This article is more than 5 years old. Information may no longer be current.
Clinical trial data provide 'divergent' conclusions about pharmacogenomic-guided warfarin dosing
Click Here to Manage Email Alerts
Two trials presented at the 2013 American Heart Association Scientific Sessions and a third trial, all simultaneously published online last year in The New England Journal of Medicine, come to divergent conclusions and leave many wondering if the data may spell the end for pharmacogenomic-guided dosing of warfarin.
Trials under examination
The first trial, EU-PACT, was presented at AHA in Dallas by Munir Pirmohamed, MD, PhD, of the University of Liverpool, United Kingdom. The study evaluated 455 patients from the United Kingdom and Sweden who were initiating warfarin. Patients were randomly assigned, in a single blinded fashion, to either a genotype-guided or standard warfarin-dosing scheme.
William L. Baker
Sarah A. Spinler
Warfarin genotyping was performed using a point-of-care system that provided results within 2 hours and were incorporated into a computerized loading-dose algorithm for the first 3 days of treatment. An INR was checked on day 4 of therapy, and the dose was individualized using the algorithm that incorporated both genetic and clinical factors. Anticoagulation control was maintained thereafter using a computerized software program according to usual clinical care in the region. The control group received a standard warfarin loading dose over the first 3 days of 10 mg, 5 mg and 5 mg (or 5 mg, 5 mg and 5 mg if aged older than 75 years).
The investigators wrote in NEJM that their “trial design was consistent with clinical practice in the United Kingdom and Sweden.” They said “clinical algorithms are not used in either country,” and “the study was designed pragmatically to reflect clinical practice.”
EU-PACT found that genotyping resulted in a significantly higher time in therapeutic range (67.4%) during the 12-week follow-up period compared with the standard-dosing arm (60.3%; P<.001). Genotyped patients also had significantly fewer INR values ≥4 (27% vs. 36.6%; P=.03) and a shorter time to reach therapeutic INR (median: 21 days vs. 29 days; P=.003). No major bleeding events occurred and only a single thromboembolic event occurred in the control group.
Pirmohamed and colleagues concluded that “genotype-guided warfarin dosing was superior to standard dosing,” but questioned “whether this will translate to improved clinical outcomes.” During the discussion following presentation of the study results at AHA, Pirmohamed said “the first thing we need to do next is to look at the cost-effectiveness,” estimating that the point-of-care genetic test evaluated in EU-PACT cost approximately €50.
The second trial, COAG, was presented by Stephen Kimmel, MD, MSCE, of the University of Pennsylvania. This study evaluated 1,015 patients from the United States who were initiating warfarin, regardless of indication. The COAG trial, unlike EU-PACT, randomly assigned patients in a double blind fashion to receive warfarin dosed according to a genotype-guided strategy or a clinically based strategy (control).
Genotyping was determined using standard platforms and was incorporated into a prespecified algorithm that also utilized clinical data to determine the warfarin dose. This provided individualized dosing during the first 3 days of therapy, and then a dose-revision algorithm was used for dosing on days 4 and 5. Standardized dose-adjustment techniques were used thereafter. The control group had its warfarin dosed without incorporation of the genotyping information into the algorithm.
According to the investigators, “the algorithms … used in the trial have been validated and account for race (specifically black vs. nonblack).” This is a design element that differentiates the COAG trial from those previously published, as the investigators noted that “previous studies enrolled either no black patients or a minimal number of black patients.” Another unique element of the COAG trial design was the use of both a dose-initiation and dose-revision algorithm, compared with prior trials that utilized only a genotype-guided algorithm at baseline.
The COAG trial showed no significant difference in time in the therapeutic range between the genotype and control groups after 4 weeks of therapy (45.2% vs. 45.4%, respectively; P=.91). Interestingly, in black patients, genotyping led to significantly lower time in the therapeutic range than the control group (35.2% vs. 43.5%; P=.01), suggesting that these patients may yield especially low value from genetic warfarin testing. Significant differences in the mean percentage of time above or below the INR range or time to first INR in the therapeutic range were not seen. The principal secondary outcome of any INR ≥4, major bleeding or thromboembolism was similar between groups (P=.93); however, the investigators noted that the COAG trial was not powered for these outcomes.
Kimmel and colleagues concluded that the findings “do not support the hypothesis that initiating warfarin therapy at a genotype-guided maintenance dose for the first 5 days, as compared with initiating warfarin at a clinically predicted maintenance dose, improves anticoagulation control during the first 4 weeks of therapy.” They said these “findings exclude a meaningful effect of genotype-guided dosing on the percentage of time in the therapeutic range during the first month of warfarin treatment.”
A third trial, also published by the EU-PACT group, studied the impact of the same genotype-guided dosing algorithm in patients receiving acenocoumarol and phenprocoumon therapy. Unlike the warfarin study, time in the therapeutic range at 12 weeks was not different between the genotype and control groups. This was true for both patients receiving acenocoumarol and phenprocoumon. Thus, it does not appear to aid initiation of dosing for these agents.
Impact of studies unknown
In an accompanying editorial, Bruce Furie, MD, of Beth Israel Deaconess Medical Center and Harvard Medical School, noted that the conclusions of the three aforementioned studies were similar, despite differences in organization and structure. He also wrote that the trials “indicate that this pharmacogenetic testing has either no usefulness in the initial dosing of vitamin K antagonists or, at best, marginal usefulness, given the cost and effort required to perform the testing.” Furie said, however, that these studies addressed the process of anticoagulant initiation and thus are not reflective of an intermediate- or long-term anticoagulation approach.
During a panel discussion following presentation of the EU-PACT and COAG trials at AHA, the discussants said genetic differences account for approximately 35% to 40% of the variability of warfarin in individuals. This information must be balanced with clinical factors such as age, race/ethnicity, smoking, body surface area and concomitant medication use. This could explain the incremental benefits of genotype-guided warfarin initiation. The concept of using the INR as a surrogate for anticoagulation control may be problematic, and the impact of genotyping on clinical outcomes such as thromboembolism and bleeding remains largely unknown.
Current state of testing
To date, clinical trials have not consistently demonstrated improvements in care, a reduction in bleeding events or lower costs and, at this time, neither the American College of Cardiology Foundation nor the AHA recommend pharmacogenetic testing for warfarin anticoagulation.
Disclosure: Baker reports no relevant financial disclosures.