The Dutch Society of Gastroenterology and Hepatology and the Society of Internal Medicine previously published separated guidelines on endoscopy in patients on antiplatelets or anticoagulant therapy in 2016. This is a joint update of the previous guidelines and a supplement of the recently updated Dutch National guideline on Antithrombotic Therapy and Interventions.

For all patients on antiplatelets or anticoagulants we state that there is an increased risk of post-procedure bleeding compared to patients not on these drugs. Therefore, the management of antithrombotic drugs is a balance of risks of bleeding from the procedure versus the risk of thrombosis. Bleeding secondary to high-risk endoscopic procedures can often be controlled by further endoscopic therapeutic measures and is rarely fatal. A thrombotic stroke may result in lifelong disability, and a major cardiac event may result in death. Not only do the risks of thrombosis vs bleeding need to be assessed on an individual patient basis, but patients should be fully informed and involved in this decision-making process. The risk of a potentially catastrophic thrombotic event such as a stroke may not be acceptable to a patient even if that risk is very low.  

Aspirin

When given as long-term secondary prevention aspirin reduces vascular events by  approximately one-third and vascular deaths by about one-sixth. In patients on long-term

Low-dose aspirin for secondary prevention, aspirin interruption was associated with a threefold increased risk of cardiovascular or cerebrovascular events, and 70% of these events occurred within 7–10 days after interruption (Biondi-Zoccai, 2006 and Maulaz, 2005). In a randomised controlled trial (RCT) of 220 patients on low-dose aspirin for secondary prevention undergoing non-cardiac surgery, patients were randomised to continuation or temporary replacement of aspirin by placebo (Oscarsson, 2010). Major cardiac events occurred within 30 days in 1.8% of the aspirin group compared with 9% in the placebo group (p=0.02). No difference in haemorrhagic complications was seen between the two groups. The risks related to continuation or discontinuation of aspirin for endoscopy are discussed in the sections on endoscopic procedures, and recommendations made accordingly. With regard to patients on a Direct Oral Anticoagulants (DOAC) and a single antiplatelet agent, such as aspirin, the AFIRE trial randomised patients with atrial fibrillation (AF) and stable coronary artery disease to rivaroxaban alone versus rivaroxaban and a single antiplatelet. They demonstrated that rivaroxaban alone was non-inferior for the primary efficacy endpoint (a composite of thromboembolic events or death from any cause) and superior for the primary safety endpoint of major bleeding. However, this trial was not sufficiently powered to address the risk of stent thrombosis (Yasuda, 2019). For patients with documented coronary artery disease and an indication for formal anticoagulation, for example, AF, then a life-long single antiplatelet such as aspirin is no longer recommended, beyond the antiplatelet duration required postcoronary stent insertion, in the latest ESC guidelines (Knuuti, 2020). However, late stent thrombosis (1–12 months poststent implantation) and very late stent thrombosis (beyond 12 months poststent implantation) occur with a reported incidence of 0.5%–1.0% and 0.2%–0.4%, respectively, and stent thrombosis is associated with a 4%–45% mortality (Gori, 2019). Furthermore, technical aspects of the procedure, for example, left main stents, may have an essential bearing on the subsequent risk of stent thrombosis. The trials addressing whether a long-term antiplatelet agent is needed in patients that have stents and require a DOAC, which suggest a DOAC alone may be safe, are not sufficiently powered to look at the important endpoint of stent thrombosis (Yasuda, 2019 and Lopes, 2019). Many consultant interventional cardiologists in the UK recommend their patients remain on a single antiplatelet agent if they have stents and need a long term DOAC, although this will very much vary on an individual case-by- case basis. Therefore, in patients with coronary stents who are taking a DOAC the cessation of the single antiplatelet agent should always be discussed in advance with the consultant interventional cardiologist responsible for the patient’s care. Following the publication of three RCTs in 2018 the use of aspirin in primary prevention in cardiovascular disease is no longer routinely recommended and should only be considered, on a case-by-case basis, in those with a very high individual risk of cardiovascular events (Dasa, 2021).

P2Y12-receptor antagonists

Antiplatelet drugs have a rapid onset of action, and irreversibly inhibit platelet activity. Clopidogrel, prasugrel and ticagrelor are antagonists of the P2Y12 receptor on platelets. Clopidogrel plus aspirin is more potent than aspirin alone (Budaj, 2002). For clopidogrel, platelet function returns to normal 5–7 days after withdrawal of the drug (Korte, 2011). Prasugrel and ticagrelor are more rapidly acting and more potent platelet receptor antagonists than clopidogrel. Clopidogrel monotherapy may be used following a thromboembolic cerebrovascular accident, or for peripheral vascular disease. P2Y12 receptor antagonists are frequently used as DAPT with aspirin in acute coronary syndrome (ACS) and after placement of coronary stents. Patients on DAPT, with aspirin and a P2Y12 receptor antagonist, particularly in the context of coronary artery stents, are at high risk of thrombosis if drug therapy is discontinued. Without antiplatelet therapy coronary stents are at high risk of occlusion due to thrombosis, and this confers an approximate 40% risk of acute myocardial infarction or death. In one study, discontinuation of DAPT was associated with an HR of 161 for these events (Iakovou, 2005). DAPT is generally prescribed for 12 months after drug eluting stents (DES) implantation, though there have been occasional reported instances of thrombosis after this duration. Prasugrel and ticagrelor are now the main drugs of choice used for ACS in the UK. Prasugrel has been shown in a large UK observational series to be associated with a lower mortality (Olier, 2018). This has been confirmed in a large randomised open-label trial (Schüpke, 2019). The latest NICE guidelines for ACS now recommend prasugrel is used as the first line agent in ACS, unless the patient is on anticoagulant therapy, in which case clopidogrel is recommended. Clopidogrel is generally the first choice of P2Y12 inhibitor in patients undergoing elective percutaneous coronary intervention (PCI) and the current guidelines recommend a 6-month course of DAPT post elective PCI (Knuuti, 2020). Where temporary cessation of P2Y12 inhibitors in patients with stents has been agreed, after discussion with the responsible consultant interventional cardiologist, then stopping the P2Y12 inhibitors 7 days pre procedure will minimize the bleeding risk. However, it should be noted that, depending on the P2Y12 agent, this time can be shortened dependent on the P2Y12 agent in use. In patients with a very high risk of stent thrombosis there may also be a role for bridging either with intravenous cangrelor and/or GIIbIIIa agents. However, the role of such bridging agents is beyond the remit of this guideline and this decision should be made along with the responsible consultant interventional cardiologist on a case-by-case basis (Valgimigli, 2018).

Bare metal stents (BMS), which only require 1 month of DAPT, are now much less frequently used and the national BCIS audit shows they are used in less than 15% of cases in the UK, and many units no longer even stock them. Their main indication is for use in patients with a high bleeding risk or requiring urgent major surgery. Many UK units now use BioFreedom DES, these are polymer-free drug coated stents, which have a license of 1 month of DAPT and are superior to BMS with respect to major safety endpoints including bleeding and restenosis rates (Urban, 2015). Finally, there is increasing evidence that with a number of third generation DESs 1 month of DAPT can safely be used in patients at high bleeding risk without an increase in ischaemic events (Kandzari, 2020; Varenne, 2017 and Watanabe, 2019). A single antiplatelet agent is continued in all cases after discontinuation of DAPT, this tends to be aspirin. In some cases, for example, patients who had a previous stroke, the patients may have aspirin stopped at the end of the DAPT course and clopidogrel continued long-term. The important issue of a single antiplatelet agent in patients on long-term DOACs is discussed in the previous section.

Finally, in patients on a DOAC for AF and stroke prophylaxis who need DAPT after stent implantation, evidence from trials such as the RE-DUAL trial indicate that much shorter courses of triple therapy followed by a course of DOAC and clopidogrel alone are safer than previous practice of triple therapy up to a year (Cannon, 2017). The European Society of Cardiology and NICE guidelines, for both ACS and elective stenting, now support shorter durations of triple therapy followed by a DOAC and clopidogrel as the current standard of care (Knuuti, 2020; Collet, 2021 and Ibanez, 2018).

The duration of DAPT post-PCI may also depend on a number of other factors, beyond the scope of this guideline. Therefore, in patients with coronary stents, we recommend that the endoscopist discusses the plan, with respect to potential DAPT cessation, with a consultant interventional cardiologist. Ideally this should be the consultant interventional cardiologist responsible for the patient’s care, who did the stent procedure.

Warfarin and heparin

Updated literature searches were conducted on the use of warfarin and heparin in patients undergoing endoscopy. There were no new data to indicate a change to the existing protocols, as described in the previous guideline (Veitch, 2016).

DOACs

DOACs have been a major advance in anticoagulant therapy. Dabigatran is an inhibitor of thrombin, and rivaroxaban, apixaban and edoxaban inhibit FXa. They do not need routine

monitoring, and INR and activated partial thromboplastin time (aPTT) are unreliable indicators of anticoagulant activity. Unlike warfarin they have a rapid onset of action and full anticoagulant activity is established within 3 hours of the first dose. They have relatively short half-lives, but these may be prolonged in renal failure, particularly for dabigatran. In patients undergoing a high-risk procedure with a low thrombotic risk we recommend

that the last dose of rivaroxaban, apixaban or edoxaban is taken 3 days before the procedure. Dabigatran may have to be stopped for longer than this, however, when renal function is significantly reduced (Schulman, 2015).

For patients on dabigatran with CrCl of 30–50 mL/min we recommend that the last dose of the drug is 5 days before the procedure. Dabigatran therapy is contraindicated in patients

with CrCl<30 mL/min. eGFR is a suitable alternative measurement of renal function, and the same numerical values apply for the purposes of these guidelines. If a patient on any DOAC is clinically deteriorating, his/her renal function should be checked before the procedure. These recommendations are supported by the findings of the PAUSE trial (Douketis, 2019). Three thousand and seven patients on apixaban, dabigatran or rivaroxaban due for elective interventional procedures had a standardised interruption of therapy: last dose of drug 2 days before low-risk procedures (including diagnostic GI endoscopy) and 3 days before high-risk procedures (including high-risk therapeutic GI endoscopy; e.g., polypectomy). This was extended to last dose 5 days before a high-risk procedure for dabigatran patients with CrCL<50 mL/min. Resumption of therapy occurred at 1 day after low bleeding risk procedures and 2–3 days after high bleeding risk procedures. Low rates of major bleeding or arterial thromboembolism were observed. Low rates of venous thromboembolism were also observed although this was not a primary study outcome. We have maintained the recommendation to omit the morning dose of DOAC prior to low-risk endoscopic

procedures, although the PAUSE trial demonstrates the safety of omitting the DOAC for 1 day before a low-risk endoscopic procedure if desired.

Bridging of anticoagulant therapy

Unfractionated heparin (UFH) and LMWH (LMWH) have short half-lives compared with warfarin and can be employed as an anticoagulation ‘bridge’ while warfarin is temporarily discontinued for endoscopic procedures considered to have a significant risk of bleeding. UFH is administered by continuous intravenous infusion and LMWH by subcutaneous injection once or two times a day. The former requires an inpatient stay in hospital while warfarin is discontinued, and then re-introduced, and also requires frequent monitoring of aPTT; the latter can often be managed in an outpatient setting without monitoring of anticoagulation levels. Some cardiologists have a preference for UFH for bridging warfarin in patients with metal heart valves. A meta-analysis demonstrated higher rates of bleeding in patients with mechanical heart valves bridged with LMWH versus UFH, but numbers were small in the LMWH studies (Passaglia, 2015). A small multicentre registry study found no difference in adverse events between patients bridged with UFH or LMWH in this context (Spyropoulos, 2008), and bridging with LMWH is now commonplace. Guidelines from the American Heart Association and American College of Cardiology do not recommend one strategy over the other (Otto, 2021). Prosthetic metal heart valves in the mitral position are at particularly high risk of thrombosis if warfarin is temporarily discontinued for 5–7 days. Heparin bridging for patients with a bi-leaflet mechanical aortic valve replacement in the absence of AF is considered unnecessary in guidelines from the British Society of Haematology (Keeling, 2016) and from the American College of Cardiology/American Heart Association (Nishimura, 2017), but it is recommended in guidelines from the ESC and the European Association for Cardio-Thoracic Surgery (Baumgartner, 2017). There are no high quality studies to inform us, and consequently levels of evidence are low quality in all three guidelines. We have now placed mechanical aortic valve replacement in the high-risk category requiring bridging, in line with European guidelines, but this should be decided in conjunction with local cardiology or cardiothoracic surgery services given the uncertainty. This should always be discussed with the consultant cardiologist responsible for the patient’s care. The risk of thromboembolism with AF increases with additional cardiovascular factors such as hypertension, heart failure and diabetes and this risk has been quantified by the CHADS2 score (annual risk of stroke increasing from 1.9% with a score of 1 to 18.2% with a score of 6). This has been updated with the CHA2DS2VASc scoring system in which the annual risk of stroke increases from 1.3% with a score of 1 to 15.2% with score of 9. A randomised, prospective, double-blind placebo-controlled trial was conducted in 1884 patients with AF undergoing a variety of operative procedures including approximately 50% GI endoscopy (Douketis, 2015). Patients were randomised to LMWH or placebo, and risk factors were well matched. Thirty-eight percent of the patients had CHADS2 scores>3, <2% had mitral stenosis and ≤3.4% had CHADS2 scores of 5 or 6. There was no significant difference in rates of thromboembolism between the LMWH and placebo groups, but there was a significant increase in major haemorrhagic events in the LMWH group versus placebo (3.2% vs 1.3%). Caution should be exercised when interpreting the results in the high-risk thromboembolic groups as the study was not designed or statistically powered to examine these categories. The previous BSG/ESGE guidelines (Veitch, 2016) do not recommend bridging for non-valvular AF, ASGE guidelines (Acosta, 2015) recommend bridging with LMWH for CHA2DS2VASc>2 and the APAGE/APSDE guidelines (Chan, 2018) recommend this for CHA2DS2VASc>5. CHADS2 scores are unfortunately not directly equivalent to CHA2DS2VASc scores. Further research on the benefits of heparin bridging is required in high-risk patients with non-valvular AF on warfarin in order to determine the optimum approach, but it would be reasonable to consider bridging patients with CHADS2 scores >5 as recommended by the British Society of Haematology (Keeling, 2016). This applies to patients with AF and a previous stroke or transient ischaemic attack (TIA), and three of the following factors: congestive cardiac failure, hypertension (>140/90 mm Hg or on antihypertensive medication), age>75 years or diabetes mellitus. Heparin bridging is also recommended for patients with AF who have had a stroke or TIA within 3 months (Keeling, 2016). Bridging with LMWH has also been studied in patients on DOACs. In a German registry, heparin bridging versus no bridging led to a higher rate of major bleeding (2.7% vs 0.5%, p=0.01) with no reduction in thromboembolism (Beyer-Westendorf, 2014). In the RELY trial bridging of dabigatran with LMWH resulted in higher major bleeding rates compared with no bridging (6.5% vs 1.8%, p<0.001) with no difference in thrombosis rates between the groups (Douketis, 2015). In a Japanese study of 16 977 patients on warfarin or DOACs undergoing a variety of high-risk endoscopy procedures a propensity matched analysis demonstrated a significant increase in postprocedural GI-bleeding and thromboembolism in patients who were bridged with heparin (Nagata, 2018). It should be noted that all patients were bridged with UFH rather than LMWH. For colonoscopy in patients on warfarin who were bridged with LMWH, several studies have demonstrated an increase in postpolypectomy bleeding without a decrease in thromboembolic events (Ishigami, 2017; Lin, 2018 and Ono, 2019). Bridging with LMWH should, of course still be advocated for those patients on warfarin at high risk of thromboembolism, but patients should be advised of the increased risk of post-procedure bleeding. The safety of temporary cessation of DOAC therapy, without bridging, is supported by the findings of the PAUSE trial (Douketis, 2019). Patients with a history of venous thromboembolism within 3 months of commencing anticoagulant therapy are at high risk of recurrent thrombosis if anticoagulation is interrupted. Most of these patients are now commenced on a DOAC rather than warfarin, and bridging would not be recommended. Ideally, we should not interrupt anticoagulation in this high-risk group due to the risk of thrombosis; an elective low-risk procedure could be performed without interrupting anticoagulation if necessary, but it may be preferable to defer a high-risk procedure beyond 3 months anticoagulation if safe to do so. Patients with thrombophilia syndromes should be discussed with a haematologist. Factor V Leiden and the common prothrombin mutation F2G20210A are low-risk thrombophilias and bridging is not required. Patients with deficiencies of antithrombin, protein C or protein S are at higher risk of thrombosis, but in most of these patients bridging therapy will not be required.

To answer the clinical question, the ESGE and BSG guideline was used: Endoscopy in patients on antiplatelet or anticoagulant therapy: British Society of Gastroenterology (BSG) and European Society of Gastrointestinal Endoscopy (ESGE) guideline update (Veitch, 2021), which is an update of the previous version (Veitch, 2016). Development of this guideline comply with the requirements of the National Institute for Health and Care Excellence (NICE). The literature review was adopted. No additional systematic review of the literature was performed.