Accuracy of the first fully automated method for anti-cardiolipin and anti-beta2 glycoprotein I antibody detection for the diagnosis of antiphospholipid syndrome.

ABSTRACT

In this study we evaluated the diagnostic accuracy for antiphospholipid syndrome (APS) of new fully automated immunoassays for anticardiolipin (aCL) and anti-beta2 glycoprotein I (anti-beta2-GPI) auto-antibody detection (EliA-Phadia), and compared the results with those obtained with Orgentec and Inova ELISA methods. Sixty-two APS patients and 123 controls (20 syphilis, 33 Lyme disease, 30 HCV infection and cryoglobulinemia, 40 healthy subjects) were studied. Using the 99(th) percentile cutoff, the sensitivity and specificity of EliA aCL IgG, aCL IgM, anti-beta2-GPI IgG, and anti-beta2-GPI IgM were 69.4% and 81.9%, 64.5% and 86.7%, 64.5% and 98.8%, and 53.2% and 92.8%, respectively. Using the Sydney criteria cutoff (>40 GPL/MPL units), sensitivity and specificity of EliA aCL IgG and aCL IgM were 45.2% and 98.8%, and 35.5% and 97.5%, respectively. The best diagnostic efficiency was obtained combining the aCL tests (>40 GPL/MPL units) with the anti-beta2-GPI tests (sensitivity 85.7%, specificity 90.4%). The area under the ROC curves for EliA, Orgentec, and Inova methods were 0.870, 0.940, and 0.850 for aCL IgG; 0.820, 0.820, and 0.820 for aCL IgM; 0.910, 0.960, and 0.920 for anti-beta2-GPI IgG; 0.840, 0.840, and 0.820 for anti-beta2-GPI IgM, respectively. Finally, the overall agreement between EliA assays and the other two ELISA methods ranged from moderate (anti-beta2-GPI IgG EliA versus Orgentec Cohen's k = 0.426) to good (anti-beta2-GPI IgM EliA vs. Inova k = 0.841). In conclusion, newly developed EliA methods for antiphospholipid antibody detection perform similarly to other ELISA assays and represent a useful tool for APS laboratory diagnosis in daily practice.