Antiplatelet agents and anticoagulants make up the larger group of antithrombotic medications. These drugs have seen increasing use worldwide as populations in developed countries age, and ischemic heart disease prevalence rises. The antiplatelet agents are comprised of aspirin, the non-steroidal anti-inflammatory drugs (NSAIDs), the thienopyridines (e.g., clopidogrel, prasugrel & ticlodipine) and the glycoprotein IIb/IIIa receptor inhibitors. Anticoagulants include, warfarin, heparin and the low-molecular-weight heparins.

Antithrombotic medications reduce the risk of thromboembolic events in susceptible individuals, but increase the risk of gastrointestinal bleeding.¹ Cessation of antithrombotic drugs prior to endoscopic therapy has been proposed in a number of studies, aimed at reducing the risk of immediate and early bleeding. Interruption of antithrombotic therapy is however associated with cardiovascular risk; it has been estimated that about 5% of hospitalizations for acute coronary syndrome are due to discontinuation of antiplatelet therapy in patients undergoing a non-cardiovascular procedure.² The peri-endoscopic management of patients at high thromboembolic risk therefore requires knowledge of both the bleeding risk associated with endoscopic procedures, and the potential risks of stopping antithrombotic therapy.³

Three major endoscopy organizations have published guidelines aimed at providing a rational strategy for the endoscopist in managing the individual patient on antithrombotic medication. This paper will compare and contrast the approach of each guideline, in an attempt at consensus.

The published guidelines, in chronological order of publication, are: (1) “Guidelines for the management of anticoagulant and antiplatelet therapy in patients undergoing endoscopic procedures” published by the British Society of Gastroenterology (BSG) in 2008⁴; (2) “Management of antithrombotic agents for endoscopic procedures”, which is an update of two guidelines and was published by the American Society of Gastrointestinal Endoscopy: (ASGE) in 2009⁵; and (3), “Endoscopy and antiplatelet agents” by the European Society of Gastrointestinal Endoscopy (ESGE) published in 2011.²

The BSG and ASGE guidelines address the use of both anti-platelet agents and anticoagulants during the peri-endoscopic period, while the ESGE guideline is focused solely on antiplatelet medication. All three guidelines stratify the risk of bleeding by procedure type (low vs. high bleeding risk) and the risk of thromboembolic events (low vs. high risk) arising from discontinuation of therapy. It is noteworthy that these guidelines were formulated with reference mainly to observational studies and expert opinion, and many of the recommendations made therefore have a limited basis in evidence.

The guidelines also vary by the depth into which specific endoscopic procedures are discussed with respect to associated bleeding risk. The BSG does not discuss individual procedures. The ASGE discusses diagnostic endoscopy, colonic polypectomy, endoscopic sphincterotomy and percutaneous endoscopic gastrostomy (PEG). The ESGE guideline carries the most detailed discussion of the bleeding risks associated with all the procedures highlighted by the ASGE, but in addition reviews endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD), EUS-FNA, endoscopic stent placement and dilation, device-assisted enteroscopy, endoscopic variceal ligation, and argon plasma coagulation (APC) hemostasis. Table 1 summarizes bleeding risk stratified by type of endoscopic procedure. Tables 2 and 3 stratify cardiovascular conditions by thromboembolic risk when antiplatelet agents and anticoagulant therapy are discontinued, respectively.

All three guidelines are in agreement that aspirin is to be continued. The BSG and ASGE recommend continuing thienopyridines. The ESGE however recommends stopping the thienopyridines for these particular scenarios although they are deemed “low-risk” for bleeding: removal of subcentimeter colonic polyps, stricture dilation, enteral stent placement, APC and EUS-FNA of solid lesions. Except for enteral stent placement (ASGE), these are considered high-risk procedures by the other two societies. For anticoagulants, BSG and ASGE guidelines advocate continuation at therapeutic international normalized ratio (INR).

In the main, these are procedures involving endoscopic resection and cautery, plus procedures with the potential to induce bleeding that is inaccessible to endoscopic therapy. The three guidelines each dichotomize the approach between conditions at low thromboembolic risk, and those at high thromboembolic risk, with respect to use of aspirin, thienopyridines and anticoagulants (the ESGE guideline deals only with antiplatelet agents). The general recommendation with regards to low thromboembolic risk conditions is to stop the therapy concerned before endoscopy (thienopyridine 5?7 days before, warfarin 5 days before). For conditions associated with high thromboembolic risk the societies advise delaying endoscopy until the thienopyridine/dual antiplatelet therapy course has ended, or stopping the medication temporarily in consultation with the managing cardiologist. Consideration should also be given to performing an alternative or temporizing procedure associated with lower bleeding risk, if possible. Aspirin should be maintained in all cases, or used in place of the thienopyridine.² In patients taking warfarin, both BSG and ASGE recommend stopping warfarin 5 days before the procedure, and bridging with low-molecular-weight heparin.

Prior to performing an endoscopic procedure for a patient on antithrombotic therapy, one should first consider the risks of a thromboembolic event related to interruption of antithrombotic medication, and second, bleeding related to endoscopic therapy while on antithrombotic medication. One should be mindful that a thromboembolic event that may occur following withdrawal of medication can be devastating, whereas bleeding after high-risk procedures, although increased in frequency, is rarely associated with significant morbidity or mortality.⁵

Conditions carrying a higher risk of thromboembolic events if antithrombotic therapy is interrupted include atrial fibrillation associated with valvular heart disease, mechanical valves in the mitral position, and mechanical valves in patients who have had a previous thromboembolic event. Study data indicate that the absolute risk of an embolic event for patients in whom anticoagulation is interrupted for 4 to 7 days is about 1%.^6,7

Patients with coronary stents are at high risk of stent thrombosis when dual antiplatelet therapy is discontinued before the minimum duration specified by the American College of Cardiology (ACC); 1 year for drug-eluting stents, and 1 month for bare metal stents.⁸ One large prospective study reported a hazard ratio of 89 for stent thrombosis when antiplatelet therapy was discontinued prematurely.⁹ Case fatality is extremely high, at 20% to 45%.^3,9 Patients who require dual antiplatelet therapy should always be kept on aspirin, and the decision to discontinue the thienopyridine should be taken in consultation with the attending cardiologist.

The peri-endoscopic bleeding risk for patients on antithrombotic therapy may be considered by procedure; this review will focus on the procedures at higher risk of bleeding for which more evidence exists. For colonoscopic polypectomy while on aspirin or NSAIDs, the bleeding risk appears to be small.¹⁰ Warfarin use is associated with increased bleeding risk, as is resumption of anticoagulation within 1 week of polypectomy.^10,11 Several studies of prophylactic endoclip application suggest that it keeps bleeding rates low in anticoagulated patients, but current evidence is insufficient for its routine use to be recommended.^12,13 The risk of bleeding after endoscopic sphincterotomy (ES) is 0.3% to 2%.¹⁴?¹⁶ Withdrawal of aspirin does not appear to reduce this risk¹⁷; anticoagulation with warfarin or heparin however increases the risk of post-sphincterotomy bleeding.¹⁸ Large balloon papillary dilation in combination with ES (to avoid mechanical lithotripsy for the removal of large biliary calculi) is associated with higher bleeding risk than ES alone. In PEG placement, the overall risk of bleeding is about 2.5%,¹⁹ but the additional risk conferred by antithrombotic therapy is not known. EMR is done less frequently in the West than in the East; Western series report higher bleeding rates of 4.6% to 12%. Duodenal polypectomy is associated with higher risk of bleeding than polypectomy at other sites, and is reported as 3.1% to 11.6% in the five most recent prospective studies using EMR techniques. The technique of endoscopic ampullectomy is similar to EMR, and is associated with bleeding risk of 5.6% from five large retrospective series.² A recent meta-analysis showed that ESD has a two-fold risk of bleeding when compared with EMR (OR 2.20; 95% CI, 1.58?3.07). Antiplatelet agents and anticoagulants were routinely stopped ahead of all studies involving these high-risk procedures. Appendix summarizes the approach of each of the three guidelines towards the management of antithrombotic agents for endoscopic procedures at high risk for periprocedural bleeding.

The ASGE guideline is the only one that discusses this frequently-encountered scenario in depth. An estimated 1% to 3% of patients with acute coronary syndrome (ACS) will have an associated gastrointestinal (GI) bleed, and these individuals are expected to have a four- to seven-fold increase in the risk of in-hospital mortality over those without GI bleeding.⁵ The overall risk of peri-procedural complications associated with upper GI endoscopy is about 1% to 2% (1% with colonoscopy),^20,21 but may be as high as 12% for endoscopy done on the same day as the acute cardiac event.²² The data in this setting however remains scanty. A decision analysis showed that upper endoscopy before cardiac catheterization was beneficial in patients who presented with overt GI bleed in the setting of ACS, significantly reducing overall mortality.²³ The ASGE suggests withholding antiplatelet agents until hemostasis is achieved, but qualifies that no strong recommendation can be made.

All three guidelines stratify patients by thromboembolic risk if antithrombotic therapy needs to be interrupted, but differ in the detail with which they discuss the individual endoscopic procedures. The ESGE guideline deals only with antiplatelet agents. There is general agreement among them with respect to anti-platelet therapy continuation for individuals at low risk of thromboembolism and for continuation of anticoagulant use, but broad differences are seen in the ESGE review of specific endoscopic therapies. The decision to interrupt antithrombotic therapy has to be individualized in consideration of the patient’s condition and procedure risks. In patients with coronary stent placement of duration shorter than the appropriate minimum stipulated by the ACC, the high risk of a thrombotic event with devastating consequences mandates that the decision to interrupt antiplatelet therapy is taken in consultation with the attending cardiologist.