Objective To study the utility of fibrin glue on prevention of bleeding after endoscopic submucosal dissection ( ESD).Methods Consecutive patients with gastrointestinal tumors who underwent ESD between July 2015 and June 2016 in the First Affiliated Hospital of Nanjing Medical University were enrolled in the study. The patients receiving ESD before December 31, 2015 were assigned into the fibrin glue group, and patients receiving ESD after December 31, 2015 were assigned into the control group. The fibrin glue group was sprayed with fibrin glue on wound followed by routine hemostasis method, and the control group was given routine hemostasis method only. The bleeding rate after ESD, mean hospital stays and cost were compared between the two groups.Results The bleeding rate after ESD in the fibrin glue group was significantly lower than that in the control group[7. 45%(12/161) VS 14. 79%(25/169), P＝0. 035]. There was no significantly difference in the mean hospital stays (9. 09±2. 65 days VS 9. 20±2. 99 days, P＝0. 744) and mean cost (24 246±5 519 yuan VS 25 214±6 258 yuan, P＝0. 138) between the two groups.Multivariate analysis revealed that the use of fibrin glue was a protective factor for bleeding after ESD. Conclusion Fibrin glue is safe, effective and economical in prevention of bleeding after ESD.

Objective:To evaluate the efficacy and prognosis of endoscopic submucosal dissection (ESD) for undifferentiated-type early gastric cancer (EGC).Methods:Data of 393 patients with 400 EGC lesions who underwent ESD between January 2010 and April 2019 in the First Affiliated Hospital of Nanjing Medical University were collected in the retrospective study. Patients were divided into undifferentiated-type group (50 cases with 50 lesions) and differentiated-type group (343 cases with 350 lesions) according to postoperative pathology. Their data including age, gender, size and location of the resected lesion, general classification, depth of infiltration, presence or absence of ulcers, and follow-up were compared.Results:Logistic regression analysis showed that age ≤60 years ( OR=2.02, 95% CI: 1.04-3.95, P=0.011), female ( OR=2.83, 95% CI: 1.41-5.68, P=0.003), gastric antrum lesions ( OR=3.92, 95% CI: 1.65-9.30, P=0.002), endoscopic depressed type ( OR=5.37, 95% CI: 2.16-13.38, P<0.001), and submucosa invasive depth ( OR=5.09, 95% CI: 2.40-10.80, P<0.001) were independent risk factors for undifferentiated EGC. The undifferentiated-type group showed a significantly higher non-curative resection rate [90.0% (45/50) VS 19.8% (68/343), χ2=104.902, P<0.001]. Of the 393 patients, 5(4.4%) died in the 113 patients with non-curative resection, while 2 (0.7%) died in the 280 patients with curative resection. Patients with non-curative resection had a higher mortality ( χ2=5.558, P=0.023). There were 27 and 51 patients undergoing additional surgery in the undifferentiated-type group and the differentiated-type group, respectively. None of them had recurrence. Among the 315 patients who did not undergo surgery, the recurrence rate of the undifferentiated-type group was significantly higher than that of the differentiated-type group [26.1% (6/23) VS 4.1% (12/292), χ2=5.560, P<0.001]. Conclusion:Young age (≤60 years), female, gastric antrum lesions, endoscopic depressed type, and submucosa invasive depth are predictors of undifferentiated EGC. Patients with undifferentiated EGC have a higher non-curative resection rate and higher possibility of recurrence after ESD, and additional operation are suggested.