Prevalence and Clinical Significance of Antiphospholipid Antibodies to Noncardiolipin Antigens in Systemic Lupus erythematosus

The specificity and immunoglobulin isotype distribution of antiphospholipid (aPL) antibodies have been evaluated in 68 patients with systemic lupus erythematosus (SLE) by ELISA which employed a panel of 7 different PL antigens. A total of 49 patients (72%) were positive for aPL antibodies of different isotypes and directed to one or more PL epitopes. Prevalence of IgG anticardiolipin (aCL, 37%) was similar to that of the other negatively charged PLs phosphatidylserine (PS, 35%), phosphatidylinositol (PI, 35%), phosphatidylglycerol (PG, 35%) and phosphatidic acid (PA, 40%); prevalence reduced to 9-12 % and 7-16 % respectively for IgM and IgA isotypes to the same antigens. aPL antibodies to the zwitterionic PLs phosphatidylcholine (PC) and phosphatidylethanolamine (PE) were also observed, though their prevalence was lower than that demonstrated for negatively charged PLs. Of the 36 SLE patients who were aCL negative (53%), 17 (25% of all patients and 47% of aCL-negative patients) were positive for aPL antibodies of different isotypes to one or more non-CL epitopes. During a mean follow-up period of 30 months, 10 patients had deep vein thrombosis (DVT) with a total of 21 events. By χ² test, a significant correlation was found between DVT and IgG aCL (p = 0.03) and between this event and the presence of lupus anticoagulant (LA) antibody (p = 0.04). However, stronger correlations were demonstrated between DVT and IgA aCL (p = 0.007), IgG anti-PS (p = 0.02), and IgA anti-PC, -PI and -PG (p = 0.02,0.003 and 0.02, respectively), whereas no correlation was found between thrombotic events and aPL antibodies with PE and PA specificities. In contrast, a significant correlation was demonstrated between aPL antibodies to the latter antigen and thrombocytopenia (p = 0.02). Moreover, a significant correlation was observed between the presence of LA and IgG and IgM aCL antibodies, IgG and IgM anti-PS, IgG anti-PI and IgG and IgA anti-PA, suggesting the possibility that LA activity could be mediated by different subsets of aPL antibodies heterogeneous for specificity and isotype. The complexity of our data indirectly explains some of the discrepancies often observed between solid-phase immunoassays and coagulation studies employed for aPL antibody detection and indicate the importance of aPL antibodies directed to non-CL epitopes as additional parameters to predict the occurrence of thrombotic events in SLE patients.