Background Antimitochondrial antibodies (AMA) are autoantibodies directed against proteins of the outer and inner mitochondrial membrane. M2 antibodies stand out because of their high sensitivity and specificity for the identification of primary biliary cirrhosis.
Objectives To analyse the prevalence of AMA and M2 autoantibodies in a large cohort of Spanish patients with primary Sjogren syndrome (SS), and to correlate positive results with epidemiological, clinical and immunological expression of the disease.
Methods The GEAS-SS multicenter registry was formed in 2005 with the aim of collecting a large series of Spanish patients with primary SS. The cumulated ESSDAI index (2010 EULAR-SS disease activity index) was retrospectively calculated at diagnosis. AMA were detected by indirect immunofluorescence in triple rat tissue and M2 autoantibodies by ELISA against M2 proteins (PDC-E2, BCOADC-E2, OGDC-E2).
Results AMA were tested consecutively in 733 patients with primary SS (688 women, mean age at diagnosis of 54.65 yrs), of whom 63 (9%) showed positive results. No significant differences were found in the main epidemiological, clinical and immunological features according to the presence or absence of AMA, except for a higher frequency of anti-Ro/SSA antibodies (83% vs 68%, p=0.015) and leucopenia (33% vs 21%, p=0.038) in AMA+ patients in comparison with those with negative AMA. M2 autoantibodies were tested in 210 patients, of whom 19 (9%) were positive. Comparison of the main epidemiological, clinical and immunological features according to the presence or absence of anti-M2 antibodies showed a higher frequency of anti-Ro/SSA antibodies (95% vs 74%, p=0.049), leucopenia (53% vs 19%, p=0.002), neutropenia (48% vs 22%, p=0.025), low C4 levels (26% vs 6%, p=0.008) and biological activity measured by the ESSDAI (84% vs 50%, p=0.007). We have the paired determination of AMA and anti-M2 in 23 patients with positive results: 12 have concordant positive results and 11 discordant (7 had negative AMA with positive M2, and 4 positive AMA with negative M2). The mean value of M2 autoantibodies in AMA+ patients was twice than in those with negative AMA, although the difference was not statistically significant (85.67 vs 41.51, p=0.126).
Conclusions The study showed three key messages: i) the prevalence of AMA/M2 in primary SS was of 9%; ii) AMA/M2 correlated with the coexistence of Ro autoantibodies but not with systemic Sjögren; and iii) nearly half the patients showed discordant results between AMA and M2 positivities. We recommend the use of M2-ELISA to test for AMA in patients with primary SS.
Disclosure of Interest None declared