Randomized controlled trial (RCT) of hemoclip closure of post-polypectomy induced ulcers (PPIU's) for prevention of delayed hemorrhage in patients on chronic warfarin: Initial results, techniques, & recommendations

Background & Purpose: Retrospective reports estimate the risk of severe delayed PPIU hemorrhage (H) in patients on chronic warfarin is about 10 X that of other patients. Although no prospective RCT data confirm this risk, many endoscopists are prophylactically using hemoclips (HC's) after colon polypectomies to potentially reduce delayed PPIUH in such patients. Our purposes in this RCT in patients on warfarin are to report: 1. prevalences with & without colon polyps after informed consent, 2. techniques & results of PPIU closure in HC patients, & 3. 30 day outcomes. Methods: Written informed consent on an IRB approved blinded RCT (e.g. patient & primary care physician blinded) was obtained in patients on chronic warfarin for severe co-morbidities prior to screening or surveillance colonoscopy at 2 referral centers. The anticoagulation schedule was managed in all patients using current ASGE guidelines. Patients with 1-6 polyps & PPIU's between 5 - 15 mm were randomized to control vs. HC closure. Location & # of polyps, PPIU size, decrease in Hgb, clinical bleeding, abdominal pain, & other complications were prospectively evaluated up to 30 days in all patients. For HC patients, # of HC's/PPIU & success/failure of HC closure were recorded. Results: 30 patients on chronic warfarin were referred for outpatient colonoscopies. 29 patients consented for the RCT & 1 refused. On colonoscopy of consented patients, 12 (41%) had no polyps & were not randomized. 17 patients (59%) had polyps & were randomized - 9 to HC closure & 8 to control - no HC's. For all patients, the mean # of polyps was 3 & mean PPIU size was 7 mm. For PPIU patients the mean # HC/PPIU was 2. A zipper technique of HC closure from each side was used to successfully close all PPIU's < 16 mm. During prospective follow-up (FU) to 30 days, no control or PPIU HC closure patient had pain, reduction in hemoglobin, clinical bleeding or other complications. Conclusions: For patients on chronic warfarin who consented for a blinded RCT study of PPIU HC closure: 1. the prevalence with polyps on outpatient colonoscopy was 59%. 2. Using ASGE guidelines for anticoagulation in all RCT patients on warfarin, none of the controls nor any HC patients had delayed bleeding or other complications within 30 days of FU. 3. The risk of delayed PPIUH appears to be much lower than reported in retrospective studies. 4. HC closure of PPIU's < 16 mm was technically feasible. 5. However, HC closure is not recommended in clinical practice after polypectomies of patients on chronic warfarin until better methods to identify subgroups at higher risk for delayed PPIUH are reported or a large RCT confirms a higher rate of PPIUH in the control group than HC group when current guidelines are followed.Funded from NIDDK CURE DDRC Grant (AM 41301) Human Studies CORE, a Clinical VA Merit Review Grant, & Boston Scientific Corp.