Clinical utility of screening for antinuclear antibodies by enzyme immunoassay--a preliminary study

To evaluate the advantages and reliability of screening for antinuclear antibodies (ANA) by enzyme immunoassay (ELISA). Sera from 96 patients comprising 51 with systemic lupus erythematosus (SLE), 11 with other systemic rheumatological diseases (SRD) and 34 with various other diseases (non-SRD) were tested using a commercial ELISA kit (ANA-Ease, Genesis Biotechnology, U.K.). These sera consisted of 53 immunofluorescence assay (IF) ANA-positive and 43 IF ANA-negative samples We observed that when compared to the IF for ANA the sensitivity, specificity, predictive values for positives (PPV) and negatives (NPV) of ELISA were 90.7%, 85.7%, 89.1% and 87.8% respectively. Exclusion of borderline ELISA positive by slightly raising the cut-off optical density (OD) increased the specificity and PPV to 93.1%, and 94.1% respectively. Importantly, none of the non-SRD sera were positive when this higher cut-off was used. ELISA was noted to be strongly positive in three IF ANA-negative SLE patients. However there was no correlation between the ELISA ANA semi-quantitative index and the IF ANA titers. ELISA appears to be suitable as a preliminary screening test for ANA. An appropriate cut-off should be identified to segregate low positive samples that could be false-positives. Nevertheless, IF will need to be performed to estimate the titers, identify patterns of ANA positive samples and confirm results of low positive "gray-zone" samples and ELISA negative sera from patients with a high index of clinical suspicion of SLE.