Diagnostic performance of phospholipid-specific assays for the evaluation of antiphospholipid syndrome.

Abstract

The diagnostic performance of commercially available nonstandard antiphospholipid (aPL) assays for the evaluation of antiphospholipid syndrome (APS) is unknown. In 62 patients with APS, 88 with recurrent pregnancy loss, 50 healthy blood donors, and 24 women with one or more successful pregnancies, we measured antiphosphatidic acid (aPA), antiphosphatidyl-choline (aPC), antiphosphatidylethanolamine (aPE), antiphosphatidylglycerol (aPG), antiphosphatidylinositol (aPI), and antiphosphatidyl-serine (aPS) IgG and IgM antibodies from 2 manufacturers. We computed the areas under the curve (AUC), sensitivities, specificities, positive and negative predictive values, and 95% confidence intervals to assess diagnostic performance. The AUC analyses of the IgM assays demonstrated significant differences (P < .01) for all markers except aPC, whereas the IgG markers showed comparable performance for most assays with the exception of aPE (P < .01) and aPS (P = .02) antibodies. Overall, the combined sensitivity of the aPL assays differed significantly between manufacturers and did not improve the diagnostic yield for APS.