Mucosal abnormalities as key predictive factors of colorectal cancer in ulcerative colitis

Abstract Although an increased risk of colorectal cancer (CRC) among patients with ulcerative colitis (UC) is well established, the magnitude and determinants of excess risk remain ambiguous. This uncertainty hampers risk stratification and evidence-based endoscopic surveillance. We empaneled a population-based cohort of incident UC cases in Scandinavia, achieved a median follow-up of 15.0 years through record linkages, designed a case-cohort study, and undertook standardized data abstraction from hospital records of 566 patients with CRC and 720 patients from a sub-cohort randomly sampled at baseline. We fitted weighted multivariable Cox proportional hazard models to estimate hazard ratios (HRs) with 95%CIs, which approximates analyzing the full cohort of 90 780 subjects. We confirmed 2-fold higher CRC risk in patients with pancolitis compared with proctitis (HR 2.3, 95%CI 1.5–3.4). We saw no association between severe macroscopic (HR 1.1, 95%CI 0.7–1.8) or microscopic (HR 1.5, 95%CI 1.0–2.3) inflammation and CRC risk. In contrast, low grade dysplasia predicted a more than 3-fold, and adenomas an almost 10-fold excess rate of CRC (HR 9.8, 95%CI 5.2–18.6). Overall, HRs did not depend on follow-up time (0–1 years, 1–8 years, and >8 years). Pharmacologic treatments with biologic drugs (HR 2.4, 95%CI 0.7–7.7) or immunomodulators (HR 1.1, 95%CI 0.7–1.7) did not significantly reduce the CRC rate. In patients with UC, mucosal abnormalities rather than disease extent and inflammation were the strongest predictive factors of CRC. Our findings do not support prevailing recommendations for endoscopic surveillance.