P95 Colonoscopic surveillance yield in autoimmune liver disease associated IBD in a tertiary centre

IntroductionAn intensive colonoscopic surveillance programme is proposed in patients with primary or autoimmune sclerosing cholangitis (PSC/AISC) associated inflammatory bowel disease (IBD) because of the high risk of colorectal neoplasia (CRN) in this group. But, evidence for this is limited and further exploration into CRN in this cohort could better inform our practice. We thus aim to describe colonoscopic surveillance in PSC/AISC-IBD in a tertiary centre.MethodsThis is a retrospective review of practice from King’s College Hospital, reporting on the yield of colonoscopic surveillance in PSC/AISC-IBD. All colonoscopy procedures for patients with PSC/AISC-IBD between 2008 and 2022 were reviewed, and data was collected on CRN (low/high grade dysplasia (LGD/HGD), colorectal cancer (CRC)).ResultsA total of 427 colonoscopies were reviewed for 131 patients (PSC: 116, AISC: 9, PSC/AIH overlap:6). 110 were diagnosed with ulcerative colitis, for whom 400 colonoscopies were performed: an average of one colonoscopy every 1.5 years. Pancolitis was seen in 66% and active disease was identified in 37%. CRN was identified in 23 procedures (5%) involving 15 patients (14%): 2 CRC (1.8%), and 13 patients (11.8%) with 28 adenomas with LGD. Only one HGD was found on random biopsy from more than 2580 biopsies, which was not seen on subsequent colonoscopy. The median duration of UC in patients with CRN was 19 years (range 3,27) and for PSC was 12 (1,20). CRC was diagnosed in two male patients (ages: 55 and 33) (1.8%). One had extensive UC and the other pancolitis. Both had been lost to follow up (UC duration 7 and 12 years). CRN location did not associate with actively inflamed segments at index procedure or preceding colonoscopy in 17/23 (74%) colonoscopies. 21 patients were diagnosed with Crohn’s disease; 14 have colonic disease with or without small bowel disease. A total of 27 colonoscopies were performed over 114 patient years, an average of one every 3 years. 3 adenomas were detected, but these patients do not undergo regular surveillance.ConclusionsThe yield of surveillance colonoscopy in this high risk cohort remains low. Further work is under way to inform optimal timepoints for surveillance in this cohort.