Periprocedural Management of Antithrombotics - Do Not Hold if you Can!

The decision to hold or not to hold antithrombotic (AT) medications for an endoscopic procedure rests on the balance of the procedural bleeding risk and the patient’s risk of suffering a thromboembolic event (TE), including stroke, pulmonary embolism, deep vein thrombosis, angina, or a myocardial infarction. Guidelines have categorized procedures into those with high and low bleeding risk. AT may be continued if the bleeding risk is low. If the bleeding risk is high, AT medications should be held or modified. Management of AT medications is not always clear and can be especially challenging for patients who undergo a high-risk procedure and who also have a high risk for TE. For instance, the optimal time to reinitiate AT medications after a resection of a large polyp is yet uclear. In contrast, periprocedural management of AT seems simple if the procedural bleeding risk is low, AT, including anticoagulants, do not need to be stopped. However, the idea to continue anticoagulants even for a low-risk procedure may be foreign to many of us. After all, we may believe that the risk of TE is low with only transient interruption of anticoagulation, and there have been little data to convince us otherwise. The study by Li et al1Li Y.K. Guo C.-G. Cheung K.S. et al.Risk of postcolonoscopy thromboembolic events: a real-world cohort study.Clin Gastroenterol Hepatol. 2022; (XX:XX-XX)Abstract Full Text Full Text PDF Scopus (1) Google Scholar in this issue of Clinical Gastroenterology and Hepatology provides important insights into the risk of TE and bleeding risk among patients on AT medications undergoing a colonoscopy. This single-center retrospective cohort study enrolled 6220 patients; of these, 1755 (28%) were on any AT, 19% were on antiplatelet agents, and 8% were on anticoagulation (4% on warfarin and 4% on direct oral anticoagulants). As the main result, Li et al1Li Y.K. Guo C.-G. Cheung K.S. et al.Risk of postcolonoscopy thromboembolic events: a real-world cohort study.Clin Gastroenterol Hepatol. 2022; (XX:XX-XX)Abstract Full Text Full Text PDF Scopus (1) Google Scholar reported a 0.3% risk of TE (n = 20) among all patients in the cohort within 30 days after the procedure. The risk was significantly greater among patients who were on AT medications compared with those who were not (0.9% vs 0.1%). Among patients with a high risk for TE (eg, congestive heart failure, hypertension, age, diabetes, stroke/transient ischemic attack, vascular disease, ≥5) a TE occurred in 2.9% (n = 10). Most importantly, among patients on AT medications, almost all TEs occurred after periprocedural interruption of AT medications (except for 2 patients on aspirin). Six patients died as a complication of a TE. Postprocedure bleeding was more frequent among patients on AT medications compared with those not on AT medications (1.1% vs 0.1%), but no patient died as a complication of bleeding. The strengths of the study include its large sample size, the use of a regional database that covers all hospitals in the region (Hong Kong) to capture all postprocedure events, individual chart review on the management of AT medications of all patients who had a TE, and the inclusion of any AT medications (antiplatelets and anticoagulants). Limitations include the retrospective study design using administrative data with its potential for selection bias and lack of information on the timing of resumption of AT medications, procedure indication (30% were defined as “other”), and mode of polyp resection (eg, hot or cold resection). Detailed management of AT medications was known only for those patients with TE who underwent chart review, for the remaining patients it was assumed that AT medications were managed according to current guidelines. Before considering, how the results may help us in our practice, it is important to review some of the reported findings. First, the rate of TE may seem high. Almost 1 in 100 patients suffered a postprocedure TE among patients on AT medications. However, these rates are in line with previous studies reporting a rate between 1% and 3%,2Lau L.H. Guo C.L. Yip T.C. et al.Risks of post-colonoscopic polypectomy bleeding and thromboembolism with warfarin and direct oral anticoagulants: a population-based analysis.Gut. 2022; 71: 100-110Crossref PubMed Scopus (17) Google Scholar,3Rodriguez de Santiago E. Sanchez Aldehuelo R. Riu Pons F. et al.Endoscopy-related bleeding and thromboembolic events in patients on direct oral anticoagulants or vitamin K antagonists.Clin Gastroenterol Hepatol. 2022; 20: e380-e397Abstract Full Text Full Text PDF PubMed Scopus (6) Google Scholar which highlights the importance of adequate periprocedural management of AT medications. Li et al1Li Y.K. Guo C.-G. Cheung K.S. et al.Risk of postcolonoscopy thromboembolic events: a real-world cohort study.Clin Gastroenterol Hepatol. 2022; (XX:XX-XX)Abstract Full Text Full Text PDF Scopus (1) Google Scholar pointed out that no TE occurred among patients who underwent a screening or surveillance colonoscopy. However, for almost one third of patients the indication for the procedure remained unclear and it therefore is difficult to generalize the results to our screening and surveillance patients. Second, most TEs occurred in patients with an indication of rectal bleeding as the only listed bleeding indication. This finding raises concern for unmeasured confounding. In clinical practice we may hold AT medications among patients with gastrointestinal bleeding until the underlying bleeding source is identified. In such situations, AT medications may be interrupted for longer than just the periprocedural time. It also is plausible that in some of these patients AT medications may not have been restarted because of the concern for rebleeding. A longer interruption of AT medications translates into a greater TE risk. Third, the investigators provided a wealth of additional results through subanalyses and adjusted analyses. However, the total number of TE events was only 20, and the number of patients in these subgroups often was low (fewer than 5). The precision of reported percentages in these subgroups therefore is poor with wide 95% CIs. For instance, the TE risk for patients on dual antiplatelets was reported to be 4.7% (2 TE events among 43 patients), with an adjusted odds ratio of 28 (95% CI, 3.8–144). Some also may question the validity of performing adjusted analyses given such low event rates with wide CIs. The generalizability of these findings therefore is limited. Finally, it is relevant to recognize what risks were compared. The chosen reference groups were patients who were not on any AT medications. In other words, the study compared the risk among patients on AT with those who were not on AT medications. The study does not compare TE risks between patients who stopped their AT medications with those who continued their AT medications for the duration of the procedure, which would be the clinically relevant comparison. Of course, the number for such comparisons are far too few to calculate adjusted risks (for the earlier-mentioned reasons). Nevertheless, and despite these issues, the study provides important albeit descriptive results. First, the risk for TE was very high among high-risk TE patients (almost 3%). And second, if antithrombotics were continued, a TE was very unlikely. These observations alone are convincing enough to not interrupt AT medications for low-risk procedures and optimally modify AT periprocedural management for high-risk procedures (eg, heparin bridge or changing from double- to single-platelet agents). The findings of the study also urge us to clarify our definition of what constitutes a low bleeding risk procedure that would allow continuation of AT medications. For instance, cold snare polyp resection may no longer be considered a high-risk intervention as it has been defined historically when hot snare polypectomy was the standard. An increasing number of studies support a low-risk determination for cold snare polyp resection; however, guidelines have not addressed this question yet.4Takeuchi Y. Mabe K. Shimodate Y. et al.Continuous anticoagulation and cold snare polypectomy versus heparin bridging and hot snare polypectomy in patients on anticoagulants with subcentimeter polyps: a randomized controlled trial.Ann Intern Med. 2019; 171: 229-237Crossref PubMed Scopus (53) Google Scholar, 5Horiuchi A. Nakayama Y. Kajiyama M. et al.Removal of small colorectal polyps in anticoagulated patients: a prospective randomized comparison of cold snare and conventional polypectomy.Gastrointest Endosc. 2014; 79: 417-423Abstract Full Text Full Text PDF PubMed Scopus (215) Google Scholar, 6Acosta R.D. Abraham N.S. Chandrasekhara V. et al.Asge Standards of Practice Committee: the management of antithrombotic agents for patients undergoing GI endoscopy.Gastrointest Endosc. 2016; 83: 3-16Abstract Full Text Full Text PDF PubMed Scopus (410) Google Scholar, 7Chan A. Philpott H. Lim A.H. et al.Anticoagulation and antiplatelet management in gastrointestinal endoscopy: a review of current evidence.World J Gastrointest Endosc. 2020; 12: 408-450Crossref PubMed Google Scholar, 8Veitch A.M. Radaelli F. Alikhan R. et al.Endoscopy in patients on antiplatelet or anticoagulant therapy: British Society of Gastroenterology (BSG) and European Society of Gastrointestinal Endoscopy (ESGE) guideline update.Endoscopy. 2021; 53: 947-969Crossref PubMed Scopus (22) Google Scholar The study by Li et al1Li Y.K. Guo C.-G. Cheung K.S. et al.Risk of postcolonoscopy thromboembolic events: a real-world cohort study.Clin Gastroenterol Hepatol. 2022; (XX:XX-XX)Abstract Full Text Full Text PDF Scopus (1) Google Scholar provides valuable data to support continuation of AT medications among patients undergoing an endoscopic procedure, whenever the procedure risk allows. For our practice it is essential to clarify, ideally with societal guidance, what constitutes a low-risk procedure and to ensure that AT medications are not held erroneously.