A comparison of anti-nuclear antibody quantification using automated enzyme immunoassays and immunofluorescence assays

Anti-nuclear antibodies (ANA) have traditionally been evaluated using indirect fluorescence assays (IFA) with HEp-2 cells. Quantitative immunoassays (EIA) have replaced the use of HEp-2 cells in some laboratories. Here, we evaluated ANA in 400 consecutive and unselected routinely referred patients using IFA and automated EIA techniques. The IFA results generated by two independent laboratories were compared with the EIA results from antibodies against double-stranded DNA (dsDNA), from ANA screening, and from tests of the seven included subantigens. The final IFA and EIA results for 386 unique patients were compared. The majority of the results were the same between the two methods (n = 325, 84%); however, 8% (n = 30) yielded equivocal results (equivocal-negative and equivocal-positive) and 8% (n = 31) yielded divergent results (positive-negative). The results showed fairly good agreement, with Cohen's kappa value of 0.30 (95% confidence interval (CI) = 0.14-0.46), which decreased to 0.23 (95% CI = 0.06-0.40) when the results for dsDNA were omitted. The EIA method was less reliable for assessing nuclear and speckled reactivity patterns, whereas the IFA method presented difficulties detecting dsDNA and Ro activity. The automated EIA method was performed in a similar way to the conventional IFA method using HEp-2 cells; thus, automated EIA may be used as a screening test.