Interference of lupus anticoagulants in prothrombin time assays: implications for selection of adequate methods to optimize the management of thrombosis in the antiphospholipid-antibody syndrome

Abstract

BACKGROUND AND OBJECTIVE: Prolonged anticoagulation aiming at International Normalized Ratio (INR) values > 3.0 has been recommended for patients with thrombosis and the antiphospholipid-antibody syndrome. We evaluated the influence of anticoagulant antibodies in two different prothrombin time (PT) assays carried out on plasma from lupus anticoagulant patients on oral anticoagulation. DESIGN AND METHODS: INR values obtained with a combined (final test plasma dilution 1:20) and a recombinant (final test plasma dilution 1:3) thromboplastin were compared in 17 patients with persistent lupus anticoagulants (LA) receiving oral anticoagulant treatment and monitored for 69.8 patient-years. Doses of anticoagulant drugs were always assigned based on the results obtained with the combined thromboplastin, aiming at a target INR of 2.5 or 3.0 for patients with venous or arterial thromboembolic disease. Paired determinations with both reagents were also obtained throughout the study period in 150 patients on stable oral anticoagulation but free of antiphospholipid antibodies. Total IgG fractions were purified from selected patients to evaluate effect in the two PT assay systems. RESULTS: No patient experienced recurrence of thrombosis or major bleeding complications (95% confidence interval: 0.1-6.5 per 100 patient-years). INR values with the recombinant reagent were significantly higher than with the combined reagent in 8 LA patients (mean DINR ranging from 0.17 to 0.54) of the degree of anticoagulation was overestimated in all but one LA patients with the recombinant reagent when compared to the DINR observed in non-LA patients (-0.64 +/- 0.42). The anti-cardiolipin IgG titer (r(2) = 0.43, p = 0.004) and the anti-b(2)GPI IgG titer (r(2) = 0.30, p = 0.023) were positively associated with the mean deltaINR observed in LA patients. When added to plasmas with different levels of vitamin K-dependent factors, total IgG fractions from 6 LA patients with significant overestimation of the INR with the recombinant reagent (mean DINR ranging from 0.17 to 0.54, group 1) and from 7 LA patients with mean deltaINR < or = 0.0 (ranging from -0.25 to 0.04, group 2) reproduced the effects observed ex vivo in the two assay systems. However, when total IgG fractions were tested at the same final concentration in the two PT assay systems, there was no difference in the clotting times determined with total IgG fractions from group 1 and group 2 LA patients. Addition of negatively charged liposomes (0.4 and 0.8 mg/mL final concentrations) to platelet free plasma from LA-free patients on stable oral anticoagulation caused a 20% to 48% prolongation of the prothrombin time determined with the recombinant reagent. In contrast, no significant prolongation of the prothrombin time determined with the recombinant reagent was observed upon addition of negatively charged liposomes to plasma from group 1 LA patients. INTERPRETATION AND CONCLUSIONS: These results confirm previous suggestions of assay-dependency of INR values in LA patients on oral anticoagulation. For these patients, accurate INR values may be obtained using combined thromboplastin reagents that permit testing at high plasma dilution.