Routine testing for antinuclear antibodies: a comparison of immunofluorescence, counterimmunoelectrophoresis and immunoblotting.

We compared the classical immunofluorescence test (IFT) and counterimmunoelectrophoresis method (CIE) with the new immunoblotting technique (IBT) for the detection of antinuclear antibodies (ANA). Sera from 200 patients were tested in all three assays. Patients were classified as having either rheumatic disease including rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), mixed connective tissue disease (MCTD), primary Raynaud's phenomenon or nonrheumatic disease. Within these broad categories, we observed that IFT and IBT showed a more or less comparable sensitivity and specificity (IFT: 0.54 and 0.82, respectively; IBT: 0.39 and 0.79, respectively). The CIE method combines a high specificity (0.99) with an extremely low sensitivity (0.08). By combining positive results obtained by IFT and IBT, a higher specificity (0.97) but a diminished sensitivity (0.24) is obtained. As IBT allows simultaneous discrimination between ANA of different specificities, we also tested for a correlation between the presence of anti-Sm, anti-RNP and anti-SS-B and the disease category. Only anti-SS-B discriminated significantly between rheumatic- and nonrheumatic disease. Anti-RNP was found in 50% of the SLE patients and in 50% of the MCTD patients; anti-Sm in 17% of the SLE patients and 25% of the MCTD patients. Anti-SS-B was found in 33% of the SLE patients. However, predictive rates of these ANA were low: 0.37 (anti-RNP), 0.67 (anti-Sm) and 0.43 (anti-SS-B). We conclude that from a practical point of view IFT is the preferable assay to screen for the presence of ANA. To characterize ANA specificities, however, the IBT is far superior to the CIE technique.