Comparison of Screening Dilution and Automated Reading for Antinuclear Antibody Detection on HEP2 Cells in the Monitoring of Connective Tissue Diseases

Background: Indirect immunofluorescence plays a major role in the detection of antinuclear antibodies (ANAs) and follow-up of their titers in the context of connective tissue diseases. Given the numerous unfavorable features of the conventional manual reading of HEP2 slides (need of time and expert morphologists for the reading, lack of standardization, subjectivity of the interpretation), the biomedical industry has developed automated techniques of slide preparation and microscope reading. Methods: We collected 49 sera beforehand analyzed by the conventional reading of slides. They were prepared again by QUANTA-Lyser (R) and reanalyzed in four different conditions: two dilutions of screening (1/40 and 1/80), two different systems of analysis, NOVA View (R) automated reading (INOVA Diagnostics), then confirmation by the operator, and conventional manual reading by two different qualified operators. The analysis was realized in blind of the first interpretation and clinical diagnosis. The sera were classified in four groups, on the basis of the results of the first analysis: negative sera (titer <1/160; 11 patients), low positives (titer at 1/160; 18 patients), moderated positives (titers between 1/320 and 1/640; 10 patients), and strong positives (titers between 1/1,280 and 1/2,560; 10 patients). Results: Among the 49 patients, 13 presented a connective tissue disease including 4 systemic scleroderma (SS), 3 rheumatoid arthritis (RA), 2 Goujerot-Sjogren (GS), 2 systemic lupus erythematosus (SLE), 1 polymyositis (PM), 1 Raynaud's syndrome (RS), and 1 CREST syndrome. One patient presented both an SLE and an SS. Regarding the screening dilution, the 1/40 dilution is less specific than the 1/80 dilution for both the systems of analysis (5.6% vs. 16.7% for the manual reading, and 27.8% vs. 50% for the automated reading). It also generates statistically more false positives (P = 0.037 for the conventional analysis and P = 0.003 for the automated system). The automated NOVA View (R) reading of slides allows a gain in specificity for both dilutions, and also statistically less false positives (P = 0.002 at the 1/40 and P = 0.0006 at the 1/80), and detriment of the sensitivity at the highest dilution (84.6% vs. 92.3% with manual reading). Thus, according to our analysis of 49 sera, the automated NOVA View (R) system of reading of slides at the dilution 1/80 seems to be a successful condition for the detection of ANAs on HEP2 cells, close to the significance (P = 0.067). Conclusion: The automated NOVA View (R) reading of slides allows saving time, and an improvement in the standardization. Nevertheless, it requires a confirmation by a qualified operator, to interpret mixed patterns in particular. (C) 2016 Wiley Periodicals, Inc.