Consultant Opinion #1

Anica Bulic, MD

Anne M. Dubin, MD

This is an interesting case that highlights several important arrhythmia and device management issues in a patient with ccTGA (l-TGA). The first issue to discuss, and one that perhaps dictated this patient’s outcome, is the timing of the double switch procedure. These patients are vulnerable to LV deconditioning and are at a high risk of RV failure and TR over time. LV “retraining” is performed by PAB in an attempt to confer a pressure load on the LV and stimulate LV hypertrophy to be able to contract against systemic vascular resistance. Our institution has published data to support an LV retraining program in patients with l-TGA with excellent midterm results. The criteria used to assess LV readiness following PAB for a double switch procedure include LV pressure 90% of systemic pressure; normal LV systolic function with an LVEF greater than 55%; LV end diastolic pressure less than 12 mmHg; mild or less mitral valve insufficiency; and LV mass by MRI greater than 50 g/m ² in children and greater than 65 g/m ² in adults. Careful monitoring of LV size and systolic function both intraoperatively and postoperatively is crucial as there is some evidence to suggest that excessive tightening can cause subendocardial ischemia and potentially myocardial fibrosis in the long term. Five-year follow-up showed that all patients demonstrated that their LVs could retrain in response to PAB and that the majority of those patients who had undergone a double switch procedure had normal LV systolic function. The influence of age on the success of LV retraining is controversial, but some evidence suggests that the older the patient, the less the likelihood of success. One could argue that this patient was not a good candidate for a double switch procedure as LV systolic dysfunction after banding is a risk factor for death following a double switch procedure.

The second issue that warrants discussion is the appropriateness of biventricular pacing, or cardiac resynchronization therapy (CRT), in a patient with repaired two-ventricle congenital heart disease. There are no randomized controlled trials in the pediatric or adult congenital heart disease (ACHD) populations, and the criteria for CRT are extrapolated from the adult literature. In adult heart failure patients with an NYHA class II–IV despite optimal medical management, LV systolic dysfunction (LVEF 35% or less), and a left bundle branch block (LBBB) with QRS 150 ms or wider, CRT has been shown to improve cardiac output, heart failure symptoms, exercise tolerance, and survival. Pediatric and ACHD CRT studies are limited but do support the hemodynamic and functional benefits seen in the adult population. In the largest multicenter pediatric study of 103 patients who underwent CRT, Dubin and colleagues showed that CRT decreased QRS duration by approximately 38 ms from a baseline of 166 ms, and improved EF from approximately a mean of 26% to a mean of 40%. Subsequent studies have suggested that the benefits of CRT in congenital heart disease may depend on the underlying anatomy of the systemic ventricle, the presence and degree of systemic atrioventricular valvar regurgitation, ventricular myocardial scarring, and the type of electrical dyssynchrony. Janousek and colleagues have reported that patients with a systemic RV, systemic atrioventricular valvar regurgitation, and a poor initial NYHA class responded less favorably to CRT. With regards to the efficacy of CRT in heart failure patients with a narrow QRS, there are studies to suggest its harmful effects. The LESSER-EARTH trial showed that in symptomatic adult heart failure patients with an LVEF of 35% or less and a QRS of 120 ms or less, CRT did not improve exercise tolerance, LV dimensions, or EF. In fact, CRT patients had a reduction in the 6-min walk test, an increase in QRS duration, and showed a trend toward heart failure–related hospitalizations. According to a published PACES/HRS expert consensus statement on the management of arrhythmias in ACHD patients, this patient with LV systolic failure, but a narrow QRS, would not meet criteria for CRT implantation.

Little is known about the risk factors for sudden cardiac death (SCD) in patients with l-TGA. The few published studies to date suggest that unrepaired l-TGA with a failing systemic RV and associated cardiac lesions are associated with a higher risk of SCD. Koyak and colleagues in their multicenter case-control study investigating the risk factors for SCD in ACHD reported that out of 12 l-TGA patients with SCD, 9 had associated defects such as a ventricular septal defect (VSD), pulmonary stenosis, and Ebstein malformation of the tricuspid valve. Heart failure symptoms, a history of atrial flutter or fibrillation, and impaired systemic ventricular function conferred a higher risk of SCD. Of the combined l-TGA and d-TGA cohort following surgical repair, 90% of those who had SCD had a systemic RV. One plausible explanation for this is impaired myocardial perfusion to a hypertrophied RV, particularly the anterior and inferior walls, as well as the interventricular septum. It remains to be seen whether the timing and type of surgical repair, a history of ventricular arrhythmias, or results from programmed ventricular stimulation can further stratify SCD risk. According to the PACES/HRS guidelines, ACHD patients with biventricular physiology, NYHA class II or III symptoms, and a systemic ventricular EF 35% or less represent a class I indication for implantable cardioverter-defibrillator (ICD) implantation, such as in this patient.

ICD implantation in l-TGA patients involves certain technical aspects that are unique to this patient population. Transvenous access through the atrial baffles can prove to be challenging, especially in the setting of superior vena cava (SVC) baffle stenosis, as in the case of this patient. Thoracotomy or hybrid procedures are used in as many as 2/3 of the cases. Subcutaneous ICDs have recently been shown to be a safe and viable alternative in ACHD patients and would be a possibility in this patient. Moore and colleagues published their retrospective multicenter studies of subcutaneous ICD implantation in ACHD patients, involving 21 patients over a 5-year period, half of whom had single ventricle physiology and only one of whom had l-TGA. Ventricular arrhythmia was induced in approximately 80% and all were rescued with 80 J or less. There was one case of device infection, which did not result in device removal. The benefits of avoiding transvenous access come at the expense of a high rate of inappropriate discharges (21%), which are from SVT, T wave oversensing, and in one case, low amplitude artifact from subcutaneous air. As in most clinical cases where there is a paucity of evidence-based literature, the decision to implant an ICD should be made on a case-by-case basis, with careful consideration of the risks and benefits for each particular patient.

Consultant Opinion #2

Frank Cecchin, MD

This is a complex case that evolved over a very long period of time. The first question I would like to address is What is the rationale for a double switch in ccTGA?

ccTGA is the term used to define the cardiac malformation associated with atrioventricular and ventriculoarterial discordance. This double discordance results in a physiologically balanced circulation. It can be an isolated anomaly (20%), although it is often associated with other anatomic anomalies; the most common of which is a VSD in 60%–80%, pulmonary stenosis in 50%, and abnormalities of the conduction system in 15%–50% of ccTGA patients. The classic surgical approach has been to repair the associated lesions and is termed a “physiologic repair,” which leaves the tricuspid valve and RV as part of the systemic circulation. However, a long-term result of this approach is progressive TR and RV failure. This led to the introduction of a true “anatomic repair” in which the LV and mitral valve support the systemic circulation. The true anatomic repair is accomplished by either a venous and arterial switch operation (double switch) or by a venous switch operation and a Rastelli procedure (RV to pulmonary artery [PA] conduit with an overriding VSD). If pulmonary stenosis or a large VSD are not present, then the LV will become deconditioned soon after birth and will need to be retrained as a ventricle able to handle the systemic circulation. The placement of a PA band will provide a high-resistive load for the LV to regain its ability to be a systemic ventricle.

The case under review involved an individual with ccTGA and no associated lesion who was diagnosed at 1 year of age. My recommendation today would be to start LV retraining prior to the development of RV failure .

It remains controversial as to whether patients who are born with ccTGA, an absence of a ventricular septal defect, and an absence of pulmonary stenosis should undergo prophylactic PAB. The reason is that patients in this subset with good tricuspid valve function have the most favorable natural history, and there is insufficient data to compare this natural history with an anatomic repair pathway. The impetus to proceed with placement of a PAB is based on the observation that the prognosis for patients undergoing a double switch is more favorable in patients who never experienced detraining compared with patients who require retraining. A group in France reported their experience with an aggressive surgical management protocol for isolated ccTGA. They reported the results of 11 infants in whom primary PAB was performed. There was no hospital mortality and all the children maintained normal RV function; tricuspid valve function was stabilized or improved, and systemic competence of the LV was maintained. They concluded that neonates with isolated ccTGA, “prophylactic PA banding is safe and carries a low morbidity.” An additional case series from Boston looked at 25 patients with ccTGA who underwent PAB for LV preparation prior to anatomic repair. For the 18 patients who underwent anatomic repair at a median of 10 months from PAB, LV dysfunction developed in 4 of 7 patients repaired after age 3 years compared with 0 of 11 repaired before 3 years. The caveat is that LV retraining should occur as early in age as possible to preserve long-term LV performance.

The clinicians in this situation chose to band the PA at 6 years of age because severe tricuspid (systemic atrioventricular) regurgitation had developed. This was done in the hope that improvement in TR would occur and the child could undergo a double switch procedure. The absence of pulmonary stenosis meant that the LV would need preparation prior to a double switch. The PA banding procedure in ccTGA can be used for two purposes. One is preparation of the LV for anatomic repair. This concept was initially described by Yacoub in 1977 and Mee in 1986 as a means to retrain the morphologic LV in patients with dTGA to allow a late arterial switch operation for patients who had previously undergone an atrial level switch. The second purpose is as a palliative procedure aimed at improving RV function by reducing TR. This reduction in TR occurs by restoring normal geometric relationships between the RV, interventricular septum, tricuspid valve, and subvalvar structures. The PA band achieves this by elevating the LV pressure which in turn reduces septal shift toward the LV. This provides better tricuspid valve coaptation, reducing RV size, systemic heart failure, and preventing further tricuspid annular dilation.

In this patient the PA band procedure was performed in two stages over a 3-month period. At this point, systolic pressures in the subpulmonic LV had risen to about 90% of systemic pressures, and due to the shift of interventricular septum and improved tricuspid valve geometry, the degree of TR had improved from severe to mild. Although LVEF was noted to be mildly impaired (45%–50%) a double switch operation was performed. The postoperative course was complicated by rapid, progressive systemic LV failure, requiring escalation of medical therapy.

That brings up the third question: Why do patients develop LV dysfunction after a true anatomic repair?

The early- and intermediate term results of the double switch operation have been favorable, although, there have been more recent concerns regarding the long-term function of the retrained LV. Brawn and colleagues published their late results in 44 double switch subjects, comparing those who required LV retraining (n = 11) and those who did not (n = 33). The 30- day hospital mortality was (4.5%) and the long-term raw mortality was (11%). The rate of death and transplantation combined was (16%). The actuarial freedom from death, transplantation, or the development of moderate-to-severe LV dysfunction was 85% at 1 year, 80% at 5 years, and 72% at 10 years. The incidence of moderate-to-severe LV dysfunction at 1 year following double switch was 39% in subjects requiring LV retraining, compared with 6% of subjects not requiring retraining. Thus, retraining an LV increases the risk of late LV dysfunction.

Preparing an LV to handle a systemic pressure load requires a delicate balance between providing enough resistance to develop hypertrophy without inducing ischemia and fibrosis. Animal models have demonstrated that retrained ventricles display subendocardial edema, myocardial necrosis, and fibrosis with reduced ventricular work index. In some patients who do not tolerate a tight band, it may be necessary to progressively tighten the band over time to avoid damaging the LV. The key to successful retraining is stepwise training and rigorous testing of the retrained ventricle to make sure it develops “clean hypertrophy” without significant fibrosis. The Stanford cardiovascular surgical team has successfully retrained the LV in 24 ccTGA patients, with 18 undergoing a double switch procedure using the following protocol.

Assessment of LV preparedness for a double switch:

• 1.
  

  LV pressure 90% of systemic pressure

  
  
• 2.
  

  LV systolic function EF >55%

  
  
• 3.
  

  LV end diastolic pressure less than 12 mmHg

  
  
• 4.
  

  Mitral valve function mild or less insufficiency

  
  
• 5.
  

  LV mass (by MRI) > 50 g/m ² (in children); >65 g/m ² (in adults)

Using the above criteria, the case study patient would not have met criteria for double switch. Either more time or a third PAB would have been needed prior to moving onto the double switch.

Then because of persistent LV dysfunction despite a narrow QRS width of 96 ms, the patient underwent implantation of biventricular epicardial pacemaker (CRT-P) in 2005 at the age of 9 years.

Was it a mistake to place a biventricular pacemaker for LV dysfunction with a narrow QRS complex? The answer to this question is YES it was a mistake to perform narrow QRS pacing, although this fact was not known at the time of the resynchronization procedure. A consensus conference in 2014 published recommendations for ACHD arrhythmia management and listed the following as a class III CRT indication:

Class III: CRT is not indicated in adults with congenital heart disease (CHD) and a narrow QRS complex (<120 ms).

CRT is a powerful tool to counter mechanical and electrical dyssynchrony associated with worsening ventricular performance. CRT was first developed to relieve electrical dyssynchrony in the setting of LBBB and LV dysfunction. The best results are achieved in those with classic LBBB pattern and widest QRS duration. The technique was expanded to those in heart block and eventually the concept of mechanical dyssynchrony came to be understood. Mechanical dyssynchrony is defined as actual discoordinated contraction of ventricular segments versus pure electrical delay as defined by prolonged QRS duration. Successful CRT occurs when optimal LV lead position is achieved, as defined as the viable myocardial region with the latest contraction onset. Thus, narrow QRS CRT is rarely going to meet these criteria because electrical delay is nearly always coupled with mechanical dyssynchrony. When there is a mismatch between electrical and mechanical dyssynchrony, typically the myocardium associated with mechanical dyssynchrony is fibrotic.

In 2008, investigators from 115 centers attempted to investigate the effect of CRT on morbidity and mortality among patients with symptomatic heart failure, a narrow QRS complex, and echocardiographic evidence of LV dyssynchrony. In 2013, after recruiting and following 809 patients for 19 months, the study called Echocardiography Guided Cardiac Resynchronization Therapy (EchoCRT) was stopped for futility on the recommendation of the data and safety monitoring board. There were 45 deaths in the CRT group and 26 in the control group (11.1% vs. 6.4%; hazard ratio 1.81; 95% CI 1.11–2.93; P = .02). The investigators concluded in the NEJM publication “that in patients with systolic heart failure and a QRS duration of less than 130 ms, CRT does not reduce the rate of death or hospitalization for heart failure and may increase mortality.”

Then in 2015, Sohaib et al. identified all trials comparing CRT with no CRT, which reported Kaplan-Meier curves in groups defined by QRS duration: narrow, non-LBBB broad, and LBBB broad. For each trial, the change in life span every 3 months was calculated. Four trials (MADIT-CRT [Multicenter Automatic Defibrillator Implantation Trial-Cardiac Resynchronization Therapy], RAFT [Resynchronization-Defibrillation for Ambulatory Heart Failure Trial], REVERSE [Resynchronization reVErses Remodeling in Systolic left vEntricular dysfunction], and EchoCRT [Echocardiography Guided Cardiac Resynchronization Therapy]), totaling 4717 patients, reported curves for mortality or heart failure–related hospitalization. In patients with LBBB broad QRS duration (within MADIT-CRT), life span gain increased in proportion to time. In contrast, in patients with non-LBBB broad QRS (within MADIT-CRT) and patients with narrow QRS (EchoCRT), life span was lost in proportion to time. The nonlinear growth of life span gained when a CRT device is implanted in patients with LBBB broad QRS is unfortunately mirrored by a similarly progressive loss in life span in narrow QRS heart failure. This suggests the culprit is a progressive physiological effect of pacing rather than implant complications. They suggested that a randomized controlled trial of deactivating CRT in patients with narrow QRS is needed, with a primary endpoint of increasing survival.

The patient in this study never demonstrated any gain in ventricular performance with narrow QRS CRT and ultimately improved when CRT was deactivated. Thus, it is likely that myocardial scar developed during PAB placement for LV retraining and that pacing causes additional electrical delay without adequate offsetting improvement in mechanical synchrony. If narrow QRS CRT was to be reconsidered in this patient, then advanced myocardial imaging techniques to guide LV lead placement position, such as tissue Doppler imaging, 2D and 3D speckle tracking echocardiography, cardiac magnetic resonance imaging, and nuclear imaging could be utilized. Myocardial viability at the optimal LV lead position is a key factor in enhancing CRT response rate. The quantification of scar burden in LV dyssynchrony is essential in the evaluation of CRT candidates because extensive scar burden can negatively impact LV functional outcomes. Recent studies show that preserved viability in the LV lead segment is related to greater LV reverse remodeling and functional benefit.

The final question is Does the patient now need a defibrillator? I would argue that this is the exact type of patient that would benefit from ICD therapy.

In the 2014 ACHD consensus statement this patient would qualify as a class I indication “ICD therapy is indicated in adults with CHD and a systemic LVEF <35%, biventricular physiology, and New York Heart Association (NYHA) class II or III symptoms.” A 2017 study by Moore et al. looked at the incidence of SCD in ACHD and found that ccTGA was one of the top four anatomic diagnosis with the highest rate of SCD. Eisenmenger syndrome was the leading diagnosis at 4.8 SCD/1000 patient years, and then ccTGA, Fontan, and dTGA ranged from 2.0 to 2.4. In this study and others, atrial arrhythmias have been found to be associated with SCD in ACHD. Based on the failed LV retraining via PAB placement, I would assume that significant fibrosis is present. This would be the substrate for malignant ventricular arrhythmias when the EF is less than 35%.

Implant considerations always play a factor in deciding whether ICD therapy is appropriate for an adult with ACHD. However, there are many options available to overcome the common technical challenges of ACHD ICD therapy, such as poor venous access, high DFT’s, valve regurgitation, and cardiac malposition. The implant choices include joint cardiac interventional procedure (SVC stent placement), hybrid procedure (use epicardial pace sense leads and add transvenous and or subcutaneous coils), epicardial system (add a combination of pericardial, pleural, or subcutaneous coils), or a total subcutaneous system. My choice in this patient would be a conventional joint interventional procedure via placement of an SVC stent, transvenous ICD lead, and atrial leads. If atrial lead placement was difficult then the atrial epicardial lead could be used. Having atrial pacing and atrial antitachycardia pacing provides a distinct advantage over a total subcutaneous device.

Recommendation: Proceed with ICD implantation in conjunction with cardiac interventional team SVC stent placement, a new transvenous ICD lead, and either a new transvenous atrial lead or reuse of the chronic epicardial atrial lead. I would do atrial and ventricular programmed stimulation at the time of implant to customize antitachycardia pacing.