A Comparative Evaluation of the Performances of Anti-MIT3, Anti-gp210, and Anti-sp100 Antibodies for the Diagnosis of Primary Biliary Cirrhosis

ABSTRACT

BACKGROUND: Anti-mitochondrial antibody (AMA) is a serological hallmark of primary biliary cirrhosis (PBC). AMAs are detected by an immunofluorescence assay (IF), which is subject to errors. We evaluated the diagnostic performances of the AMA ELISA test (the anti-MIT3 antibody) and PBC-associated antinuclear antibody (ANA) tests (the anti-gp210 and anti-sp100 antibodies). METHODS: AMA, anti-gp210, and anti-sp100 were measured in the sera of 130 subjects including patients for whom the AMA test was requested with the clinical suspicion of PBC, patients with other autoimmune diseases, and those undergoing health check-ups. AMA was detected by both IF and ELISA (anti-MIT3 antibodies), and anti-gp210 and anti-sp100 were detected by ELISA. The diagnostic performances of the anti-MIT3, anti-gp210, and anti-sp100 were compared with that of the AMA IF test. Associations between the presence of anti-sp100 or anti-gp210 and the diagnosis and biochemical abnormalities of PBC were investigated. RESULTS: The area under the curve of anti-MIT3 for the diagnosis of PBC was 0.934 (95% confidence interval, 0.877-0.970), and the agreement between anti-MIT3 and AMA IF was 93.8% (kappa, 0.82). The sensitivities of anti-MIT3 and AMA IF were both 100%, and the specificities were 83.1% and 81.4%, respectively, whereas the sensitivities of anti-gp210 and anti-sp100 were 41.7% and 16.7%, and their specificities were 94.9% and 97.5%, respectively. The presence of anti-gp210 was associated with the diagnosis of PBC (P=0.0001), but that of anti-sp100 was not. CONCLUSIONS: The diagnostic performance of anti-MIT3 is comparable to that of AMA IF. Anti-gp210 seems to be complementary to AMA for the diagnosis of PBC.