A novel automated indirect immunofluorescence autoantibody evaluation

Abstract

Autoantibodies (AAb), especially antinuclear (ANAs) and anticytoplasmatic antibodies (ACyA), are essential diagnosing markers for several autoimmune diseases. The current gold standard method for ANA detection is manual indirect immunofluorescence (IIF) on human epithelial-2 (HEp-2) cells. However, this technique is cost and time consuming, and characterized by considerable intra- and interlaboratory variability. Thus, an automated IIF-HEp-2 reader has been developed recently. In the current study, we compared the performance of the automated AAb IIF-HEp-2 interpretation to conventional detection methods. Autoantibody detection by IIF on HEp-2 cells was performed in a total of 260 sera of patients, including 34 with systemic lupus erythematosus, 111 with dermatomyositis or polymyositis, 74 with systemic sclerosis, 41 with rare AAb patterns, and 137 healthy individuals. Visual interpretation and routine immunoassays were compared with a novel automated IIF-HEp-2 system using Aklides pattern recognition algorithms. Positive AAbs were detected in 95-100% of rheumatic patients by automated interpretation, in 74-100% with manual reading, and in 64-100% by immunodot assay. Receiver operating characteristic curve analysis of fluorescent intensity revealed a high sensitivity and specificity for automated reading of AAb with an agreement ranging from 90% to 95% between manual and automated interpretation (kappa 0.554-0.69) for systemic sclerosis and myositis, respectively. This study demonstrates a good correlation between manual and automated interpretation of AAb including ANA and ACyA in patients with autoimmune diseases. Full automation of HEp-2 cell assay reading may minimize errors in ANA pattern interpretation and thus help in the standardization of ANA assessment.