Introduction

Patients with systemic inflammatory conditions such as systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) have been shown to develop premature and accelerated atherosclerosis [1, 2], but it remains unclear whether this occurs in systemic sclerosis (SSc). Involvement of the microvasculature is one of the earliest features of SSc, preceding and potentially contributing via tissue ischemia to the widespread fibrosis characteristic of this condition. Pathological changes include disruption of the endothelium, mononuclear cell infiltration of the vessel wall, frank obliterative lesions, and progressive loss of capillaries. There is also disruption of neuroendothelial control mechanisms of vascular tone. These microvascular abnormalities contribute to the pathogenesis of pulmonary arterial hypertension (PAH), scleroderma renal crisis, Raynaud's phenomenon, and digital ulceration.

Although macrovascular disease was not originally considered a feature of SSc, multiple studies have revealed an increased prevalence of large-vessel disease of the upper and lower limbs in patients with SSc [3, 4]. The prevalence of coronary artery and cerebrovascular disease in SSc, however, remains unclear. The purposes of this review are to summarize the evidence for macro-vascular disease, cardiovascular risk factors, and surrogate measures of cardiovascular disease in SSc and to explore potential pathogenic mechanisms of accelerated atherosclerosis in patients with SSc.

Clinical cardiovascular disease

Peripheral vascular disease

Peripheral vascular disease in patients with SSc has been reported in uncontrolled observational studies that used techniques such as the ankle brachial pressure index (ABPI), lower-limb Doppler ultrasound, and angiography [5–7]. Two studies have compared the prevalence of peripheral vascular disease in patients with SSc with that in controls. Veale and colleagues [3] measured the prevalence of intermittent claudication in a cohort of 46 patients with SSc by using a validated World Health Organization questionnaire and compared the findings with those of the Edinburgh Artery Study, a cross-sectional population-based study of a neighboring region. The rates of prevalence of intermittent claudication were 21.7% in the SSc cohort and 4.5% in the Edinburgh Artery Study. Although the two groups were not directly comparable, the authors suggested that the size of the difference and the fact that the SSc cohort was younger and predominantly female strengthened the significance of their findings. The authors went on to perform a case control study of 54 patients with SSc and 43 age- and sex-matched controls [4]. Peripheral vascular disease, as diagnosed by the ABPI, was present in 17% of patients with SSc and in none of the controls ( P = 0.003) [4]. There was no difference in the traditional cardiovascular risk factor profile between the two groups.

The contribution of traditional cardiovascular risk factors to peripheral vascular disease was examined in a retrospective review of 26 angiograms performed in a single SSc cohort [5]. The authors demonstrated a significant association between the presence of traditional cardiovascular risk factors and proximal, but not distal, vascular disease in the lower limb. Though limited by a small sample size, this study suggests that, in at least some cases, peripheral vascular disease in SSc is not atherosclerotic but related to the vasculopathy of SSc itself. In keeping with this possibility, Youssef and colleagues [8] reported chronic obliterative thrombo-angiitis on histological examination of an amputated limb in a woman with limited SSc. Involvement of the vasa vasorum has been suggested as a potential cause of macrovascular disease in SSc [9].

Cerebrovascular disease

Few studies have investigated the prevalence of cerebrovascular disease in SSc. Youssef and colleagues [8] performed a retrospective cohort study comparing the prevalence of macrovascular disease in 31 patients with limited SSc and 31 matched controls. Cerebrovascular disease was defined as a transient ischemic attack or completed stroke diagnosed by a physician, the presence of carotid or vertebral artery bruits, Doppler evidence of carotid or vertebral artery disease, or angiographic evidence of carotid artery stenosis. The investigators found no significant difference in cerebrovascular disease between patients with SSc and controls (26% versus 19%, respectively; P = 0.544). In contrast, Ho and colleagues [4] examined the prevalence of carotid artery disease by using B mode and color Doppler ultrasound and demonstrated carotid artery stenosis in 64% of patients with SSc and 35% of age- and sex-matched controls (P = 0.007).

Cardiac disease

Coronary artery disease

Early autopsy studies of patients with SSc suggested that atherosclerotic disease of the coronary arteries was infrequent. Several case series demonstrated myocardial infarction (MI) in the presence of patent coronary arteries, suggesting primary myocardial disease or vasospasm [10, 11]. The most quoted of these studies was a retrospective review of 52 autopsies of patients with SSc; in that review, 26 (50%) patients were found to have foci of myocardial necrosis [11]. Of these 26 patients, 23 had widely patent extramural coronary arteries. A common pattern in those with myocardial involvement was contraction band necrosis, which is a histopathological correlate of reperfusion injury. The authors therefore hypothesized that the observed myocardial damage was due to intermittent hypoperfusion of the myocardium secondary to Raynaud's phenomenon of the intramural coronary arteries. It is important to note that these autopsy studies were performed in an era when SSc was largely untreated and patients were more likely to succumb to SSc-related complications such as sclero-derma renal crisis. This temporal change is reflected in an analysis of mortality over a 30-year period from 1972 to 2002, in which 10-year survival improved from 54% to 66% and non-SSc-related deaths increased from 20% to 50% of all deaths [12].

There are no large cohort studies examining the prevalence of coronary artery disease in living patients with SSc. Most studies to date have reported the presence of coronary atherosclerosis in SSc patients who undergo coronary angiography for clinical indications. For example, Akram and colleagues [13] reviewed all 172 coronary angiograms performed in a single SSc referral center from 1998 to 2004 and found coronary artery disease in 22%. These authors calculated standardized prevalence ratios based on age, gender, and presenting cardiac symptom and, using published probability tables relating pretest likelihood of coronary artery disease to angiographic abnormalities, concluded that the prevalence of coronary artery disease was no greater than that expected in individuals without SSc. This was, however, a retrospective study with no control group. Furthermore, patients included in the study had differing indications for angiography and the number of patients undergoing angiogram for suspected coronary artery disease was not reported. When grouped according to predominant cardiac symptom, only 27 patients had typical angina, 29 had atypical angina, and 116 had dyspnea on exertion. There was also a very high prevalence of PAH (63% of patients with limited disease and 85% of patients with diffuse disease), perhaps accounting for a significant proportion of reported symptoms. This was therefore a study of a highly selected group rather than consecutive patients with SSc.

The relationship between clinical events and coronary angiogram findings was examined in a retrospective review of all SSc patients admitted to a single institution with acute MI over a 20-year period [14]. MI was defined as electrocardiographic changes and an at least twofold increase in serum myocardial creatinine kinase (CK) level. The authors identified 11 patients, 3 of whom had normal coronary arteries on cardiac catheterization. The odds ratio of finding normal coronary arteries in acute MI was 34 (95% confidence interval of 14 to 81) for patients with SSc compared with the general population. Of the 11 cases, however, 7 had significant renal impairment and 4 had frank scleroderma renal crisis. This may affect the validity of these results given that the clearance of CK is reduced in renal failure and an elevated CK formed part of the definition of MI.

Another angiographic study assessed the overlap between PAH and coronary artery disease [15]. One hundred twenty consecutive patients with SSc were assessed clinically for cardiorespiratory disease. Both right heart catheterization and coronary angiography were performed for suspected PAH in 20 patients and for suspected coronary artery disease in 10 patients. There was considerable overlap of diagnoses; PAH was found in 12 of the 20 suspected PAH cases and in 2 of the 10 suspected coronary artery disease cases. Similarly, coronary artery stenosis was found in 9 of the 20 suspected PAH cases and in 6 of the 10 suspected coronary artery disease cases. All eight of the patients who underwent revascularization for coronary artery disease improved significantly in 6-minute walk distance and the Borg dyspnea index. The authors concluded that coronary artery disease symptoms may be atypical in patients with SSc and that revascularization may improve symptoms and physical activity in selected patients.

Few angiographic studies have been performed in asymptomatic patients with SSc. Tarek and colleagues [16] performed coronary catheterization in 9 patients with diffuse SSc and 5 patients with limited SSc, all of whom were female and asymptomatic. The authors detected a total of 19 angiographic abnormalities, with significant coronary artery stenosis in 3 patients with limited disease. Other angiographic findings included coronary artery ectasia, slow flow, tortuosity, calcification, and spasm. The prevalence of traditional cardiovascular risk factors in this cohort was low, with hypertension in 3 patients but no other risk factors of note. While the study was small and uncontrolled, the findings suggest that angiographic abnormalities may be higher than previously thought in asymptomatic patients with SSc.

Primary myocardial disease

Microvascular disease of the myocardium can also result in angina pectoris or acute MI (or both) in SSc. It has been proposed that recurrent vasospasm, together with irreversible structural lesions, leads to repeated focal ischemia and eventual myocardial fibrosis [17]. When advanced, this manifests as systolic and diastolic dysfunction, leading in some cases to overt congestive cardiac failure. Myocardial perfusion in SSc has been found to be impaired by using cardiac catheterization [18], radio-nuclide imaging [19], and echocardiographic techniques [20]. One study of 26 unselected patients with diffuse SSc found evidence of cardiac abnormalities on thallium-perfusion scanning in 20 patients, only 6 of whom had symptomatic cardiac disease [19]. Of the 7 patients who had exercise-induced defects and who underwent subsequent coronary angiography, all had normal coronary arteries. Furthermore, patients with exercise-induced defects that were more significant had lower left and right ventricular ejection fractions on radionuclide ventriculography. This led the authors to conclude that abnormalities of myocardial perfusion in SSc are due to a disturbance of the myocardial microcirculation and that this, in turn, contributes to ventricular dysfunction.

In recent years, modalities that are more sensitive have been developed to detect primary myocardial involvement in SSc. Non-standard echocardiographic techniques such as pulsed tissue Doppler and strain rate imaging better quantify regional myocardial function than standard echocardiography. Cardiac magnetic resonance imaging (MRI) has been used to image myocardial fibrosis in SSc [21]; however, histological confirmation of areas of MRI signal abnormality is lacking. N-terminal pro-brain natriuretic peptide (NT-proBNP) is an amino acid produced from cleavage of the prohormone proBNP, which is released from the heart in response to pressure or volume overload or both. In the general population, it has diagnostic and prognostic value in heart failure and other cardiovascular diseases [22]. In SSc, NT-proBNP has been investigated as a biomarker of primary myocardial disease and is elevated in patients with primary myocardial dysfunction as determined by tissue Doppler echocardiography [23]. It has the advantage of being easily performed and widely applied; however, it is not specific for primary myocardial disease, as it is also elevated in PAH.

Surrogate markers of atherosclerosis

Various surrogate markers of atherosclerosis have been investigated in SSc but with conflicting results.

Carotid intima-media thickness

Carotid intima-media thickness (CIMT) as measured by high-resolution ultrasound is a well-validated marker of subclinical atherosclerosis. Increased CIMT has been shown to correlate with traditional cardiovascular risk factors and to predict future vascular events in healthy individuals [24]. A meta-analysis of CIMT in rheumatic diseases, including RA, SLE, and SSc, found that CIMT was significantly increased in this population compared with healthy, age- and sex-matched controls [25]. The pooled result of the SSc studies demonstrated a greater CIMT in patients with SSc than controls, suggesting an increased prevalence of subclinical atherosclerosis. The effect size seen in SSc was also greater than those in RA and SLE. Despite this finding, a number of individual studies have found no increase in CIMT in patients with SSc (Table 1) [26–31].

Table 1 Studies of carotid intima-media thickness in systemic sclerosis
Full size table

Correlates of increased CIMT have also been assessed in several studies. Bartoli and colleagues [32] demonstrated increased CIMT in patients with SSc and found an association between higher CIMT and the deletion (D) polymorphism of the angiotensin-converting enzyme (ACE) gene. The ACE DD genotype has been shown to be associated with atherosclerosis in the general population [33]. Another group has demonstrated an association between increased CIMT in patients with SSc and the presence of antibodies to heat-shock protein-65 [34], which are also implicated in the pathogenesis of atherosclerosis.

Flow-mediated dilatation

Flow-mediated dilatation (FMD) is a marker of endothelium-dependent vasodilatation and is measured using high-resolution ultrasound at the brachial artery. In healthy arteries exposed to a short period of occlusion by tourniquet, restoration of blood flow results in transient vasodilatation. Impaired FMD is associated with the presence of traditional cardiovascular risk factors [35] and is independently predictive of incident cardiovascular events [36]. FMD is usually performed in concert with nitrate-mediated dilatation, which is assessed after sublingual nitroglycerin administration and reflects endothelium-independent vasodilatation. Many, but not all, studies have found FMD to be decreased in SSc compared with controls (Table 2) [27, 28, 37–44]. Some groups have found nitrate-mediated dilatation to be reduced also [40, 41, 45], and this could suggest a co-existing functional or structural abnormality of arterial smooth muscle, adventitia, or both. Interestingly, in the studies in which no difference in FMD was detected, the majority of patients had limited rather than diffuse disease [27, 44], suggesting more endothelial dysfunction in diffuse SSc.

Table 2 Studies of flow-mediated dilatation in systemic sclerosis
Full size table

Arterial stiffness

Arterial stiffness is measured by the techniques of pulse wave analysis (PWA) and pulse wave velocity (PWV), and carotid-femoral PWV is considered the current 'gold-standard' measurement of arterial stiffness [46]. PWA, expressed as the augmentation index (AI), reflects the stiffness of the aorta, whereas carotid-femoral PWV reflects the velocity of the pulse wave along the aortic and aorto-iliac pathways. Increased arterial stiffness results in premature return of reflected waves in late systole, causing increased load on the left ventricle and increased myocardial oxygen demand. Arterial stiffness is increased in the presence of cardiovascular risk factors [47] and is an independent predictor of cardiovascular events and cardiovascular and all-cause mortality across a wide range of patient populations [48].

Arterial stiffness has been examined in SSc but with varying results (Table 3) [27, 31, 37, 39, 44, 49, 50]. Cypiene and colleagues [37] found arterial stiffness as measured by PWA and PWV to be elevated in 17 patients with diffuse SSc compared with 34 healthy controls. Timar and colleagues [49] found that AI and PWV were both elevated in 40 patients with SSc and in 35 controls. The investigators also found that PWV was higher in patients with limited disease than in those with diffuse disease and that PWV was positively correlated with disease duration. The authors therefore postulated that PWV may be a better measure of arterial stiffness than AI in SSc. Another group found arterial stiffness to be elevated in patients with SSc or mixed connective tissue disease and that arterial stiffness correlated with elevated levels of soluble markers of endothelial activation, including plasma nitrate, soluble E-selectin, and soluble vascular cell adhesion molecule-1 (VCAM-1) [44]. Liu and colleagues [31] found regional differences in arterial stiffness with increased PWV in the forearm and arm but not the upper arm, aorta, or leg in patients with SSc compared with controls. This suggests preferential involvement of the muscular arteries of the forearm [31]. Several other studies have been unable to demonstrate a significant difference in arterial stiffness between patients with SSc and controls [27, 50].

Table 3 Studies of arterial stiffness in systemic sclerosis
Full size table

Coronary calcium score

A novel method of assessing coronary artery disease is the coronary calcium score, as determined by multi-detector computed tomography. This technique measures coronary artery calcification that occurs in atherosclerotic plaque. In a cross-sectional study of 17 patients with SSc free from clinical cardiovascular disease and 17 controls, coronary calcification was present in 9 patients with SSc and 3 controls (P = 0.03) [51]. The correlation between coronary calcification and angiographic findings in the SSc population is unknown and larger studies are required to investigate this association.

Cardiovascular risk factors in systemic sclerosis

There is limited information regarding the prevalence of traditional cardiovascular risk factors in SSc. A study of 40 patients with SSc and 45 controls demonstrated no difference in blood pressure on 24-hour ambulatory blood pressure monitoring [29]. Significantly lower levels of high-density lipoprotein were recorded in one study of 24 female patients with limited SSc compared with 24 healthy controls [52], whereas another study found significantly elevated lipoprotein [a] in 31 female patients with SSc compared with 33 healthy controls but no significant difference in other cholesterol parameters or triglycerides [53]. Other novel cardiovascular risk factors that have been reported as elevated in SSc include oxidized low-density lipoprotein and endothelin [54].

Potential mechanisms of cardiovascular disease in systemic sclerosis

Inflammation

Inflammation is a key component of atherosclerosis [55], and it is well established that even a relatively minor elevation of inflammatory markers (such as C-reactive protein) is predictive of cardiovascular events in the general population [56]. Cardiovascular disease occurs more frequently in patients with chronic inflammatory diseases such as SLE [1] and RA [2] than in the general population, and although the precise mechanism of accelerated atherosclerosis in these conditions has not been fully elucidated, various cellular and cytokine pathways have been implicated. Excess cytokine production and release are also key events in the pathogenesis of SSc. Numerous inflammatory mediators that have been implicated in the pathogenesis of atherosclerosis, including tumor necrosis factor-alpha, interleukin-6, and high-sensitivity C-reactive protein, have been demonstrated to be increased in patients with SSc compared with controls [57]. The relationship between these mediators and cardiovascular disease in SSc is unclear; however, it is possible that chronic systemic inflammation could promote accelerated atherosclerosis in patients with SSc.

Endothelial dysfunction

Endothelial dysfunction is a component of the pathophysiology of both SSc and atherosclerosis (Figure 1). In SSc, the initiating injury is unknown, but endothelial cell damage leads to enhanced expression of adhesion molecules and elevated levels of circulating soluble adhesion molecules. Soluble E-selectin, intercellular adhesion molecule-1, and VCAM-1 levels are all significantly increased in SSc, reflecting endothelial activation [58]. Enhanced endothelial cell expression of adhesion molecules results in adhesion of inflammatory cells, transmigration across the vessel wall, and infiltration of the extracellular matrix. Another important component of endothelial dysfunction in SSc is derangement of vasoactive mediators, with an increase in vasoconstrictive endothelin and a decrease in the vasodilator nitric oxide (NO).

Figure 1
figure 1

Potential factors contributing to the pathogenesis of cardiovascular disease in systemic sclerosis. EPC, endothelial progenitor cell; MP, microparticle; NO, nitric oxide.

Full size image

In atherosclerosis, endothelial cell dysfunction is the common pathway by which factors such as elevated low-density lipoprotein, elevated plasma homocysteine, various infectious agents, and exposure to free radicals from smoking, hypertension, and diabetes mellitus are proposed to contribute to pathogenesis [55]. Endothelial dysfunction leads to upregulation of adhesion molecules on the endothelium and increased vessel wall permeability. Lipid-laden monocytes and macrophages known as foam cells then accumulate. Migration and proliferation of vascular smooth-muscle cells follow, leading to remodeling of the vessel wall and atherosclerotic plaque formation.

Endothelial progenitor cells

A reduction in the density of blood vessels is one of the hallmarks of vascular disease in SSc. Paradoxically, this occurs in the setting of increased circulating angiogenic factors, such as vascular endothelial growth factor [59]. Several theories have been advanced to account for the deficient angiogenesis in SSc despite the presence of permissive vascular growth factors. One explanation is a reduction in circulating endothelial progenitor cells (EPCs). Asahara and colleagues [60] isolated a population of progenitor cells from human peripheral blood which were incorporated into sites of new blood vessel formation when adoptively transferred to an animal model of ischemia. Whereas Kuwana and colleagues [61] initially showed that EPCs were decreased in SSc, others [62] have since demonstrated that these cells are elevated in early disease and decrease with increasing disease duration. EPCs have also been the subject of study in cardiovascular disease, in which a decreased number in the peripheral circulation has been shown to be predictive of recurrent acute coronary artery events [63].

Nevskaya and colleagues [64] examined the relationship between cardiovascular disease and EPCs in 40 patients with SSc and found that a decreased EPC count correlated with endothelial dysfunction, as measured by FMD. Decreased EPC number did not, however, correlate with subclinical atherosclerosis as measured by CIMT [64]. There was also no relationship between EPC count and the Framingham risk factor score or the presence of traditional cardiovascular risk factors.

Although EPCs are a promising potential biomarker of both cardiovascular risk and SSc disease activity, there is some debate over their true significance. Some groups have questioned whether the cells isolated by Asahara and colleagues [60] and others are true progenitor cells that incorporate into new blood vessels or rather are cells of hemotopoietic lineage which have a paracrine effect on blood vessel formation [65]. The development of cell lineage markers that are more specific will help elucidate the biological properties of these cells.

Microparticles

Microparticles (MPs) are small circulating membrane-coated vesicles that are important mediators of intercellular signaling. MPs contribute to the immunopathogenesis of various thrombotic and rheumatic diseases via their role in the regulation of inflammation, thrombosis, and angiogenesis. MPs arise from a variety of cell types and are formed by the process of blebbing during cell activation and apoptosis. MP numbers have been shown to be elevated in patients with SSc and to correlate with the presence of interstitial lung disease [66]. In SSc, MPs have been identified to arise from platelets, endothelial cells, monocytes, and T cells, reflecting activation of these cells. MPs have also been suggested as a biomarker of coronary artery disease. High levels of endothelial MPs are found in patients with acute MI and result in severe endothelial dysfunction by selectively impairing the production of NO [67].

Angiotensin-converting enzyme insertion/deletion polymorphism

Activation of the renin-angiotensin-aldosterone axis is central to the pathogenesis of scleroderma renal crisis, as demonstrated by the dramatic improvement in survival since ACE inhibitors have been used to treat the condition. Over a 30-year period that spanned the introduction of ACE inhibitors, deaths due to renal crisis decreased from 42% to 6% (P <0.005) [12].

An insertion/deletion (I/D) polymorphism of the ACE gene has been identified, and the highest levels of plasma ACE are associated with the DD genotype and lowest levels are associated with the II genotype [68]. The original report of the relationship between this polymorphism and acute MI found a risk ratio of 1.34 (95% confidence interval of 1.05 to 1.70) for MI in those with the DD genotype compared with those with either the II or ID genotype [33]. A later meta-analysis of studies concluded that the risk conferred by the DD genotype was more modest (risk ratio of 1.0 to 1.1) [69]. As described earlier, the ACE gene polymorphism has been examined in patients with SSc; in this study, the presence of a D allele correlated with increased CIMT [32]. These investigators also reported a correlation between the presence of a D allele and an increased risk of SSc in the Italian population [70]; however, this finding has not been confirmed in other populations [71].

Antiphospholipid antibodies

The presence of anticardiolipin (aCL) or anti-β2-glycoprotein I (anti-β2GPI) antibodies or both in the absence of antiphospholipid syndrome or other autoimmune disease has been examined as an independent risk factor for ischemic heart disease. A study of incident cardio-vascular events in a population of Hawaiian men of Japanese descent found that anti-β2GPI in the presence of aCL was independently predictive of incident ischemic stroke and MI over 20 years of follow-up [72]. In a study of recurrent cardiac events post-infarction, elevated IgG aCL and low IgM aCL, but not anti-β2GPI, were independent risk factors for recurrent events [73].

An increased prevalence of aCL and anti-β2GPI antibodies occurring in the absence of typical clinical manifestations of antiphospholipid syndrome has been demonstrated in patients with SSc compared with controls [74]. Boin and colleagues [75] showed that anti-β2GPI is associated with both higher mortality and vascular disease, including digital ischemia and PAH, in SSc. The authors did not report on the cause of mortality or the prevalence of cardiovascular disease in their cohort.

Conclusions

Since the introduction of ACE inhibitors for the management of scleroderma renal crisis in the 1980s, the overall survival of patients with SSc has improved significantly and the relative contribution to mortality of non-SSc comorbidities has increased [12]. With better therapies for PAH, we are likely to see further improvements in survival and non-SSc-related organ dysfunction become more apparent. Patients with other chronic inflammatory diseases such as SLE and RA experience excess cardiovascular disease, predominantly due to accelerated atherosclerosis, and this may occur in SSc also. Given the prominent involvement of the vasculature in SSc and the role of inflammation in the development of atherosclerosis, overlap between these disorders might not be that surprising. However, there is a relative paucity of data regarding clinical and preclinical cardiovascular disease in SSc, whether due to atherosclerosis or microvascular involvement. In particular, adequately sized population-based cohort studies of the incidence and prevalence of coronary artery disease in SSc relative to the general population are lacking. Surrogate markers of cardiovascular disease may assist in quantifying the burden of subclinical atherosclerosis and may also provide insights into the mechanisms underlying the development of macrovascular disease in SSc. Recognition of such an association could have important clinical ramifications.