New Data Suggests Bridging Patients on Anticoagulants During Surgery Should Stop
ARTICLE IN BRIEF
Investigators reported that bridging atrial fibrillation patients on anticoagulant therapy who are undergoing elective surgery did not reduce the risk of stroke compared with those who went into surgery without bridging (receiving saline).
Bridging patients on anticoagulant therapy who are headed into elective surgery has been standard practice for decades. It was thought, and then routinely accepted, that replacing oral therapy of warfarin with subcutaneous doses of low-molecular-weight heparin a few days before and after surgery would reduce the risk of stroke and bleeding.
But some clinicians thought the evidence was weak and wondered whether this very common intervention was really helping, or whether it was just creating more burden for patients, who needed to learn how to inject themselves with the heparin. It may also have increased their fears of stroke and bleeding problems during and following surgery.
Now, the decades-old practice has finally been tested in patients with atrial fibrillation on anticoagulant therapy, and the results showed that it did not reduce the risk of stroke compared with those who went into surgery without bridging (receiving saline).
Surprisingly, the scientists also found that bridging significantly increased the rate of serious bleeding. These findings, published in the June 22 online edition of The New England Journal of Medicine (NEJM), should put an end to this practice, at least for patients with atrial fibrillation heading into surgery, said Thomas L. Ortel, MD, PhD, chief of the division of hematology at Duke University and principal investigator of the clinical coordinating center for the study.
“This whole concept of bridging has been around for decades, but it hasn't been clear whether we really needed to do it,” said Dr. Ortel. “Now, we have definitive data that we do not, at least in patients with atrial fibrillation.”
The study only focused on patients with atrial fibrillation, as this group represents the largest number of people on anticoagulants. It is estimated that 250,000 patients with atrial fibrillation have elective surgery every year. The BRIDGE study did not enroll patients with deep vein thrombosis, prosthetic heart valves, or those who have had previous strokes, so the jury is still out on whether it is necessary to use a bridging method to reduce the risk of stroke or bleeding during surgery in these patients.
STUDY METHODOLOGY, FINDINGS
Dr. Ortel and his colleagues applied for a federal grant a decade ago, and it took almost six years to enroll 1,884 patients at 108 centers in the United States and Canada. Patients were blindly randomized to receive bridging anticoagulation therapy with low-molecular-weight herparin (100 IU of dalteparin per kilogram of body weight) or placebo administrated subcutaneously twice a day from three days before surgery until 24 hours before the actual procedure. Then, the subcutaneous doses were continued for five to 10 days after the procedure. Warfarin was stopped five days before the surgery and resumed again 24 hours after surgery. The primary outcomes (assessed at post-op day 30) were stroke, systemic embolism or transient ischemic attack, and major bleeding.
The investigators reported that there was no significant difference in incidence of thromboembolism in those who received bridging with active heparin and those given a placebo (0.4 percent in the no-bridging group and 0.3 percent in the bridging group). The p value was 0.01 for non-inferiority. The incidence of major bleeding was 1.3 percent in the non-bridging group and 3.2 percent in the bridging group (p=0.005 for superiority). There were 12 serious bleeding events in the 950 patients assigned to receive no bridging and 29 events in the 934 patients randomized to the bridging arm of the study. There was also less minor bleeding in patients who were not bridged.
“These findings are very important because they show that bridging does not improve the stroke outcome and it increases the risk of bleeding,” said Dr. Ortel. “Bridging has no benefit and it can potentially do significant harm.”
“The rationale for the use of bridging anticoagulation therapy has been anchored on the premise that the associated higher bleeding risk was clinically acceptable because it would be offset by a lower risk of perioperative arterial thromboembolism,” the study authors wrote. This risk, they added, “may have been overstated...The premise that warfarin interruption leads to rebound hypercoagulability and that the milieu of the procedure confers a pro-thrombotic state, which in turn leads to arterial thromboembolism, is not supported by the results of the trial.”
The results of the study are specific to patients with atrial fibrillation who take warfarin, Dr. Ortel noted. A growing number of clinicians are now prescribing newer generations of blood thinners, and this study does not address the question of whether these patients will also be at higher risk for serious bleeding problems.
While many clinicians opt to bridge, a growing number have not been convinced that it protects patients against strokes and bleeding. Clinicians have used CHADS2 prediction scores to help stratify their patients. This risk assessment tool gives one point for congestive heart failure, one point for hypertension, one point for age over 75, one point for diabetes mellitus, and two points for stroke. The risk assessment often includes the underlying disease requiring blood thinners, the presence of cardiovascular risk factors, and the type of surgery. Patients enrolled in the current study had a CHADS2 score of two or three.
“We believe that the findings from the BRIDGE study will lead to a change in clinical practice,” said Alex C. Spyropoulos, MD, an associate professor of medicine at Hofstra North Shore-LIJ School of Medicine, a co-investigator who collaborated on the study when he was at McMaster University in Canada. The first author of the study is James D. Douketis, MD, a professor in the department of hematology and thromboembolism at McMaster.
The study investigators said that a similar study should be conducted on patients taking the newer generation of blood thinners, as well as patients with other medical conditions who are prescribed anticoagulant therapy.
“There are many things that are done in medicine because it made sense to do them,” said Gregory W. Albers, MD, the Coyote Foundation professor of neurosurgery at the Stanford University Medical Center. “Bridging seemed very reasonable. Patients were nervous when they thought that we were stopping their anticoagulants. They worried that they would have a stroke. Now, we have information that will help ease these fears. New guidelines will come out and there will be very different recommendations than we have now.”
He noted, however, that the NEJM study should not change the practice of bridging for patients at high risk for stroke and bleeding. The Stanford group does not bridge patients with a CHADS2 score of one to two, he said.
Robert McBane, MD, a professor of medicine at the Mayo Clinic in Rochester, MN, said that clinicians at Mayo have not been bridging patients with a low risk for stroke either. “In the NEJM study, the average CHADS2 score was two. We would not have done bridging on these patients,” he said. Still, he added, “this is an important paper because it shows what we have thought. We looked at our experience in 2007 and could show no benefit of bridging with a CHADS2 score of two. Ours was a cohort study, and this is the first randomized trial.”
These findings are “consistent with meta-analyses of observational studies,” said Larry B. Goldstein, MD, a professor and chairman of neurology at the University of Kentucky and co-director of the Kentucky Neuroscience Institute. “The BRIDGE trial found that bridging therapy did not reduce embolic complications but did lead to an increased risk of major bleeding in patients with atrial fibrillation who required temporary interruption of warfarin treatment for an elective operation or invasive procedure.
“There are, however, important caveats that need to be considered,” he said. “Patients who had stroke or transient ischemic attack [TIA] within 12 weeks of the procedure were excluded. Only 8-10 percent of subjects had a history of prior stroke and only 8 percent a history of TIA. Almost no subjects had a CHADS2 score of 5-6. The balance between the risks of stroke and the risks of bleeding with interruption of anticoagulation in these groups of patients at very high stroke risk is, therefore, uncertain.”