Background/Purpose: The oral pathogen Aggregatibacter actinomycetemcomitans (Aa) generates citrullinated proteins targeted by autoantibodies in RA through its pore-forming toxin leukotoxin A (LtxA).  Aa-derived LtxA is implicated in atherogenesis and atherosclerotic plaque instability, conditions over-represented in RA.  No prior studies have explored the potential link between Aa, Aa-derived LtxA, and atherosclerosis in RA.

Methods: RA patients underwent cardiac computed tomography (CT) with coronary arterial calcification (CAC), a measure of coronary atherosclerosis, assessed.  Serum levels of IgG antibodies against Aa serotype b and purified LtxA were measured by ELISA.  Healthy controls cutoffs for seropositivity were used.   Multivariable robust regression was used to model the associations of anti-Aa and anti-LtxA with CAC, adjusting for pertinent confounders.

Results: A total of 194 RA patients [mean age=59 years; 60% female: median RA duration=9 years; 79% RF or anti-CCP seropositive] were studied.  Anti-Aa was detected in n=41 (21%) and anti-LtxA in n=82 (42%).  Adjusting for relevant confounders (listed in Fig), those with anti-LtxA had a mean CAC score 65% higher than those without anti-Aa or anti-LtxA (61 vs. 37 units, respectively; p=0.005 (Fig1a)].  Those anti-Aa⁺/anti-LtxA^– had a mean adjusted CAC score 262% higher than those with neither antibody (134 vs. 37 units, respectively; p<0.001) and 120% higher than those with anti-LtxA, with or without concomitant anti-Aa.  Among CAC-associated features, swollen joint count (SJC) was differentially associated with CAC depending on the context of anti-LtxA.   In anti-LtxA^– patients, SJC was not associated with CAC (Figure 1B, open circles).  However, in anti-LtxA⁺ patients, each swollen joint was associated, on average, with a 6 unit higher CAC score (p<0.001).  The interaction of anti-LtxA and swollen joints was significant (p=0.002).  The association was not linear, with a marked difference noted in those with>10 swollen joints (Figure 1B, solid diamonds).  Combining the two above groups with higher CAC scores [i.e. anti-Aa⁺/anti-LtxA^– and anti-LtxA⁺ and>10 swollen joints; n=32 (16% of the cohort)], the adjusted mean CAC was 171% higher vs. those without these features (114 vs. 42 units, respectively; p<0.001).

Conclusion: Infection with the periodontal pathogen Aa may contribute to atherosclerosis in RA. Anti-LtxA, a marker of exposure to leukotoxic Aa strains, was significantly associated with more CAC and identified patients in which synovitis was robustly and independently associated with the extent of CAC.  A subgroup of RA patients with isolated antibody positivity to the leukotoxic Aa serotype b showed markedly higher risk of atherosclerosis. Additional studies are warranted to elucidate the underlying mechanism, and whether Aa-specific screening and treatment may reduce CVD risk in RA patients.