OUP user menu

Non-contact mapping guided cardiac resynchronization therapy for a failing systemic right ventricle

Kevin A. Michael , Gruschen R. Veldtman , John R. Paisey , Stephen Robinson , Stuart Allen , Nadia S. Sunni , Paul R. Roberts , John M. Morgan
DOI: http://dx.doi.org/10.1093/europace/eum076 880-883 First published online: 19 June 2007


Aims Progressive systemic right ventricular (sRV) dysfunction, atrial and ventricular arrhythmias and sudden cardiac death are well-recognized late sequelae of atrial redirection surgery in which the right ventricle is left connected to the systemic circulation. Although cardiac resynchronization therapy (CRT) poses an attractive therapeutic option, little is known about indications, patient selection, and technical aspects of best lead placement.

Methods and results We undertook CRT in a 27-year-old female patient post-Mustard correction for d-transposition (d-TGA) with New York Heart Association (NYHA) grade III disability with QRS duration measuring 130 ms. There was also echocardiographic (TTE) evidence of severe sRV dysfunction. Non-contact mapping (NCM) was used to define sites of late activation within the sRV and the acute intra-arterial blood pressure (BP) response was assessed during implantation of a 4 french (F) lead onto the endocardial surface of the sRV. At 4 weeks post-implant sRV lateral wall motion had improved and the ejection fraction (EF) rose from 23 to 33%. The patient has been successfully anticoagulated and improved to NYHA II status after 6 months.

Conclusion The use of NCM proved safe and effective and provided a qualitative assessment of electrical viability of the sRV complimenting the measurement of mechanical function provided by TTE. The favourable clinical response in the above case justifies a prospective evaluation of this strategy.

  • Non-contact mapping
  • cardiac resynchronization therapy (CRT)
  • heart failure
  • congenital heart disease
  • systemic right ventricle


Progressive sRV dysfunction, atrial and ventricular arrhythmias and sudden cardiac death are well-recognized late sequelae of atrial redirection surgery in which the right ventricle is left connected to the systemic circulation.1 Although CRT poses an attractive therapeutic option, little is known about indications, patient selection, and technical aspects of best lead placement. We previously attempted cardiac resynchronization/defibrillator therapy in 2 patients (both males aged 24 and 22 years) who bear resemblance to the index case. Both presented with severe sRV dysfunction and had pre-existing dual chamber pacemakers for sinus bradycardia with pacing leads implanted transvenously into the sub-pulmonary ventricles (pLV). A hybrid form of CRT was achieved in both patients by surgical placement of epicardial pace/sense leads onto the sRV guided by pacing thresholds. Pacing was optimized using standard doppler echocardiography. Interim follow-up at 6 and 12 months was disappointing. Only one patient demonstrated clinical improvement (NYHA III to II) whereas the second deteriorated. His pacemaker was converted to atrial pacing alone (AAIR) as this achieved better doppler parameters and a narrower QRS width compared to biventricular pacing (BVP). These mixed results prompted a more refined strategy to assess mechanical and electrical ventricular asynchrony and to attempt a fully percutaneous means of achieving CRT as is illustrated by the following case.

Case presentation


A 27-year-old female with a Mustard procedure for d-TGA, and had a dual chamber (DDDR) pacemaker implanted for sinus node dysfunction at age 17 years, presented with progressive effort intolerance. Transthoracic echocardiogram and angiography revealed a dilated sRV with severe systolic dysfunction (EF = 23%) and predominantly lateral wall hypokinesia. Holter monitoring revealed episodes of non-sustained ventricular tachycardia. The 12-lead electrocardiogram (ECG) revealed a fused paced/sinus rhythm QRS width of 130 ms. She achieved 4.8 min on treadmill testing per Bruce protocol compared to 8.4 min documented 10 years earlier.

Diagnostic cardiac catheterization revealed severe stenosis of the inferior baffle limb and a minor leak in the superior portion communicating with the pulmonary venous atrium. Stent angioplasty of the inferior baffle was performed to relieve the obstruction (Figure 1).

Figure 1

Panel A displays the leak in the superior limb of the atrial baffle. The existing dual chamber pacemaker leads are also noted. Panel B shows the stenotic inferior baffle limb. Panel C displays the post-stent result.

A multi-polar electrode array (MEA) catheter was then inserted retrogradely across the aortic valve into the sRV and ventricular activation maps were created during various pacing modes using Ensite (St. Jude Medical, Inc., USA). Biventricular pacing was simulated by stimulating the pLV using the existing implanted pacemaker. An impulse was then conveyed to the sRV via a roving ablation catheter (7F Stinger. Bard, Medical, USA) from an external pacing system using a triggered function at a maximal delay of 20 ms (Figure 2, panel A). Adjustment of the V–V pacing offset was therefore not possible.

Figure 2

Panel A shows the MEA catheter deployed within the sRV with a roving ablation catheter used for geometry and pacing. The pre-existing atrial and pLV pacing leads are also visible. Panel B depicts the sheathed delivery of the 4F pacing lead through the atrial baffle defect into the sRV. Panel C shows the final lead positions: (a) sRV pacing lead, (b) defibrillator lead in the pLV, (c) pre-existing pLV pacing lead, (d) pre-existing atrial pacing lead, (e) subcutaneous array posteriorly located around the left chest wall, and (f) left sub-pectoral ICD position.

Device implantation was performed at a separate procedure. A 4F lumenless pace/sense active fixation lead (SelectSecure™, Medtronic Inc., USA) was deployed using a steerable delivery sheath via a standard left sub-clavian vein approach while anticoagulated (heparin 1000 units/kg maintaining an activated clotting time of ± 300 s). The superior limb of the systemic baffle was crossed through the baffle leak and positioned in a basolateral segment of the sRV (Figure 2, panel B). Arterial blood pressure (BP) response was assessed during isolated right atrial (RA), pLV, sRV pacing, and simulated BVP (pLV + sRV).2

Defibrillation incorporated pLV and anterior active can electrodes with a posterior single finger subcutaneous array (model 6996-58 cm, Medtronic, Inc.) implanted horizontally around the left chest wall (Figure 2, panel C). Sequential TTEs were used to monitor ventricular function over a 6 month follow-up period.


A maximal mean blood pressure response of 73 mmHg (average 60 ± 9 mmHg) occurred during BVP (Table 1). Activation time of the sRV was a 102 ms during AAI pacing but Wenckebach's phenomenon of the atrioventricular node was observed at a rate of just 70 b.p.m. which may have contributed to the blunted BP response (Figure 3). Direct endocardial pacing within the sRV also yielded a diffuse activation pattern of equally short duration (100 ms) but was also accompanied by a significant BP rise during BVP. Pacing within the pLV resulted in the voltage wave front initially appearing on the interventricular septum of the sRV, then progressing to the apex and then back along the superior septal surface to the basal segment. The activation wave front failed to involve the sRV free wall accounting for a truncated activation duration of 88 ms.

Figure 3

Sequence of voltage maps of the sRV depicting the motion of endocardial depolarization (white = maximum voltage, purple = minimum voltage) during various pacing configurations. The sRV is shown in right and left oblique views in each frame. Top row of images during RA pacing shows activation of the entire sRV starting at the interventricular septum (A), then apex (B), progressing toward the basal segments (C). The middle row during pLV endocardial stimulation shows an incomplete ventricular activation pattern: the peak depolarization wave front remains over the septal and apical regions (DF) and fails to involve the sRV free wall and all basal regions before dissipating. Pacing endocardially on the sRV free wall (bottom row) shows activation originating in the inferior region of the sRV (G), spreading toward the apex (H) to involve the entire endocardial surface in the direction of the basal segments (I).

View this table:
Table 1

Peak intra-arterial blood pressure response during various pacing configurations with corresponding voltage wave front durations recorded across the endocardium of the sRV

Paced chamberPeak blood pressure (mmHg)Mean blood pressure (mmHg)sRV activation timea (ms)
BVP(pLV + sRV)100/5873100
  • aFrom onset of sRV activation to dissipation of voltage wavefront.

Post BVP insertion, an early (1 week), improvement was noted in the EF (23–33%). This was sustained at the 6 month follow-up TTE. Effort tolerance improved from NYHA III to II accompanied by a reduction in QRS width from 130 to 120 ms during consistent, simultaneous (no offset) BVP in DDDR mode (paced atrioventricular delay = 110 ms). No antitachycardia therapies have been documented thus far.


We have presented here a novel strategy using ventricular activation mapping to direct lead placement in an effort to achieve successful BVP in patients with a Mustard procedure and a failing sRV.

The application of CRT has been extrapolated to the adult population with congenital heart disease following success in treating patients with cardiomyopathies. Although no randomized controlled data exists for these patients, there has been documentation in the form of multi-centre case series reports.3 What is apparent is that conventional selection criteria may not apply and that there is variation in technical approaches sometimes requiring a combined endocardial and epicardial system.

Conventional CRT has a non-responder rate of between 20 and 30%.4 This could potentially be higher in this exclusive group and hence the need for a more rigorous selection process. Non-contact mapping has been used to define areas of slow conduction guiding left ventricular pacing in cardiomyopathies.5 The MEA catheter was easily deployed into the sRV in our patient under protection of anticoagulation. The late and possible non-activation of the sRV lateral wall during pLV pacing was made apparent by this three-dimensional electrical mapping technique. This correlated with mechanical evidence of severe hypokinesia in these segments on TTE. This observation suggests that pacing from the apex of the pLV not only results in delayed and asynchronous sRV activation, but may worsen sRV dysfunction—a phenomenon described previously in the non-paced ventricle.6 The acute haemodynamic response during pacing is well documented and correlates with optimal pacing selection.7 The BP response was used as a surrogate marker for the cardiac output during the implant procedure. This proved effective, fast and clinically relevant evidenced by a significant rise in the mean BP of 18 mmHg during BVP as opposed to isolated single chamber pacing (P = 0.001).The baffle leak provided a portal of access to the sRV without causing a disruption to baffle integrity. An epicardial position may have provided similar benefit but we are currently not equipped to do this as a minimally invasive procedure. The sheathed delivery system used allowed the operator manoeuvrability and the narrow diameter of the lead helped minimize obstruction to the conduit and systemic atrioventricular valve function. However this favourable lead profile, in our opinion, does not reduce the risk of thrombo-embolism. We have thus elected to fully anticoagulate our patient with lifelong warfarin.


Given the relative ease of accomplishing CRT guided by NCM and intra-procedural haemodynamic monitoring, we propose that a prospective evaluation of this approach is warranted. Long-term follow-up is also needed to see if the above strategy affords any advantage over current practice and a reduction in the non-responder rate. Advances in minimally invasive epicardial lead placement will also provide greater access to pacing sites over the sRV when an endocardial approach is not feasible.

Conflict of interest: Dr J. M. derives consultant fees from Medtronic Inc. and Dr P. R. receives research funding from Medtronic Inc. The remaining authors have no conflict of interest to declare.


View Abstract