Analytical and clinical performance of a new, automated assay panel for the diagnosis of antiphospholipid syndrome

Background: Anticardiolipin (aCL) and anti-beta(2)glycoprotein I (a beta(2)GPI) antibodies are part of the criteria for antiphospholipid syndrome (APS). Therefore they are widely measured and about 30 commercial kits are available. Objectives: To investigate the analytical and clinical performances of four fully automated, chemiluminescent assays: HemosIL AcuStar aCL IgG, HemosIL AcuStar aCL IgM, HemosIL AcuStar a beta(2)GPI IgG, and HemosIL AcuStar a beta(2)GPI IgM. Methods: Cut-off values were assessed by testing 250 blood donors. Total coefficients of variation (CV) were determined with six plasma pools and two controls ranging from 4.3 to 2694 U mL-1 depending on the assay. Samples from 218 well-characterized patients and 103 controls were measured in three laboratories to determine inter-laboratory variation. The results of the 321 samples were compared with three commercial assays (REAADS, INOVA and VARELISA). Results: Cut-off values were assigned to 20 arbitrary units for all the tests. Total CV ranged from 4.3 to 11.2%. No interference of hemoglobin, bilirubin, triglycerides, heparins and rheumatoid factor was observed. Inter-laboratory variability was low and no sample changed status. Overall status agreement between HemosIL assays and the comparator kits ranged from 82 to 96%. Sensitivity, specificity, agreement when predicting APS and the odds ratios when predicting a thrombotic or obstetric event gave comparable results between HemosIL AcuStar and the three other assays. Conclusions: Our study demonstrates that the fully automated HemosIL AcuStar aPL assay panel showed similar performances to the three commercial ELISAs commonly used by various laboratories to detect antiphospholipid antibodies.