The mortality of patients with rheumatoid arthritis (RA) is increased compared with the general population and most of this increase can be attributed to excess cardiovascular (CV) deaths 1. Immune dysregulation and systemic inflammation have important roles in the accelerated atherosclerosis of RA. Serologic markers of inflammation, such as C-reactive protein (CRP) and serum amyloid A, are strongly associated with CV disease in the general population, in addition to systemic inflammatory diseases, including RA, systemic lupus erythematosus and inflammatory bowel disease 234. Although the INTERHEART study showed 90% to 94% of the risk of myocardial infarction could be accounted for by conventional risk factors (abnormal lipids, smoking, hypertension, diabetes, abdominal obesity and inactivity), in addition to psychosocial factors and diet 5, the contribution of inflammation is associated with these risk factors. Inflammation might potentiate a prothrombotic inflammatory state, or promote metabolic and psychosocial factors, which affect disease 67. For example, smoking is a risk factor for both CV disease and RA and itself associated with the promotion of inflammation, characterized by elevated CRP levels 8. For patients with inflammatory polyarthritis or RA, the baseline level of CRP or peak erythrocyte sedimentation rate (ESR) within the first year after onset has been independently associated with the risk of CV death 910. Thus, the progress of CV disease might depend on the burden of inflammation.

The receptor for advanced glycation end products (RAGE) is a multiligand member of the immunoglobulin superfamily of cell-surface receptor molecules encoded by chromosome 6 at the major histocompatability locus (MHC) class II/III junction 11. When ligated, RAGE is implicated in atherogenesis and chronic inflammatory diseases by amplifying inflammatory events in tissues previously sensitized by lipid deposition or inflammatory triggers 12. RAGE is expressed at low levels in normal tissues but upregulated at sites of vascular pathology or severe injury 1314. RAGE ligands include the calgranulin family of S100 proteins, the cytokine amphoterin (high-mobility group box chromosomal protein (HMGB1)), β-pleated sheet fibrils of amyloid and advanced glycation end products (AGEs) 15. AGEs are the products of nonenzymatic glycation and oxidation of proteins that form with aging, diabetes and renal failure, and at sites of inflammation (by oxidation) 16.

Levels of S100 proteins and HMGB1 are increased in both adult and juvenile RA 17, and inhibition of the RAGE has been shown to suppress inflammation in animal models of arthritis 18. There are four polymorphisms of the RAGE, of which the Gly82→Ser functional dimorphism shows the highest prevalence 19. Ser82 enhances receptor signaling through mitogen activated protein (MAP) kinases and nuclear factor of kappa B (NF-κB) 18. The increased prevalence of Ser82 in RA might be accounted for by linkage disequilibrium with the RA-associated HLA–DRB1*0401 allele 111820. However, although RAGE has been implicated in atherosclerosis, and the vascular component of diabetes mellitus and renal disease, RAGE Gly82→Ser polymorphism showed no association with CV disease in the general population 21. Here, we tested whether the RAGE Gly82→Ser polymorphism was associated with CV events in RA.

CV disease is a frequent and serious accompaniment for RA. We show here that in our population diagnosed with RA the risk of CV events was associated with history of statin therapy for hyperlipidemia and shorter RA disease duration, suggesting that the risk of vascular events is greatest in early disease. Association of CV events with shorter disease duration is consistent with several previous studies. Goodson and colleagues recently demonstrated 10% mortality, predominantly owing to CV events, within 10 years of onset of inflammatory polyarthritis 27. In a large study of CV mortality in RA spanning 40 years of follow-up, disease duration was found to have no influence, suggesting that the increased risk of CV death was already present at RA diagnosis. Our data also demonstrate that the RAGE Ser82 allele, which is present in one-fifth of patients, does not contribute to the risk of CV events in patients with RA.

Laboratory, clinical and epidemiologic evidence indicates that immune dysregulation, in addition to ongoing systemic inflammation, is crucial to the excess atherogenesis and CV mortality in RA compared with age-matched individuals without RA 282930. Thus, baseline or average ESR or CRP levels over the course of RA have been shown to predict subsequent CV death 910. However, inflammatory disease and its treatment, in turn, might affect either the prevalence or the contribution of traditional CV risk factors. Thus, cholesterol and inflammation tend to be inversely associated, owing to a disproportionate reduction of HDL and an adverse lipid profile 3132. Statins are known to be anti-inflammatory and have been shown to both improve vascular function in RA patients with a normal lipid profile and have additional immunomodulatory and RA disease-modifying benefits, including reduction of CRP levels, serum TNF concentrations and indices of plasma viscosity, in addition to short-term improvements in endothelial function and arterial stiffness 3334353637. Of clinical importance, this is the first study to show that statins significantly influence the outcome of atherosclerosis in RA. Although no direct indicator of inflammatory disease, such as CRP, persisted in the multivariate model, our study was limited in its ability to discern the influence of inflammation on CV events because we examined only RA patients and not matched non-RA controls and inflammatory markers and radiographs were examined at only a single time point, which occurred at variable times after commencing a variety of antirheumatic drugs rather than at baseline.

In RA, proinflammatory ligands of the RAGE, including S100A12 and HMGB1, are reported to be expressed at high levels, levels of soluble RAGE are reduced and RAGE expression is increased by synovial macrophages 173839. At least one RAGE allele was present as Ser82 in 10% to 20% of RA patients in the present prospective and retrospective cohorts, figures similar to previous studies of RA patients and higher than those in unselected population-based studies or control cohorts 18192140. We confirm here that the RAGE Ser82 allele is in linkage disequilibrium with HLA–DRB1*0401 404142. Because the RAGE Ser82 isoform is a functional polymorphism enhancing receptor signaling through MAP kinases and NF-κB, which activate an array of proinflammatory genes 18, we predicted an association of this polymorphism with CV events in RA, potentially through a more pronounced inflammatory response in these patients. However, the RAGE Ser82 isoform was not associated with more severe disease by any criteria in this RA cohort. Furthermore, no study has shown an independent association of the RAGE Ser82 isoform and RA. Thus, although certain RAGE ligands are highly expressed in RA and might contribute to the inflammatory process, the influence of the gain-of-function RAGE Ser82 isoform on RA expression is still unclear. Nevertheless, for similar reasons to those outlined above, the relationship between the RAGE Ser82 isoform and both inflammation and damage outcome in RA must be formally tested prospectively in RA inception cohorts.

Furthermore, in the current study of RA patients, the RAGE Ser82 isoform was not associated with CV events. This result supports the observation that the RAGE Ser82 isoform was not associated with CV disease in the general population 21, suggesting that the simple hypothesis that enhanced inflammatory signaling, mediated through the RAGE Ser82 isoform, predisposes to CV events is unlikely.

A Ser82 polymorphic allele of RAGE associated with enhanced RAGE signaling is present in 20% of patients with RA in linkage disequilibrium with HLA–DRB1*0401. In a cohort of established RA patients attending a tertiary referral center, a history of statin use and shorter duration of RA, but not RAGE genotype, were significantly associated with CV events. Prospective trials to determine the role of statins on CV outcome in RA patients will be of great interest.

ACR = American College of Rheumatology; AGE = advanced glycation end product; BMI = body-mass index; CCP = cyclic citrullinated peptide; CrCl = creatinine clearance; CRP = C-reactive protein; CT = computed tomography; CV = cardiovascular; ECG = electrocardiogram; ESR = erythrocyte sedimentation rate; HDL = high-density lipoprotein; HMGB1 = high-mobility group box chromosomal protein; HR = hazard ratio; LDL = low-density lipoprotein; MAP = mitogen activated protein; MHC = major histocompatability; MI = myocardial infarction; NF-κB = nuclear factor of kappa B; PCR = polymerase chain reaction; PVD = peripheral vascular disease; RA = rheumatoid arthritis; RF = rheumatoid factor; RAGE = receptor for advanced glycation end products; SCr = serum creatinine; TG = triglyceride; TIA = transient ischemic attack; TNF = tumor necrosis factor; VLDL = very-low-density lipoprotein.

The authors declare that they have no competing interests.

LC, IF, MT, TM and RT were involved in conception, design, acquisition, analysis and interpretation of data. LC, IF, TM and RT drafted and revised the manuscript. All authors read and approved the final manuscript.