Ischemic arterial events and atherosclerosis in patients with systemic sclerosis: a population-based case-control study

All participants were >18 years old and recruited from the adult population in Stockholm County (N = 1,534,272) between August 2006 and December 2009. Prevalent cases with SSc were identified at the three Rheumatology clinics associated with general/university hospitals and from all rheumatologists with private practice. All dermatologists, gastroenterologists, specialists in respiratory medicine, hand surgeons and general practitioners were contacted twice and asked to inform us about their patients with SSc. We identified 149 cases fulfilling the American College of Rheumatology (ACR) criteria for SSc [9], which corresponds to a prevalence of 97 adult SSc cases/million. A total of 74% participated in the study. According to our review of medical records, the remaining 26% did not differ from the included patients regarding age, disease duration or severe organ manifestations. Control subjects were recruited from the same population through use of the national registration number, which includes date of birth and is coded for sex. The sex-matched person with the birth date closest to the patient was contacted and asked to participate. Some 65% of these 'first choice' controls accepted to participate. When the 'first choice' declined, the second-closest person was asked, until a control subject gave his/her consent. A diagnosis of SSc was the only exclusion criteria. A total of 105 controls were included.

The local ethics committee of the Karolinska University Hospital approved the study and all participants gave written informed consent.

All participants underwent a structured interview and a physical examination by a rheumatologist (AN, LB).

The patients were classified as lcSSc or dcSSc using the LeRoy classification [10]. Skin score was measured with the modified Rodnan skin score [11] and disease activity with the preliminary European Scleroderma Study Group (EScSG) activity index [12]. The Medsger scale of disease damage was used, with the exception that the joint/tendon and muscular item was not performed [13]. Organ involvement was defined as: suspected pulmonary hypertension (sPH): tricuspidalis V-max >2.9m/s on Doppler ultrasound. Pulmonary fibrosis: signs of fibrosis on X-ray or high-resolution computed tomography (HRCT). HRCT was performed in 90% of the patients. Myositis: muscular weakness and elevated creatine kinase and signs of inflammation on magnetic resonance imaging (MRI), electromyography (EMG) or muscular biopsy. Arthritis: swollen joints on examination or arthritis documented in the journals. Kidney involvement: a history of scleroderma renal crisis or >1 on the Medsger scale for kidney damage [13].

Ischemic arterial events were defined as:

1. Ischemic heart disease (IHD): myocardial infarction (confirmed by electrocardiography and a reversible rise in plasma creatine kinase, muscle and brain fraction (CK-MB) or troponin T) or angina pectoris (confirmed by an exercise stress test).

2. Ischemic cerebrovascular disease (ICVD): cerebral infarction (confirmed by computed tomography) or transitory ischemic attacks (TIA, defined as transient focal symptoms from the brain or retina with a maximum duration of 24 hours).

3. Ischemic peripheral vascular disease (IPVD): intermittent claudication + ankle-brachial index (ABI) <0.9 or peripheral arterial thrombosis/embolus (confirmed by angiogram or Doppler flow studies).

4. Any ischemic arterial event that includes 1 to 3 above

The left and right common carotid arteries and bifurcation areas were scanned for presence of plaque and images for IMT measurements were received using a duplex scanner (Siemens Acuson Sequoia, Mountain View, CA, USA) with a 7.0 MHz linear array transducer. Scans were digitalized for offline analysis. The subject's head was tilted to get the picture of the common carotid artery (CCA) just proximal to the bulb placed horizontally across the screen. Pictures were frozen synchronously with the R wave on the electrocardiogram. The IMT was defined as the distance between the leading edges of the luminal echo and the media/adventitia echo [14]. The IMT was calculated as the intima-media area divided by the measured length (10mm) on one scan. Plaques were defined as a local increase in wall thickness of >1 mm and 100% increase in wall thickness compared to the adjacent wall. A technician recorded all measurements and a single experienced reader (KJU) interpreted the registrations without knowledge of patient/control status or other test results.

To obtain the ABI, the patients were placed in a supine position. The highest systolic pressure of the tibialis posterior or dorsalis pedis for each foot were divided by the highest brachial systolic pressure to obtain an ankle-brachial pressure ratio.

Antinuclear antibodies (ANA) were analyzed by immunofluorescence (IFL) on sections of rat liver and HEp-2 cells (Immunoconcepts, Sacramento, CA, USA). Anticentromere antibodies (ACA) and antitopoisomerase 1 antibodies (ATA) were analyzed by the BioPlex 2200 ANA screen system (Bio-Rad, Hercules, CA, USA). High-sensitivity C-reactive protein (hsCRP), α-1 antitrypsin, orosomucoid and fibrinogen were measured using BN ProSpec System (Dade Behring, Deerfield, IL, USA). Intercellular adhesion molecule 1(ICAM-1), vascular cell adhesion molecule 1(VCAM-1), interleukin 6 (IL-6) and vascular endothelial growth factor (VEGF) were analyzed by sandwich ELISA kits (DY720, DY206, DY137 and DY293B, R & D Systems, Minneapolis, MN, USA). Cutoff values were 15 pg/mL for ICAM-1, VCAM-1 and VEGF and 6 pg/mL for IL-6.

Cystatin C (reagent: 1014, Gentian, Moss, Norway) was analyzed on an Architect Ci8200^™ analyzer (Abbott, Abbot Park, IL, USA). The total analytical imprecision of the cystatin C method was 1.1% at 1.25 mg/L and 1.4% at 5.45 mg/L. The equation used for calculating glomerular filtration rate (GFR) in mL/min/1.73 m² from the cystatin C results in mg/mL was y = 79.901x^-1.4389[15].

von Willebrand factor (vWF) was measured by sandwich ELISA (Dako, Glostrup, Denmark) [16].

Characteristics were summarized for all patients and controls. Skewed continuous variables were log transformed to obtain a normal distribution, if possible. Groups were compared using ANOVA, Mann-Whitney U test and X² test as needed. Odds ratio (OR) and 95% confidence intervals (CI) were calculated from 2 × 2 contingency tables or from nominal logistic regression models. For continuous variables, we used standard least squares linear regression to calculate standardized regression coefficients (β) and P values. Variables were sorted into functional groups (traditional risk factors, and so on. see Table 1). After age adjustment, the variables within each group, which were representative and most significantly, as determined by lowest P value, associated with plaque occurrence were entered into a multivariable-adjusted model. Due to limited number of plaques, we restricted the number of variables to six. Calculations were done using JMP software (SAS Institute, Cary, NC, USA). A P value < 0.05 was considered statistically significant.

At inclusion patients with SSc had lower body mass index, lower diastolic blood pressure, but higher triglycerides and biomarkers of systemic inflammation/endothelial activation as compared to controls. Baseline characteristics and comparisons between patients and controls are given in Table 1.

Ischemic arterial events were more common in patients than in controls (Table 2), due to more prevalent IHD and IPVD, while ICVD did not differ between groups (events are tabulated in detail, see Additional file 1,Table 1). Of participants with IHD, seven patients and three controls were diagnosed with myocardial infarction. The remaining IHD subjects had angina pectoris verified with an exercise test, of these two patients and one control had undergone an angiogram, all with normal cardiac vessels. Five patients and one control had had multiple events.

The majority of the IPVD patients (78%) in our study had severe macrovasculopathy requiring surgical intervention. One SSc patient and the only control had IPVD based on ABI <0.9 together with typical symptoms of claudication (see Additional file 1, Table 1).

Plaque occurrence, IMT and ABI did not differ between patients and controls (Table 2). Six patients and two controls had an ABI below 0.9 and one patient and one control had an ABI higher than 1.4.

As expected, all ischemic arterial events were positively associated with age. We therefore adjusted all further calculations for age. All variables presented in Table 1 were investigated for associations with ischemic arterial events. All variables, which after age adjustment remained associated with at least one of the respective cardiovascular outcomes (any event, IHD or IPVD), are presented in Table 3. Age, kidney involvement, cystatin C estimated (e)GFR were associated with any ischemic arterial event and with IHD. ACA+ and skin score were associated with any ischemic arterial event and IPVD. Interestingly, patients with ATA had less total events and less IHD. IHD was also associated with endothelial biomarkers. Due to few events, we were not able to perform further multivariable calculations.

The investigated measures of atherosclerosis were all positively associated with age. Thus, all further calculations were age-adjusted. Each variable presented in Table 1 was investigated for association with plaque occurrence, IMT and ABI. All variables, which after age adjustment remained associated with at least one of the respective atherosclerosis outcomes (plaque, IMT or ABI), are presented in Table 4.

Plaque occurrence was associated with smoking, higher blood pressure, impaired kidney function, high levels of IL-6 and VCAM-1 and with ACA+.

As expected, we noted that IMT was strongly associated with systolic blood pressure and with male gender. ABI was associated with disease activity, waist-hip ratio and smoking.

Several variables were associated with plaque occurrence and we included age and the variables with the lowest P values in each 'functional group' in a multivariable-adjusted logistic regression. In this model only ACA+ and high levels of IL-6 remained convincingly associated with plaque occurrence, while the association with diastolic blood pressure was of borderline significance (Table 5).

Finally, due to the observed association between ACA+ and any ischemic arterial event and plaque occurrence, we stratified by ACA status and compared the groups with respect to all variables in Table 1. There were 34 ACA+ patients and 77 ACA-patients. In the ACA+ group disease duration was longer (P = 0.04) and dcSSc, pulmonary fibrosis, and ATA antibodies were less common (P < 0.05) compared with the ACA-group. Of traditional risk factors, waist-hip ratio, P-glucose and triglycerides levels were lower in ACA+ patients (P ≤ 0.05). Levels of orosomucoid and ICAM-1 were also lower in the ACA+ group (P ≤ 0.05) (see Additional file 1, Table 2).

Occurrence of vascular events, measures of atherosclerosis and levels of endothelial markers in ACA+ and ACA-patients (adjusted for age, sex and disease duration), and in ACA+ patients and controls (adjusted for age and sex) are demonstrated in Table 6, Figure 1, Figure 2 and Figure 3.

A higher total of ischemic arterial events, IPVD, and more plaques were observed in ACA+ patients as compared to both ACA-patients and controls.