Introduction Biliary anastomotic strictures (AS) occur in 5%-15% of liver transplant (LT) recipients and cause graft dysfunction and morbidity. The main alternative to surgical reconstruction is stenting at endoscopic retrograde cholangiopancreatography (ERCP), using plastic polyethylene or fully-covered self-expanding metal stents (fcSEMS). We retrospectively reviewed outcomes after ERCP for biliary AS, aiming to assess rates of success and complications. Methods Records for all patients

more »

... oing ERCP post LT between 2013-16 were reviewed. AS were classified as early or late (<3 or!3 mths post LT) and cases of diffuse cholangiopathy excluded. Data collected included graft characteristics, timing/nature of ERCPs and complications. Stricture resolution was determined from graft function, imaging and ERCP findings. Results In total 252 ERCPs were performed (median 4/ patient, range 2-10) in 62 patients, 72.6% male, age at LT 54 years (range 20-71). 48 grafts (77.4%) were donations after brain death, 13 (21.0%) after cardiac death and 1 live donor. At first endoscopy 53 (85.4%) had a native papilla. Median time from LT was 28 days (4-75) in early AS (n=29, 46.7%, 9 presenting with bile leak) vs 11.5 mths (3.4-251) in late AS (n=33, 53.2%). Biliary stents were placed in 172 ERCPs (63.4% plastic vs 36.6% fcSEMS) and AS dilated in 68 (27.0%; 60 in conjunction with stenting). Stricture resolution was achieved endoscopically in 47/55 patients with complete outcome data (85.5%). Plastic stents were placed at index ERCP in 23/26 of early AS, in whom sequential plastic 'upstenting' was ultimately successful in 9 (39.1%) and 12 (52.2%) proceeded to fcSEMS (stricture remodelled in all). Of late AS, 9/29 were managed with fcSEMS (88.8% successfully), 8 plastic 'upstenting' (75% success), 8 fcSEMS after initial plastic stent (all successful) and 2 resolved with dilatation only. The overall rate of post-ERCP pancreatitis (PEP) was 7.1% (none severe by Cotton criteria) and was higher after fcSEMS (15.9%) than plastic stenting (5.5%, p=0.024), despite similar use of NSAID prophylaxis. Compared with plastic stents, fcSEMS were more likely to migrate (36.5% vs 12.8%, p<0.001) and embed (6.3% vs 0%, p=0.008) but stent occlusion was numerically lower (0% vs 4.6%, p=0.084). Of 8 endoscopic failures (14.5%), 3 (5.5%) required percutaneous stenting and 2 (3.6%) biliary reconstruction, one after duodenal perforation by a migrated plastic stent. Conclusions Endoscopic management is effective in treating 85% of biliary AS. fcSEMS appear superior to sequential plastic 'upstenting' but are associated with higher rates of PEP and migration.