Relationship between interleukin-6 and cardiac involvement in systemic sclerosis

Abstract

Objectives. To explore the relationship between IL-6 levels and echocardiographic abnormalities, and N-terminal probrain natriuretic peptide (NT-proBNP) levels in SSc patients and to correlate tested parameters with European Scleroderma Activity (EUSTAR) score.

Methods. This case–control study included 31 SSc patients with preserved left ventricular ejection fraction (LVEF) and 32 matched healthy controls. Serum IL-6 and NT-proBNP levels were measured and subjects were evaluated by conventional and pulsed-wave tissue Doppler echocardiography.

Results. The level of IL-6 was significantly increased in patients with SSc (3.2 vs 2.2 pg/ml, P < 0.001). SSc patients had significantly lower values of LV systolic (7.7 vs 9.25 cm/s, P < 0.001) and early diastolic (8.7 vs 10.3 cm/s, P = 0.014) myocardial velocities and higher E/e′ (9.04 vs 7.37, P = 0.001) ratio, although there was no between-group difference according to LVEF (68% vs 65%, P = 0.248). On evaluating the right ventricle there was no significant between-group difference in systolic tricuspid annular velocity (13 vs 13.9 cm/s, P = 0.105), but the peak early diastolic velocity was significantly lower (11.7 vs 13.6, P = 0.044) and E/e′ was significantly higher (4.3 vs 3.38, P = 0.008) in SSc patients. IL-6 level showed correlation with LV mean e′ (r = −0.57, P = 0.001), E/e′ (r = 0.55, P = 0.001) and NT-proBNP (r = 0.52, P = 0.003). EUSTAR score correlated with LV E/e′ (r = 0.48, P = 0.006), mean e′ (r = −0.67, P < 0.001), mean s′ (r = −0.51, P = 0.004), NT-proBNP (r = 0.60, P < 0.001) and IL-6 (r = 0.79, P < 0.001).

Conclusion. IL-6 level is increased in patients with SSc and significantly correlates with LV diastolic dysfunction, NT-proBNP and EUSTAR score. These results support the role of IL-6 in the development of cardiac disease in SSc patients.

systemic sclerosis, cardiac involvement, interleukin-6, tissue Doppler echocardiography

Introduction

Myocardial involvement, often clinically occult, is common in SSc and is recognized as a poor patient prognostic factor [1]. Although the pathogenesis of SSc remains unclear, a variety of cells contribute to the fibrosis process via mutual interactions and production of cytokines, including IL-6 [2]. Previous studies on SSc patients reported on elevated IL-6 serum levels, which correlated positively with the severity of skin lesions and were related to lung fibrosis and pulmonary hypertension [3, 4]. The role of IL-6 in the development of cardiac manifestations of SSc has not yet been explored.

The aims of our study were to determine myocardial dysfunction in SSc patients with preserved left ventricular ejection fraction (LVEF) by using conventional and pulsed-wave tissue Doppler echocardiography, assess correlation between IL-6 levels and echocardiographic abnormalities and N-terminal probrain natriuretic peptide (NT-proBNP) levels, and evaluate the relationship between all tested parameters and disease activity.

Patients and methods

This case–control study included 31 patients with SSc diagnosed according to the criteria of the ARA [5] who presented for follow-up care. The study was approved by the ethics committee of the Split University Hospital Center, Split, Croatia, and all participants gave their informed consent.

A complete history was taken from each patient, followed by physical examination and laboratory evaluation. The patients were not taking any disease-modifying drugs. Vasoactive drugs were stopped at least 5 half-lives before evaluation. The patients were selected after the exclusion of overlapping syndromes, diabetes mellitus, permanent atrial fibrillation, diagnosis of heart failure, LVEF <55%, known pulmonary arterial hypertension (PAH, confirmed by right heart catheterization), malignancy and active chronic or acute infection.

Disease activity (EUSTAR score) was determined according to the European Scleroderma Activity criteria [6].

The control group consisted of 32 healthy subjects matched for age, sex, BMI and smoking status. IL-6 was determined by high-sensitivity ELISA (Bender, Vienna) and NT-proBNP serum concentration by electrochemiluminescence immunoassay (Roche, Mannheim).

Echocardiographic examinations were performed by the same cardiologist blinded for clinical data, using an ultrasound system (Xario Ultrasound prime, Toshiba) equipped with a 2 MHz transducer. Conventional echocardiography was performed according to the American Society of Echocardiography (ASE) recommendations [7–9]. LVEF was estimated by Simpson's method. LV hypertrophy was defined as LV mass index >115 g/m² for men and >95 g/m² for women [7]. Right ventricular (RV) wall thickness >5 mm was the indication of RV hypertrophy [8]. The early diastolic/late diastolic (E/A) velocity ratio and E-wave deceleration time (DT) were measured from the mitral and tricuspid velocity profiles [8, 9]. Myocardial velocities were recorded using a standard pulsed-wave tissue Doppler imaging (PW TDI) technique [8, 9]. The sample volume was placed at the junction of the LV wall with the mitral annulus of the septal and lateral myocardial segments from the four-chamber view and the inferior and anterior myocardial segments from the two-chamber view. Peak systolic velocity (s′) and peak early diastolic velocity (e′) were measured. Mean s′ and mean e′ represented the average of the four measured sites. The E/e′ ratio (Mit E/e′) was calculated by dividing the peak E velocity by the mean e′ velocity. PW TDI, used to measure tricuspid annular velocities and the E/e′ ratio (Tr E/e′), was obtained from the junction of the RV free wall with the tricuspid annulus.

Septal s′ was a marker of LV contractility and a value of septal s′ <7.5 cm/s was defined as abnormal [10]. RV systolic dysfunction (SD) was defined by Tr s′ <10 cm/s [8]. Diastolic dysfunction (DD) of both ventricles was defined and classified according to the ASE recommendations [8, 9]. Intraobserver variability was assessed in 10 selected subjects at random from the SSc study group by repeating the measurements under the same conditions.

Statistical analysis

Continuous variables were expressed as the mean (±s.d.) or median (Q1–Q3). Categorical variables were expressed as number and/or percentage. Patients with SSc were compared with healthy controls by Mann–Whitney test and by χ² test. Spearman’s correlation tests were performed to examine correlations between variables. P < 0.05 was considered statistically significant. All analyses were performed using SPSS 15.0 for Windows.

Results

Although the values of echocardiographic parameters were within normal ranges in most SSc patients, they significantly differed from the respective values in the control group (Table 1). Five of 31 SSc patients had left ventricle hypertrophy vs two of 32 control subjects (P = 0.397). Seven of 31 SSc patients had RV hypertrophy vs none of the control subjects (P = 0.014). Although all subjects had an LVEF ≥55%, 18 SSc patients had septal s′ <7.5 cm/s vs 5 control subjects (P = 0.010). RVSD was observed in six SSc patients and none of the control subjects (P = 0.03). Of Doppler parameters evaluating LV and RV diastolic filling, the E/A was similar between the groups, whereas significant between-group differences were found in myocardial diastolic velocities and filling pressure, estimated from E/e′ (Table 1).

Table 1

General characteristics, conventional echocardiography and pulsed-wave tissue Doppler parameters in SSc and control groups

Parameter	SSc patients (n = 31)	Controls (n = 32)	Pa
General characteristic
Age (years)	64 (49–71)	63.5 (49–68.5)	0.831
Gender (F/M)	30/1	31/1	—
BMI (kg/m2)	24.7 (22–27.5)	25.9 (24–28.3)	0.522
Systolic blood pressure (mmHg)	130 (120–150)	137 (124.5–153)	0.531
Diastolic blood pressure (mmHg)	78 (70–85)	80 (73.5–86.5)	0.367
Heart rate (beats/min)	77 (70–82)	72 (78–81)	0.995
NT-proBNP (pmol/l)	11.6 (5.23–30.6)	9.59 (4.29–16.7)	0.132
IL-6 (pg/ml)	3.2 (2.4–6.2)	2.2 (1.7–2.8)	<0.001
SSc type (diffuse/limited)	28/3	—	—
Disease duration	12 ± 9.9	—	—
EUSTAR score	4.48 ± 2.12	—	—
Pulmonary fibrosis, CT scan, n (%)	11 (35.4)	—	—
Conventional echocardiography
LVDd (mm)	50 (46–52)	46.8 (44–50)	0.033
IVSd (mm)	9 (8.4–10.4)	9 (7.6–10.2)	0.563
LVPWd (mm)	9.3 (8–10.5)	8.5 (7.5–9.5)	0.110
LVEF (%)	68 (64–73)	65 (60.5–71.5)	0.248
LA (mm)	38 (35–43)	36.5 (32–38)	0.025
RV (mm)	24.3 (22–26)	21.75 (21–23)	0.001
RVWd (mm)	3.5 (3–4.2)	3 (2.8–3.5)	0.030
RA (mm)	34 (30–40)	29.5 (27–32)	<0.001
E/A MV	1 (0.81–1.2)	1.11 (0.88–1.42)	0.312
DT MV(ms)	180 (160–213)	180 (150–210)	0.725
E/A TV	1.28 (1.06–1.44)	1.2 (1.13–1.38)	0.815
DT TV (ms)	213 (180–227)	182.5 (163.5–208.5)	0.016
PASP (mmHg)	36.5 (31–44.5)	26 (22–29)	0.003
Pulsed-wave tissue Doppler
Septal e′ (cm/s)	7.2 (6–9.4)	8.55 (7.25–10.4)	0.035
Lateral e′ cm/s)	9.3 (8.2–14)	11.9 (9.75–14.9)	0.013
Mean e′ (cm/s)	8.7 (7.15–12)	10.3 (8.54–12.7)	0.014
Mit E/e′	9.04 (7.2–11.6)	7.37 (6.2–7.99)	0.001
Septal s′ (cm/s)	7.2 (6.6–7.5)	8.1 (7.8–9.5)	<0.001
Mean s′ (cm/s)	7.7 (6.67–7.5)	9.25 (8.11–10.5)	<0.001
Tr e′ (cm/s)	11.7 (9.7–14.6)	13.7 (12.3–15)	0.044
Tr E/e′	4.3 (3.25–5.15)	3.38 (2.93–3.87)	0.008
Tr s′ (cm/s)	13 (11.3–15)	13.9 (12.5–15.2)	0.105

Values are expressed as median (Q1–Q3) or mean (±s.d.). ^aMann–Whitney test. LVDd: left ventricular end-diastolic diameter; IVSd: interventricular septum thickness; LVPWd: posterior wall thickness; LVEF: left ventricular ejection fraction; RVDd: right ventricular end-diastolic diameter; RVWd: right ventricular wall thickness; LA: left atrial diameter; RA: right atrial diameter; E/A MV: transmitral Doppler ratio; DT MV: mitral deceleration time; E/A TV: transtricuspid Doppler ratio; DT TV: tricuspidal deceleration time; PASP: pulmonary arterial systolic pressure; septal e′: septal annulus early diastolic velocity; lateral e′: lateral annulus early diastolic velocity; mean e′: mean of the four measured sites of early diastolic mitral annular velocities; Mit E/e′: ratio of peak E velocity and mean e′; septal s′: septal annulus systolic velocity; mean s′: mean of the four measured sites of mitral annulus systolic velocities; Tr e′: early diastolic tricuspid annular velocity; Tr E/e′ ratio: ratio of peak E velocity and Tr e′; Tr s′: systolic tricuspid annular velocity.

According to the ASE recommendations, 18 SSc patients and 9 controls had LVDD (P = 0.032). Impaired LV relaxation was found in 4 SSc patients, a pseudonormal pattern in 12 and a restrictive pattern in 2 of 31 SSc patients. Impaired LV relaxation was recorded in nine control subjects.

Five SSc patients and none of the control subjects had RVDD (P = 0.024). Two SSc patients had abnormal RV relaxation, and another three had a pseudonormal pattern of RVDD. Pulmonary arterial systolic pressure (PASP) was significantly higher in SSc patients (Table 1), but there was no correlation with RV function parameters. NT-proBNP levels were correlated with PW TDI parameters of LV systolic and diastolic function in SSc patients (Table 2).

Table 2

Correlation coefficients between features of SSc patients and echocardiographic parameters of LV systolic/diastolic function

Parameter	Septal e′ (cm/s)	Lateral e′ (cm/s)	Mean e′ (cm/s)	Mit E/e′	Septal s′ (cm/s)	Mean s′ (cm/s)
Age (years)	−0.61**	−0.83**	−0.86**	0.65**	−0.61**	−0.60**
Disease duration (years)	−0.35	−0.31	−0.36	0.32	−0.39*	−0.37*
EUSTAR score	−0.66**	−0.57**	−0.67**	0.48**	−0.56**	−0.51**
IL-6 (pg/ml)	−0.61**	−0.40**	−0.57**	0.55**	−0.31	−0.21
NT-proBNP (pmol/l)	−0.56**	−0.53**	−0.51**	0.59**	−0.47**	−0.42**

*P < 0.05, **P < 0.01; Spearman’s correlation test. EUSTAR score: disease activity score; septal e′: septal annulus early diastolic velocity; lateral e′: lateral annulus early diastolic velocity; mean e′: mean of the four measured sites of early diastolic mitral annular velocities; Mit E/e′: ratio of peak E velocity and mean e′; septal s′: septal annulus systolic velocity; mean s′: mean of the four measured sites of mitral annulus systolic velocities.

SSc patients showed significantly increased serum IL-6 levels in comparison with those in healthy individuals (Table 1). IL-6 showed correlation with PW TDI parameters of LV diastolic function in SSc patients (Table 2). In control subjects there was no correlation between IL-6 and any of the echocardiographic parameters.

IL-6 level significantly correlated with the presence (r = 0.46, P = 0.010) and severity (r = 0.54, P = 0.002) of LVDD and NT-proBNP level (r = 0.52, P = 0.003) in the SSc group, whereas no correlation was observed in the control group. Correlation was also found between IL-6 level and EUSTAR score (r = 0.79, P < 0.001) and disease duration (r = 0.36, P = 0.044).

EUSTAR score correlated with LV diastolic and systolic function parameters (Table 2) and NT-proBNP (r = 0.60, P < 0.001). The correlation between EUSTAR score and LVSD presence (r = 0.363, P = 0.044) and LVDD presence (r = 0.58, P = 0.001) and severity (r = 0.621, P < 0.001) was observed. Intraobserver variability was < 5% for all echocardiographic measurements.

Discussion

Our results suggest that myocardial dysfunction in SSc patients is common despite normal echocardiographic LVEF and that IL-6 serum levels correlate with the parameters of LVDD and NT-proBNP level. We also found a correlation between disease activity (EUSTAR score) with LV dysfunction and IL-6 levels.

IL-6, a pleiotropic Th2 immune response cytokine, has been implicated as the possible target in the treatment of SSc. Myocardial fibrosis is the main histopathologic feature of SSc cardiomyopathy, which differs from atherosclerotic fibrosis. The fibrosis is considered to result from recurrent ischaemia due to microvascular lesions [1, 11]. It has been postulated that hypoxia influences IL-6 transcription in SSc [2], whereas recent in vitro studies point to the role of IL-6 in the activation and apoptosis of endothelial cells [12].

In our study, IL-6 serum levels were significantly higher in SSc patients than in healthy controls, showing strong positive correlation with the EUSTAR score and moderately positive correlation with disease duration. Among the echocardiography parameters, those pointing to LVDD showed the highest correlation with IL-6.

The prevalence of cardiac involvement in SSc varies according to the definition and sensitivity of the diagnostic method. Previous studies using conventional echocardiography have reported a low prevalence of depressed LVEF and LVDD [13, 14]. In contrast, Meune et al. [15], using PW TDI in their controlled study, found a higher prevalence of LVSD, RVSD and LVDD.

PW TDI is a more sensitive, less load dependent and currently widely available echocardiographic method to measure peak velocities of the mitral and tricuspid rings and longitudinal myocardial fibre [9]. Since subendocardial fibres, being most susceptible to microvascular ischaemia, are primarily longitudinally oriented, longitudinal systolic and, particularly, diastolic dysfunction is generally the first sign of a myocardial lesion in SSc [16].

Although we excluded patients with LVEF ≤55% because of its potential influence on IL-6 levels [17], based on PW TDI findings from our study, both LVSD and LVDD were significantly more frequent in SSc patients than in control subjects. RVSD and RVDD were also found in some of the SSc patients as opposed to none of the controls. The prevalence of myocardial dysfunction in our study was higher than that reported by Meune et al. [15], which may be partly explained by the higher mean age, potentially longer duration and higher disease activity in our study population. Like Meune et al. [15], we found no correlation between RV function parameters and PASP and pulmonary involvement, a finding that supports the hypothesis of intrinsic myocardial involvement. Also, a positive correlation between NT-proBNP, an indicator of overall cardiac dysfunction [18], and IL-6 levels in SSc patients was observed, suggesting that IL-6 might be an additional cardiac biomarker in SSc.

In our study the prevalence and severity of LVDD positively correlated with serum IL-6 levels. Consistently with previous descriptions of SSc cardiomyopathy characteristics [19], we found a significant difference in the presence and severity of LVDD between SSc patients and healthy subjects, along with a non-significant difference in the prevalence of LV hypertrophy. The impact of age, blood pressure and heart rate was obviated by selecting a closely matched control group. The significant difference in LVDD values could be partly explained by a difference in PASP values, i.e. the possible presence of PAH in SSc patients. We did not analyse in detail the causes of elevated PASP values in SSc patients by differentiating precapillary and post-capillary pulmonary hypertension. However, while the decreased values of septal e′ might be attributed to the concurrence of PAH, the decreased values of lateral e′ and mean e′ and elevated values of Mit E/e′ in our SSc patients were suggestive of primary myocardial disease [9]. These very parameters of myocardial dysfunction yielded a correlation with IL-6 level, while no correlation was recorded in healthy subjects. As myocardial abnormalities first reflect through diastolic dysfunction, the results of our study support the potential role of IL-6 in the development of these lesions in SSc patients.

In our study the prevalence of LVSD and LVDD and the severity of LVDD showed a positive correlation with EUSTAR score. Data obtained from the EUSTAR database also indicate an association between LVSD and EUSTAR score [13]. Rosato et al. [16] found no correlation between the presence of inverted e′/a′ (index of LVDD) and disease score. In their study, longitudinal and radial myocardial velocities of basal and mid regions demonstrated that the myocardial damage first involved subendocardial fibres and the basal segments of LV. For this reason we preferred to use only longitudinal myocardial velocities to define LV systolic and diastolic function and found their negative correlation with EUSTAR score. This finding indicates a need for more precise cardiologic evaluation of SSc patients with increased disease activity.

The observed correlation between IL-6 and EUSTAR score suggests that IL-6 might contribute to the active stage of the disease, as postulated previously by Bosello et al. [20].

The limitations of our study include a small sample of patients to evaluate the predictive power of IL-6. Also, our findings were observational in nature. Furthermore, we relied on patient and earlier medical documentation of medication history, concomitant disease and disease duration. We also focused on PW TDI assessment, because of its availability, although MRI is a more sensitive method of detecting cardiac manifestations of SSc.

In conclusion, we observed a high prevalence of myocardial dysfunction in SSc patients with preserved LVEF and found that IL-6 serum levels correlated with the presence and severity of LVDD in SSc patients. A positive correlation between IL-6 and EUSTAR score and their association with echocardiographic abnormalities and NT-proBNP may open new possibilities for the treatment of SSc cardiomyopathy.

Rheumatology key messages

• IL-6 is correlated with the presence and severity of LV diastolic dysfunction in SSc patients.

• IL-6 is correlated with NT-proBNP in SSc patients.

• IL-6 is associated with disease activity of SSc assessed by EUSTAR score.

Disclosure statement: The authors have declared no conflicts of interest.