Author + information
- Received January 4, 2018
- Revision received June 25, 2018
- Accepted July 6, 2018
- Published online October 15, 2018.
- Edward T. O’Leary, MDa,
- Kimberlee Gauvreau, ScDa,
- Mark E. Alexander, MDa,
- Puja Banka, MDa,
- Vassilios J. Bezzerides, MDa,
- Francis Fynn-Thompson, MDb,
- John K. Triedman, MDa,
- Edward P. Walsh, MDa and
- Douglas Y. Mah, MDa,∗ ()
- aDepartment of Cardiology, Boston Children’s Hospital, Boston, Massachusetts
- bDepartment of Cardiac Surgery, Boston Children’s Hospital, Boston, Massachusetts
- ↵∗Address for correspondence:
Dr. Douglas Y. Mah, Department of Cardiology, Boston Children’s Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115.
Objectives This study sought to determine the long-term effect of dual-site ventricular pacing (i.e., cardiac resynchronization therapy [CRT]) versus single-site pacing in patients with Fontan physiology and high-grade atrioventricular block (AVB).
Background Chronic single-site ventricular pacing in Fontan patients is associated with significant morbidity and mortality.
Methods The study conducted a retrospective review of all Fontan patients with high-grade AVB who received pacemakers at a single institution between 1990 and 2016 with follow-up of 12 months or greater. The primary study outcome was a composite of death or heart transplantation.
Results Nineteen patients received CRT devices and 43 patients received single-site ventricular pacemakers (SVPMs), with a median follow-up of 4.9 years (range 1.1 to 11.3 years) and 10.4 years (range 1.1 to 26.8 years), respectively. One (5.3%) CRT patient and 11 (25.6%) SVPM patients reached the composite endpoint of death or transplant (odds ratio: 0.16; 95% confidence interval: 0.02 to 1.36; p = 0.09). Kaplan-Meier analysis showed similar rates of freedom from death or transplant between the 2 study groups over a 5-year follow-up period (p = 0.08). The proportion of SVPM patients with abnormal ventricular systolic function before device implantation (9.5%) and at follow-up (33.3%) significantly increased (p=0.009). The CRT cohort had less change in their ventricular function (42.1% before device implant, 36.8% at follow-up; p = 1.00); however, a Kaplan-Meier analysis over a 5-year follow-up period found no difference in the presence of abnormal ventricular systolic function between groups (p = 0.27).
Conclusions There was no significant difference in long-term outcomes between Fontan patients who were single- or dual-site paced for high-grade AVB.
Cardiac resynchronization therapy (CRT) is a well-established treatment for adult patients with refractory systolic heart failure and evidence of ventricular dyssynchrony (1,2). The evidence behind CRT, or dual-site ventricular pacing, in the pediatric and congenital population, while limited to retrospective studies, suggests similar benefits when looking at heterogeneous populations of congenital heart disease (CHD) patients (3). Less is known about the effects of dual-site ventricular pacing in CHD patients undergoing single-ventricle palliation to the Fontan circulation, although subgroup analyses from prior studies have demonstrated a positive acute and short-term response to CRT (4–6).
Despite improvements in clinical outcomes over the past several decades, Fontan patients remain at risk for numerous complications and considerable morbidity (7). High-grade second- and third-degree atrioventricular block (AVB) necessitating permanent pacemaker implantation can be a sequela of direct surgical trauma or developmental abnormalities in the conduction systems of certain CHD lesions (8–10). Associations between long-term single-site ventricular pacing and worsened clinical outcomes have been documented in adults, and more recently, Bulic et al. (11) and others (12,13) have highlighted its detrimental effect on patients with single ventricle physiology. Although dual-site ventricular pacing has shown to be beneficial in subtypes of CHD complicated by high-grade AVB, it remains unclear if its benefits are consistently seen in single-ventricle patients. The objective of our study, therefore, was to examine the impact of dual-site pacing in single-ventricle Fontan patients in comparison to those with single-site pacing for high-grade AVB.
Patient records at Boston Children's Hospital were queried from 1990 to 2016. Patients with single ventricles palliated to Fontans who had pacemakers for high-grade second- or third-degree AVB were included. Patients who underwent dual-site (CRT) device implantation were compared with those with single-site ventricular pacemakers (SVPMs). Patients were considered to have received a CRT device if dual-site ventricular epicardial leads were placed on the ventricle with the goal of providing resynchronization to the single, systemic ventricle; additional ventricular leads placed prophylactically for future backup dual-chamber pacing were not considered CRT. Patients were excluded if they had follow-up for <12 months. The study protocol was approved by the Committee on Clinical Investigation at Boston Children's Hospital.
Standard pacemaker and CRT implantation procedures
Historically at our institution, epicardial pacing systems were placed via a midline sternotomy or thoracotomy approach at the discretion of the operating surgeon. For CRT implants, the ventricular leads were ideally placed in an anteroposterior orientation on the single ventricle at the midventricular level, if possible, based on surgical exposure. The optimal balance of lead placement and excessive dissection through adhesions or abnormal anatomy was determined by the operating surgeon. Representative examples of an anteroposterior lead configuration are shown in Figure 1, along with their respective electrocardiograms (ECGs). Pre- and post-CRT echocardiograms are included in Online Videos 1, 2, 3, and 4. After implantation of a CRT device, ventricular lead optimization was performed before discharge, and then at the discretion of the primary electrophysiologist in the outpatient setting using 12-lead ECG. The programmed delay in pacing between the 2 ventricular leads was varied, with the optimal setting defined as the one that provided the shortest QRS duration (14).
The primary outcome variable was a composite of death or heart transplantation (death or transplant). The main secondary outcome variables were the presence of abnormal ventricular systolic function and significant atrioventricular valve regurgitation (AVVR) at time of follow-up. The primary predictor variable was pacemaker device type (CRT vs. SVPM). Secondary predictor variables included systemic ventricular morphology (i.e., single left vs. right ventricle), abnormal ventricular systolic function (defined as qualitatively “moderate” or “severe” dysfunction) before device implantation, significant AVVR (defined as qualitatively “moderate” or “severe” regurgitation) before device implantation, and, within the CRT cohort, whether leads were placed in the planned anterior-posterior orientation—assessed by reviewing the location of the ventricular leads on a posteroanterior and lateral chest x-ray performed after CRT implantation.
Baseline demographics, congenital heart lesion, percent of time with ventricular pacing at last device interrogation, perioperative complications, and medical comorbidities were analyzed as potential confounders. Perioperative complications were defined as a history of requiring (during any surgical procedure) extracorporeal membrane oxygenation or cardiopulmonary resuscitation, or history of coronary arterial injury. Medical comorbidities were defined as renal insufficiency (serum creatinine level above laboratory upper limit of normal) at time of follow-up, sepsis requiring vasoactive medications during any hospitalization, history of myocarditis (as documented in the medical record), and history of arrhythmias requiring antiarrhythmic medications or ablation.
Ventricular systolic function was assessed by reviewing the echocardiogram before device implantation and again at the most recent follow-up date after device implantation. If patients received a heart transplantation or died, the last echocardiogram performed before transplantation or death was used as the most recent echocardiogram. Qualitative echocardiographic measurements of normal, mild, moderate, or severe dysfunction were used. Quantitative measurements of ventricular systolic function were not routinely calculated by our institution's echocardiography laboratory in single-ventricle patients due to limitations from their complex ventricular anatomy (15). In an effort to both capture clinically meaningful changes in ventricular systolic function, echocardiograms were then grouped as “normal” for normal systolic function or mild dysfunction and “abnormal” for moderate or severe dysfunction. Using these definitions, ventricular systolic function before device implantation was compared with that at most recent follow-up, with patients then categorized as having an improvement, no change, or worsening of their function. Serial echocardiograms were reviewed at 0 to 6 months, 6 to 12 months, 1 to 4 years, and ≥4 years from the date of device implantation. Echocardiograms obtained during the implantation hospitalization were not included. When multiple echocardiograms were available, the echocardiogram closest to the upper limit of the respective time interval was submitted for analysis.
Continuous variables were summarized as median and range or mean ± SD, and categorical variables as number and percentage. The Fisher exact test and univariate logistic regression analysis were used to compare binary predictor and outcome variables. Paired and unpaired t tests were used to compare continuous predictor and outcome variables as appropriate. All p values ≤0.05 were deemed statistically significant. Kaplan-Meier analyses were performed for the outcomes of death or transplant and the presence of abnormal ventricular systolic function at follow-up, with follow-up capped at 5 years after the onset of pacing in both analyses to allow for equivalent follow-up between groups.
From 1990 to 2016, 105 Fontan patients required ventricular pacing for high-grade AVB at our institution. Forty-three patients were excluded for having a follow-up period of <12 months. Of the remaining 62 patients, 43 received SVPMs and 19 received CRT devices. There were no significant differences between the groups with regard to baseline characteristics, segmental anatomy, systemic ventricular morphology, percent of time ventricular pacing, or prevalence of perioperative complications or medical comorbidities (Table 1). Table 2 outlines the location of the pacing leads in patients with CRT devices, the surgical approach to placing the leads, and the QRS morphology and duration after dual-site pacing was instituted. A significantly higher proportion of CRT patients had abnormal ventricular systolic function before device implant (p = 0.006) (Table 3). The median duration of cross-sectional follow-up was 4.9 (range 1.1 to 11.3) years for the CRT group and 10.4 (range 1.1 to 26.8) years for the SVPM group.
One (5.3%) CRT patient and 11 (25.6%) SVPM patients reached the composite endpoint of death or transplant. Of the 11 SVPM patients that died or underwent heart transplantation, 3 received a heart transplant, 2 died secondary to decompensated heart failure, and 1 died secondary to complications from Ewing sarcoma; the causes of the remaining deaths were not documented in the medical record. There were no deaths in the CRT group, with 1 patient receiving a heart transplant for medically refractory heart failure. Ventricular morphology, ventricular systolic function and AVVR pre-device implant, and lead location in the anteroposterior configuration (in CRT patients) were not associated with a higher frequency of death or transplant (Table 3).
Evaluating outcomes based on time since the onset of pacing, there was no difference in the freedom from death or transplant between groups (Figure 2). Although the Kaplan-Meier curve appeared to show a worse outcome for SVPM patients, this was ultimately not found to be a statistically significant difference (p = 0.08).
Ventricular systolic function
Seven (36.8%) CRT patients and 14 (33.3%) SVPM patients had abnormal ventricular systolic function at latest follow-up. Dual-site ventricular pacing was not associated with the presence of abnormal ventricular systolic function at time of cross-sectional follow-up (odds ratio [OR]: 1.17; 95% confidence interval [CI]: 0.38 to 3.60; p = 0.79) (Table 4). Evaluating the entire cohort of paced patients, patients with single left ventricles had a lower (but not statistically significant) frequency of abnormal ventricular systolic function at follow-up (OR: 0.34; 95% CI: 0.11 to 1.04; p = 0.06). Moreover, the effect of dual-site ventricular pacing was not dependent on ventricular morphology, as neither single left ventricle (OR: 1.30; 95% CI: 0.29 to 5.60; p = 0.75) nor single right ventricle (OR: 1.69; 95% CI: 0.22 to 12.80; p = 0.61) patients treated with dual-site pacing had a lower frequency of abnormal ventricular systolic function at follow-up.
The ventricular systolic function of patients at device implantation and at most recent follow-up is shown in Figure 3. In the CRT group, the proportion of patients with abnormal ventricular systolic function (predefined as moderate or severe dysfunction) did not change significantly over time (42.1% vs. 36.8%; p = 1.00), with 1 patient worsening (5%) and 2 patients improving (10%). This appeared to be unrelated to QRS duration, as there was no statistically significant difference in the mean QRS duration before CRT device implantation (140.5 ± 34.5 ms) and at time of follow-up (146.0 ± 27.8 ms; p = 0.46). In contrast to the CRT cohort, however, there was a significant increase in the proportion of SVPM patients with abnormal ventricular systolic function over time, with 9.5% having abnormal function before device implantation compared with 33.3% having abnormal function at follow-up (p=0.009). This included 10 patients (24%) who had worsening ventricular systolic function and no patients with improved function. There was no significant change in the proportion of patients with or without significant AVVR over time in the SVPM and CRT groups (p = 0.69) (Figure 4). Figure 5 shows the Kaplan-Meier curve for freedom from abnormal ventricular systolic function between groups (limited to 5 years of follow-up). There was no significant difference in the worsening of ventricular systolic dysfunction in those with SVPM and CRT (p = 0.27).
CRT, or dual-site ventricular pacing, has become an important tool in the treatment of heart failure within the adult population. Its role in the pediatric and congenital population is less well defined, with conflicting results in the literature. However, there is increasing evidence that chronic single-site pacing is detrimental to patient outcomes (11,13). This is the first study comparing the long-term outcomes of single-site versus dual-site pacing in single-ventricle Fontan patients with high-grade AVB. In this population, dual-site ventricular pacing was not associated with lower odds of death or heart transplantation. More SVPM patients had worsening of their ventricular systolic function at follow-up, whereas those with CRT had no significant change; however, there was no significant difference when comparing the 2 groups over time. Overall, our study did not clearly demonstrate a significant difference in outcomes between SVPM and CRT patients at long-term follow-up.
The negative effects of single-site ventricular pacing in our study are consistent with that published in the literature (11,12). With a median follow-up of 10 years, 26% of SVPM patients in our study died or were transplanted, and 24% developed worsening ventricular systolic function. Bulic et al. (11) recently compared Fontan patients with chronic ventricular pacing with nonpaced Fontan patients over a median follow-up of 6.6 and 7.6 years, respectively. They found that 68% of paced versus only 15% of nonpaced Fontan patients developed moderate or severe ventricular systolic dysfunction at time of follow-up.
Patients who had dual-site ventricular pacing, however, may have received some protective effect from their pacing strategy. Only 5% of patients died or were transplanted, and 5% had worsening ventricular function. This was despite the fact that a larger proportion of patients in the CRT cohort had worse baseline ventricular systolic function before pacing than in the SVPM group. Part of this protective effect may be due to the initial improvement in ventricular function after dual-site pacing is initiated, as shown by Cecchin et al. (4), whose median follow-up was 1 year. Whether dual-site ventricular pacing should be initiated from the outset for any patient with heart block is currently of debate. The BLOCK-HF (Biventricular versus Right Ventricular Pacing in Heart Failure Patients with Atrioventricular Block) trial was a prospective study that randomized patients with new heart block and left ventricular dysfunction to right ventricular or biventricular pacing, and noted a reduction in mortality and heart failure hospitalization in those with biventricular pacing (16). This is in contrast, however, to the BioPace (Biventricular Pacing for Atrioventricular Block to Prevent Cardiac Desynchronization Study) trial, whose published results are still pending, but whose results show that outcomes are no different in those with right ventricular or biventricular pacing (17,18). The difference between the 2 studies is unclear, although the BioPace trial cohort had better left ventricular function at baseline.
Further studies will be required before the use of biventricular or dual-site ventricular pacing is routinely adapted in children with single-ventricle heart disease. The desire to avoid pacing-induced dysfunction should be balanced with the complications inherent to epicardial pacing. Epicardial leads are prone to malfunction, and additional lead revisions would require another invasive open surgical procedure (19). Moreover, there is a small but present risk of myocardial and coronary artery compression from epicardial leads; as more leads are placed in younger patients, the risk of cardiac strangulation could potentially increase, especially if they are located on the posterior aspect of the heart (20).
It remains unclear why patients with single ventricles do not receive the same benefit from CRT compared with patients with biventricular circulations. Ventricular scar burden may be one potential nonmodifiable explanation for this, as this has been associated with a lower response rate to CRT (21–23). The staged palliation to the Fontan circulation usually requires 3 operations to effectively separate the systemic and pulmonary circulations, and may involve a ventriculotomy if a conduit is placed between the right ventricle and the pulmonary artery during the stage 1 surgery. Quantifying ventricular scar burden in single-ventricle patients with prior ventriculotomies using imaging technology such as cardiac MRI before pacemaker implantation may have future prognostic value. The dyssynchrony that leads to dysfunction in single ventricles may also be different than that in biventricular hearts. Advancement in imaging techniques, and a better understanding of the mechanics in single-ventricle function, may allow providers to tailor their pacing strategies for each patient. This may also allow for more novel lead placements, as demonstrated by Adelstein et al. (23), who described a case report of successful right ventricular resynchronization in a patient with hypoplastic left heart syndrome by placement of epicardial leads on the right atrium, right ventricular apex, and right ventricular free wall.
The retrospective, single tertiary center design limits its generalizability to other populations. Many patients were also lost to follow-up, which may contribute to selection bias. Additionally, the retrospective investigation spanned nearly 3 decades and the lack of standardized indications for CRT implantation likely resulted in varied practice patterns among electrophysiologists and surgeons at our institution. Advancements in surgical technique and postoperative care have also changed over the study’s timeframe; its effect may be hard to quantify, but highlight the importance of prospective multicenter trials in the future to study this unique patient population. Finally, the study period predated our current electronic medical record and digital image library, and thus precluded our ability to perform blinded reviews of echocardiograms.
Based on current methods of implantation, there is no significant difference in outcomes between Fontan patients who receive single-site versus dual-site pacing for AVB. Although there may be some indication that CRT has the potential to mitigate the negative effects of ventricular pacing, longer follow-up is required. More exact techniques for patient selection and lead positioning need to be explored to truly optimize pacing strategies in patients with single-ventricle physiology. Larger multicenter trials with increased power to detect more subtle outcome differences are warranted for this unique patient population.
COMPETENCY IN MEDICAL KNOWLEDGE: In single-ventricle patients palliated to a Fontan circulation with high-grade heart block, cardiac resynchronization does not significantly reduce mortality compared with standard single-site ventricular pacing. Single-site dual-chamber pacing in Fontan patients is associated with the development of abnormal ventricular systolic function, however, and dual-site ventricular pacing may mitigate this adverse outcome.
TRANSLATIONAL OUTLOOK: Larger, multicenter studies are necessary to properly investigate the potential clinical benefits of cardiac resynchronization therapy in the small but tenuous Fontan population. These may provide both stricter indications for resynchronization in the single-ventricle population as well as more systematic optimization protocols. More exact techniques for lead positioning need to be developed (i.e., intraoperative mapping), which may require a more personalized, patient-specific approach.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- atrioventricular block
- atrioventricular valve regurgitation
- congenital heart disease
- confidence interval
- cardiac resynchronization therapy
- odds ratio
- single-site ventricular pacemaker
- Received January 4, 2018.
- Revision received June 25, 2018.
- Accepted July 6, 2018.
- 2018 American College of Cardiology Foundation
- Motonaga K.S.,
- Dubin A.M.
- Dubin A.M.,
- Janousek J.,
- Rhee E.,
- et al.
- Janousek J.,
- Gebauer R.A.,
- Abdul-Khaliq H.,
- et al.
- Khairy P.,
- Fernandes S.M.,
- Mayer J.E.,
- et al.
- Connelly M.S.,
- Liu P.P.,
- Williams W.G.,
- Webb G.D.,
- Robertson P.,
- McLaughlin P.R.
- Dick M.,
- Norwood W.I.,
- Chipman C.,
- Castaneda A.R.
- Huhta J.C.,
- Maloney J.D.,
- Ritter D.G.,
- Ilstrup D.M.,
- Feldt R.H.
- Bulic A.,
- Zimmerman F.J.,
- Ceresnak S.R.,
- et al.
- Punn R.,
- Hanisch D.,
- Motonaga K.S.,
- Rosenthal D.N.,
- Ceresnak S.R.,
- Dubin A.M.
- Curtis A.B.,
- Worley S.J.,
- Chung E.S.,
- Li P.,
- Christman S.A.,
- St John Sutton M.
- Blanc J.J. F.R.,
- Lunati M.
- Carreras E.M.,
- Duncan W.J.,
- Djurdjev O.,
- Campbell A.I.