Introduction: Anti-centromere antibodies (ACA) are useful biomarkers in the diagnosis of systemic sclerosis (SSc). ACA are found in 20 to 40% of SSc patients and, albeit with lower prevalence, in patients with other systemic autoimmune rheumatic diseases. Historically, ACA were detected by indirect immunofluorescence (IIF) on HEp-2 cells and confirmed by immunoassays using recombinant CENP-B. The objective of this study was to evaluate a novel CENP-A peptide ELISA.
Methods: Sera collected from SSc patients (n=334) and various other diseases (n=619) and from healthy controls (n=175) were tested for anti-CENP-A antibodies by the novel CENP-A enzyme linked immunosorbent assay (ELISA). Furthermore, ACA were determined in the disease cohorts by IIF (ImmunoConcepts, Sacramento, CA, USA), CENP-B ELISA (Dr. Fooke), EliA CENP (Phadia, Freiburg, Germany) and line-immunoassay (LIA, Mikrogen, Neuried, Germany). Serological and clinical associations of anti-CENP-A with other autoantibodies were conducted in one participating centre. Inhibition experiments with either the CENP-A peptide or recombinant CENP-B were carried out to analyse the specificity of anti-CENP-A and -B antibodies.
Results: The CENP-A ELISA results were in good agreement with other ACA detection methods. According to the kappa method, the qualitative agreements were: 0.73 (vs. IIF), 0.81 (vs. LIA), 0.86 (vs. CENP-B ELISA) and 0.97 (vs. EliA CENP). The quantitative comparison between CENP-A and CENP-B ELISA using 265 samples revealed a correlation value of rho=0.5 (by Spearman equation). The receiver operating characteristic analysis indicated that the discrimination between SSc patients (n=131) and various controls (n=134) was significantly better using the CENP-A as compared to CENP-B ELISA (P<0.0001). Modified Rodnan skin score was significantly lower in the CENP-A negative group compared to the positive patients (P=0.013). Inhibition experiments revealed no significant cross reactivity of anti-CENP-A and anti-CENP-B antibodies. Statistically relevant differences for gender ratio (P=0.0103), specific joint involvement (Jaccoud) (P=0.0006) and anti-phospholipid syndrome (P=0.0157) between ACA positive SLE patients and the entire SLE cohort were observed.
Conclusions: Anti-CENP-A antibodies as determined by peptide ELISA represent a sensitive, specific and independent marker for the detection of ACA and are useful biomarkers for the diagnosis of SSc. Our data suggest that anti-CENP-A antibodies are a more specific biomarker for SSc than antibodies to CENP-B. Furthers studies are required to verify these findings.