Pharmacogenetic testing for guiding de novo phenprocoumon therapy in stroke patients.

Background: For many conditions causing transient ischemic attack or minor stroke, secondary prevention with early initiation of oral anticoagulation is indicated. The individual response to coumarins is known to vary widely and is not well predicted by clinical variables. Patients' discharge from hospital care is often delayed only because the therapeutic target range has not been reached yet. A feasible tool to guide coumarin dosing and thereby safely shortening time in hospital is required.

Methods: We established a polymerase chain reaction technique for rapid genotyping of the vitamin K epoxide reductase complex (VKORC1), which is the pharmaceutical target of the coumarins. C283 + 837C -> T (rs2359612) genotypes were determined in 49 patients who underwent de novo oral anticoagulation with phenprocoumon for cerebrovascular disease. Other variables potentially affecting phenprocoumon sensitivity were systematically evaluated.

Results: Of 49 genotyped patients, 47 were treated in hospital until an international normalized ratio (INR) of 2-3 was reached. The time and the cumulative dose of phenprocoumon necessary to achieve the target INR both were strongly dependent on the individual C283 + 837C -> T genotype (Kruskal-Wallis test p = 0.0002, and p < 0.0001, respectively). Carriers of the TT genotype reached an INR of 2-3 after a mean time of 3.2 days (n = 5), CT carriers after 4.4 days (n = 27), and CC carriers after 6.5 days (n = 15). No other variable, including body weight, was significantly correlated with the treatment response.

Conclusion: In patients with cerebrovascular disease, genotyping for VKORC1 alone can strongly predict the individual response to de novo phenprocoumon treatment. The size of the pharmacogenetic test's potential effect on a more efficient use of hospital capacities remains to be shown by a controlled interventional study.