Comparison of native prothrombin antigen with the prothrombin time for monitoring oral anticoagulant prophylaxis.

BACKGROUND

Oral anticoagulation is most frequently monitored using the prothrombin time, but an alternative approach is measurement of native, fully carboxylated, prothrombin antigen (NPA). We have correlated results of the prothrombin time and NPA with development of venous thrombosis or bleeding complications in a clinical trial of warfarin prophylaxis following total hip arthroplasty to determine the potential value of NPA measurement for monitoring oral anticoagulation.

METHODS AND RESULTS

Patients in one arm of a prospective, randomized trial received warfarin prophylactically beginning 10 to 14 days before total hip arthroplasty in a dose adjusted to prolong the international normalized ratio (INR) to 1.5 on the day of surgery and 2.5 after surgery. NPA was measured by ELISA, and the prothrombin time was measured using rabbit brain thromboplastin. Samples were tested from 97 patients, and data from 81 patients who had adequate venography were analyzed to correlate test results with occurrence of thrombosis. The prothrombin time and INR were less sensitive than NPA to the lowest intensities of anticoagulation, with the prothrombin time index increasing from 1.0 to 1.3 and the INR increasing from 1.0 to 2.0, whereas the NPA concentration decreased fourfold, from 200 to 50 micrograms/mL. There was little correlation between either the prothrombin time index or the INR and the development of thrombosis, whereas NPA concentrations were significantly higher on the day of surgery and on postoperative days 1, 3, 5, and 7 in patients who developed venous thrombosis. Higher concentrations of NPA were associated with an increased risk of venous thrombosis, but there was no relation between thrombosis and the prothrombin time index or INR. There was no significant correlation between surgical blood loss and prothrombin time index, INR, or NPA concentration. However, patients who received the largest number of transfusions on the day of surgery had significantly lower NPA concentrations than patients who required no transfusion.

CONCLUSIONS

These results indicate that the NPA concentration more accurately reflects the antithrombotic effect of warfarin than does prothrombin time and may be superior in monitoring prophylactic oral anticoagulation.