Procoagulant effect of anti-beta2-glycoprotein I antibodies with lupus anticoagulant activity

Prothrombin time (PT) is routinely used to monitor oral anticoagulant treatment in patients with the antiphospholipid antibody syndrome (APS). The fact that PT is a phospholipid (PL)-dependent coagulation test raises the possibility that lupus anticoagulant (LA) might interfere with this test, thus complicating the control of anticoagulant treatment. The effect of 6 affinity-purified preparations of anti- (a)beta2-glycoprotein I (GPI) antibodies with LA activity on the PT was tested. Instead of prolonging PT as expected, the abeta2-GPI antibodies reduced the PT of both normal plasma and anticoagulated plasma by a mean of 2.4 seconds and 5.6 seconds, respectively. This effect was also observed using other 5 commercially available preparations of thromboplastin. The abeta2-GPI-mediated reduction in PT was dose-dependent and was lost upon removal of beta2-GPI. The failure of abeta2-GPI antibodies to express LA activity in PT was found to depend on the fact that calcium ions were added together with PL at the beginning of the assay. In fact, modification of the standard diluted Russell viper venom time (dRVVT) test by adding calcium ions together with PL resulted in a loss of abeta2-GPI anticoagulant activity. The procoagulant effect was not as evident in an assay that used stimulated monocytes as a source of thromboplastin. These results show that abeta2-GPI antibodies exhibit an 'in vitro' procoagulant effect in PT and an anticoagulant effect in dRVVT only when the interaction with their antigen and PL occurs in the absence of calcium ions.