Multicentre randomised controlled trial of a self-assembling haemostatic gel to prevent delayed bleeding following endoscopic mucosal resection (PURPLE Trial)

Abstract

Background Prophylactic application of a haemostatic gel to the resection field may be an easy way to prevent delayed bleeding, a frequent complication after endoscopic mucosal resection (EMR). Objective We aimed to evaluate if the prophylactic application of a haemostatic gel to the resection field directly after EMR can reduce the rate of clinically significant delayed bleeding events. Design We conducted a prospective randomised trial of patients undergoing hot-snare EMR of flat lesions in the duodenum (≥10 mm) and colorectum (≥20 mm) at 15 German centres. Prophylactic clip closure was not allowed, but selective clipping or coagulation could be used prior to randomisation to treat intraprocedural bleeding or for prophylactic closure of visible vessels. Patients were randomised to haemostatic gel application or no prophylaxis. The primary endpoint was delayed bleeding within 30 days. Results The trial was stopped early due to futility after an interim analysis. The primary endpoint was analysed in 232 patients (208 colorectal, 26 duodenal). Both groups were comparable in age, sex, comorbidities and lesion characteristics. Preventive measures, such as selective clipping or coagulation, were applied prior to randomisation in 51.9% of cases, with no difference between groups. Delayed bleeding occurred in 14 cases (11.7%; 95% CI 7.1% to 18.6%) after Purastat and in 7 cases (6.3%; 95% CI 3.1% to 12.3%) in the control group (p=0.227), with no difference between colorectal and duodenal subgroups. Conclusion The application of a haemostatic gel following EMR of large flat lesions in the duodenum and colorectum does not reduce the rate of delayed bleeding. Data are available on reasonable request. Data and full trial protocol are available on reasonable request. All data relevant to the study are included in the article of uploaded as online supplemental information.