This article is more than 5 years old. Information may no longer be current.
Role of genetics in warfarin dosing a source of debate
Click Here to Manage Email Alerts
DALLAS — Two trials presented at AHA 2013, EU-PACT and COAG, yielded conflicting data on the utility of genetics in achievement of optimal warfarin dosing.
Clinicians often have difficulty determining the optimal dose of warfarin for a particular patient. Research shows that CYP2C9 and VKORC1 account for a large proportion of inter-individual variability seen with dose requirements, but there is currently little understanding of how to incorporate these data into clinical practice, which served as the motivation for the EU-PACT and COAG trials, two presenters said at AHA 2013.
EU-PACT
Munir Pirmohamed, MD, PhD, National Health Services chair of pharmaco-genetics at the University of Liverpool, U.K., and colleagues conducted the pragmatic, single blind, randomized EU-PACT trial to evaluate whether genotype-guided dosing was superior to standard clinical care over 3 months.
Munir Pirmohamed
The standard care arm included a warfarin loading dose for 3 days. International normalized ratio (INR) was checked on day 4 and the next dose was adjusted based on an algorithm used by standardized software. From day 6 onward, dose adjustment was based on INR. For the genotype-guided dosing arm, the researchers used a point-of-care assay to perform genetic testing. Patients received an individualized loading dose for 3 days; INR was checked on day 4 and dose adjustments were made based on clinical and genetic factors; and on day 6, patients received standard care.
A total of 455 patients (mean age, 67.3 years; 61% men; 98.5% white) with atrial fibrillation (72.1%) or venous thromboembolism were randomly assigned to genotype-guided warfarin dosing (n=222) or standard clinical care (n=228). The primary endpoint was percent time in the therapeutic INR range of 2 to 3 during 3 months after initiation of warfarin therapy.
The primary analysis included 211 patients in the genotype-guided dosing arm and 216 in the standard care arm. For the primary outcome, percent time in the therapeutic INR range was 60.3% in the standard care arm vs. 67.4% in the genotype-guided dosing arm, with a statistically significant adjusted difference of 7% (P<.001).
The researchers observed no major bleeds; three clinically serious bleeds, all of which occurred in the standard care arm; and no difference in minor bleeds between study arms. Only one thromboembolic event occurred.
“Genotyping before initiation of warfarin increased the percent time in the therapeutic range by approximately 7%; reduced overanticoagulation by 69%; reduced time required to reach therapeutic INR by 28%; improved the time to reach required stable dose by 25%; and reduced the number of warfarin dose adjustments by 9%,” Pirmohamed said at a press conference.
COAG
Results from the COAG trial yielded different conclusions, Stephen Kimmel, MD, MSCE, professor of medicine at University of Pennsylvania Perelman School of Medicine, said at the press conference.
For the randomized, double blind, multicenter COAG trial, researchers compared two approaches to warfarin dosing: initiation of therapy based on clinical information only vs. initiation of therapy based on clinical information and a patient’s individual genetic information using CYP2C9 and VKORC1. All patients were genotyped at the beginning of the trial.
Stephen Kimmel
The intervention occurred over 5 days. In both arms, a dosing algorithm was used from the first 3 days and then a dose revision algorithm was used on days 4 and 5. The pharmacogenetic-guided arm used clinical and genetic data while the clinical-guided arm used only clinical data. Similar to EU-PACT, the primary endpoint was percent time in the therapeutic INR range, but at 1 month rather than 3 months.The researchers randomly assigned 1,015 patients (mean age, 57 years; 27% black) with stroke, venous thrombosis or AF to pharmaco-genetic dosing (n=514) or clinically-guided dosing (n=501). About two-thirds of all patients began as inpatients, with a little more than half having had deep vein thrombosis or VTE.
At 45% in each group, results revealed no difference in percent time in the therapeutic range at 4 weeks. However, data revealed a significant difference in effects of the algorithms by race, according to Kimmel. Black patients in the pharmacogenetic group fared worse than non-black patients and had better rates of the primary outcome with clinically-guided dosing vs. pharmaco-genetic dosing.
Adverse events did not differ between groups at 4 weeks.
“The COAG trial does not support the hypothesis that genetic information provides added benefit above and beyond clinical information on anticoagulation control over the first 4 weeks of warfarin therapy. In addition, there was no effect of pharmacogenetic-based dosing in those expected to have benefit based on predicted dose differences. … COAG highlights the importance of performing randomized trials for pharmacogenetics, particularly for complex medicine regimens such as warfarin,” Kimmel said. – by Melissa Foster
For more information:
Kimmel S. LBCT.05. New strategies for atrial fibrillation patients: Rhythm and thrombosis.
Pirmohamed M. LBCT.05. New strategies for atrial fibrillation patients: Rhythm and thrombosis.
Both presented at: the American Heart Association Scientific Sessions; Nov. 16-20, 2013; Dallas.
Disclosure: Kimmel reports financial ties to Johnson & Johnson and Pfizer. Pirmohamed reports receiving grants from the EU FP7 framework program. Additional funding for COAG was provided by AutoGenomics Inc, Bristol-Myers Squibb and GenMark Diagnostics.
Perspective
Back to Top
Patrick T. Ellinor, MD, PhD
Warfarin is incredibly commonly prescribed, but it also has a narrow therapeutic window and serious and sometimes deadly bleeding complications. There’s also a wide range, potentially as much as a 10-fold variation, in dosing of warfarin in the population. Many variables are known to predict warfarin dosing including two polymorphisms or genetic variants that were looked at in these studies, VKORC1 and CYP2C9, contribute to anywhere from 35% to 50% of that variability in warfarin dosing. The CYP2C9 is a major enzyme that metabolizes warfarin and VKORC1 is the enzyme that warfarin directly inhibits. Putting this together, the sum total of what you get is that the more of these genetic variants that you have, the lower the dose of warfarin that's ultimately needed for that particular patient.
So it sounds like a perfect fit. We have ways to predict the variability in dosing, and it would be a natural extension to then apply this to our patient populations, and it has been done. This is also, given the challenges in warfarin dosing, one of the prototypic examples of the application of pharmaco-genetics. This has been extensively tested and is an area of active investigation. Therefore, with this backdrop in mind, how do we reconcile the data from these two trials?
In the EU-PACT trial, the genotyping did better than the standard care, but in the COAG trial, there really was no difference between implementation of the clinical data and the genetic data. And so, I think EU-PACT is a nice example of what is common practice in the world, whereas the COAG trial is probably what we ought to be striving for. But it’s important to emphasize that this is not routinely done in most centers.
It's important that, although the VKORC1 and CYP2C9 captured a lot of information, we're leaving a lot of information on the table when we're dealing with the initiation of warfarin. And when you factor in the other clinical variables that are available to each of us, such as age, race, etc., you add a lot of additional information. This is helpful, as is the genetic information that we ought to be using. But it’s important to say that in many cases, we’re not.
I don't think that we really ought to be using a fixed dose for all of our patients for warfarin initiation at this point. When you factor in these other variables, we can reliably predict the dose of warfarin that we want to use for most patients. We shouldn't rely just on genetic data for warfarin initiation in African-American patients, and I think that it was important to see the inclusion of the African-American patients in this particular study.
We should be using a clinical algorithm at warfarin initiation, and again, that is not where we are at the moment, in most centers. One clear message from the studies presented today is that when you prescribe warfarin, it should be customized based on the clinically available data that can help to predict the correct initial dose.
Click Here to Manage Email Alerts