Anti-beta2-glycoprotein I autoantibodies, in vitro thrombin generation, and the antiphospholipid syndrome

OBJECTIVE: Thrombin plays a pivotal role in the regulation of hemostasis. We examined the effect of antiphospholipid (aPL) antibodies, in particular those with specificity for beta2-glycoprotein I (beta2-GPI), on in vitro thrombin generation, and examined the association with clinical manifestations of the antiphospholipid syndrome (APS). METHODS: We studied plasma samples from 59 patients with aPL antibodies determined by the presence of either elevated anticardiolipin (aCL) antibodies or lupus anticoagulant (LAC). Direct antibody binding of IgG, IgM, and IgA to beta2-GPI and prothrombin (PT) was determined by ELISA. Affinity purification of total IgG and IgG anti-B2-GPI antibodies was performed using staphylococcal protein A and phospholipid liposomes. A chromogenic assay was used to determine the effect of plasma samples and purified autoantibodies on in vitro thrombin generation. RESULTS: Thirty-three of 59 (56%) plasma samples inhibited in vitro generation of thrombin and 7/59 (12%) accelerated thrombin formation. There was a strong negative correlation between thrombin generation and IgG aCL (r = -0.72, p < 0.001) and IgG anti-beta2-GPI (r = -0.71, p < 0.001) antibody levels, and a weaker correlation with LAC (r = -0.46, p = 0.001). This association was not found with anti-PT antibodies and could not be attributed to concurrent therapy with warfarin. Additional experiments with affinity purified IgG antibodies indicated a dose dependent inhibition of thrombin generation, which was restricted to anti-beta2-GPI antibodies. Patients with a history of core clinical manifestations of the APS [venous and arterial thrombosis, recurrent (> or = 2) fetal loss] had significantly greater inhibition of in vitro thrombin generation (mean +/- SEM Z score: -3.38 +/- 0.51 vs -1.42 +/- 0.56; p = 0.01) and higher levels of IgG aCL (mean +/- SEM Z score: 8.39 +/- 1.12 vs 5.39 +/- 0.88; p = 0.04) and IgG anti-beta2-GPI antibodies (mean +/- SEM Z score: 4.49 +/- 0.69 vs 2.26 +/- 0.54; p = 0.01). Odds ratios for these variables and clinical manifestations of the APS were 5.43, 4.17, and 3.28, respectively. CONCLUSION: aPL antibodies may accelerate or inhibit the rate of in vitro thrombin formation. The predominant effect is inhibition that is restricted to IgG anti-beta2-GPI antibodies and it is strongly associated with clinical manifestations of the APS.