Hemorrhage Following Small Polyp Resection in the Colon in Anticoagulated Patients

Oral anti thrombotic medications are commonly prescribed for patients with venous thromboembolism, chronic atrial fibrillation, prosthetic heart valves, etc. As gastrointestinal endoscopy in often required in these patients, one should consider the urgency of the procedure, the risk of bleeding (related both to antithrombotic therapy itself and to the type of endoscopic procedure) and the risk for thromboembolic events related to interruption of antithrombotic therapy.Current ASGE guidelines state that aspirin may be continued for all diagnostic and most therapeutic endoscopic procedures . As to oral anticoagulants (ie. warfarin), a distinction is made as to the risk of thromboembolism (high vs. low patient risk of thrombosis) and the risk of bleeding ( high vs. low procedure risk of bleeding). ASGE and ACCP Guidelines suggest discontinuing anticoagulation in the periendoscopic period in patients with low risk of thromboembolic events (appendix I), while continuing anticoagulation in patients with high risk of thromboembolic events. Switching to LMWH , termed "bridging therapy", is suggested when anticoagulation need to be continued. However there is little evidence to support such an approach. In addition, there are a number of potential drawbacks to bridging therapy with heparin in the periprocedural period. This approach consumes considerable healthcare resources, involves a short period of normal coagulability (if not hypercoagulability) with an associated risk of thromboembolism. Finally, In a recent systematic review of 34 studies of anticoagulated patients (19 of which enrolled endoscopic procedures), bridging therapy was found to be associated with increased rates of overall and major bleeding, and no difference in thrombotic events, as compared to no bridging. Colonic polypectomy is considered to be a procedure with high risk for bleeding. The reported risk of polypectomy related bleeding ranges in various reports from as low as 0.3% to as high as 6.1 %, averaging in most studies 1-2.5% Bleeding can be immediate or delayed, with the latter being a bit more frequent, occurring in up to 2% of patients, as early as few hours and as late as 29 days after the procedure, though mostly in post procedure days. Polyp size greater than 1-2 cm was found to be associated with increased risk for immediate and delayed postpolypectomy bleeding. As an example, bleeding occurred in ~1% of polypectomies < 1 cm and in 6.5% of polypectomies > 2 cm.Other factors found to be related to bleeding risk are age, location (Cecal and right colon > rest of colon), cardiovascular and renal comorbidities, and use of anticoagulation. The continuation of anticoagulation further increases bleeding risk. In a 2004 retrospective study of 1657 patients undergoing polypectomy, post procedural bleeding rate was 2.2% (86% immediate, 14% delayed), and Warfarin was found to be associated with an Odds ratio of 13.37 for bleeding. In contrast to other studies, age, size and location of the polyp were not found to be associated with risk of bleeding. On the other hand, temporary suspension of anticoagulation therapy is associated, at least theoretically, with a risk of thromboembolic complications. The magnitude of that risk was reported to be low (~ 0.7%) in 2 retrospective studies and one meta-analysis but admittedly high risk patients were underrepresented. In fact, actual overall rates (with and without bridging) of perioperative thrombotic complications are higher and range from 0.9-1.8% , with a corresponding bleeding rate of 1.9%-2.7%.As to bridging therapy, evidence comes almost exclusively from retrospective studies. Recent meta-analysis showed that, compared to no anticoagulation, perioperative bridging therapy with heparin increased the overall risk of major bleeding without a significant decrease in the risk of thromboembolic events. Later on, a prospective, randomized study comparing bridging therapy with continued warfarin in patients with moderate-to-high risk of thromboembolism undergoing defibrillator or pacemaker surgery (a high bleeding risk procedure by definition) , found more major bleeding in the bridging group as compared to continued warfarin, with no difference in thrombotic outcome.Apart from the above mentioned paper, as far as the investigators know, no comparative prospective study compared bridging with continued warfarin.

Polyps up to 1 cm in diameter have been safely removed in patients on warfarin therapy. In one retrospective series, warfarin was discontinued 36 hours prior to colonoscopy to avoid supra therapeutic INR. 3 out of 123 patients who underwent 225 polypectomies on warfarin had bleeding, only one of which required treatment. All patients were prophylactically treated with clips.19 In a recent randomized trial, 70 patients with a total of 159 polyps up to 1 cm in diameter underwent polypectomy while taking warfarin. Patients were assigned to have their polyps removed either eith a cold snare technique or electrocautery. Immediate bleeding occurred in 10 of 70 patients (14%) and was more common in patients who had their polyps removed using electrocautery (23% vs 6%). No delayed bleeding occurred in the cold group whereas 5 (14%) patients required endoscopic hemostasis in the electrocautery group.

As immediate bleeding can be effectively managed endoscopically, the investigators should question if there is any potential benefit of periprocedural bridging in terms of preventing delayed bleeding (occurring during warfarin-LMWH overlap or following warfarin re-loading).

Considering the fact that ~ 25% of patients will be found to have colorectal polyps in screening colonoscopy, and that the majority of these polyps are < 20 mm, A cardinal question is whether the investigators should keep choosing among the two traditional policies: advise all our patients to stop warfarin ahead of screening colonoscopies,Or keep them anticoagulated and reschedule them to a therapeutic colonoscopy in the face of discovering colonic polyps (an approach that was found to be cost effective in one study comparing these two policies).

The investigators think a third option might be better : keeping patients anticoagulated with warfarin, and perform polypectomies for polyps up to 20 mm, while re-scheduling patients with larger polyps.

To summarize, the three most important conclusions from the data presented in this introduction are:

1. Small polyps was shown to be safely removed under Warfarin
2. Delayed bleeding occurs after bridging is finished
3. Bridging therapy is probably related to increased bleeding, theoretically related to increased thromboembolic risk and is logistically complex Taken together, these background data underlies our logic to perform a prospective study comparing bleeding rates following polypectomies of small polyps in orally anticoagulated vs. bridged patients.

Hypothesis:In patients with moderate to high risk of thromboembolic events undergoing elective colonoscopy, a strategy of uninterrupted oral anticoagulation will lead to equivalent bleeding rates compared with conventional bridging anticoagulation.

Study Objectives:Primary study Objective : to evaluate the safety of polypectomy under Warfarin therapy compared to currently recommended bridging therapy with heparin or LMWH Study Design:Following signing on informed consent, patients on chronic Warfarin therapy meeting inclusion and exclusion criteria , who are scheduled for elective diagnostic colonoscopy will be randomly assigned to one of two groups, the first of which continue Warfarin treatment as usual while the second is switched to bridging therapy with LMWH according to recommendations of the ACCP guidelines. INR will be checked on the day before the colonoscopy (day -1) in both groups. Most recent CBC value (from recent 3 months) will serve as baseline value. If not available - CBC will be done during the recruitment period. On the day of colonoscopy (day 0), patients with INR values within therapeutic range or supratherapeutic value up to 0.5 units will proceed through colonoscopy. In case no INR value from day -1 is available, patient ambulatory INR values from the last three months will be reviewed and patient enrolled if >=80% of values were within therapeutic window. The warfarin dose on day 0 will be taken only following the supervision period after the colonoscopy. In case the INR was more than 0.5 above therapeutic range on day -1, therapy will be resumed on day1. As to the colonoscopy, both groups will be scheduled to early morning and will be monitored for clinical follow up for 6 hours at Endoscopy unit. Polyps up to 10 mm (estimated using the open-forceps technique, forceps span=8mm) will be removed with cold or hot methods (according to physician preferences). Hemostatic clips will be applied if deemed necessary according to the performing physician judgement . In case of multiple polyps, the number of polypectomies done will be determined by the endoscopist. For polyps measuring > 10 mm, colonoscopy will be rescheduled (but small polyps will still be removed during the index colonoscopy).Following the procedure, patients will be monitored for 6 hours at the Endoscopy unit. Delayed bleeding will be monitored by the patient on a daily follow up sheet. Ambulatory CBC will be done a week following the colonoscopy . Investigators will make Phone calls for follow up at day 5,10,14,30 . Outpatient clinic follow up will take place on day 30 ± 7 days.Patients will be instructed to contact the investigating physician in case of black stools or BRBPR or otherwise come to ER in case of any continuous bleeding or bleeding perceived to be significant by patient or family.

Study Procedures:Recruitment visit at GI institute outpatient clinics performed at least 1 week ahead of the Colonoscopy;Patient's primary care physician will be contacted by the study coordinator upon signing the informed consent, in order to perform necessary blood test before and after the procedure.; colonoscopy (day 0);Follow up period of 14 days via phone calls and patient sheets;Clinical follow up visit at day 30 Sample Size:The bleeding rate post polypectomy in orally anticoagulated patient undergoing polypectomy of small polyps on Warfarin was reported to be as low as 1% in recent trials (see background). The rate of postpolypectomy bleeding in the general population is 1-2.5% . As bridging was found to increase bleeding compared to no bridging in the largest available meta-analysis, the investigators assume a bleeding rate of 4% in the bridging group. The sample size was estimated based on the complicated rates of delayed bleeding within 2 weeks after polypectomy and was based on detecting equivalence in proportions at the 5% level of significance with a power of 80%, yielding a sample size of 286 (143 each group) .

Statistical Analysis:Descriptive statistics including 95% CIs will be calculated for all baseline variables using means, medians, SDs and interquartile ranges for continuous variables and rates and proportions for discrete variables for each treatment arm. For the primary outcome, clinically significant bleeding, treatment arms will be compared using chi-square test. Baseline characteristics will be compared and, if any clinically significant differences are identified, a logistic regression analysis will be conducted to compare clinically significant bleeding in between the two treatment arms adjusting for these differences. Same analysis will be followed for the secondary outcomes.

Ethical issues:The investigator will ensure that this study is conducted in accordance with the principles of the "Declaration of Helsinki" guidelines and with the laws and regulations of the state of Israel.

Written informed consent must be obtained prior to participation in the study.