LAUSR

BACKGROUND  Endoscopic resection has developed over the years. The main complications are perforation and bleeding. This study aimed to evaluate safety and effectiveness of digestive endoscopic resection in patients with cirrhosis. METHODS  This retrospective, open-label, single-center study included all consecutive patients with cirrhosis who were admitted for endoscopic resection between 2009 and 2016. Safety, efficacy, and risk factors for delayed bleeding were analyzed. RESULTS  126 patients undergoing 164 procedures were included: 65 endoscopic resections (49 patients) in the upper gastrointestinal tract (esophagus 34, stomach 20, duodenum 11) and 99 in the lower gastrointestinal tract (77 patients). Mean Model for End-Stage Liver Disease score was 9.9 (standard deviation 4.5). Esophageal varices were present in 50 patients, and 21 patients had decompensated cirrhosis. The overall curative rate of endoscopic resection was 84.0 %. No patients died during 30-day follow-up. Immediate overall morbidity was 6.1 %, with two postoperative fevers and eight bleeds. Risk factors for delayed bleeding were duodenal location (P < 0.01), antiplatelet medication (P = 0.02), and lower glomerular filtration rate (GFR) (P = 0.01) in univariate analysis. Duodenal location and lower GFR remained statistically significant in multivariate analysis, with respective odds ratios for bleeding of 52.12 and 1.04. No liver decompensation occurred after endoscopic resection. CONCLUSIONS  Endoscopic resection was safe and effective in patients with mild (Child - Pugh class A/B) cirrhosis, and should be proposed as a first option for treatment of superficial neoplasia. Additional data in patients with severe cirrhosis are needed to confirm the safety in this population.