Comparison of plasma prothrombin and factor VII and urine prothrombin F1 concentrations in patients on long-term warfarin therapy and those in the initial phase.

Control of warfarin anticoagulation during the initial phase of therapy is difficult and empirically based. Plasma and urine samples were obtained from normal controls, patients under stable anticoagulation, and patients in the initial phase of anticoagulation. Total plasma prothrombin, des-carboxy (non-adsorbable with barium chloride) prothrombin, and native (total minus non-adsorbable) prothrombin were quantitated using Echis carinatus venom activation. Functional plasma factor VII (VII) was measured using a one-stage clotting assay. Total and des-carboxy urine prothrombin F1 (F1) were measured by ELISA. All urine F1 in normals and both anticoagulated groups was adsorbed by barium chloride. Plasma des-carboxy prothrombin concentration was similar for the two anticoagulated groups and did not correlate with 1/INR. Native prothrombin correlated with 1/INR in both the stable (r = 0.76) and initial phase (r = 0.74) groups. For any given INR, the subjects on stable anticoagulation had lower native prothrombin concentrations than the initial phase patients. Functional factor VII concentration also correlated significantly with 1/INR in both the stable (r = 0.64) and initial phase (r = 0.76) patients. Unlike native prothrombin, VII concentrations did not vary between the two cohorts for any given INR. Previous studies indicate that native prothrombin is a superior predictor of both hemorrhagic and thromboembolic complications during warfarin therapy. Our findings indicate that VII, and not prothrombin, may be the predominant factor monitored by the INR. This further supports the need to reevaluate the usefulness of the INR in the monitoring of warfarin therapy during the initial phase.