Fragmented QRS and subclinical left ventricular dysfunction in individuals with preserved ejection fraction: A speckle‐tracking echocardiographic study

Abstract Introduction Fragmented QRS (fQRS) complex on routine 12‐lead electrocardiogram (ECG) predicts adverse outcomes in patients with cardiovascular diseases. In addition, it has been found to be associated with subclinical myocardial dysfunction in chronic diseases. We sought to investigate the relationship between the presence of fQRS with the myocardial functions in individuals free from known systemic cardiovascular diseases. Methods In a case‐control study, we evaluated normal individuals from March 2017 to February 2018. All participants underwent a 2‐dimensional transthoracic echocardiographic examination using tissue Doppler imaging (TDI) and speckle‐tracking echocardiography. In addition, all participants were examined using a 12‐lead surface ECG, and patients with fQRS and a group of age‐ and sex‐matched controls without fQRS were enrolled in our study. Results The patients' mean age was 40.3 ± 10.7 and 35.4 ± 11.2 years in fQRS‐positive and fQRS‐negative groups, respectively (P = .110). Patients with fQRS had significantly lower values of apical left ventricular global longitudinal strain (LV GLS) in 2‐chamber (16.9 ± 2.5 vs. 20.5 ± 3.3, P < .001), 4‐chamber (16.9 ± 3.4 vs. 20.1 ± 3, P = .001), LAX views (17.7 ± 2.8 vs. 20.8 ± 3.5, P = .001), and averaged LV GLS (17 ± 2.6 vs. 20.4 ± 2.7, P < .001) values compared to patients without fQRS. In a multivariate analysis, averaged LV GLS and smoking history were independent predictors for positive fQRS. Conclusion The presence of fQRS on 12‐lead ECG in healthy population was associated with lower values of LV GLS compared to normal individuals without fQRS.


| INTRODUC TI ON
Fragmented QRS (fQRS) complex is detected in a 12-lead electrocardiogram (ECG), which has enormously been used for the prediction of cardiovascular outcomes. 1 The fQRS indicates altered ventricular depolarization caused by myocardial fibrosis and scar in patients with ischemic heart diseases. [2][3][4][5] It is not only observed in patients with coronary artery diseases, but is also found in cardiomyopathies, structural heart diseases, heart rhythm disturbances, cardiac sarcoidosis, and even healthy population. 1,[6][7][8] The fQRS has been found to be a marker of clinical outcomes in different cardiovascular diseases. Several studies have shown that the fQRS is a predictor of sudden cardiac death in patients with cardiomyopathy and heart failure, 6,9 a predictor of cardiac events and mortality in patients with coronary artery disease, 10-12 a marker for nonresponsiveness to cardiac resynchronization therapy in heart failure patients with ventricular dyssynchrony, 13 a prognostic marker for dysrhythmia in patients with Brugada syndrome, 14 and short-term prognosis of patients undergoing transcatheter aortic valve implantation. 15 The standard 12-lead ECG can be suggestive of abnormal findings during echocardiographic findings and patients should be regularly followed up. 16 In addition to the prognostic value of fQRS in patients with cardiovascular diseases, it has also been found to be associated with subclinical myocardial dysfunction in patients with coronary artery disease, 17 diabetes mellitus, 18 and chronic kidney disease. 19 The assessment of myocardial tissue by conventional echocardiography is load-and angle-dependent, but 2-dimensional speckle-tracking echocardiography (2D-STE) lacks such a limitation for the evaluation of myocardial mechanics. [20][21][22][23] Although left ventricular ejection fraction (LVEF) is commonly used for the evaluation of LV function, 2D-STE parameters, mainly global longitudinal strain (GLS), have been found to be a feasible modality for examination of left ventricle (LV) function. 24 In this case-control study, we sought to determine the relationship between the presence of fQRS in apparently healthy individuals without cardiovascular diseases and preserved LVEF and subclinical LV dysfunction using 2-dimensional transthoracic echocardiography (2D-TTE) with the implementation of tissue Doppler imaging (TDI) and speckle-tracking echocardiography.

| Study protocol and population
In a case-control study, we prospectively enrolled normal individuals with or without fQRS on their 12-lead surface ECG at rest, from March 2017 to February 2018 in Urmia city, Iran. All participants underwent a comprehensive 2D-TTE examination using TDI and STE modalities to evaluate the structural and functional features of their heart. The study was approved by the institutional review board in Seyyed-al-Shohada Heart Center and the local ethics committee of Urmia University of Medical Sciences, Urmia, West Azerbaijan province, Iran. In addition, informed consent was obtained from all participants at baseline.
Twenty-six consecutive patients who visited our outpatient clinic with positive fQRS were recruited into this study. In addition, 28 age-and sex-matched individuals who had negative fQRS in baseline ECG were also enrolled. Inclusion criteria included healthy individuals who were referred to be evaluated for probable cardiac diseases before military service entrance, employment in governmental organizations, a requirement for insurance validity, and participation in sport events. In addition, some people who visited our outpatient clinic for palpitation and atypical chest pain with a low probability of coronary artery diseases whose stress tests and/or cardiac scans had been negative for any ischemic heart diseases. Exclusion criteria included patients with known history of coronary artery diseases, any signs of ischemic heart diseases in stress tests and cardiac scan, heart failure, diabetes mellitus, hypertension, hypercholesterolemia, LVEF lower than 55%, any significant valvular diseases (ie, greater than mild involvement), any congenital heart diseases, atrioventricular conduction abnormalities, QRS duration longer than 120 milliseconds, pulmonary arterial hypertension, chronic obstructive pulmonary diseases, consumption of any medications for chronic diseases, and bundle branch block in ECG trace.

| Electrocardiographic examination
All participants were examined by a 12-lead surface ECG at rest (0.5 Hz to 150 Hz, 25 mm/s, 10 mm/mV). ECGs were analyzed by two clinicians who were blinded to the impression of cases. In case of disagreement, the mutual agreement was used to reach a final diagnosis. The fQRS was defined as the detection of RSRʹ pattern, which can be observed as follows: (i) an additional R wave; (ii) a notching in nadir of the S wave; (iii) a notching of the R wave; (iv) Rʹ wave >1 millimeter; and (v) the presence of fragmentation, more than one Rʹ wave, in two contiguous ECG leads. 1,25

| Echocardiographic examination
All echocardiographic measurements were performed based on the recommendations of the American Society of Echocardiography and the European Association of Cardiovascular Imaging, 26 using an ultrasound scanner (Vivid S6). A 2D-TTE was used to evaluate the presence of any structural heart diseases, valvular and congenital heart diseases. The LVEF was also measured using the biplane Simpson's method. All echocardiographic examinations were carried out by a single echocardiographer in our echocardiography laboratory.
A standard M-mode, 2D, and color-coded TDI images were obtained during breath hold, as an average of three consecutive beats.
The TDI was provided from the annular and septal mitral valve area.
The harmonic image recordings of apical and short-axis views at mitral valve and papillary muscle levels (30-90 frames/second) were provided and stored to be analyzed offline using AFI (automated functional imaging) software. The left ventricular GLS (LV GLS) was calculated by averaging apical 4-chamber, 2-chamber, and LAX views.
Based on a meta-analysis, 27 the ranges of GLS in normal population is 15.9% to 22.1%; therefore, the GLS was considered abnormal when it was smaller than 16 in our cohort. The evaluation of LV diastolic dysfunction was also performed using the latest American Society of Echocardiography guideline for LV diastolic dysfunction as well. 28

| Statistical analysis
All variables are presented as mean ± standard deviation, and t test was used to compare continuous variables between groups. Chisquared tests or Fisher's exact test was implemented for comparing categorical variables as appropriate. A multivariate regression analysis was conducted to evaluate the relationship between fQRS and LV GLS values. The model was also adjusted for age, gender, smoking history, QRS duration, QT interval corrected for heart rate, body mass index, and total cholesterol level. Adjusted odds ratios (ORs) and corresponding confidence intervals (CIs) were reported. A two-tailed P < .05 was considered statistically significant. All statistical analyses were performed using the IBM SPSS software version 22.0 (IBM, Armonk).
The number of smokers was significantly higher in positive fQRS compared to negative fQRS group (30.8% vs 7.1%; P = .026). The half of patients with fQRS (13%, 50%) had fQRS in less than three leads and others had fQRS in three or more than three leads. Fifteen patients (57.7% of fQRS patients) had fQRS in inferior leads, five patients (19.3% of fQRS patients) in precordial leads, and others (23% of fQRS patients) in a combination of precordial and limb leads. The QRS duration was comparable between individuals with or without fQRS (93.5 ± 6.3 and 93.6 ± 6.2 msec, respectively; P = .949). All laboratory data and other parameters of ECG were comparable between groups (Table 1).
Due to a high number of smokers among individuals with fQRS, we compared echocardiographic values among individuals with or without fQRS in subgroups by smoking history. All echocardiographic TA B L E 1 Baseline characteristics in patients with or without fQRS  Table 3.

| D ISCUSS I ON
In this case-control study, we showed that among apparently healthy individuals with normal findings in routine cardiovascular screening using a 12-lead standard ECG and a 2D-STE, the LV GLS values were significantly lower in individuals with fQRS than those without fQRS.
Although those with fQRS smoked more than non-fQRS participants, there was no significant difference with regard to the echocardiographic parameters between individuals with or without smoking history. In addition, on multivariate analysis, averaged LV GLS, inversely, and smoking, directly, were associated with the presence of fQRS in our cohort. analysis, this association remained significant even after adjustment for other factors; however, gender modified this association, so that GLS did not predict outcomes in women rather than men. The present study was a case-control study and participants were matched by sex and age, so we could not find any effect of sex on our findings. Given previous studies and our study, we believe that it can be postulated that among apparently healthy population, the presence of fQRS might be considered as a surrogate of subclinical myocardial dysfunction, which can be evaluated by 2D-STE. The impact of this association on the outcomes of apparently healthy population warrants further large-scale studies to assess whether concomitant fQRS and decreased LV GLS in a healthy population are associated with worse outcomes at long-term follow-up or not.

| Study limitations
This study suffers from some shortcomings needed to be addressed in future studies. Firstly, it was a case-control study with small sample size. Secondly, all participants with atypical chest pain were categorized as low-risk group and they did not undergo coronary angiography, so we could not definitely say that all patients were free of coronary artery disease. Thirdly, we performed neither noninvasive imaging for the evaluation of myocardial scar/fibrosis nor a pathologic evaluation of myocardial tissue to show that which pathogenesis might contribute to such a phenomenon in our population, fQRS and lower LV GLS. Finally, we did not follow-up individuals to pursue the prognostic value of concomitant fQRS and decreased LV GLS in the healthy population. Given these limitations in this study, it seems that future studies should focus on these issues.

| CON CLUS ION
The presence of fQRS on standard 12-lead ECG in the apparently healthy population is significantly associated with lower values of LV GLS compared to individuals without fQRS, indicating a subclinical myocardial dysfunction.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interests for this article.