Background/Purpose:

Among rheumatoid arthritis (RA) patients, antibodies to citrullinated protein antigens (ACPA) have been detected in atherosclerotic plaques. In particular, antibodies to citrullinated fibrinogen (cit-fib) have been associated with prevalent coronary artery calcium (Sokolove, Arth Rheum 2013). In the absence of RA, ACPA have been associated with cardiovascular disease (CVD) (Cambridge, Atherosclerosis 2013), suggesting that ACPA themselves may directly contribute to accelerated CVD recognized in RA. Therefore, we hypothesized that among RA-free first-degree relatives (FDRs) of patients with RA, ACPA are associated with markers of endothelial dysfunction, an indicator of initial vascular injury in CVD.

Methods:

From the Studies of the Etiology of RA (SERA) (a multicenter prospective study of preclinical RA), 113 FDRs who had been positive for any of 5 RA-related antibodies (Abs):  rheumatoid factor (RF), RF isotypes – IgM, IgG, IgA, or anti-cyclic citrullinated peptide (anti-CCP2) on at least one of their visits, and 100 FDRs who had never been Ab positive were selected, frequency matched on age, sex, and field center site. In cross-sectional testing of single samples obtained at the baseline exam, the following were measured: endothelial activation/injury markers: soluble intracellular adhesion molecule-1 (sICAM), soluble vascular cell adhesion molecule-1 (sVCAM) and E-selectin; and a panel of 15 ACPA (using Bio-Plex bead-based assay). ACPA were dichotomized as positive/negative based on cut-offs in 200 RA patients and 98 blood bank controls, and were developed using receiver operating characteristic (ROC) curves giving >90% specificity. Soluble VCAM was analyzed as a continuous variable; sICAM and E-selectin were log-transformed to approximate normality and were analyzed in log units. ANCOVA was used to evaluate associations between ACPA positivity and markers of endothelial dysfunction, adjusting for age, sex, race, body mass index, pack-years of smoking, and high sensitivity C-reactive protein.

Results:

Among 193 FDRs with complete data and presented in detail in Table 1, sICAM was significantly higher in FDRs who were positive for ACPA to cit-enolase compared with those who were negative, with anti-cit–vimentin showing similar magnitude and marginal significance. sVCAM  was significantly higher in those who were positive for antibodies to cit-fibrinogen and cit-vimentin. No significant associations were observed between ACPA and E-selectin.

Conclusion:

In FDRs without classified articular RA, sICAM and sVCAM were higher in those who were ACPA positive. While cit-fib and cit-vimentin have been implicated in subclinical CVD in RA, and antibodies to a-enolase have been associated with vasculitis, our findings support the potential role of these ACPA in the initial insult in the development of CVD, even in the absence of clinical RA.