We read with interest the article by Nakamura et al. 1 reporting clinical and prognostic significance of autoantibodies to nuclear pore gp210 antigen and centromere in primary biliary cirrhosis (PBC). By studying 276 Japanese patients, the authors found that anti-gp210 antibody positivity was a risk factor for progression to end-stage hepatic failure and related to disease severity whereas anti-centromere antibody positivity predicted progression to complications of portal hypertension, such as esophageal varices, in the absence of jaundice.We tested 332 PBC patients including 170 from Barcelona, Spain and 162 from Larissa, Greece for gp210 and centromere autoantibody reactivity, using the same commercial enzyme-linked immunosorbent assay kits used by Nakamura et al. 1 Biochemical, clinical, and immunological data were obtained at presentation and during follow-up.Compared to the Japanese study, which reported a prevalence of 26% for both autoantibodies, we found this to be much lower in our cohort with anti-gp210 being present in 38 of 362 (10.4%) and anticentromere in 39 of 362 patients (10.8%). There was no difference between Spanish and Greek patients.Compared to seronegative patients, those who were positive for anti-gp210 had higher baseline (895 Ϯ 340 versus 612 Ϯ 423, P ϭ 0.006) and 12-month (573 Ϯ 213 versus 359 Ϯ 292, P ϭ 0.01) alkaline phosphatase levels, higher baseline bilirubin levels (1.4 Ϯ 0.7 versus 0.78 Ϯ 0.4, P Ͻ 0.001), and higher Mayo risk score (5.2 Ϯ 1 versus 4.2 Ϯ 0.7, P ϭ 0.012). Anti-gp210 positivity neither predicted outcome (survival, transplantation, liver-related death) nor defined histologic disease severity. Anti-centromere seropositivity predicted neither clinical or biochemical disease severity nor outcome. Akin to Nakamura et al., 1 we found that fewer than 15% anti-gp210 -seropositive patients progress to hepatic failure, but at variance with them, we are hesitant at attributing prognostic value to what appears to be a relatively weak association.In summary, our data highlight 3 points. They show that antigp210 is a biomarker of disease severity in PBC, 2-5 partially concurring with the findings from Japan. They also show that the autoantibody prevalence is much lower-less than half-in Mediterranean populations, in whom it also fails to act as a prognostic marker. Differences in results between the 2 studies are likely to derive from ethnic/geographical variations, because the methodological approach was identical. To define more confidently which characteristics of anti-gp210 are universal and which are ethnically/geographically determined, further studies on large PBC cohorts are needed.
Reply:We are pleased that Bogdanos and colleagues have undertaken an analysis of gp210 and centromere autoantibodies in their combined cohort of 332 patients with primary biliary cirrhosis (PBC) from Spain and Greece.The authors stated they used the same commercial anti-gp210 and anti-centromere enzyme-linked immunosorbent assay (ELISA) kits used by us. This is incorrect. We used the same...