Electrocardiographic features of the presence of occult myocardial disease in patients with VPD‐induced cardiomyopathy

Abstract Background Frequent ventricular premature depolarizations (VPDs) can cause reversible cardiomyopathy (CMP). However, many patients maintain a normal left ventricular (LV) function with a high VPD burden. The electrocardiographic characteristics of VPD‐induced CMP have not been elucidated. Methods One hundred and eighty (91 men, age; 51 ± 15 years) patients with frequent idiopathic VPDs (>10% VPDs/day or >10 000 VPDs/day) were studied. All patients underwent successful ablation and were then divided into two groups according to the echocardiographic findings before and after the ablation procedure. Results Group A (n = 139) had a normal LV function with VPD frequencies, and Group B (n = 41) had reversible LV dysfunction after ablation. The VPD QRS duration (QRSd) was wider in patients with CMP (Group A vs Group B; 137.2 ± 12.0 milliseconds vs 159.7 ± 5.3 milliseconds, P < .001). VPDs with a terminal QRS delay marked by a notch followed by a discrete lower amplitude signal after the peak R wave in any precordial lead were identified. The incidence of terminal signals was higher in the CMP group (Group A vs Group B; 2.1% vs 53.6%, P < .001). Conclusions The wider VPD QRSd and terminal QRS delay in patients with VPD‐induced CMP suggest subclinical cell‐to‐cell conduction abnormalities as a potential factor predisposing VPD‐induced CMP.


| INTRODUC TI ON
Idiopathic ventricular premature depolarizations (VPDs) are usually considered a benign condition, even when the VPDs are frequent. 1,2 However, the VPD burden is one of the main causes leading to left ventricular (LV) dysfunction. 3,4 Previous several studies have described that a high burden of VPDs (>24%) is associated with LV cardiomyopathy (CMP) that resolves after successful VPD ablation. [4][5][6][7][8] However, in 25%-30% of patients the status of the LV systolic function cannot be explained only by the VPD burden. 9 The mechanisms underlying the development of VPD-induced CMP are not completely understood. In our experience, some VPD patients maintained normal ventricular function even with persistent ventricular bigeminy, while a significant number of patients had a depressed LV function, continuously, even with a lower VPD burden.
Nishikawa et al reported that mild to severe interstitial fibrosis was observed by myocardial biopsy performed in patients with various arrhythmias. 10 It could be that the extent of microscopic myocardial disease at baseline varies, which would not be detected by imaging studies in patients with idiopathic VPD-induced CMP. In this study, we sought to find the useful electrocardiographic (ECG) characteristics of patients with VPD-induced cardiomyopathy by analyzing and comparing the clinical and ECG parameters between patient groups with normal LV function and VPD-induced CMP after undergoing successful radiofrequency catheter ablation (RFCA).

| The inclusion criteria were as follows
All patients were enrolled based on the following criteria: (a) frequent VPDs (>10% or 10 000 VPDs/day); (b) no evidence of structural heart disease; (c) history of a successful VPD ablation (>80% suppression of the VPD burden); (d) baseline and follow-up transthoracic echocardiography (TTE), 24 hours Holter monitoring data; (e) had no evidence of sustained ventricular tachycardia (VT); and (f) stopped all anti-arrhythmic drug (AAD) use for at least five half-lives.
And the patients with normal LV ejection fraction (EF) were defined as Group A. And those with depressed LV EF at baseline and with normalized LV EF after a successful VPD ablation were defined as Group B. Successful ablation was defined as having at least an 80% reduction in the 24-hour burden of VPDs, based on our previously published experience. 11 We did coronary angiogram or computed tomographic coronary angiogram for evaluation of coronary lesions in all patients.

| Assessment of the LV function
TTE was performed before the ablation procedure using the Simpson formula. For assessment of the LV EF, the second of three consecutive sinus beats was used to avoid any post-extra-systolic potentiation. An LV EF of <50% was considered abnormal. TTE with a quantitative assessment of the LV function was repeated 3-6 months post-ablation.

| Follow-up
Patients were seen in an outpatient clinic 3, 6, and 12-48 months post-ablation. 24 hours Holter monitoring was performed prior to the ablation procedure to measure the VPD burden (% and numbers/day).
Follow-up Holter monitoring was repeated 3-6 months post-ablation and again, later, if palpitations recurred. All anti-arrhythmic drug therapy was discontinued if the ablation was effective. β-Blockers(BB) and heart failure medications were continued initially and were discontinued if and when the LV function and dimensions normalized. No new medications were added after an effective ablation procedure.

| ECG measurement
Sinus rhythm and the VPD ECG morphology were measured on the same 12-lead ECG with electric calipers on the Prucka Cardiolab recording system (GE Healthcare; Figure 1). A standard 12-lead ECG electrode placement was used. The lead gain was uniform with a paper speed of 100 mm/s. Additional clinical and electrophysiologic parameters were assessed by a detailed retrospective review. All electrocardiographic measurements were performed, blinded to the TTE outcomes, by one of the two authors (SII or KMP) using digital calipers at 100 mm/s, on the CardioLab® (version 6.5.4.1858; GE Medical Systems). To distinguish between the true J point and presence of retrograde p waves in the measurement of the VPD QRS duration (QRSd), we examined the intracardiac electrograms (after catheters were in place) to evaluate any retrograde conduction. Assessment of the interobserver variability was performed for the measurements of the VPD QRSd and VPD coupling interval (CI; Figure 1A-C).

| Durations (width) and intervals
During the clinical VPD, the following measurements were obtained during both sinus rhythm (SR) and VPDs; (a) sinus cycle length (SCL); The VPD QRSd was defined as the interval from the earliest ventricular activation to the offset of the QRS in any of the 12 leads. The CI was defined as the interval from the onset of the "q" or "R" wave of the previous sinus rhythm to the onset of the VPD QRS. The CI ratio was defined as the CI divided by the SCL. The post-VPD CI was defined as the interval from the VPD onset to the onset of the "q" or "R" wave of the next sinus rhythm. The post-VPD CI ratio was defined as the post-VPD CI divided by the SCL. The "qR" width was defined as the width from the initial onset of the "q" wave of the VPD to the highest peak of the "R" wave in the precordial and limb leads. The "Rs" width was defined as the width from the peak "R" at the highest amplitude, in the same lead as the qR interval, to the terminal s wave in the precordial and limb leads. The "r/ R" transition interval was defined as the interval from the peak point of the earliest "r/ R" to the peak point of the latest "r/ R" in the precordial and limb leads. The definition of an "r" wave was one with an amplitude of over 0.1 mV. If there was notching in the summit of the VPD, the dominant peak was measured.

| Parameters
The following parameters were obtained during normal SR and VPDs: (a) highest "RS" amplitude on the precordial and limb leads; (b) maximal "R" and "S" amplitude on the precordial and limb leads; (c) initial and terminal angles measured using the Pythagorean theorem from the isoelectric line to the peak of the "R" or "S" wave in the lead with the highest amplitude in the precordial and limb leads ( Figure 1D).
The highest RS amplitude was defined as the highest amplitude from the peak "r/ R" to the peak "s/ S" in the precordial and limb leads.
The maximum "R" amplitude was defined as the amplitude from the maximum peak "R" point to the isoelectric line in the precordial and limb leads. The angles were measured using the Pythagorean theorem in the lead with the highest "R" amplitude. If there was no "R" wave in either the precordial or limb leads, then the "S" wave was used instead of the "R." The initial angle was defined as the angle between the QRS onset to the peak "R" or "S" peak and the isoelectric line. The terminal angle was defined as the angle between the terminal point of the QRS to the peak "R" or "S" peak and the isoelectric line.

| Terminal signals
Terminal signals were defined as the terminal QRS delay marked by a notch followed by a discrete lower amplitude signal after the peak R wave in any precordial lead. If there were any potential-like terminal signals in the precordial leads, we considered as those patients with terminal signals (Figure 2).

| Statistical analysis
The Pearson's product moment correlation coefficient was calculated to quantify the inter-variability in the measurement of both the VPD QRSd and coupling interval. Differences in the baseline characteristics across the groups of interest were carried out first in a univariate fashion using a Fisher's exact test for categorical variables and the Kruskal-Wallis test for continuous variables. The percent of VPDs F I G U R E 1 Eletrocardiographic (ECG) parameters. A, CI and post-VPD CI; B, QR and RS interval and amplitude; C, r/R transition interval; D, initial and terminal angle of VPD. CI, coupling interval; SCL, sinus cycle length; VPD, ventricular premature depolarization over 24 hours was a priori included in the preliminary main effects model given its clinically plausible influence on the outcomes follow-

| Baseline characteristics
There was no difference in both groups except age, gender, implanted cardiac defibrillator (ICD) implantation, use of BB, angiotensin-converting enzyme inhibitor (ACEi), and angiotensin receptor blocker (ARB; Table 1

| Interobserver reliability of the ECG measurements
The correlation coefficient for the measurements of the VPD QRSd was 0.871, and that for the VPD coupling interval was 0.936.

| Terminal signals
Potential-like signals were found at the terminal portion of the clin-  b Defined as any area of delayed gadolinium enhancement or regional wall motion abnormality.
The proportions presented are the number of abnormal exams over the number of subjects who underwent MRI.

| Main findings of the study
It is increasingly recognized that idiopathic VPDs may cause LV dysfunction that is reversible after a successful ablation. 5,9 Recent studies suggest that a VPD frequency of more than 24% during 24hour Holter monitoring is a risk factor for VPD-induced CMP. 3,6,7,12 However, 20%-25% of the VPD patients did not meet this cutoff value in those studies. In fact, many patients maintained a normal LV function even with a high VPD burden. Inversely, some patients have reduced ventricular function with a lower VPD burden. One study to date has examined the longitudinal impact of VPD burden on the LV function, and a subclinical deterioration in the LV function was found with a high burden of VPDs over 5 years. 13 Finally, the paradigms of VPD-induced CMP could not be explained with the VPD burden alone. Recently, some authors have questioned the preexistence of occult structural heart disease as one of the mechanisms of VPD-induced CMP. 13,14 Interestingly, in this study, the number of asymptomatic patients was significantly higher in Group B than in Group A (A vs B; 10% vs 41.4%, P < .001). This finding is consistent with previous study that a lack of symptoms could be associated with a greater risk of VPCinduced cardiomyopathy. 15,16 In this study, we examined the clinical and electrocardiographic features in patients with VPD-induced CMP compared with normal control patients. The VPD QRS duration of the CMP patients was significantly more prolonged than that of the normal patients, even after adjusting for the sites of origin of the VPDs, LV dimension, and body surface area. Among the ECG parameters, Rs width was significantly wider in Group B. Furthermore, only terminal peak angle was significantly lower in Group B. We also found abnormal distorted potential-like signals within the Rs segment of the VPD, and were found predominantly more often in patients with LV dysfunction than in the normal control patients. From these results, we could infer carefully that Rs segment of VPD is more important to identify the presence of occult myocardial disease in the patients with VPDinduced CMP than qR segment. They explained that a broadly notched VPD with a long duration was a useful marker for a dilated globally hypokinetic left ventricle. They described the mechanisms of this distorted VPD as dilatation of the T-tubule system due to altered microanatomy and as an abnormality in the cell-to-cell communication by desmosomes.
Even though 61% of the patients had coronary artery disease, it was a meaningful result in that the prolonged VPD QRS duration and notch on the ECG were important as indirect evidence of the myocardial status. Aizawa Y et al also reported that tachycardia-dependent augmentation of "notched J waves" in a general patient population without ventricular fibrillation or cardiac arrest was augmented at shorter RR intervals, but not at prolonged RR intervals.
Mechanistically, conduction delay is most likely responsible for this change. 30 However, we do not know the meaning of these signals in our study and could not perform myocardial biopsy to confirm the results in the tissue. However, the signals occurred more frequently in CMP patients than in normal patients. These findings may suggest some occult microscopic myocardial disease in the myocardium. If the occult microscopic myocardial disease was distributed diffusely, a myocardial conduction delay would occur globally and the VPD QRSd would be significantly wider than normal.
Whatever the mechanisms of these signals, a delayed slow potential or dyssynchrony, these results suggest a new direction for the electrophysiological and pathophysiological mechanisms that lead to VPD-induced CMP.

| Study limitations
There were some limitations in this study, First, this study was a single-center, retrospective study derived from a real-world practice with inherent limitations and we could not assess quantitative analysis of dyssynchrony by echocardiography. Hence the results of our study should be considered as hypothesis generating, and future prospective studies are warranted to confirm our results. Second, in our study, among all enrolled patients, 36% of the patients underwent cardiac MRI while 13% underwent CAG to rule out any structural heart disease. Therefore, we could not rule out with certainty the existence of minimal structural heart disease that could be detected by using TTE. Third, the accurate measurement of the LV dysfunction may have been compromised by frequent VPDs, particularly in patients with incessant ventricular bigeminy, who never have two simultaneous sinus beats.

| CON CLUS ION
In patients with idiopathic VPDs, the presence of wider VPD QRS duration and potential-like signals at the terminal portion of VPD may be indirect evidence of the pre-existence of occult microscopic myocardial disease with reversible CMP.

ACK N OWLED G EM ENTS
We thank all members of Cardiology at Samsung Medical Center and Kosin university gospel hospital for their assistance and support with data collection.

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