LAUSR

BACKGROUND/AIMS Although Dieulafoy's lesion (DL) is an important cause of nonvariceal upper gastrointestinal (GI) bleeding, few studies have investigated the clinico-epidemiological outcomes due to its rarity. Here, we investigated clinical features of upper GI bleeding caused by peptic ulcer (PU) or DL and compared endoscopic treatment outcomes. METHODS Patients with upper GI bleeding resulting from PU or DL who visited emergency room between January 2013 and December 2017 were eligible. Clinical features and treatment outcomes were retrospectively investigated. RESULTS Overall, 728 patients with upper GI bleeding due to PU (n=669) and DL (n=59) were enrolled. The median age was 64 years (interquartile range [IQR], 56-75 years), and 74.3% were male. Endoscopic intervention was performed in 53.7% (n=359) and 98.3% (n=58) of the PU and DL groups, respectively (p<0.0001). Patients were matched by sex, age, body mass index, comorbidity, and past medical history, and 190 PU and 52 DL were finally selected. The rebleeding rates within 7 (7.37% vs. 17.31%, p=0.037) and 30 (7.37% vs. 26.92%, p<0.001) days after initial endoscopy were significantly lower in the PU than in the DL group after propensity score matching. During the median follow-up period of 52 months (IQR, 34-70 months), there was no difference in overall survival rate (67.9% vs. 82.7%, p=0.518). CONCLUSIONS Although DL is a rare cause of upper GI bleeding, it requires endoscopic hemostasis more frequently and has a higher rate of rebleeding than PU even after therapeutic endoscopy. Endoscopists should pay attention and perform active endoscopic hemostasis for DL bleeding.