Author + information
- Received March 6, 2015
- Accepted April 16, 2015
- Published online August 1, 2015.
- Faisal F. Syed, MBChB∗,
- Christopher V. DeSimone, MD, PhD∗,
- Elisa Ebrille, MD∗,
- Prakriti Gaba, BS†,
- Dorothy J. Ladewig, BA‡,
- Susan B. Mikell, BA‡,
- Scott H. Suddendorf, RT§,
- Emily J. Gilles, MS‡,
- Andrew J. Danielsen, MS‡,
- Markéta Lukášová, MSc‖,
- Jiří Wolf, MSc‖,
- Pavel Leinveber, MSc‖,
- Miroslav Novák, MD, PhD‖,
- Zdeněk Stárek, MD, PhD‖,
- Tomas Kara, MD, PhD∗,‖,
- Charles J. Bruce, MD∗,
- Paul A. Friedman, MD∗ and
- Samuel J. Asirvatham, MD∗,¶∗ ()
- ∗Division of Cardiovascular Diseases, Department of Internal Medicine, Mayo Clinic, Rochester, Minnesota
- †Mayo Medical School, Rochester, Minnesota
- ‡Mayo Clinic Ventures, Mayo Clinic, Rochester, Minnesota
- §Department of Cardiovascular Surgery, Mayo Clinic, Rochester, Minnesota
- ‖ICRC, Department of Cardiovascular Diseases, St. Anne's University Hospital, Brno, Czech Republic
- ¶Department of Pediatrics and Adolescent Medicine, Mayo Clinic, Rochester, Minnesota
- ↵∗Reprint requests and correspondence:
Dr. Samuel J. Asirvatham, Division of Cardiovascular Diseases, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905.
Objectives This study hypothesized that shielded electrodes could capture myocardium without extracardiac stimulation.
Background Epicardial cardiac resynchronization therapy (CRT) permits unrestricted electrode positioning. However, this therapy requires surgical placement of device leads and risks unwanted phrenic nerve stimulation.
Methods In 6 dog and 5 swine experiments, we used a percutaneous approach to access the epicardial surface of the heart and deployed novel leads housing multiple electrodes with selective insulation. Bipolar pacing thresholds at pre-specified sites were tested to compare electrode threshold data, facing both toward and away from the epicardial surface.
Results In 151 paired electrode recordings (70 in 6 dogs; 81 in 5 swine), thresholds facing myocardium were lower than those facing away (median threshold of 0.9 [interquartile range (IQR): 0.4 to 1.6] mA vs. 4.6 [IQR: 2.1 to >10.0] mA, respectively, for dogs, p < 0.0001; and 0.5 [IQR: 0.2 to 1.0] mA vs 2.5 [IQR: 0.5 to 6.8] mA, respectively, for swine, p < 0.0001). Myocardial capture was feasible without extracardiac stimulation at all tested sites, with mean ± SE threshold margin of 3.6 ± 0.7 mA at sites of high output extracardiac stimulation (p = 0.004).
Conclusions Selective electrode insulation confers directional pacing to a multi-electrode epicardial pacing lead. This device has the potential for a novel percutaneous epicardial resynchronization therapy that permits placement at an optimal pacing site, irrespective of the anatomy of the coronary veins or phrenic nerves.
- bioelectrical therapy
- biventricular pacing
- cardiac resynchronization therapy
- epicardial mapping
- epicardial pacing
- minimally invasive
- multisite pacing
- pericardial intervention
- phrenic nerve stimulation
- steerable pericardial sheath
Cardiac resynchronization therapy (CRT) results in a significant reduction in mortality and symptoms in selected patients with heart failure (1). However, challenges remain with current approaches to resynchronization, including a high rate of non-responders due in part to anatomical limitations in negotiating the coronary venous system, resulting in suboptimal lead placement, phrenic stimulation precluding therapy delivery, and the risks and challenges associated with endovascular lead placement in specific patients such as those with congenital heart disease and patent foramen ovale (2,3). The alternative is the use of epicardial leads placed directly onto the left ventricle. However, this method currently requires a surgical approach, and there may need to be multiple revisions over time as these leads have an increased failure rate in long-term follow-up (4). With 2% of the adult population in the developed world currently suffering from heart failure, a rising prevalence as longevity increases, and an estimated 5% to 10% of heart failure patients having an indication for CRT (5), there is an increasing role for more versatile pacing approaches to help overcome some of these limitations.
The technique for percutaneous subxiphoid access to the pericardial space, initially described by Sosa et al. (6) for the purpose of epicardial mapping of ventricular tachycardia, lends itself well to epicardial pacing, as much of the cardiac surface is reachable through the anterior approach (7,8). However, the phrenic nerves are also intimately related, and pacing in the pericardial space risks inadvertent phrenic capture.
We developed percutaneously delivered leads with multiple electrodes arranged in a forked configuration (Figure 1), allowing for multiple pacing vectors over a wide area of myocardium. We hypothesized that use of selective insulation on the pericardial-facing aspect of the lead would direct the electric field toward cardiac tissue, permitting pacing without extracardiac stimulation and defibrillation without pain. The purpose of this initial study was to confirm that these self-expanding novel leads could be placed percutaneously in the pericardial space and thus could selectively pace myocardium while avoiding extracardiac stimulation, regardless of their position.
Lead design and prototyping
The novel partially insulated multi-electrode lead was designed to be introduced percutaneously into the pericardial cavity and permit multiple pacing vectors to be applied over a variably wide epicardial surface area. This was accomplished through use of an expanding design, which could be introduced into a sheath in a low-profile state and then expanded once deployed in the pericardium to permit pacing and sensing from desired regions (Figures 1, 2, and 3). Electrodes were selectively insulated to direct the electric field toward the myocardium and shaped to a convexity closely representing that of the epicardial surface to maximize contact.
The prototype lead has a 50-cm long, 16-F central body containing a central 8F channel (Online Figure 1) and is wire in nylon insulation. The distal end of the device has a forked design, with two arms 5.4 cm long and 1.5 mm in diameter and curved to a convexity similar to that of the epicardial surface. Each arm hosts 3 to 4 stainless steel circumferential electrodes placed 7 mm apart and covered circumferentially by polyether block amide insulation (Pebax, Arkema, King-of-Prussia, Pennsylvania) on all sides, except at the exposed surface (3.2 × 1.5 mm) intended to be placed onto the epicardium.
The lead is deployable in the pericardial space through a percutaneously introduced steerable sheath (Figure 2). The sheath has a round caliber measuring 24-F in external diameter, with a 20-F central channel sealed with a one-way valve. In certain prototype iterations, it is designed with distal mapping electrodes and a detachable proximal handle to enable it to function as a platform that directs therapies toward the myocardium and enable instrument exchanges. A proximal steering mechanism allows for flexion and extension of the distal end (Online Videos 1 and 2), with an additional central port for aspiration or injection of fluid.
Prototypes were refined in the first three canine experiments. A nitinol loop was introduced between the distal tips of the lead arms. This prevented spontaneous drift of the pacing lead arms within the pericardial space and electrical crosstalk between the electrodes on each arm, as well as allowing the arms to be deployed in a controlled fashion at variable separations (maximum separation at tips was 31 mm) through a pulley at the proximal (operator’s) end (Figure 1, Online Video 3). An additional nitinol loop leveraged the pericardium for further support to resist migration of the lead itself within the pericardial space. A radio-opaque marker was added to help identify the attitude of the electrodes with respect to the myocardium (Figure 3).
Electrophysiology laboratory setup
All experiments were carried out in customized large-animal translational cardiac catheterization laboratories (Siemens, Erlangen, Germany) at the Cardiovascular Innovation Laboratory at Mayo Clinic, Rochester, Minnesota (canine studies) and at the Cardiovascular Animal Research Center in the Laboratory for Advanced Cardiovascular and Central Nervous System Interventions at the School of Veterinary Medicine, University of Veterinary and Pharmaceutical Sciences, Brno, Czech Republic (porcine studies). Electrophysiological studies were performed using the Cardiolab System (version 6.8.1 release 2, GE Healthcare, Wauwatosa, Wisconsin) using standard cardiac stimulators. Regular bipolar configurations between electrodes arranged in parallel between lead arms (1–2, 3–4, 5–6, and 7–8) or serially along each arm (1–3, 2–4, and 1–5) were used (Figure 3). Sensed signals were gained 500 times and processed with 30- to 500-Hz band pass and 60-Hz notch filters.
The study was approved by the Mayo Clinic Animal Care and Use Committee and the Ethics Committee of the University of Veterinary and Pharmaceutical Sciences, Brno, Czech Republic. Six male mongrel dogs (mass of 30.2 ± 3.6 kg) and 5 female swine (47.6 ± 4.8 kg), all in sinus rhythm on study initiation, were maintained under general anesthesia using ketamine (10 mg/kg) and diazepam (0.5 mg/kg) and mechanical ventilation with 1% to 3% isoflurane. Paralytic agents were not used. Continuous surface electrocardiography (ECG) was monitored with placement of standard limb leads and a V1 chest lead. Continuous intra-arterial blood pressure was monitored using a 9-F left femoral artery sheath. An intracardiac echocardiography (ICE) catheter was introduced through a 1-2F sheath placed in the right external jugular vein. Two 9-F sheaths were placed, one in each of the right and left femoral veins for additional endovascular access. Sheaths were placed using Seldinger technique.
Percutaneous pericardial access
Percutaneous pericardial access of the anterior surface of the heart was obtained using the technique described by Sosa (6) and, upon progressive dilatation of the access site with serially larger diameter dilators, 1 to 2 of the prototype steerable sheaths were deployed in the pericardial space. Pneumopericardium and pericardial effusions were monitored for using ICE and fluoroscopy and controlled using pericardial aspiration from the sheath. The novel leads were introduced into the sheath, which was steered under fluoroscopic guidance to guide the lead to standardized locations over the right ventricular outflow tract (RVOT), anterolateral left ventricular free wall (LV), and left atrial appendage (LAA). The radiopaque marker was used to ascertain which direction the exposed electrode surfaces faced (Figure 3).
Epicardial pacing and extracardiac stimulation
Bipolar pacing was tested in regions with adequately sensed signals and no significant mechanical or electrical interference. Pulsed square waves of 2-ms duration were delivered at a regular frequency of 15% above basal heart rate, starting at an output of 10 mA and reducing gradually until capture threshold was reached. Capture was confirmed by a concordant change in surface ECG and arterial pressure waveform rates. Paced QRS complex and epicardial electrograms were compared to those sensed in intrinsic rhythm to differentiate atrial and/or ventricular capture (Online Figure 2).
Sensed signals were analyzed offline for slew, ensuring that signals of <0.5 mV from uninsulated electrode surfaces facing the myocardial surface were excluded from analysis. Slew was measured at the highest frequency near-field electrogram deflection.
Extracardiac stimulation (both phrenic and chest wall musculature) was noted using inspection, palpation, and fluoroscopy while pacing. If stable extracardiac stimulation was noted, the capture threshold was determined as the pacing output at which no further extracardiac stimulation occurred. If no extracardiac stimulation was noted during standard pacing maneuvers, then phrenic location and integrity were confirmed by high output pacing with electrodes facing away or using a pacing probe through the lead’s central channel.
Using side-by-side fluoroscopic images to ensure comparable catheter location, leads were rotated such that the electrodes were turned from facing the heart to facing the serosal surface of the parietal pericardium, and the above-described protocol was repeated to generate paired sensing and pacing data. At the end of the procedure, a standard endovascular electrophysiology catheter was introduced into the ventricle, and the animal was sacrificed by inducing ventricular fibrillation through either rapid ventricular pacing or application of direct current. Necropsy was performed immediately upon death to assess pericardial access and document injury to cardiac and extracardiac structures.
Statistical analysis was performed using JMP version 10.0.0 software (SAS Institute Inc., San Francisco, California). Non-normally distributed data were expressed as medians and interquartile ranges (IQRs). Paired readings of slew and pacing thresholds were assessed using the Wilcoxon signed rank test. Correlation between slew and threshold was compared using Spearman rho. Differences in lead parameters between sites and electrodes were compared using the Wilcoxon rank sum test. The null hypothesis was rejected at a 2-tailed p value of <0.05.
Epicardial lead placement
Successful epicardial access and placement of 1 to 2 epicardial sheaths was achieved for all animals. The steerable sheaths allowed for easy maneuverability of the pacing leads within the pericardial space and for placement over the RVOT, anterolateral LV, and LAA, while the nitinol loops leveraged the pericardium to provide lead stability.
Insulated sensing and pacing
Data from a total of 151 paired electrode recordings were analyzed (70 in dogs and 81 in swine) to compare pacing thresholds when the exposed electrode surfaces were directed toward the myocardial surface versus facing away. In both of the animal models, the pacing threshold was significantly lower when facing the myocardium than when facing away (median dog threshold of 0.9 [IQR: 0.4 to 1.6] mA vs. 4.6 [IQR: 2.1 to >10.0] mA, p < 0.0001; and median swine threshold of 0.5 [IQR: 0.2 to 1.0] mA vs. 2.5 [IQR: 0.5 to 6.8] mA, respectively, p < 0.0001) (Figure 4).
With electrodes directed toward myocardium, capture was feasible without extracardiac stimulation (either phrenic or chest wall musculature) at the pacing sites in all experiments, including sites in juxtaposition with the phrenic nerve. With high output, extracardiac capture with electrodes facing myocardium was noted at 4 of 70 sites (5.7%) in canine and 5 of 81 sites (6.2%) in porcine experiments. Thresholds were significantly lower for myocardial than extracardiac capture (0.3 [IQR: 0.1 to 1.5] mA vs. 5 [IQR: 1.7 to 7.0] mA; threshold margin of 3.6 ± 0.7 mA, p = 0.004).
Analysis of sensed epicardial signals in canine experiments demonstrated a clear, high frequency signal after an initial low frequency signal, which presumably represented near-field epicardial activation following initial far-field endocardial activation, permitting analysis of slew on the near-field component (Figure 5). There were significant differences between sensed signals obtained from electrodes directed onto the myocardium and those obtained from electrodes facing away (median of 3.4 [IQR: 1.4 to 6.5] V/s toward vs. 1.7 [IQR: 1.0 to 2.7] V/s away; p < 0.0001) and an inverse correlation between slew and myocardial pacing threshold (Spearman’s rho = −0.401, p = 0.0002). In contrast, sensed signals obtained in porcine experiments were multicomponent, with multiple discreet, high-frequency components, which rendered slew analysis unreliable, despite a visible reduction in signal frequency when electrodes were directed away (Figure 6).
Effect of lead position and interelectrode distance on lead function
Depending on electrode position, pacing captured the atrium (when electrodes were over the appendage or atrial tissue) or the ventricle. The lower pacing thresholds with electrodes facing the myocardium were consistently present, regardless of pacing site or electrode vector. In dogs, there was more variability in thresholds when pacing thresholds were obtained at the LAA than at the RVOT and LV; whereas in swine, there was less variability at the LAA as compared than at the RVOT and LV (Figure 7). We noted a difference in anatomical approach in navigation to the prespecified pacing sites between the two species. To reach the LAA in dogs, the lead had to be maneuvered to a leftward posterolateral position, whereas in swine, a more anterior approach was often sufficient and required less distortion of the lead, particularly the distal-most electrode bipole, which had the least accuracy, particularly in canine experiments (Figure 7). To confirm that the increased thresholds noted on the distal-most electrodes were not due to increased interelectrode distance, we repeated threshold testing for electrodes with arms apart vs. together over the RVOT in a single canine experiment. There were no significant differences between thresholds of parallel bipoles (apart: 0.55 [IQR: 0.35 to 1.4] vs. together: 0.95 [IQR: 0.3 to 3.2], p = 0.31).
Complications and necropsy findings
Successful entry into the anterior pericardial space through the connective tissue superficial to the diaphragm was confirmed in all animals (Figure 8). There were no significant arrhythmias caused by lead placement and pacing. Sheath- and access-related trauma were seen in 3 experiments. In one, inadvertent laceration of the distal left anterior descending artery during serial sheath dilatation was tolerated hemodynamically once pericardial blood was aspirated, allowing for successful completion of the study. In another, the study was performed with a wire entrapped in the left ventricle after inadvertent entry and knotting therein after ensuring that it was not causing ventricular ectopy, mitral regurgitation, or ventricular dysfunction. In a third, there was inadvertent puncture and exit of the sheath through the pericardium overlying the left atrial appendage, which resulted in the inability to manipulate the lead over the appendage for sensing and pacing due to exit from the pericardium each time. Isolated minor abrasions on the inferior surface of the heart were noted in four experiments and were of uncertain significance, possibly created during sheath placement, with no bleeding or evidence of myocardial injury during the procedure. There were no pacing-related complications in any experiment.
The 2 main findings of our study are: 1) epicardial pacing via subxiphoid percutaneous access is feasible using novel self-expanding leads that collapse for sheath placement and expand in the pericardium; and 2) insulated electrodes confer directionality and preference to pacing the myocardial surface and permitting epicardial pacing without extracardiac stimulation, even when the electrodes are positioned directly opposite the phrenic nerve. A standardized experimental protocol and prototypes yielded comparable results at 2 separate experimental sites located on separate continents, with multiple operators involved, and with different large animal models used, suggesting a robustness to this approach and a shallow learning curve. To the best of our knowledge, this is the first reported use of percutaneously placed, insulated epicardial electrodes to confer directional pacing for these purposes.
Previous studies have reported the use of subxiphoid and subcostal surgical approaches for epicardial pacing (9–11) and, together with our report on a completely percutaneous approach, support the utility of the pericardial space as a vantage point for bioelectrical therapies for heart failure. Specifically for pacing for CRT, these therapies include access to left-sided chambers without the limitations set by coronary sinus anatomy, avoid bloodstream infections due to endovascular foreign material, or lead thromboembolism (3). This may be particularly relevant to patients with an especially high risk of adverse events from endocardial leads, such as those with cyanotic congenital heart disease, intracardiac shunts, or blood stream infections (2). Furthermore, coronary sinus leads have a significant failure rate due to both dislodgement and from revisions necessitated by symptomatic extracardiac stimulation, although similar long-term experience using percutaneous epicardial leads is lacking in comparison.
We designed our leads to harness the pericardium for mechanical support, introduced supportive nitinol loops to provide additional lead and lead arm stability, and contoured the lead to the natural curvature of the cardiac surface. These actions together resulted in good lead stability and function at the areas of interest. A combination of fluoroscopic and electrical navigation was sufficient for accurate placement, and ICE was used to monitor for complications to chambers, valves, and pericardium. The use of steerable sheaths greatly facilitated maneuverability of leads, and the successful use of this strategy for epicardial ablation has been previously reported (12). We investigated whether our data were influenced by differences in interelectrode distances between the parallel bipolar configuration, with the more distal electrode pair being farther apart than the proximal pair (Figures 1 and 3) and were not able to demonstrate this. We did find, however, that, owing to the size and shape of the forked design, the distal-most pair had a higher threshold, most likely due to interaction with the valvular annuli and great vessels and over corners that needed to be negotiated, a finding which is being incorporated into designing the next prototypes for chronic, indwelling use.
Because of the course of the phrenic nerves in intimate proximity to the pericardium (13) and other electrically active thoracic structures in close vicinity (nerves and skeletal muscle), preventing phrenic and extracardiac stimulations are important considerations as epicardial pacing strategies are developed. With coronary sinus lead placement, phrenic stimulation affects approximately 15% to 30% of patients (14,15), requiring either repositioning the lead or electrically reprogramming another vector, both of which may result in lead dysfunction (16,17). Detecting phrenic stimulation at implantation has poor sensitivity due to positional dependency (18) calling for alternative strategies for preventing stimulation. We demonstrated that our insulated lead has differential sensing and pacing function depending on the side insulated, resulting in a significant threshold margin between myocardial and extracardiac stimulation. In clinical studies, a threshold difference <3 V has been associated with clinically significant phrenic stimulation (19). Other strategies that have been reported include using leads with multiple electrodes, allowing for reprogramming between electrodes (20), using closely spaced bipoles (21), and having longer pulse duration (22), although we did not attempt to replicate these findings for epicardial pacing or investigate whether they have incremental effect.
In clinical practice, resynchronization therapy fails for 25% to 40% of patients (23–25). An important mechanism for failure to response may be absence of pacing at a late electrically activated site (26–28). The ability to freely position electrodes anywhere over the left ventricular free wall, regardless of coronary venous anatomy, holds promise for increasing response rates by overcoming that limitation. The ability to do so without regard to the position of the phrenic nerve due to selective insulation may be an important advance.
Another factor contributing to the lack of response may be impaired hemodynamics due to poor atrioventricular coordination. All current CRT systems sense or pace the right atrium, so that left atrial-to-left ventricular timing depends on the propagation of a wavefront from the right to the left atrium, often through diseased and variably conducting tissue. We found that the expandable leads were able to pace both left atrium and the lateral wall of the left ventricle. By independently energizing distal and then proximal electrodes on an intrapericardial device, left atrial-to-left ventricular timing may be optimized, thus affording another level of resynchronization and potential mechanism for response. Last, it has been proposed that CRT has failed to evoke response from some patients despite improved left ventricular contraction, due to the development of tricuspid regurgitation, caused by a lead in the right ventricle that mechanically disrupts tricuspid valve function. Development of new tricuspid regurgitation leads to adverse hemodynamic effects that counteract the benefit of resynchronization when present (29,30). A totally percutaneous epicardial system would also eliminate the deleterious effects of a right ventricular lead, affording another benefit of such a strategy.
These preliminary findings demonstrate proof of concept but require additional studies prior to human implementation. Based on comparably similar performance in both of the large-animal species in this study, we predict that the currently prototyped platform would require minimal modification for clinical use. The complications seen were all related to pericardial access and sheath deployment, which suggests that an alteration to sheath profile may be required. In addition, future studies are ongoing and/or planned to establish the medium to long-term effectiveness, stability, safety, and tolerability of a chronic lead design. Prototyping and testing of a screw-in mechanism and imaging guidance to identify and thereby avoid coronary vasculature are currently under way.
We did not investigate the effects of pacing on cardiac activation and recovery (31) or the various forms of dys-synchrony (atrioventricular, interventricular, and intraventricular), nor have we fully explored the utility of this approach for ablation, defibrillation, or autonomic modulation (32,33). Given the growing understanding of autonomic modulation of arrhythmogenesis, selectively insulated electrodes facing away from the myocardium may be useful for adjunctive autonomic stimulation as part of an implanted system. Long-term human application will also have to consider the relative risks and benefits of such an approach once prototype refinement is finalized, as well as the challenges in pericardial access imparted by cardiac surgery, thoracic irradiation, and prior pericarditis.
This novel multi-electrode lead with insulated electrodes demonstrates effective myocardial pacing in preclinical large animal studies without extracardiac stimulation. This device has the potential for a novel percutaneous epicardial resynchronization therapy that permits placement at an optimal pacing site, irrespective of the anatomy of the coronary veins or phrenic nerves.
A novel, percutaneously placed multi-electrode epicardial pacing lead confers directional pacing to the myocardial surface. Combined with the use of a steerable sheath, effective myocardial pacing at multiple sites was possible with good lead stability and without extracardiac stimulation. This strategy may prove promising for a percutaneous epicardial atrial and ventricular pacing device.
COMPETENCY IN MEDICAL KNOWLEDGE: A novel, self-expandable lead designed for percutaneous epicardial pacing with multiple electrodes, shaped and directed to maximize myocardial contact and selectively insulated to minimize extracardiac pacing, allowed for directionality and preference to pacing the myocardial surface. This permitted epicardial pacing without extracardiac stimulation, even when the electrodes are positioned directly opposite the phrenic nerve.
TRANSLATIONAL OUTLOOK: This innovation provides a therapeutic approach for epicardial pacing which overcomes current limitations of cardiac resynchronization therapy, including a high rate of non-responders due in part to suboptimal lead placement afforded by coronary venous anatomy; phrenic stimulation; and vascular lead complications.
Dr. Kara is currently affiliated with the Department of Cardiology, University Hospital Olomouc, Czech Republic. This work was supported by European Union Regional Development Fund Project FNUSA-ICRC (CZ.1.05/1.1.00/02.0123). Mayo Clinic owns intellectual property related to the investigated technology reported in the manuscript under U.S. Patent 7620458B2 and U.S. Patent Application 61/968977. Neither Mayo Clinic nor the inventors (Patent 7620458B2 issued to Drs. Asirvatham, Friedman, and Bruce; Patent application 61/968977 issued to Drs. Syed, Bruce, DeSimone, Friedman, Asirvatham, Kara, Leinveber, Novák, Stárek, and Wolf) will receive any compensation for the use of this product with Mayo Clinic patients. Dr. DeSimone received a U.S. National Institutes of Health training grant, HL007111. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- cardiac resynchronization therapy
- intracardiac echocardiography
- interquartile range
- left atrial appendage
- left ventricle
- right ventricular outflow tract
- standard error
- Received March 6, 2015.
- Accepted April 16, 2015.
- American College of Cardiology Foundation
- Epstein A.E.,
- DiMarco J.P.,
- Ellenbogen K.A.,
- et al.
- DeSimone C.V.,
- Friedman P.A.,
- Noheria A.,
- et al.
- Brignole M.,
- Auricchio A.,
- Baron-Esquivias G.,
- et al.
- Syed F.,
- Lachman N.,
- Christensen K.,
- et al.
- Mizuno H.,
- Maccabelli G.,
- Della Bella P.
- Biffi M.,
- Exner D.V.,
- Crossley G.H.,
- et al.
- Goetze S.,
- Defaye P.,
- Bauer A.,
- et al.
- Biffi M.,
- Moschini C.,
- Bertini M.,
- et al.
- Sangaralingham S.J.,
- Burnett J.C. Jr..,
- McKie P.M.,
- Schirger J.A.,
- Chen H.H.
- Chung E.S.,
- Leon A.R.,
- Tavazzi L.,
- et al.
- Ypenburg C.,
- van Bommel R.J.,
- Borleffs C.J.,
- et al.
- Delgado V.,
- van Bommel R.J.,
- Bertini M.,
- et al.
- Abu Sham'a R.,
- Buber J.,
- Grupper A.,
- et al.