|
ATRIAL FIBRILLATION AND STROKE—WHAT ARE THE RISKS?
ANAHEIM, CALIF—New data from the Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA) study confirm that in “real-world” practice, international normalized ratios (INRs) between 2 and 3 provide optimal protection from stroke and minimize the risk of intracranial hemorrhage for patients with atrial fibrillation taking anticoagulant therapy.
In a second report from the ATRIA cohort, prevalence of the factor V Leiden mutation—thought to increase thrombotic risk—was found to be lower than expected and was not significantly associated with an increased risk of stroke in patients with nonvalvular atrial fibrillation. Both reports were presented at the American Heart Association’s 2001 Scientific Sessions.
WARFARIN AND STROKE
The ATRIA study is a collaborative effort between Massachusetts General Hospital in Boston and Kaiser Permanente of Northern California, based in Oakland. The study includes more than 13,000 patients with nonvalvular atrial fibrillation at Kaiser, a cohort assembled between July 1996 and December 1997 and followed for about five years. Investigators were funded by the National Institute of Aging for a four-year study, and that funding has recently been extended for another five years, said Daniel Singer, MD, Chief of the Clinical Epidemiology Unit at Massachusetts General Hospital and Professor of Medicine at Harvard Medical School. With this extension, investigators are using the cohort to answer a variety of questions related to risk factors and treatment of these patients. Among these, for example, is the effect of warfarin for the treatment of atrial fibrillation in the community.
“In trials you try to select patients reasonably carefully, and you have excellent control of a tricky medicine, warfarin, but what’s the net impact in usual clinical care?” Dr. Singer asked. “It’s very hard to study that except in a system like Kaiser, where they have excellent automated databases, access to records, and virtually 100% follow-up.”
Trials such as the Stroke Prevention in Atrial Fibrillation (SPAF) trial, reported in the early 1990s, showed a profound benefit from anticoagulation with warfarin in patients with atrial fibrillation. So substantial was the benefit that many of the trials were actually stopped early, “because nearly all the strokes were occurring in those not taking anticoagulation,” Dr. Singer said. Because of that, a relatively small number of events was accumulated among patients on anticoagulation, he added.
Since there was no ethical way to continue, stopping the trials meant that studying aspects of treatment among those on anticoagulation in relation to stroke risk was not possible. “If you want to study things like the precise relationship between INR, stroke risk, and hemorrhagic risk, particularly intracranial hemorrhage risk, you need to have a fair number of events among people who are taking warfarin,” Dr. Singer pointed out.
INR ANALYSIS
The ATRIA cohort now has data on 169 stroke events among patients taking anticoagulants, allowing analysis of factors contributing to the strokes such as anticoagulation intensity. Elaine Hylek, MD, of Massachusetts General Hospital, and colleagues attempted to determine INR levels that appear to maximize stroke protection and minimize the risk of hemorrhage. They compared INRs among the 169 patients who had stroke and 55 patients who had intracranial hemorrhage while on anticoagulation to those among controls. Four controls were examined for each case, and the INR measure taken at the visit closest to the stroke admission date was assigned to each subject.
Analysis showed that compared to an INR of 2 to 3, an INR of 1.7 to 1.9 was associated with an odds ratio of stroke of 1.75, and as high as 4.94 when the INR fell below 1.5. INRs above 3 did not appear to confer any further benefit.
In terms of safety, however, higher INRs were associated with an increased risk for intracranial hemorrhage. Patients with an INR between 3 and 4 had an odds ratio for intracranial hemorrhage of 2.4 against the 2 to 3 INR base, and this ratio climbed steeply, reaching 16 for INRs between 4 and 6, and 27 for an INR above 6. “So we see a U-shaped curve of INR versus efficacy in preventing stroke or toxicity in producing intracranial hemorrhage,” Dr. Singer said. These data provide further evidence in favor of INR 2 to 3 as the optimal target for patients with atrial fibrillation.
In the current phase of the ATRIA study, he added, investigators are emphasizing follow-up of elderly patients to plot a similar risk-versus-benefit curve. While the fastest-growing group of patients with atrial fibrillation is those older than 80, with about 10% of that population developing the problem, trials examining this therapy did not include many patients older than 75, Dr. Singer pointed out, “yet we’ve generalized the results of the trials to elderly people.”
GENETIC ANALYSIS AND FACTOR V LEIDEN
According to Dr. Singer, another strength of the database rests in the fact that investigators are able to periodically “run” the data for new strokes, and then collect DNA from these patients, eventually accumulating a genetic database for patients with stroke. As Alan S. Go, MD, principal investigator of the ATRIA cohort at Kaiser, points out, “atrial and venous pressures are similar in atrial fibrillation patients, suggesting that genes that promote venous thromboembolism may also increase the risk for stroke.”
The focus of Dr. Go’s work was on factor V Leiden, the “most notorious and common genetic risk factor for thrombosis,” Dr. Singer said. In atrial fibrillation, strokes arise from a low-pressure stasis clot in the atrial appendage, a mechanism similar to deep venous thrombosis. “Factor V Leiden is clearly related to the risk of deep vein thrombosis, so it may also be related to the risk of stroke in patients with atrial fibrillation,” he noted.
Researchers examined 137 patients with incident ischemic stroke and 214 control patients without stroke. Genomic DNA was taken from leukocytes and examined for the factor V Leiden polymorphism. They found a “tantalizing” trend toward a relationship, but it did not reach statistical significance, Dr. Singer said.
Eight of 137 cases (5.8%) and eight of the 214 controls (3.7%) were heterozygous for factor V Leiden, an odds ratio of 1.6. When analysis was restricted to those not on anticoagulation at the time of the stroke, seven of 96 cases (7.8%) and three of 83 controls (3.6%) were found to be heterozygous for factor V Leiden, an odds ratio of 2.1. In this large sample, Dr. Go and colleagues concluded, “the prevalence of factor V Leiden was lower than anticipated and was not significantly associated with risk of stroke in patients with nonvalvular atrial fibrillation. However, given the suggestive observed odds ratio, further research in larger numbers of patients off anticoagulation is needed to clarify the impact of factor V Leiden on stroke risk in atrial fibrillation.”
Going forward, Dr. Singer said, involves planning further such analyses to provide even more precise risk stratification for atrial fibrillation patients. “The ATRIA cohort has the capacity to provide long-term follow-up of patients with atrial fibrillation both on and off anticoagulants, generate new information about risk factors, and in particular, give us a sense of how warfarin works in the real world,” he concluded.
NR
—Susan Jeffrey
Return to table of contents
|
|
