Impact of Cardiac Resynchronization Therapy on Heart Transplant-Free Survival in Pediatric and Congenital Heart Disease Patients
Chubb H, Rosenthal DN, Almond CS, Ceresnak SR, Motonaga KS, Arunamata AA, Long J, Trela AV, Hanisch D, McElhinney DB, Dubin AM.
Circ Arrhythm Electrophysiol. 2020 Apr;13(4):e007925. doi: 10.1161/CIRCEP.119.007925. Epub 2020 Mar 22.
Take Home Points:
- Cardiac Resynchronization Therapy (CRT) has become a mainstay in adult heart failure management and a potential option in children with and without congenital heart disease.
- CRT is associated with improved transplant-free survival in those patients with symptomatic heart failure (AHA Stage C or D), ventricular dysfunction (systemic ventricular EF <45%), and electrical dyssynchrony (QRS duration Z-score ≥ 3 or ventricular paced ≥ 40%).
- Heart transplantation or death was significantly lower in the CRT group compared to controls, 19% vs 58% (hazard ratio, 0.24, p<0.001).
- CRT was effective in pediatric patients with congenital heart disease, cardiomyopathy, and systemic RV failure.
- CRT was associated with sustained improvements in ejection fraction at 5 years.
Comment from Dr. Akash Patel (San Francisco), section editor of Congenital and Pediatric Electrophysiology Journal Watch. Heart failure among pediatric patients with and without congenital heart disease continues to provide challenges due to varying efficacy of anti-congestive therapy and limitations on advanced therapies such as heart transplantation. Cardiac resynchronization therapy (CRT) for the treatment of heart failure in adults without congenital heart disease (ACHD) is well established and has shown improvements in ventricular function, functional status, and mortality. The role of CRT in pediatric patients with and without congenital heart disease remains limited due to the small numbers of patients. In particular, the pediatric cohort is impacted by the large heterogeneity of congenital heart disease lesions, morphology of the systemic failing ventricle (right vs. left), variations in circulation (univentricular vs. biventricular), variations in patient size during childhood, and limited follow-up. In addition, the primary outcomes of CRT studies in pediatrics has focused on improvements in ventricular function and not on long term survival. This study aimed to assess the impact of CRT in pediatric patients with and without congenital heart disease on transplant-free survival.
This was a retrospective case-control single center study of pediatric patients (≤ 21 years) with and without congenital heart disease from 2004-2017 who had reduced systemic ventricular function (ejection fraction < 45%), symptomatic heart failure (AHA Stage C or D), and electrical dyssynchrony (unpaced QRS duration z-score ≥ 3 or ventricular pacing ≥ 40%). Cases were enrolled at time of implant and defined as patients who underwent CRT implantation during this time frame with ≥ 1 lead on the systemic ventricle. Controls were enrolled at first outpatient visit meeting inclusion criteria and were subsequently matched using propensity scoring. Exclusion criteria included Eisenmenger syndrome, current VAD, previous heart transplant, and weight <4 kg. See Figure Below
The clinical practice for determining who underwent CRT placement during this time period was made on an individual basis through a multidisciplinary approach involving the heart failure, electrophysiology, and surgical teams. After implantation, optimization of device programming was based on routine clinical practice which changed in 2016. Pre-2016, optimization was focused primarily on echocardiographic assessment for mechanical dyssynchrony and cardiac output. Post-2016, optimization was focused primarily on electrocardiographic assessment of electrical dyssynchrony. Baseline clinical, device, electrocardiographic, and echocardiographic data were obtained. Follow-up data was obtained at 6 (±3) months, 1 (±0.5) years, 2 (±0.5) years and 5 (±1) year when available. Outcome measures included death or heart transplantation, overall survival, time to first transplant listing, and time to first heart failure hospitalization.
The study group included 86 with CRT, 133 controls, and 63 propensity score matched (PSM) pairs. Of note, 6 cases (10%) had CRT turned off in > 1month for reason other than death or transplant. These patients were included in the CRT group for analysis. See Figure Below.
The median age of the CRT-PSM cohort was 11 years with 62% male and 81% with congenital heart disease. The systemic ventricle was left in 81% with a mean systemic EF of 32%. The mean QRS duration z-score was 8.1. The median NHYA Class was 2. Only 5% were listed for transplant at baseline. There was no significant difference with controls across a total of 21 indices used match cases and controls (only 10 shown in table below).
There was no difference in type of circulation (biventricular vs univentricular) and systemic ventricle (left vs right) between the PSM-CRT and control groups.
The median follow-up was 2.7 (0.8 – 6.1) years overall and 2.4 (0.6 – 5.1) years for non-transplanted survivors.
The CRT approach was affected by size and anatomy with the majority of devices implanted being epicardial CRT-P. See Figure Below.
|Device Type in CRT Group|
|CRT-P at baseline||44 (70%)|
|CRT-D at baseline||19 (30%)|
|CRT-P to D upgrade during follow-up||1 (2%)|
|Device Type in Control Group|
|ICD at baseline||12 (19%)|
|ICD during follow-up||2 (3%)|
Overall, heart transplantation or death was significantly lower in the CRT group compared to controls, 12 (19%) vs. 37 (58%), [HR of 0.24, (95% CI, 0.12–0.46), p <0.001]. In addition, the CRT group had a higher rate of overall survival and survival without heart failure hospitalization or transplant listing. See Figures Below.
Risk factors associated with death or transplantation after multivariate analysis included increased risk with higher NHYA/Ross Class and decreased risk with use of CRT and presence of congenital heart disease.
Subgroup analysis showed CRT was effective in improving transplant-free survival in those with and without congenital heart disease, in those without and without the need for bradycardia pacing, and those with systemic right ventricles. In this cohort, CRT in univentricular anatomy did not demonstrate significant improvement upon transplant-free survival.
Longitudinal follow-up showed the median QRS duration decreased 23 msec (95% CI, −38 to −6 msec) in the PSM-CRT group and increased 1 msec (95% CI, −4 to +8 msec) in the PSM-control group at 6 months (P<0.001). Of note, there was a disproportionate decline in PSM controls over time, but a significant increase in absolute QRS duration from baseline over time. There was no mention of Z-score change over time. See Figure Below.
Longitudinal follow-up showed a significant increase in median systemic ventricular EF at 6 months, 11% (95% CI, +0.5% to +21%) in the PSM-CRT group compared to 0.1% (95% CI, −9.8% to +3.2%) in the PSM-control group (P<0.001). Of note, there was a disproportionate decline in PSM controls over time, but a significant increase in absolute and change from baseline EF over time.
Device complications were noted in 5% (acute) and 31% (chronic) with lead issues as the most common complication. There was no mortality with CRT procedures. See Table below.
Device therapy was noted in both groups. Appropriate shocks for VF/VT was seen in 4 patients in the CRT-D group and 2 in the control group with ICDs. Inappropriate shocks were seen in 2 patients in CRT-D group and none in the control group with ICDs.
This study demonstrates the improved transplant-free survival for pediatric patients with and without congenital heart disease who received CRT for symptomatic heart failure (AHA Stage C or D and systemic EF < 45%) with electrical dyssynchrony (QRS duration z-score ≥ 3 or ventricular pacing ≥ 40%). The design of this study allowed for a more robust retrospective analysis of CRT therapy in pediatric patients than prior studies using propensity score matching and providing a longer term of follow-up with a median of 2.7 years
Overall, there was improvement in ventricular function seen at 6 months and throughout follow-up to 5 years. However, a limitation of this study was the assessment of ventricular function with LV EF extrapolated from shortening fraction when EF by Simpson method was not available. In addition, the RV EF was estimated from fractional area change for the systemic RV. In addition, there was attrition during follow-up limiting data over time. Of further interest would be the assessment of CRT on longitudinal function (EF) in the univentricular heart which was not associated with survival benefit and in the systemic right ventricle which was associated with survival benefit in this cohort.
Heart failure functional status also plays a key role due to its impact on management decisions and survival outcomes. A lower function status at baseline was associated with a reduced risk for death or transplantation in this study. In addition, the presence of CRT was associated with improved survival to first heart failure hospitalization or transplant listing. Due to sample size and objectives, subgroup analysis was not performed to assess risk factors for heart failure hospitalization or transplant listing.
QRS duration has been an important predictor of response to CRT in adults without congenital heart disease. All patients in this study had significant conduction disease and CRT resulted in a significant decrease in QRS duration. However, it appears that the paced QRS duration remained >120 msec in the majority potentially impacting the degree of response. The varied response to CRT is multifactorial based on variations in bundle morphology (i.e. left vs right ventricular dyssynchrony), anatomy, lead locations, and optimization protocols. In this study, the location of multisite ventricular pacing positions was not standardized or analyzed. In addition, the optimization prior to 2016 was based on echocardiographic assessments thus resulting in limited follow-up using electrocardiographic optimization. Despite these variations, CRT showed improvement in overall transplant-free survival.
Overall, this study showed that 81% of those who had CRT were alive without heart transplantation compared to the 42% in the control group. This association between CRT and transplant-free survival remained significant when controlling for confounders. However, due to the heterogeneity of patients and approaches in this study, not all groups may demonstrate similar findings in a generalizable population.
This study provides additional data regarding the potential benefits of CRT in pediatric heart failure with and without congenital heart disease. However, more data is needed with homogenous congenital populations, standardized approaches, and longer term follow-up to determine those who will respond to CRT that can be used to ultimately refine existing guidelines.
Management of heart failure in pediatric patients is important to reduce mortality, transplantations, hospitalizations, and comorbidities. As mentioned in this study, a multidisciplinary approach is needed when determining who and how one should receive CRT therapy. In addition, medical and device optimization management will continue to require an individualized approach.