Electromechanical Dyssynchrony and Clinically Silent Ventricular Dysfunction in Young Subjects with Ventricular Pacing for Congenital and Early Acquired Av Block

Purpose: Single site ventricular pacing can result in electromechanical dyssynchrony and LV dysfunction. A classic pattern by strain echocardiography pattern analysis has been defined that specifically identifies electromechanical dyssynchrony that leads to LV dysfunction in other populations. Our purpose is to evaluate for the presence of the classic pattern and an association with LV dysfunction in asymptomatic young subjects with chronic pacing.

While there are many echocardiographic dyssynchrony indices, recently, regional strain pattern analysis has been used to more specifically identify a classic pattern of electromechanical dyssynchrony. This classic pattern has been identified in children [11] and adults [14] with native LBBB and in congenital heart disease patients [12,13,15]. This pattern is characterized by paradoxical wall motion resulting from early contraction of the septal wall versus late contraction in the activation-delayed lateral wall.
Cardiac resynchronization therapy (CRT) can treat electromechanical dyssynchrony when properly diagnosed and has been shown to reduce heart failure hospitalizations and deaths, improve functional status, and cause favorable ventricular reverse remodeling in the majority of patients [16,17]. While many echo dyssynchrony criteria, such as the time to peak (TTP) criteria, have not proven effective in improving CRT response rates as they evaluate for general mechanical discoordination rather than electromechanical dyssynchrony, [8] the classic pattern has sensitivity and specificity > 90% for predicting CRT response in adults with native LBBB [14]. This observational study is the first to identify the classic pattern in young subjects with chronic RVP and assess the relationship with LV dysfunction. Identification of pacing induced dyssynchrony is important in this AV block population due to their life-long pacing requirements and the fact that progressive dysfunction is often seen in patients with electromechanical dyssynchrony.

Study Population
We retrospectively identified subjects from the Duke Pediatric  [16].
Bundle branch block was determined based on standard criteria [18]. Pacemaker type and settings were recorded from pacemaker interrogation reports. Implantation dates of each pacemaker system and the percentage of ventricular paced beats as a percentage of total ventricular beats at least since the previous device interrogation were also reported. Pacing system duration defined as time from initial or current pacer insertion until date of echocardiogram.

Echo Image Acquisition
Retrospectively identified subjects had studies obtained on Simpson's Rule [19].

Velocity Vector Imaging Tomtec Strain Analysis
All prospective and retrospective studies were converted to DI-  [20,21].
Contraction timing was measured using segmental time to peak (TTP) intervals as the time from the onset of the QRS to the primary distinct peak on a strain curve. The standard deviation of the six segmental TTP values for each patient was calculated as an index of mechanical discoordination. A standard deviation of TTP (TTPsd) ≥ 60 milliseconds has been used as a traditional marker of significant dyssynchrony in adults with LBBB-associated cardiomyopathy [22].

Regional Strain Pattern Analysis to Identify the Classic
Pattern Methodology for regional strain pattern analysis has been previously published in multiple populations to identify a classic pattern consistent with electromechanical dyssynchrony based on longitudinal strain analysis. [14,23] The following three criteria define the classic pattern: a) early terminated contraction of at least one septal segment simultaneous with early stretch in at least one opposing lateral wall segment b) early terminated peak contraction in the septum must occur in the first 70% of the systolic ejection phase; and c) at least one lateral wall segment demonstrates early stretch followed by late contraction (peaking after AVC) [14] ( Figure 1).

Figure 1:
The classic pattern of electromechanical dyssynchrony. Numbers identify the three criteria to define the pattern on two representative strain curves (septum: yellow, lateral wall: red): a) early contraction of at least one septal segment (yellow tracing) and early stretching in at least one opposing lateral wall segment (red tracing) b) early peak contraction must occur in the first 70% of the systolic ejection phase (from aortic valve opening to closure) and c) late post-systolic peak contraction occurring after aortic valve closure (AVC) in the segment which had demonstrated early stretch. ECG tracing (black) provided for timing. Aortic valve opening (AVO).

Cardiac Resynchronization Therapy
Cardiac resynchronization therapy has only been performed in four of the study subjects. In these subjects, primary strain analy-sis was performed on an echocardiogram prior to CRT. Follow-up echocardiograms were obtained on all four CRT subjects between 4 months and 5.3 years after CRT. CRT response was defined as an improvement in LVEF > 10%.

Statistical Analysis
Continuous variables were reported as the median value (

Classic Pattern of Electromechanical Dyssynchrony
The classic pattern was seen in 38% of subjects (13/34). Of those with a classic pattern, all (13/13) demonstrated BBB on EKG.
LV function was diminished and the QRSd adj was longer in those with a classic pattern ( Table 2). In the classic pattern sub-group, the earliest contracting region was the mid or basal septum in 4/13, the apex in 7/13, and both the septum and apex in 3/13. Pacing lead site was associated with the earliest contracting LV segment (p=0.04).
An RV apical septal lead site typically led to earliest contraction in the LV apical septum, whereas an RV mid/basal septal lead position would cause earliest contraction in the mid or basal LV septum.
However, lead position was not associated with the development of the classic pattern in the overall cohort (p=0.76). In subjects with the classic pattern, there was a correlation between total duration paced and LV ejection fraction (r=-0.68, p=0.01), but in those without the classic pattern, there was no such association (r=0.072, p=0.76). Overall, the TTP sd ranged from normal to dyssynchronous (56ms, 38 -129 ms). Fourteen subjects (41%) met the threshold for significant dyssynchrony by this TTP sd criteria (> 60 milliseconds).
The classic pattern and TTP sd > 60ms criteria were strongly associated and the groups overlapped significantly, but there were three total subjects who met one but not both criteria (Table 2).

Cardiac Resynchronization Therapy
In the four subjects (ages 6, 12, 26, and 27 years) who underwent cardiac resynchronization therapy, the single ventricular lead

Chronic Pacing and LV Dysfunction
Although chronic ventricular pacing is known to be associated with LV dysfunction [5,6,10], we observed that the majority of our subjects fell into the abnormal range and that 17% of our population had at least moderate LV systolic dysfunction. None of the subjects having LV dysfunction, even those with moderatesevere dysfunction, reported any symptoms of heart failure, potentially due to the gradual nature of this process. In a large multicenter study of pediatric AV block subjects, Janousek et al. [5] found that > 50% of those with chronic RV pacing had diminished LVEF < 55%. Our observations in this asymptomatic cohort agree with previous literature and serve to emphasize the importance of routine echocardiographic surveillance in those with chronic ventricular pacing.

Classic Pattern of Electromechanical Dyssynchrony
This observational study demonstrates that identification of the classic pattern using strain analysis was associated with more significant cardiac dysfunction. In adults with native LBBB, the classic pattern criteria were based on longitudinal strain and designed to represent the consequences of a pathophysiological sequence in which electrical activation delays lead to electromechanical dyssynchrony, systolic dysfunction, and eventually heart failure ( Figure 3) [14]. The classic pattern, found in subgroups of various populations with electromechanical dyssynchrony [11,14,15,[23][24][25][26][27], was also identified in 38% of the current ventricular paced young patient cohort having normal cardiac anatomy. In subjects who require lifelong RVP, the classic pattern may help to identify a subgroup at increased risk for more severe dysfunction who may benefit from more frequent echo surveillance. Identification of the classic pattern may also more specifically identify those who will respond to an upgraded biventricular pacing system, but that was not systematically evaluated in this study.

Figure 3:
The pathophysiologic sequence of electrical activation leading to dyssynchrony, ventricular remodeling and eventually heart failure.

Non-Classic Pattern Population
In our study population, those without the classic pattern (62%) had significantly better LV function, but some had low-normal or mildly diminished systolic function. Nearly all of the subjects who did not have the classic pattern, but did have mild LV dysfunction, had congenital AV block. Previous studies have shown that primary myocyte dropout and fibrosis secondary to in utero immunemediated mechanisms from maternally acquired autoimmune antibodies may contribute to fetal and infantile cardiomyopathy [20][21]. Such a mechanism could result in a regionally discoordinated LV strain pattern (positive for TTP criteria) that would not meet classic pattern criteria as it does not arise from electromechanical dyssynchrony. This may explain the two subjects who were positive for the TTP sd criteria but negative for the classic pattern. Additionally, this pro-fibrotic mechanism may account for LV ventricular dysfunction through a different mechanism than electromechanical dyssynchrony in this population.

Duration of Pacing and Ventricular Dysfunction
A recent multi-center study demonstrated that total duration of pacing did not predict LV dysfunction [4]. In our study, the total

Site of Pacing and Abnormal LV Mechanics
The location of a single ventricular pacing lead has been shown to be associated with increased dysfunction, although the location associated with more severe dysfunction has varied by study [5,7,10]. A recent multi-center study that considered lead location and LV dysfunction demonstrated that the RVOT leads were associated with the poorest LV function, although there were only 9 subjects in this subgroup [5]. All RV lead locations were associated with diminished LV function compared to LV epicardial locations. There was no association between RV pacing site and the development of the classic pattern in our study. This may suggest that the abnormal activation from single site pacing with any RV lead location may cause this pattern of electromechanical dyssynchrony given the right circumstances. However, this analysis is underpowered to answer this question, and any other necessary circumstances required for the development of the classic pattern are unknown. Left ventricular pacing from the apex or lateral wall has been shown to better maintain normal LV function in multiple studies [5,28]. In our study, there were only two subjects with epicardial LV pacing. One had normal function and no dyssynchrony.
The other had mildly diminished function and a classic pattern with early apical and late basal contraction. This type of classic pattern with early LV apical contraction and late basal contraction is similar to the classic pattern seen with pacing from the apical RV septum [6]. The interplay between the type of pacer, the pacing lead site and the development of electromechanical dyssynchrony and progressive dysfunction will require future research.

CRT Response
Only four of our study subjects had undergone CRT with placement of a coronary sinus or LV epicardial lead at the time of this study. All four subjects had the classic pattern and met TTP criteria prior to undergoing CRT. They all demonstrated significant beneficial LV reverse remodeling post-CRT. None of these subjects ever reported symptoms of heart failure before or after CRT. Biventricular CRT is known to improve ventricular function compared to RV pacing in the AV block population [29].
However, it is unknown whether the classic pattern can improve CRT response rates over EKG only or EKG and TTP criteria in this paced population as it has in native LBBB cardiomyopathy [14,23].
This study was not designed to evaluate the classic pattern or TTP indices to predict CRT response in this AVB population.

Limitations
The echo lab standard changed in 2007 when 4 chamber apical views became consistently of adequate quality for strain analysis.
Because only one echocardiographic evaluation of adequate quality was performed for many of our subjects, this study could not determine when the classic pattern or ventricular dysfunction first developed. An effort to analyze echocardiograms immediately after pacemaker implantation was made impossible due to the variation in echo image quality in many of those studies that were performed before the current echo lab standard was in place and using ultrasound machines and imaging standards from a previous era. Furthermore, the LV3 and LV2 apical and parasternal short axis views were not consistently obtained at adequate quality for strain analysis during the study period, so evaluation of the 18 segment GLS or other strain planes were not feasible. The lack of LV3 and LV2 apical views may have led to a slight under-identification of the classic pattern, but Risum et al. [14] demonstrated that for nearly all subjects with the classic pattern, it is present in the LV four chamber view. Our study provides early data identifying a previously unrecognized association between the classic pattern and progressive dysfunction in this population that will require future prospective study. This study is also not designed to evaluate CRT response in subjects with the classic pattern. Due to incomplete records, the maternal immunologic status (anti-Ro/La) in our study population was not available.

Conclusion
This study demonstrates a high frequency of clinically silent but significant LV dysfunction in young subjects having CAVB who require chronic single site ventricular pacing. In subjects with a classic pattern of electromechanical dyssynchrony by strain analysis, a strong correlation was found between duration of pacing and LV dysfunction that was not present in those without the classic pattern. The classic pattern identifies a group of RV paced subjects who may require closer echo follow-up. The findings from our study should lead to a larger study with serial echocardiograms prospectively protocoled for strain analysis to more thoroughly evaluate the development and consequences of the classic pattern in this young paced population.