Early and Late Effects of Cardiac Resynchronization Therapy in Adult Congenital Heart Disease.

Yin Y, Dimopoulos K, Shimada E, Lascelles K, Griffiths S, Wong T, Gatzoulis MA, Babu-Narayan SV, Li W.

J Am Heart Assoc. 2019 Nov 5;8(21):e012744. doi: 10.1161/JAHA.119.012744. Epub 2019 Oct 28. Erratum in: J Am Heart Assoc. 2019 Dec 17;8(24):e014507.

PMID:  31657270

Free PMC Article

Similar articles


Take Home Points:

  • Cardiac resynchronization therapy (CRT) appears to provide early and sustained late positive impact on measures related to heart failure in carefully selected patients with adult congenital heart disease (ACHD).
  • While findings similar to conventional CRT benefits in ischemic/non-ischemic cardiomyopathy can be achieved in ACHD patients with systemic morphologic LV’s, the same does not appear to be true in ACHD with systemic morphologic RV’s.


Comment from Dr. Philip Chang (Gainesville, FL), section editor of Congenital Electrophysiology Journal Watch:  This study by Yin et al retrospectively reviewed outcomes in ACHD patients referred for CRT at a single center between 2004-2017.  Factors considered in the application of CRT included reduced LVEF or RV fractional area change (FAC), NYHA functional class, QRS duration, and single-site ventricular pacing burden.  General demographic, cardiac, cardiac surgical, and clinical data; device implant-related variables; cardiothoracic ratio on x-ray, echo and surface ECG data; and follow-up assessment and outcomes variables were collected by chart review.  The authors defined bundle branch block as QRS duration ≥120 ms and positive CRT response as ≥5% increase in LVEF or RVFAC following CRT initiation.

A total of 54 patients with comprehensive data sets and follow-up duration were included for analysis.  Patient age varied widely (18-73 years, mean 46 years) and most patients were male (74%).  Mean follow-up duration from CRT was 5.7 years, with early and late follow-up arbitrarily delineated as mean 1.8±0.8 years and 4.7±0.8 years, respectively, after CRT device implant.  A majority of patients had systemic morphologic LV’s (39 patients, 72%) comprised of several forms of CHD while the remaining patients with systemic morphologic RV’s had either D- or L-TGA variants (the majority being L-TGA, 13/15 patients).  Of the entire cohort, 33/54 patients met a primary ventricular pacing indication prior to CRT, either due to high-grade AV block or following AV node ablation.  Over half of the cohort had preexisting cardiac implantable devices (21 pacemakers, 10 ICD’s).  In total, 46/54 patients were implanted with a CRT-D device with the remaining 8 patients implanted with CRT-P systems.  The vast majority of patients received transvenous systems only (52/54); 2 patients underwent hybrid transvenous and surgical epicardial lead placement.  19% of patients experienced complications including infection (5 patients, 9%) and lead dislodgement (3 patients, 6%).

Overall, a positive response to CRT was noted in 65% of the entire cohort.  This was primarily driven by favorable early and late responses in ACHD patients with systemic LV’s.  Figure 1 from the manuscript depicts the changes in CT ratio (B), QRS duration (C), NYHA class (D), systemic LVEF (E), and systemic RVFAC (F) over the time points of the analysis.  In general, while there was overall improvement across all measures of CRT response at early follow-up, only NYHA class improvement and LV functional measures were sustained at late follow-up.  Among systemic LV recipients, most (74%) had improvement in LVEF that was sustained out to nearly 2 years following CRT.  Systemic LV patients with history of LVOT obstructive CHD showed the highest response (15/17 patients, 88%).  Only 40% of systemic RV patients exhibited early-term positive response.  All-cause mortality occurred in 20% of patients during follow-up.  Multi-variate analysis showed QRS duration to be the only independent predictor of CRT response irrespective of permanent pacing or QRS morphology.  Kaplan Meier analysis shows somewhat contrary data with systemic RV patients actually exhibiting a longer duration of freedom from death or transplant compared to systemic LV patients following CRT.  No comparison to non-CRT cohorts was made.



Though there are substantial limitations to this study, it provides some data and insights into longer-term outcomes in ACHD patients receiving CRT, which is still insufficiently understood given low cohort sizes across multiple studies.  The rather high complication rate following CRT device implants reported in this study likely reflects the contributions of device-related re-interventions in patients with pre-existing cardiac implantable devices and other comorbidities, which are recognized to increase the risk of infections, as well as the technical challenges of CRT implant in CHD, even among patients with systemic LV’s.  From studies and trials on CRT in ischemic and non-ischemic cardiomyopathy in baseline structurally normal hearts, positive CRT response is linked to LBBB-type wide QRS, wider QRS duration (paced or non-paced), and defined evidence of LV dyssynchrony.  Favorable response to CRT can generally be expected when application of biventricular or even multi-site ventricular pacing in these scenarios effectively targets the area of latest electrical activation.  Positive CRT response has also been associated with “ideal” target vessel lead implantation (usually lateral/free wall LV in a basilar/mid-LV location) as well as very high percentage biventricular pacing (>98%).  This study failed to mention consistent achievement of these variables, which certainly could affect CRT response.  With that said, the findings from this study are consistent with the notion that systemic LV ACHD appears to exhibit a similar response to CRT compared to conventional ischemic and nonischemic cardiomyopathy substrates with similar heart failure indices and variables.  The lower CRT response in systemic RV’s may reflect implant variables and lead placement along with the unique morphologic, structural, contractile, and electrophysiologic (i.e.. single RBB vs. left anterior and posterior fascicles in LBB) differences between RV’s and LV’s.  Ventricular inter-dependence was not mentioned in this study, which could potentially be of impact in patients where the subpulmonary ventricle also exhibits significant dysfunction (as in the case of tetralogy of Fallot patients with biventricular dysfunction).  Wider baseline QRS as an independent predictor of positive CRT response can likely be understood by viewing wider QRS duration as reflecting greater ventricular electrical delay and dyssynchrony, which should stand to improve if the region of greatest delay is properly resynchronized.  As the authors mentioned, there remains a lack of solid, consensus guidelines regarding patient selection and consistent clinical variables to guide CRT implant among the heterogeneous ACHD population.