Antiphospholipid (aPL) antibodies include anticardiolipin (aCL) and lupus anticoagulant (LA) antibodies. LA antibodies recognize the complex of lipid-bound (human) prothrombin, in this way inhibiting the phospholipid-dependent coagulation reactions, whereas aCL antibodies are directed towards beta 2-glycoprotein I (beta 2-GPI) bound to an anionic lipid surface. According to their behavior in coagulation reactions, we have divided aCL antibodies into two groups: aCL-type A, which inhibit the phospholipid-dependent coagulation reactions because they enhance the binding of beta 2-GPI to the procoagulant phospholipid surface; and aCL-type B antibodies, which are devoid of anticoagulant properties. We report the distinctive laboratory and clinical profiles of 25 patients with well-characterized, phospholipid-dependent inhibitor of coagulation. Fourteen patients had LA antibodies (aCL-type B were concomitantly present in 10 cases, while in the other four, aCL titer was normal), and the other 11 had aCL-type A antibodies. The laboratory evaluation of the two groups showed the dilute Russell viper venom time (dRVVT) to be the most abnormal coagulation test in the aCL-type A-positive group, whereas the kaolin clotting time (KCT) was the most abnormal assay in the LA-positive group. In fact, the ratios of the coagulation times of patient plasma over normal pooled plasma (mean +/- standard deviation) for LA versus aCL-type A antibodies were 1.48 +/- 0.27 versus 2.20 +/- 0.42, P = .0001, and 2.22 +/- 0.42 versus 1.50 +/- 0.42, P = .0003, for the dRVVT and KCT, respectively. No differences were observed either in the ratios of the activated partial thromboplastin times and the prothrombin times or the plasma levels of beta 2-GPI and prothrombin. Conversely, aCL titers were significantly higher in aCL-type A-positive patients (147 +/- 44 U) than in the LA-positive group (61 +/- 55 U; P = .0003). We ruled out the possibility that platelet contamination of plasma could account for the observed coagulation profiles, as the two patterns were reproduced in platelet-free plasma. In addition, we performed clotting tests in plasma in the presence of phospholipids and calcium after addition of factor IXa or factor Xa. The assay performed with factor Xa was more sensitive to the presence of aCL-type A antibodies, while the assay performed with factor IXa was preferentially sensitive to LA-containing plasmas, supporting the earlier findings with the dRVVT and KCT assays.(ABSTRACT TRUNCATED AT 400 WORDS)