Objectives: This study aimed to retrospectively examine the performance of EUROPattern (EPa), an automated immunofluorescent antinuclear antibody (IF-ANA) analyzer, in detecting serum samples containing disease-specific autoantibodies. Methods: Four independent experts visually assessed IF-ANA images from 429 routine samples captured by EPa at a cutoff value of 40-fold. The results of human interpretation were compared with those generated by the automated EPa system. Thirteen disease-specific autoantibody components were measured in all samples using a multiplex flow assay (ANA Screen). Results: Disease-specific autoantibodies were detected in 155 of 429 (36.1%) samples. The sensitivity of EPa for detecting these autoantibodies was 82.6%, which was lower than the 93.2% sensitivity achieved by expert evaluation. Conversely, the specificity was 74.1%, which was significantly higher than the 46.4% reported for experts. EPa demonstrated a high probability of detecting individual components of these autoantibodies; however, the modality exhibited slightly lower IF-ANA positivity in samples positive for antibodies against Jo-1 (66.7%), Scl-70 (79.3%), double-stranded deoxyribonucleic acid (82.1%), ribosomal protein (83.3%), and SS-A 60 (87.7%). Conclusions: EPa demonstrated reduced sensitivity in detecting serum samples containing disease-specific autoantibodies. These findings indicate that visual interpretation by experts remains essential, particularly for patients with a high pretest probability of systemic autoimmune rheumatic disease.
Mori et al. (Fri,) studied this question.