Arterial Events and Atherosclerosis in Systemic Sclerosis
The first major finding in this population-based study is that ischemic arterial events, specifically IHD and IPVD, are more common in patients with SSc, while ICVD and measures of atherosclerosis did not differ on a group level as compared to population controls. Second, the ACA+ SSc subgroup was more affected with ischemic arterial events and with carotid plaques in comparison both to other SSc patients and to controls.
The observed high prevalence of IPVD in patients with SSc highlights an important clinical problem recently also reported by Man et al.. Youssef et al. found an even higher prevalence of IPVD in females with lcSSc compared with controls (58% vs.10%). More claudication, as a sign of IPVD, was also reported in SSc patients by Veale and collaborators.
Despite the higher prevalence of IPVD, SSc patients did not, as a group, have lower ABI than controls. This is in line with three previous studies, but contrasts with Ho et al.. In our study, ACA+ patients had more IPVD compared to both ACA-patients and controls and there was also a trend toward lower ABI in the ACA+ group. Wan et al. examined ABI in 119 SSc patients and found no difference between patients with lcSSc and dcSSc, but interestingly ACA+ seemed to prevail among patients with symptomatic ischemic disease. Previous studies have identified ACA+ as a predictor for severe digital loss. Thus, in addition to microvascular disease, the large arteries of the extremities seem to contribute significantly to the severe distal ischemic damage seen in many SSc patients, especially in the ACA+ subgroup.
We report a three-fold increased occurrence of IHD in SSc patients and even higher figures in the ACA+ subgroup. This is in line with the recently reported incidence of myocardial infarction and self-reported IHD, both of which were two-fold enhanced in SSc patients versus the general population. Older autopsy studies described that myocardial infarctions were common in SSc patients, despite normal coronary arteries. Akram et al. evaluated 179 angiograms from SSc patients but did not find an increased prevalence of coronary atherosclerosis and normal angiograms were more common than expected among 11 SSc patients with myocardial infarction. These studies suggest that vasospasm rather than atherosclerosis is a major pathogenic mechanism behind SSc-related heart disease. On the other hand, MRI studies have found that calcification of cardiac vessels are common in SSc. The inconsistent results of previous work may partly be explained by our observation that atherosclerotic plaque occurrence is selectively enhanced in the ACA+ subgroup of SSc patients.
In our study, impaired kidney function was highly associated with IHD, a finding that is supported by Derk et al., who reported that the majority of SSc patients with IHD had renal failure.
Cerebrovascular disease was, in accordance with Youssef et al., equally common in SSc patients and controls. However, recently Man et al. reported enhanced incidence of stroke in a large epidemiological study.
On a group level, there was no difference between patients and controls concerning atherosclerosis, measured as plaques occurrence, IMT or ABI. Our study population was fairly old and the plaque frequency was high in both patients and controls. Previous studies are smaller and the majority measured only IMT. Some report, in agreement with us, similar IMT in patients and controls, others found thicker IMT in SSc patients. Only three studies examined the frequency of plaques, but their definitions of plaque differ. We had no exclusion criteria, while some studies excluded SSc patients with manifest cardiovascular disease (CVD). Furthermore, the selection of comparators is important. The healthy controls in previous studies are likely to have less atherosclerosis than our population-based controls. Methods, definition and selection procedures of both patients and controls are likely to contribute to the heterogeneity of previous studies, reviewed in a recent meta-analysis.
Generally, levels of endothelial and inflammatory biomarkers were higher in patients compared to controls, demonstrating that SSc is a low-grade inflammatory disease. Of investigated inflammatory biomarkers, IL-6 remained independently associated with plaques. For SSc this is a new observation but IL-6 has been identified as a risk factor for subclinical coronary atherosclerosis in RA, and it has also been linked to poor prognosis in patients with IHD from the general population.
VCAM-1 was associated with both IHD and plaques, though the latter association did not remain in multivariable analysis. In a previous study VCAM was associated with arterial stiffness in patients with SSc. We reported recently in a prospective study that VCAM-1 levels predicted cardiovascular mortality in SLE. The present observation adds further support for a pivotal role of the endothelium in autoimmune vascular disease.
Plaques were more common in ACA+ SSc patients, compared both to ACA-patients and controls. This difference remained after adjustment for disease duration, gender and age.
It is of note that, despite the enhanced occurrence of ischemic arterial events and plaques, several traditional CVD risk factors (waist-hip ratio, plasma glucose and triglycerides) had a less 'atherogenic' profile in the ACA+ as compared to the ACA-patients. Another important observation is that ischemic events were rare in the ATA+ group. These patients are part of the ACA-group, which consists of patients with different autoantibody profiles. In previous studies, ACA+ patients had a different disease presentation with more localized skin involvement, calcinosis, digital loss and also a different genetic background. Cheng et al. previously noted that the carotid wall of patients with SSc was stiffer than that of controls, and in subgroup analysis stiffness was more pronounced in dcSSc than in lcSSc. Together these results imply that the antibody profiles and associated disease subsets contribute to vascular vulnerability in SSc.
To date this is the largest epidemiologically based study investigating both ischemic macrovascular disease and atherosclerosis in patients with SSc and in comparable population-based controls. Nevertheless, we are underpowered for more detailed subgroup analysis and for multivariable analysis of events. The cross-sectional design is a weakness restricting our investigation to survivors of vascular events. Case definitions in the national registries are based on ICD codes and differ from the definitions used in this study. We were thus restricted to compare event rates to our own controls, and could not reliably use the general population/national registries.
Discussion
The first major finding in this population-based study is that ischemic arterial events, specifically IHD and IPVD, are more common in patients with SSc, while ICVD and measures of atherosclerosis did not differ on a group level as compared to population controls. Second, the ACA+ SSc subgroup was more affected with ischemic arterial events and with carotid plaques in comparison both to other SSc patients and to controls.
The observed high prevalence of IPVD in patients with SSc highlights an important clinical problem recently also reported by Man et al.. Youssef et al. found an even higher prevalence of IPVD in females with lcSSc compared with controls (58% vs.10%). More claudication, as a sign of IPVD, was also reported in SSc patients by Veale and collaborators.
Despite the higher prevalence of IPVD, SSc patients did not, as a group, have lower ABI than controls. This is in line with three previous studies, but contrasts with Ho et al.. In our study, ACA+ patients had more IPVD compared to both ACA-patients and controls and there was also a trend toward lower ABI in the ACA+ group. Wan et al. examined ABI in 119 SSc patients and found no difference between patients with lcSSc and dcSSc, but interestingly ACA+ seemed to prevail among patients with symptomatic ischemic disease. Previous studies have identified ACA+ as a predictor for severe digital loss. Thus, in addition to microvascular disease, the large arteries of the extremities seem to contribute significantly to the severe distal ischemic damage seen in many SSc patients, especially in the ACA+ subgroup.
We report a three-fold increased occurrence of IHD in SSc patients and even higher figures in the ACA+ subgroup. This is in line with the recently reported incidence of myocardial infarction and self-reported IHD, both of which were two-fold enhanced in SSc patients versus the general population. Older autopsy studies described that myocardial infarctions were common in SSc patients, despite normal coronary arteries. Akram et al. evaluated 179 angiograms from SSc patients but did not find an increased prevalence of coronary atherosclerosis and normal angiograms were more common than expected among 11 SSc patients with myocardial infarction. These studies suggest that vasospasm rather than atherosclerosis is a major pathogenic mechanism behind SSc-related heart disease. On the other hand, MRI studies have found that calcification of cardiac vessels are common in SSc. The inconsistent results of previous work may partly be explained by our observation that atherosclerotic plaque occurrence is selectively enhanced in the ACA+ subgroup of SSc patients.
In our study, impaired kidney function was highly associated with IHD, a finding that is supported by Derk et al., who reported that the majority of SSc patients with IHD had renal failure.
Cerebrovascular disease was, in accordance with Youssef et al., equally common in SSc patients and controls. However, recently Man et al. reported enhanced incidence of stroke in a large epidemiological study.
On a group level, there was no difference between patients and controls concerning atherosclerosis, measured as plaques occurrence, IMT or ABI. Our study population was fairly old and the plaque frequency was high in both patients and controls. Previous studies are smaller and the majority measured only IMT. Some report, in agreement with us, similar IMT in patients and controls, others found thicker IMT in SSc patients. Only three studies examined the frequency of plaques, but their definitions of plaque differ. We had no exclusion criteria, while some studies excluded SSc patients with manifest cardiovascular disease (CVD). Furthermore, the selection of comparators is important. The healthy controls in previous studies are likely to have less atherosclerosis than our population-based controls. Methods, definition and selection procedures of both patients and controls are likely to contribute to the heterogeneity of previous studies, reviewed in a recent meta-analysis.
Generally, levels of endothelial and inflammatory biomarkers were higher in patients compared to controls, demonstrating that SSc is a low-grade inflammatory disease. Of investigated inflammatory biomarkers, IL-6 remained independently associated with plaques. For SSc this is a new observation but IL-6 has been identified as a risk factor for subclinical coronary atherosclerosis in RA, and it has also been linked to poor prognosis in patients with IHD from the general population.
VCAM-1 was associated with both IHD and plaques, though the latter association did not remain in multivariable analysis. In a previous study VCAM was associated with arterial stiffness in patients with SSc. We reported recently in a prospective study that VCAM-1 levels predicted cardiovascular mortality in SLE. The present observation adds further support for a pivotal role of the endothelium in autoimmune vascular disease.
Plaques were more common in ACA+ SSc patients, compared both to ACA-patients and controls. This difference remained after adjustment for disease duration, gender and age.
It is of note that, despite the enhanced occurrence of ischemic arterial events and plaques, several traditional CVD risk factors (waist-hip ratio, plasma glucose and triglycerides) had a less 'atherogenic' profile in the ACA+ as compared to the ACA-patients. Another important observation is that ischemic events were rare in the ATA+ group. These patients are part of the ACA-group, which consists of patients with different autoantibody profiles. In previous studies, ACA+ patients had a different disease presentation with more localized skin involvement, calcinosis, digital loss and also a different genetic background. Cheng et al. previously noted that the carotid wall of patients with SSc was stiffer than that of controls, and in subgroup analysis stiffness was more pronounced in dcSSc than in lcSSc. Together these results imply that the antibody profiles and associated disease subsets contribute to vascular vulnerability in SSc.
To date this is the largest epidemiologically based study investigating both ischemic macrovascular disease and atherosclerosis in patients with SSc and in comparable population-based controls. Nevertheless, we are underpowered for more detailed subgroup analysis and for multivariable analysis of events. The cross-sectional design is a weakness restricting our investigation to survivors of vascular events. Case definitions in the national registries are based on ICD codes and differ from the definitions used in this study. We were thus restricted to compare event rates to our own controls, and could not reliably use the general population/national registries.