| Abstract: |
Background: Primary sclerosing cholangitis with co-existent inflammatory bowel disease (IBD-PSC) confers elevated colorectal cancer (CRC) risk, but modern cohort-level incidence data reflecting changing practice remain limited. We examined CRC occurrence and associations with clinical factors in a large multicentre UK IBD-PSC cohort. Methods: We analysed 472 patients with IBD-PSC from 7 UK centres with IBD diagnosed in 2000 or later. Demographic, clinical, and treatment data were extracted. Follow-up was defined as time from IBD diagnosis to the earliest of CRC, colectomy, or last follow-up. We report descriptive statistics, univariate logistic regressions, and CRC-free survival. Small event numbers precluded multivariable modelling. Results: Among 472 PSC-IBD patients 65% were male, and 367 (78%) UC, 81 (17%) CD, 24 (5%) IBD-u. Most had large-duct PSC (83%) and extensive colitis (E3, 68%). Total follow-up summed 5176 patient-years, median 11 years (IQR 6–16) per patient. Median age at IBD diagnosis was 23 years (15–35). Ten patients developed CRC (2%), an incidence of 1.93 per 1000 patient-years. Among CRC cases median age at IBD diagnosis was 24 years (20–40), time from IBD diagnosis to CRC was 9 years (8–12), and age at CRC was 35 years (33–45). Those with CRC were older at PSC diagnosis with an age of 36 (30-45) compared to those without CRC (25, 16-40) though not significantly (p=0.094). There was a trend towards higher advanced therapy use in CRC cases with 1.5 per patient (0.25–2) and 70% receiving at least one, compared to non-CRC cases with 0 (IQR 0–2) and 45% (p=0.116). Of the 10 CRCs: 9/10 were right sided (one had synchronous sigmoid CRC) and 1 had unknown site; 9/10 were found in actively inflamed segments, 3 of which were found incidentally following colectomy for refractory disease. Endoscopic activity at prior colonoscopy (median interval 2 years): 7/10 moderate-severe activity, 1/10 quiescent disease, 1/10 no colonoscopy for >8 years, 1/10 unknown. Univariate logistic regressions did not identify statistically significant associations, but small event numbers limit inference (table 1). Cumulative CRC risk was low: 10-year 1.8% (0.3–3.2) and 20-year 4.0% (1.4–6.6, figure 1). Conclusion: In this large modern IBD-PSC cohort CRC incidence was lower than previously described. The majority occurred in inflamed segments also inflamed at prior colonoscopy, with a trend towards higher advanced therapy use in CRC patients. Small event numbers limit inference, but there is a suggestion of a relationship between inflammatory burden and CRC risk emphasising importance of surveillance in active or poorly controlled disease. Additional analyses is underway to review surveillance frequency, prior dysplasia, and cumulative inflammatory burden. [ABSTRACT FROM AUTHOR] |