studies (four cohort studies, 12 case-control studies; nine prospective studies and seven retrospective studies) were selected for further study. These studies included 13 379 IBD patients, of whom 1022 also had PSC. Patients with IBD and PSC were at an increased risk of colorectal dysplasia and cancer compared with patients with IBD alone [OR 3.24; 95% confidence interval (CI): 2.14-4.90]. This increased risk was present even when the risk of colorectal cancer alone was analysed (OR 3.41; 95% CI: 2.13-5.48). Data only from patients with UC showed that PSC was associated with an increased risk for the development of colorectal neoplasia and cancer in patients with UC (OR 2.98; 95% CI: 1.54-5.76) (OR 3.01; 95% CI: 1.44-6.29), but there were high heterogeneity among studies (I=76.9 and 62.8%, respectively). Heterogeneity of the studies was affected by the study design (prospective or retrospective). The OR of colorectal neoplasia was 2.32 (95% CI: 0.70-7.70, P=0.133) and that of cancer was 2.91 (95% CI: 0.84-10.16, P=0.388) for patients with CD and concurrent PSC. Patients with IBD and PSC have a markedly higher risk for the development of colorectal neoplasia than patients with IBD, but not PSC. Stratification by IBD type show that the presence of PSC is associated with an increased risk for the development of colorectal neoplasia in patients with UC; however, there is a nonsignificant association in CD patients. When the risk of colorectal cancer alone is analysed, the conclusion does not change.