The Couma-Gen Study did not Demonstrate a Reduction in Out-of-range INRs in Patients with Pharmacogenetic-guided Dosing of Warfarin
November 7, 2007 By Lauren Ciaglo 
Orlando, FL: Results from the Randomized Trial of Genotype-guided versus Standard Warfarin Dosing in Patients Initiating Oral Anticoagulation: The Couma-Gen Study was presented at the American Heart Association Scientific Sessions 2007 by Dr Jeffery L. Anderson. This is also published on-line in Circulation.
The authors demonstrate the feasibility of randomized trials to test pharmacogenetic-based algorithms. Recent studies have shown that pharmacogenetics, the study of interactions of genetics with pharmacotherapy, promises to be a premiere application of genetics to personalized medicine. The Couma-Gen trial studied the pharmacogenetic-guided dosing of warfarin.
Warfarin is prescribed to over 2 million people and is subject to 20-fold dose variability. Half of that dose variability is attributable to variants in the CYP2C9 and VKORC1 genes along with age and weight. Although the FDA has recently changed the labeling of warfarin to recognize its genetic effects on dosing, there have been no prospective clinical trials testing the impact of PG-guided dosing.
This study was a prospective, blinded, randomized comparison of PG-guided vs standard (STD) dose initiation of warfarin. A total of 206 patients were enrolled, with 200 evaluable cases (PG = 101; STD = 99), and a mean study follow-up of 46 days. Patients 18 years of age and older were eligible for the study and were buccal swabbed and randomized in permuted blocks of 5. DNA was amplified by polymerase chain reaction (PCR) and genotypes were identified using the derivative peak analysis module. Genotypes CYP2C9 (*1, *2, *3) and VKORC1 (C1173T), age, weight, and sex were included in the study. The standard dosing protocol was followed as previously described*: 10mg on days 1 and 2 followed by 5mg daily, which was modified based on INR results. Patient INRs were measured on days 0, 3, 5, 8, 21, 60, and 90, and according to physician discretion. Both the patients and clinicians were blinded with assignment known only to the research pharmacist. All patients were followed for up to 3 months.
The study failed to achieve its primary endpoint of demonstrating a reduction by PG dosing in the per-patient average % of INRs outside the therapeutic range. The primary endpoint of percent out-of range (OOR) INRs (<1.8, >3.2) was 30.7% in the PG group compared with 33.1% in the STD group (p=0.47; 54% of OOR values were subtherapeutic; 46% were supratherapeutic). Within the primary endpoint subset analysis the percent of OOR INR by variant status was as follows: in the multiple variant subset 31% in the PG group compared with 40.4% in the STD group (p=0.14), wild type subset 28.1% in the PG group compared with 36.9% in the STD group (p=0.21), multivariant plus wild type subset 29.3% in the PG group compared with 39.1% in the STD group (p=0.03), and single variant subset 33.6% in the PG group compared with 27.0% in the STD group (p=0.14).
The Couma-Gen study demonstrated the feasibility of prospective, randomized trials with PG-guided therapy and the possibility of genotyping in clinical “real-time” (1 hour). The PG-calculated dose was indeed a better predictor of the stable maintenance dose and PG-guidance was associated with the need for fewer INRs as well as dose adjustments. Despite the trial’s inability to meet the primary endpoint of reduction in the percent of patients whose INR were out-of range, pharmacogenetics is a promising technique that will likely have an important contribution to personalized medicine. Although the clinical benefit of PG-guidance remains unproven, promise was shown in wild-type patients (greater dose-requirements) and multiple variant carriers (lower dose-requirements). Inpatient initiation with careful management by a dedicated anticoagulation service, leading to rapid correction of inappropriate doses, likely contributed to better-than expected results in the standard arm.
Future studies should focus on including a larger patient population (at least 2,000 patients); the differential benefit early after initiation; consider additional genetic factors; and target out-patient, less aggressively monitored anticoagulation initiation scenarios. *Ann Intern Med 2003; 138-714.
References: Anderson JL, Horne B, Stevens S, Grove A, Barton S, Nicholas ZP, Kahn SF, May H, Samuelson K, Muhlestein, JB, Carlquist JF and Couma-Gen Investigators. Randomized trial of Genotype-guided versus standard warfarin dosing in patients initiating oral anticoagulation. Circulation published online Nov 7, 2007.