Abstract
There are many studies that are available on the Internet that attempt to standardize the assay for anticardiolipin antibody evaluation because of the variability of results. The aim of this study was to evaluate simultaneously the role of different microplates and the importance of sample nonspecific binding in determining different results in anticardiolipin antibody detection. Sera from 8 patients with raised levels of IgG anticardiolipin antibodies and 10 control sera were assayed by enzyme-linked immunosorbent assay in the presence (specific binding) or in the absence of cardiolipin (sample blank) with four different microplates, that is, NUNC PolySorp, FALCON ProBIND, Greiner 655061 (high binding), and Greiner 655001 (medium binding). Results were expressed as optical densities or net-optical densities (following sample blank subtraction) as well as international IgG anticardiolipin units (GPL) or net-GPL. A wide interplate variability of optical densities was found. When results were expressed as GPL, significant differences were only found between Greiner 655061, FALCON ProBIND, and NUNC PolySorp (P < .05 and P < .001, respectively) whereas differences were not statistically significant if interplate variability was analyzed as net-GPL. Results expressed as categorical variables (ie, positive/negative, according to a GPL cut-off and net-GPL cut-off, obtained with sera from 100 apparently healthy blood donors) showed a good or excellent Cohen's κ coefficient of concordance among plates when positivity was evaluated on net-GPL. Our data strongly suggest that quantification and subtraction of sample blank may improve both interlaboratory agreement and reliability of anticardiolipin assay and minimize false-positive results.