SAN DIEGO – Patients who receive heparin bridging during an interruption of oral anticoagulation appear to be at a 5.4-fold increased risk of overall bleeding and a 3.6-fold increased risk of major bleeding, without a reduction in risk of thromboembolic events.
Those are key findings from a systematic review and meta-analysis of recently published medical literature presented by Dr. Jovana Yudin at the annual meeting of the American Society of Hematology.
Antithrombotic and thrombolytic therapy guidelines published by the American College of Chest Physicians in 2008 recommended bridging according to an individualized approach (Chest 2008;133[suppl. 6]:299S-339S).
"They suggested bridging according to patients’ bleeding and thromboembolic risk," said Dr. Yudin, a fellow in the hematology residency program at McMaster University, Hamilton, Ont. "Within the last decade, several new studies have been published using periprocedural bridging. In these studies, low-molecular-weight heparin has been used with increased frequency. However, optimal strategies for bridging remain unclear. Our objective was to do a systematic review and meta-analysis of bridging trials published in the last decade to look at thromboembolic risk as well as bleeding risk."
Dr. Yudin and her associates searched the MEDLINE, EMBASE, and Cochrane Collaboration databases for systematic reviews and meta-analyses of studies published between Jan. 1, 2001, and July 31, 2010, that examined bleeding and thromboembolic events in patients receiving bridging therapy during temporary oral anticoagulation interruption for elective surgical or invasive procedures. Studies were excluded if the reporting of thromboembolic or bleeding outcomes was unclear, or if they focused exclusively on patients with renal failure (ASH 2011; abstract 545). All studies were reviewed by two independent investigators.
The researchers identified and screened 1,164 studies for review. Of these, 35 studies that included 7,169 bridged patients were selected for the final review. Most of the studies (33) were observational, and only two were randomized. The median follow-up was 30 days.
The most common indication for anticoagulation was atrial fibrillation (44%), followed by mechanical valve (24%), prior venous thromboembolism (22%), and other (10%).
The most common preoperative strategy was to discontinue oral anticoagulation more than 3 days in advance. Low-molecular-weight heparin was most commonly used, both preoperatively and postoperatively.
Dr. Yudin reported that thromboembolic events, which were primarily arterial in nature, occurred in 73 of the 7,169 patients (mean rate, 0.96%). The mean rate of overall bleeding was 13.01%, whereas the mean rate of major bleeding was 4.32%.
Eight of the studies included in the final analysis had control groups from which the researchers were able to pull data to determine an odds ratio for thromboembolism with bridging vs. no bridging. These studies included 1,691 bridged patients and 3,493 nonbridged patients. The odds ratio for thromboembolic events was 0.80, with a 95% confidence interval (CI) of 0.42-1.54, "suggesting no risk reduction for thromboembolic events with heparin or low-molecular-weight bridging," Dr. Yudin said. "There was also no difference between these two groups in the risk for arterial or venous thromboembolism."
To determine the risk of overall bleeding, the researchers pulled data from 13 studies that included control groups. These studies included 1,935 bridged patients and 5,160 nonbridged patients. The odds ratio for overall bleeding with bridging was 5.40 (95% CI, 3.00-9.74). "This suggested an increased risk of overall bleeding with bridging anticoagulation, but there was significant heterogeneity noted across these studies."
For major bleeding, five studies with control groups were assessed. These included 1,397 bridged patients and 2,104 nonbridged patients. The odds ratio for major bleeding was 3.60 (95% CI, 1.52-8.50), "again suggesting an increased risk in major bleeding with bridging," she said. "There was significant heterogeneity noted across studies."
Dr. Yudin acknowledged that the review had certain limitations: Most of the studies included in the analysis were observational, and only about a third had control groups. "Our control groups consisted largely of low-thromboembolic-risk patients, or patients who were not chronically on vitamin K antagonists, suggesting that they had a different risk profile for thromboembolism than many of the bridged patients," she said.
The findings "underline the need for studies of higher [methodological] quality in periprocedural bridging," she concluded. "It also tells us that there is a need for standardized definitions in terms of outcomes. We suspect that much of our heterogeneity had to do with varying definitions for outcomes such as major bleeding."
Dr. Yudin said that she had no relevant financial conflicts to disclose.