An automated chemiluminescence immunoassay may detect mostly relevant IgG anticardiolipin antibodies according to revised Sydney criteria.

BACKGROUND: Detection of anticardiolipin antibodies (ACA) is an independent laboratory criterion for diagnosis of antiphospholipid syndrome (APS). Alternative methods to ELISA were recently developed such as automated chemiluminescence immunoassay (CLIA). PATIENTS AND METHODS: We compared a CLIA to an ELISA kit for the detection of IgG isotype of ACA. 87 routine samples from 75 patients suspected of having APS were tested using each method. Cut-off values were calculated in our laboratory for each test using 99th percentile of 50 normal controls. RESULTS: Cut-off values were >20 GPL for ELISA and > 2 GPL for CLIA. Overall agreement (OA), agreement for positive (AP) and agreement for negative (AN) cases were 56.3%, 49.2% and 77.2% respectively. Most discrepant results were positive with ELISA and negative with CLIA. However, OA, AP and AN increased to 82.1%, 84.6% and 80% respectively when CLIA was compared to the repeated ELISA performed at least 12 weeks later. When correlated with APS-related clinical background, CLIA showed lower sensitivity, higher specificity and higher likelihood ratio (LR) as compared to first ELISA whereas these parameters were similar to those of the repeated ELISA. No association was found between any test results and APS-related clinical background of the patients. Using our own cut-off value (> 2GPL), sensitivity, specificity and LR of CLIA to identify patients with APS were respectively 100%, 72.3% and 3.6. A ROC curve showed that at 7.5 GPL cut-off value, specificity and LR improved to 91.1% and 11.25 respectively, without affecting sensitivity. A strong correlation was observed between CLIA results and APS (Chi2 = 12.25; p < 0.001). CONCLUSION: The performance of CLIA is as good as a repeated ELISA test to detect IgG ACA in suspected APS patients. It is fully automated, which represents several advantages over semi-manual ELISA techniques for its implementation in a routine laboratory.