Detection and identification of antinuclear antibodies (ANA) in a large community hospital.

Objective: Antinuclear antibodies (ANA) and their identification are important diagnostic tools in rheumatic diseases. We aimed to determine their prevalence in samples referred for ANA testing and to identify factors predicting more specific reactivities.

Methods: We analyzed the first sample of 6422 consecutive patients for ANA. Positive samples were analyzed by indirect immunofluorescence (IIF) on Crithidia luciliae and by line immunoassay. We used multivariate logistic regression to detect predicting variables.

Results: 42.6% of all patients were ANA positive of which 13.0% showed > or = one extractable nuclear antigen (ENA) reactivity with anti-SSA/Ro (5.5%), anti-SSB/La (2.9%), anti-Cenp-B (2.5%) and anti-histones (2.2%) as the most prevalent antibodies. Anti-double-stranded DNA antibodies (dsDNA) were present in 1.0%. The strongest overall predictor was ANA intensity regardless of pattern. Cenp-B however was best predicted by pattern. Anti-dsDNA and anti-histone were more frequent in samples with a homogenous as compared with a speckled pattern. Anti-SSA and anti-SSB were more frequent in females and anti-Sm in patients < or = 30 years.

Conclusions: The best overall predictor of antibodies to ENA or dsDNA is ANA intensity. Anti-Cenp-B is however best predicted by pattern. Samples with low ANA intensity (1+) may not need further testing unless a high clinical suspicion of ANA-associated disease.