QT variability index in patients with systemic sclerosis

Abstract

Objective

Systemic sclerosis (SSc) may affect the heart with microvascular dysfunction and lead to an early cardiac death, but the association between certain repolarization indexes and SSc heart disease remains controversial. Our goal was to evaluate a previously unstudied marker of repolarization dynamics, i.e., QT variability, in patients with SSc and to assess its prognostic implications.

Methods

A total of 17 patients with SSc and 21 healthy controls were included into this prospective study. Electrocardiograms were conducted under strict standards. The QT variability index (QTVI), normalized QT variability (QTVN), and power spectral analysis of QT dynamics, considered as markers of ventricular arrhythmias in a number of other disorders, were determined using designated computer software.

Results

There was no significant difference in demographic and cardiac important clinical parameters between the groups. Also, the mean QTVI, QTVN, and power spectral analysis parameters were comparable between the patients with SSc and control subjects. At baseline, the QTVI values of 1 patient with SSc, who experienced ventricular arrhythmia prior to inclusion in the study, were considerably higher compared to other patients with SSc. None of the remaining patients with SSc or the control subjects developed arrhythmia during the follow-up of 8 years.

Conclusion

Higher than normal QTVI may be found in the minority of patients with SSc. The prognostic significance of this finding is unknown, but it may entail an increased risk of ventricular arrhythmias. Therefore, the value of QTVI as a tool for arrhythmia risk stratification in SSc merits further research.

Introduction

Systemic sclerosis (SSc) is a multisystem, autoimmune connective tissue disease of unknown etiology, characterized by skin fibrosis, microvascular abnormalities, and inflammation. Both subsets of SSc, namely, diffuse cutaneous SSc (dcSSc) and limited cutaneous SSc (lcSSc) are associated with internal organs involvement (1). Severe organ involvement, particularly SSc-related cardiopulmonary disease, is associated with poor prognosis and death. Based on autopsy studies demonstrating a “patchy” distribution of cardiac fibrosis, inflammation, and contraction band necrosis (CBN), the heart may be affected in approximately 80% of patients (2, 3).

Cardiac fibrosis and restrictive cardiomyopathy in SSc are caused by coronary microcirculation impairment (4), and also by intermittent spasm of the coronary arteries (i.e. intramyocardial Raynaud phenomenon) (2, 5). SSc-associated chronic pulmonary hypertension may lead to remodeling (6), thereby further compromising cardiac function. SSc-related macrovascular abnormalities may also be manifested by endothelial dysfunction and vascular atherosclerosis, as suggested by some studies evaluating the intima media thickness, flow-mediated vasodilation, nitroglycerin-mediated dilation, and ankle brachial pressure index (7). In addition, Kahan et al. (8), and later other authors as well, reported that patients with dcSSc suffering from symptomatic cardiac disease may also have increased cardiac vulnerablity to ischemia, as indicated by a significantly reduced coronary flow reserve (CFR) (8–13).

According to the EULAR Scleroderma Trials and Research Group (EUSTAR) large database, arrhythmias lead to 6% of all-cause mortality in SSc (14). Several studies have suggested that a prolonged QTc interval and QT dispersion may predispose to arrhythmia and thereby to sudden death in some SSc patients (15–17). Another parameter of QT dynamics (QT/RR slopes) was found to be significantly steeper in dcSSc compared to controls (18). Yet, other studies argue that increased QT dispersion is not a feature of SSc (19) and that a prolonged QTc, although found in some SSc patients, is not associated with major arrhythmic complications (20). This controversy further supports the need for other arrhythmic markers in SSc. The QT variability index, reflecting repolarization variance, was found to be a powerful marker of ventricular arrhythmias and sudden death in some other diseases (21), but to the best of our knowledge, it has never been studied in SSc. Therefore, here we endeavor to evaluate the possible prognostic role of QTVI in SSc as well.

Methods

Study design

Using a prospective study design, we attempted to determine whether abnormal QTVI may predispose to ventricular arrhythmia. QTVI parameters were determined at baseline, and the possible or actual occurrence of arrhythmias during the study period was considered, at each regular follow-up visit, based on clinical parameters and medical reports of acute episodes experienced by the patients and control subjects. The research protocol was approved by the institutional (Sheba Medical Center) review board. All participants signed informed consent.

Study subjects

The study cohort consisted of 17 patients with SSc recruited from the outpatient clinic. All patients with SSc were seen regularly every 12 months and performed at least once a year follow-up tests including ECG, tissue Doppler echocardiogram, pulmonary function tests, including diffusion capacity, routine laboratory blood tests, and the SSc autoantibody profile. More studies were performed if needed. The diagnosis of SSc was based on the ARA classification criteria for SSc, which were the standard classification criteria at subject recruitment (22). Patients were recruited on December 2010 and have been followed since then. Twenty-one unaffected control subjects who were followed at the Executive Health Screening Program clinic of the hospital (a preventive program for early detection and treatment of health hazards) and who were found not to be affected by cardiopulmonary disease, SSc, or another autoimmune or inflammatory disorder (based on medical history, physical examination, cardiac stress test, resting ECG, chest X-ray, complete blood count, and blood chemistry), were randomly recruited in 2010 and served as a reference group for normal ECG parameters.

Procedure

Patients and control subjects were requested to refrain from caffeine and drugs on the morning of the study. Participants were instructed to lie motionless in the supine position for 10 minutes. Electrodes were placed in anatomical positions, according to the standard ECG procedure. Five-minute ECG strips were recorded with a standard device at a sampling rate of 2000 Hz. ECGs with technical errors or inadequate quality were repeated. Moreover, since abnormal depolarization might be associated with abnormal repolarization, ECGs with bundle block and altered depolarization were excluded. QTVI variables were calculated using the commercial software (Norav Medical, version 5.514, Yokne’am, Israel). The RR interval was measured between two consecutive beats. The QT interval was measured according to accepted standards (23). The mean QT interval (QTm) and QT variance (QTv), mean heart rate (HR) interval (HRm), and HR variability (HRv) were computed from the respective time series (24). The mean corrected QT (QTc) was calculated for all patients using Bazett’s Formula. The calculated time domain parameters include the root mean square of the successive differences of QT intervals (RMSSD-QT) and QT triangular index (i.e. the integral of the density of the QT interval histogram divided by its height).

The power spectral analysis was conducted using the nonparametric fast Fourier transform. Calculations were made in absolute values of power (ms²). The spectral components were categorized as very low frequency (VLF, 0.003–0.04 Hz), low frequency (LF, 0.04–0.15 Hz), and high frequency (HF, 0.15–0.4 Hz). The total power was computed as well.

QTVI was computed using a log ratio adjusted to heart rate, according to the following equation:

$${QTVI}_{(HR)}={log}_{10}\left[\frac{{QT}_v/{{QT}_m}^2}{{HR}_v/{{HR}_m}^2}\right]$$

QTVN is not adjusted either to the HR variability or to RR interval changes. Therefore, we calculated QTVN for the mean QT (21):

$$QTNV=\frac{{QT}_v}{{{QT}_m}^2}$$

Determination of arrhythmia

The occurrence of arrhythmias over the study period was determined every 6–12 months during regular clinic visits. To be accepted, arrhythmia had to be noted in a medical report, or the ECG record was viewed by the treating physician. For the purpose of the study, an unexplained sudden death or fainting were considered to be related to arrhythmia.

Statistical analysis

Data were analyzed using the JMP version 7.0 (SAS Institute, Cary, NC, USA). Results are presented as mean and standard deviations. Abnormal results were defined as more than two standard deviations from the normal range. Findings were compared between the groups using the Kruskal-Wallis one-way analysis test and the Fisher’s exact test. A p-value <0.05 was considered statistically significant.

Results

The demographic and clinical parameters of the 17 patients with SSc and 21 healthy controls included in the study are presented in detail in Table 1. The mean disease duration of the SSc patients group at inclusion was 8.2±8.8 years. There were no differences between the groups regarding the age, male-to-female ratio, body mass index, and rates of cardiovascular risk factors (i.e. smoking, diabetes mellitus, dyslipidemia, hypertension, or a family history of heart disease). Also, there were no differences in the medications used, except for calcium-channel blockers and immunosuppressive drugs, which were consumed more commonly by patients from the SSc group. Although more patients with SSc were treated with antidepressants, the difference between groups did not reach statistical significance. Two patients with SSc were subtyped with lcSSc, and the remaining 15 were subtyped with dcSSc. One patient was diagnosed with mild pulmonary hypertension, and 52.9% of patients with SSc had pulmonary involvement, including abnormal diffusion capacity, and/or restrictive lung disease.

Table 1.

Patients and control subjects characteristics at baseline.

Parameter	SSc	Control subjects
Age (years)	44.6±11.8	38.4±13.8
Gender (F/M)	16/1	18/3
BMI (kg/m2)	22.5±5.1	22.4±2.1
Overt cardiac disease (%)^	5.9	0
Active smoking (%)	29.4	19.1
Former smokers (%)	11.8	0
Family history of IHD (%)	64.8	42.9
Diabetes mellitus (%)	5.9	0
Hypertension (%)	5.9	0
Dyslipidemia (%)	17.6	9.5
Asthma (%)	0	4.8
Aspirin intake (%)	11.7	0
ACEI intake (%)	0	0
ARB intake (%)	5.9	0
CCB intake (%)*	41.2	0
Beta blockers intake (%)	11.8	0
Statins intake (%)	5.9	4.8
Fibrates intake (%)	5.9	0
Insulin intake (%)	5.9	0
Other anti-diabetic drugs (%)	5.9	0
Anti-depressants intake (%)	23.5	4.8
Levothyroxine intake (%)	4.8	0
Immunosuppressive therapy (%)*	47.1	0
Prostacyclin therapy (%)	11.8	0

BMI: body mass index; IHD: ischemic heart disease; ACEI: angiotensin-converting-enzyme inhibitors; ARB: angiotensin II receptor blockers; CCB: calcium channel blockers; Immunosuppressive therapy refers to intake of either methotrexate, prednisone or immunoglobulins.

p>0.05 for all parameters except for those marked with ^*which were with p<0.01.

^{^}Overt cardiac disease is defined as congestive heart failure, ischemic heart disease and arrhythmia.

The QT variability indexes are displayed in Table 2. There were no differences in average QTc, average RR, RMSSD-QT, and QT triangular index between the groups. Also, QTVI (Figure 1), QTVN, and power spectral parameters were similar in SSc and the control groups. Over a follow-up duration, there was 1 patient with SSc who died of malnutrition. Five other SSc were lost for follow-up. Also, 1 patient with SSc developed pulmonary hypertension (mean of 59 mmHg). One patient with SSc (that was 58 years old at inclusion) has had recurrent episodes of AV nodal reentry tachycardias, and an additional event of wide complex nonsustained ventricular tachycardia 1 year prior to her inclusion. The patient was continuously treated with Flecainide (3M pharmaceuticals, Minnesota, USA) and had no arrhythmia recurrence. Interestingly, at recruitment, this patient had a considerably higher QTVI (−0.94) than all other patients with SSc (−1.36±0.13, Figure 1). None of the remaining patients with SSc developed arrhythmia prior to inclusion and during the 8-year follow-up.

Table 2.

QT variability results in SSc patients and healthy controls at baseline.

	SSc	Control
Average QTc (ms)	428.9±26.4	421.5±23.5
Average RR (ms)	862.8±123.9	902.1±120.9
RMSSD-QT	10.5±6.1	8.5±3.7
QT triangular index	2.8±0.7	2.6±1.0
QTVI(HR)	−1.32±0.16	−1.31±0.13
QTVN	0.33±0.08	0.33±0.09
QT VLF (ms2)	82.7±42.8	93.6±42.6
QT LF (ms2)	104.1±21.8	104.1±22.4
QT HF (ms2)	223.3±87.5	203.2±50.5
QT total power (ms2)	454.3±77.6	446.4±97.2

RMSSD: Root mean square of the successive differences; QTVI: QT variability index; QTVN: normalized QT variability; VLF: very low frequency; LF: low frequency; HF: high frequency.

p was insignificant in all parameters computed or measured.

Figure 1.

QTVI in SSc patients and healthy controls

The black dots represent raw values. The horizontal lines of the box represent the median and the 25 to 75 percentiles of the QTVI value distribution. The arrow indicates the patient who experienced ventricular arrhythmia.

Discussion

Overt SSc-related cardiac disease is uncommon, usually apparent in only 10% of patients with SSc (25, 26), and it emerges at a late stage in the course of the disease, being associated with poor prognosis (8, 27). In contrast, subclinical cardiac involvement in SSc is quite common. In 436 patients with SSc included in a meta-analysis by Follansbee et al. (12). ECG abnormalities were recorded in 46% of the patients. These included supraventricular and ventricular tachyarrhythmias, as well as bradyarrhythmias and interventricular conduction abnormalities (12, 28). Sudden cardiac death and arrhythmias affects 5–6% of patients with SSc (14, 29) but the role of ECG markers in its prediction is not yet established. Thus, the association between prolonged QTc, commonly found in SSc, and SSc prognosis remains unclear (15–17, 30, 31)and a prolonged QT interval corrected for heart rate (QTc, and QT dispersion, a marker of repolarization heterogeneity, is reported to be either increased (17) or normal (19).

An increased QTVI, a marker of repolarization lability, was reported in patients with congestive heart failure, and it was found to be associated with a higher risk of ventricular arrhythmias (21). In an attempt to evaluate novel markers of adverse cardiac outcomes in SSc, we aimed to study QTVI with certain other variability parameters. However, in our cohort, we found high QT variability values in only 1 of 17 SSc patients who have had an episode of ventricular arrhythmia recorded during hospitalization for palpitations and presyncope 1 year prior to our study. This may suggest that the overall risk for repolarization-associated ventricular arrhythmias during the first few years after diagnosis in patients with SSc, who do not have overt cardiac involvement, is low. Yet, in SSc patients with cardiac disease, abnormal repolarization may mark an increased propensity for ventricular arrhythmias, thereby proposing QTVI as a screening test to determine the risk for arrhythmias in SSc with overt cardiac disease. Of note, a negative QTVI study does not exclude the possibility that early myocardial involvement in SSc may interfere with depolarization rather than repolarization.

In our hands, the QTc results of patients with SSc were comparable to those of the control subjects (428.9±26.4 vs. 421.5±23.5 ms). This finding conflicts with the findings by Massie et al. (30) Morelli et al. (31) and De Luca et al. (20), who reported an increased QTc in 11%–25% of patients with SSc. Rosato et al. (16) also reported a median QTc of 447 ms in 20 patients with SSc. Morelli et al. (31) included the largest cohort of patients with SSc (689 participants) to study cardiac repolarization. They used a cutoff of 440 ms to define a prolonged QTc, found in 25% of their patients. Nevertheless, a cutoff of 450 ms for females (the majority of SSc patients) was suggested by others to be a more appropriate criterion of prolonged repolarization (23).

The difference between the previous results and our findings may also stem from a longer disease duration and greater SSc severity in the patients included in earlier studies, since these factors were found to be associated with QTc prolongation in SSc (30). For instance, Rosato et al. (16) reported that the late development of nailfold capillaroscopic changes and the presence of digital ulcers were associated with a higher QTc. It remains to be determined if the inclusion of patients with overt cardiac involvement and prolong repolarization will also be associated with more pronounced QTVI abnormalities. Also, different methodologies for QTc evaluation (manual vs. automatic, in-clinic ECG measurements vs. 24 h-ECG-Holter) may affect the results.

Notably, antidepressants were prescribed to the minority of patients in both groups. Yet, most serotonin reuptake inhibitors do not affect QTc (32). Specific reports were linked to repolarization abnormalities and Citalopram (Forest Laboratories, New York, USA), Escitalopram (Allergan, Dublin, Ireland), Venlafaxine (Wyeth Pharmaceuticals Company, New Jersey, United States), or tricyclic antidepressants (33), which were not prescribed to any of our patients. Accordingly, the mean QTc values were similar in both study groups and within normal limits.

The influence of antidepressants on QTVI is mainly unknown. None of the included patients was treated with Pemoline (Abbott Laboratories, Illinois, United States; discontinued), which has been also reported to increase QTVI (34). Interestingly, an anecdotal report has shown that in patients with panic disorder, treatment with cognitive-behavioral therapy, with and without sertraline, is paradoxically associated with lower QTVI values and that it therefore provides a cardioprotective effect (35). It remains to be determined if specific antidepressants might have an adverse effects on QTVI in patients with SSc.

In conclusion, QT variability indexes in SSc without overt cardiac disease are similar to those found in normal controls. Yet, an abnormal QTVI may be found in the minority of patients with SSc and possibly mark an increased risk for ventricular arrhythmias. Arrhythmia risk stratification of QTVI requires a longer follow-up, inclusion of patients with overt cardiac involvement, and larger cohorts.

Limitations

The studied group was small due to the overall low prevalence of SSc, and it included only 1 patient with seemingly abnormal QTVI. Also, the normal thresholds for QT variability parameters remain to be determined in a large cohort of healthy population. Yet, our study is, to the best of our knowledge, the first to examine QTVI as a marker predicting arrhythmia in SSc, and the results should prompt additional investigation of this tool.

Main Points.

• Despite increased prevalence and poor prognosis of cardiac involvement in SSc, higher than normal QTVI, an index of repolarization lability, is found only in the minority of SSc patients.

• It remains to be determined if increased QTVI has a prognostic significance in SSc.