Author + information
- Received March 5, 2018
- Revision received April 16, 2018
- Accepted April 19, 2018
- Published online September 17, 2018.
- Jackson J. Liang, DO∗,
- Benjamin A. D’Souza, MD∗,
- Brian P. Betensky, MD,
- Erica S. Zado, PA-C,
- Benoit Desjardins, MD, PhD,
- Pasquale Santangeli, MD, PhD,
- William W. Chik, MD, PhD,
- David S. Frankel, MD,
- David J. Callans, MD,
- Gregory E. Supple, MD,
- Mathew D. Hutchinson, MD,
- Sanjay Dixit, MD,
- Robert D. Schaller, DO,
- Fermin C. Garcia, MD,
- David Lin, MD,
- Michael P. Riley, MD, PhD and
- Francis E. Marchlinski, MD∗ ()
- Cardiovascular Division, Electrophysiology Section, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania
- ↵∗Address for correspondence:
Dr. Francis E. Marchlinski, Cardiovascular Division, Electrophysiology Section, Hospital of the University of Pennsylvania, 9 Founders Pavilion, 3400 Spruce Street, Philadelphia, Pennsylvania 19104.
Objectives This study sought to characterize septal substrate in patients with nonischemic left ventricular cardiomyopathy (NILVCM) undergoing ventricular tachycardia (VT) ablation.
Background The interventricular septum is an important site of VT substrate in NILVCM.
Methods The authors studied 95 patients with NILVCM and VT. Electroanatomic mapping using standard bipolar (<1.5 mV) and unipolar (<8.3 mV) low-voltage criteria identified septal scar location and size. Analysis of unipolar voltage was performed and scars quantified using graded unipolar cutoffs from 4 to 8.3 mV were correlated with delayed gadolinium-enhanced cardiac magnetic resonance (DE-CMR), performed in 57 patients.
Results Detailed LV endocardial mapping (mean 262 ± 138 points) showed septal bipolar and unipolar voltage abnormalities (VAs) in 44 (46%) and 79 (83%) patients, most commonly with basal anteroseptal involvement. Of the 59 patients in whom the septum was targeted, bipolar and unipolar septal VAs were seen in 36 (61%) and 54 (92%). Of the 35 with CMR-defined septal scar, bipolar and unipolar septal VAs were seen in 18 (51%) and 31 (89%). In 12 patients without CMR septal scar, 6 (50%) had isolated unipolar septal VAs on electroanatomic mapping, a subset of whom the septum was targeted for ablation (44%). In the graded unipolar analysis, the optimal cutoff associated with magnetic resonance imaging septal scar was 4.8 mV (sensitivity 75%, specificity 70%; area under the curve: 0.75; 95% confidence interval: 0.60 to 0.90).
Conclusions Septal substrate by unipolar or bipolar voltage mapping in patients with NILVCM and VT is common. A unipolar voltage cutoff of 4.8 mV provides the best correlation with DE-CMR. A subset of patients with septal VT had normal DE-CMR or endocardial bipolar voltage with abnormal unipolar voltage.
Electroanatomic mapping (EAM) in patients with monomorphic ventricular tachycardia (VT) and nonischemic left ventricular cardiomyopathy (NILVCM) typically demonstrates basal/perivalvular predominant areas of endocardial bipolar and unipolar electrogram (EGM) abnormalities, with isolated interventricular septum involvement previously reported as an uncommon occurrence (1–5). More commonly, septal involvement occurs with concomitant free-wall abnormality, but a detailed characterization has not been performed (6). Detailed characterization of the arrhythmogenic substrate may enhance ablation outcome in this patient population (7). The location and extent of dense scar can be delineated pre-ablation with noninvasive studies such as delayed gadolinium-enhanced cardiac magnetic resonance (DE-CMR). However, the precise role of this modality in defining NILVCM VT substrate remains unclear (8,9). The purpose of this study was to use bipolar and unipolar voltage mapping and DE-CMR to characterize septal scar in patients with NILVCM and VT and to compare the information acquired from the different scar-assessment modalities.
Patients >18 years of age with NILVCM who underwent endocardial voltage mapping during VT ablation at the University of Pennsylvania over a 3-year period from January 2011 to December 2013 were eligible. Patients with histories of myocardial infarction or obstructive coronary artery disease, congenital heart disease, primary valvular heart disease, arrhythmogenic right ventricular (RV) dysplasia, and hypertrophic cardiomyopathy were excluded. The study was approved by the University of Pennsylvania Institutional Review Board. All patients provided written informed consent. Medical records were reviewed for baseline characteristics and clinical and demographic data.
Acquisition of periprocedural magnetic resonance imaging
Magnetic resonance imaging (MRI) was performed in a subgroup of patients prior to ablation (n = 57). Electrocardiogram-gated images were acquired on a 1.5-T MRI scanner (Siemens Magnetom Avanto, Erlangen, Germany). Fifteen minutes after administration of 0.20 mmol/kg of intravenous gadolinium DTPA (diethylenetriaminepentaacetic acid) (Magnevist, Berlex Pharmaceuticals, Wayne, New Jersey), 2-dimensional delayed enhancement imaging was performed using an inversion-recovery sequence in the left ventricle (LV) short and long axes. Inversion time (250 to 350 ms) was optimized to null the normal myocardium. Nonischemic scar was defined as focal or diffuse areas of delayed enhancement in a distribution inconsistent with scar attributable to previous infarction. MRI reviewers were blinded to the results of EAM or any additional clinical history. LV endocardial, epicardial, septum, and scar contours were manually traced using 3-dimensional post-processing software (Osirix, Geneva, Switzerland). Scar distribution was refined using a full width at half-maximum approach. All measurements were performed automatically, using customized software on MATLAB (MathWorks, Natick, Massachusetts). Endocardial, epicardial, and scar 3-dimensional surface contours were processed by the Visualization Toolkit (Kitware, Clifton Park, New York) to more precisely correlate scar architecture. Prespecified variables—including septal scar volume and thickness, wall thickness, and distance from the endocardium to scar border—were assessed. For scar distribution, the septum was divided into thirds from base to apex (basal, midwall, apex) and from LV to RV (LV endocardium, intramural, and RV endocardium). Free-wall scar was not quantified in this cohort.
Endocardial electroanatomic mapping
Patients were taken to the electrophysiology lab in the post-absorptive state and studied under conscious sedation or general anesthesia. Access to the LV was achieved by a transseptal or retrograde aortic approach. Access to the pericardial space and epicardium was obtained with the technique described by Sosa et al. (10). Patients were anticoagulated with heparin to target activated clotting time >300s. Point-by-point EAM was performed in sinus rhythm, RV, or BiV-paced rhythm when necessary, using the CARTO system (Biosense Webster, Diamond Bar, California) and a 3.5-mm tip open-irrigated noncontact-force catheter, containing a 2-mm ring electrode and a 1-mm interelectrode distance (NaviStar ThermoCool, Biosense Webster) or a solid-tip 4-mm catheter (NaviStar, Biosense Webster). Bipolar EGMs were filtered at 30 to 500 Hz, displayed at 200 mm/s sweep speed and stored for offline analysis. Unipolar EGMs were filtered at 1 to 240 Hz. Wilson’s central terminal was used as an indifferent electrode to record unipolar EGMs. Acquired EGMs were reconstructed into 3-dimensional voltage maps and area measurements made with the incorporated CARTO software. Previously validated cutoffs of 1.5 and 8.3 mV were used for bipolar and unipolar EAM of the LV endocardium, respectively (4,11). Epicardial scar was defined as <1.0 mV with concomitant split, fractionated, and late potentials (3). Readers were blinded to the results of the DE-CMR, clinical history, or details of the ablation procedure at time of analysis.
For adequate sampling density and a complete representation of voltage distribution, a fill threshold of ≤20 mm was maintained. Valvular sites were identified, and intracavitary points with poor contact (based on EGM analysis and intracardiac echocardiography visualization of the catheter tip) were deleted. Unipolar low voltage within 1 cm of the mitral or aortic valve was excluded in defining scar. Scar was defined by the presence of at least 3 contiguous EGMs with defined low voltage. For scar distribution, the LV septum was divided into thirds from base to apex (basal, midwall, apex) as well as anterior or inferior septum. Scar was also defined as present or absent on the free wall.
Patients were categorized into MRI-positive septal scar versus MRI-negative septal scar, based on the presence of gadolinium enhancement as determined by expert review. Analysis of bipolar and unipolar voltage abnormalities was then performed, and scar surface area was quantified, using standard bipolar and unipolar amplitude cutoffs. Patients were separated into the presence versus absence of septal scar on CMR imaging, and graded EAM unipolar cutoffs from 4 to 8.3 mV were used with scar-area quantifications performed at each cutoff. This was used to calculate a receiver-operating characteristic (ROC) curve that best correlated with CMR imaging–defined septal scar.
VT mapping and ablation
Ventricular programmed stimulation was performed with up to 3 extrastimuli at 2 drive cycle lengths. Induced VTs were displayed at 100 mm/s and analyzed for cycle length, morphology, and hemodynamic tolerability. Mappable VTs were then targeted, using entrainment techniques, and ablated, using an irrigated ablation catheter with power of 30 to 50 W titrated to an impedance drop of 10 to 15 Ω. Unmappable tachycardias were ablated using substrate-based ablation strategies with linear or clustered lesions targeting putative exit sites with good pace maps, sites of long stimulus to QRS, highly fractionated EGMs, and isolated late potentials within the scar. The acute procedural endpoint was noninducibility of all targeted VTs.
Continuous variables are expressed as mean ± SD and categorical variables are expressed as percentages. The independent samples Student t test and Pearson chi-square test were used to compare normally distributed continuous and dichotomous variables, respectively. The Mann Whitney U test was used to compare non-normally distributed continuous variables. Analyses were performed using SPSS software (version 20.0, IBM Corporation, Armonk, New York). We considered p values ≤0.05 to indicate statistical significance.
Ninety-five patients with NILVCM (mean age 62 ± 14 years, 84.2% men, mean left ventricular ejection fraction 34 ± 15%) who underwent VT mapping and ablation between January 1, 2011, and December 31, 2013, were analyzed. Eleven additional patients were excluded due to incomplete EAM, precluding detailed scar assessment. The epicardium was mapped in 41 (43.2%) patients. The RV was mapped in 30 (31.6%) patients, but detailed voltage mapping of the RV septum before ablation from the LV side of the septum was not available for review in the majority of these patients, so RV septal voltage analysis was not performed. Table 1 shows basic demographic data for all 95 patients. The majority (93.7%) of patients had implantable cardioverter-defibrillators (ICDs), and 33 (34.7%) patients had biventricular ICDs at the time of procedure. Sixty-five (68.4%) patients had failed treatment with amiodarone. Antiarrhythmic drugs were typically stopped more than 3 half-lives before the ablation if without recurrent VT, and 42 (44.2%) patients had at least 1 previous ablation (range 0 to 6).
EAM or endocardial substrate analysis
The majority of EAM was performed during intrinsic atrioventricular conduction (n = 67, 70.5%), and the remaining were obtained during RV (n = 12, 12.6%) or biventricular (n = 16, 16.8%) pacing. Detailed LV endocardial mapping (mean 262 ± 138 points) showed the presence of bipolar and unipolar septal voltage abnormalities in 44 (46%) and 79 (83%) patients, respectively. Of the patients with any bipolar voltage abnormalities (n = 70), the majority had combined septal and free-wall abnormality (n = 29, 41.4%), and only 15 (21.4%) had isolated septal voltage abnormalities (Figure 1). The presence of any unipolar voltage abnormalities on EAM was nearly uniform in this patient population (n = 89, 94%). Of the 6 patients with no evidence of unipolar voltage abnormalities on EAM, 4 had epicardial mapping done, and 3 of these had scar found on the free wall. Of the 89 patients with any unipolar abnormality, 73 (82%) had combined septal and free-wall voltage abnormalities, whereas only 6 (7%) had isolated septal voltage abnormalities. Scar on both bipolar and unipolar EAM was further delineated as basal, midwall, or apical and either anterior or inferior septum (Figure 2). Basal anteroseptal involvement was the most common.
CMR imaging and EAM analysis
There were 57 patients in the cohort who underwent DE-CMR. Of these, 10 had extensive ICD or motion artifact precluding scar analysis. Table 2 shows scar quantification data for CMR imaging and EAM. Of the 47 patients with interpretable DE-CMR and EAM, 35 (74.5%) had evidence of septal scar on DE-CMR. Of these 35 with MRI-defined septal scar, bipolar and unipolar septal voltage abnormalities were present in 18 (51.4%) and 31 (88.6%), respectively. Importantly, 13 (37.1%) of the 35 patients with MRI-detected septal scar had evidence of unipolar but not bipolar septal voltage abnormalities, suggesting the presence of midmyocardial scar based on EAM.
There were 12 patients with no septal scar identified on DE-CMR: 3 (25.0%) had both bipolar and unipolar septal voltage abnormalities present on EAM, 6 (50.0%) had isolated unipolar septal voltage abnormalities on EAM (consistent with voltage-mapping defined midmyocardial scar), and 3 (25.0%) had no septal abnormality on EAM.
Area of scar was larger as calculated by EAM versus CMR imaging, with average area of septal scar on bipolar and unipolar EAM of 20.9 ± 23.1 cm2 and 37.4 ± 28.2 cm2, respectively, versus 14.4 ± 10.5 cm2 on CMR imaging (p = 0.07 for bipolar and p = 0.0001 for unipolar). Scar extension from the base to the midseptum on CMR imaging was present in 71% compared with 68% of patients by unipolar EAM. However, there was more apical involvement on unipolar EAM, with 41% exhibiting apical involvement versus 4% on CMRI (p = 0.0007).
In the graded unipolar cutoff analysis of septal voltage abnormalities in patients with known CMR imaging septal scar, with measurements of scar area done at each unipolar cutoff (4.0 to 8.3 mV), there was a mean change in scar area of 83.2 ± 17.6% at 4.0-mV versus 8.3-mV cutoffs. ROC analysis identified the optimal unipolar cutoff for correlation with CMRI septal scar to be 4.8 mV (sensitivity 75%, specificity 70%; area under the curve: 0.75; 95% CI: 0.597 to 0.902).
Ventricular tachycardia ablation
In the overall cohort (N = 95), the LV septum was targeted for ablation in 55 (57.8%) patients, the RV septum in 20 (21.1%), and both in 16 (16.8%). Of the 59 patients in whom the RV VT or LV septum were targeted with ablation, there were bipolar and unipolar septal voltage abnormalities seen in 36 (61.0%) and 54 (88.1%), respectively (p =0.0002). In the 5 patients with normal unipolar and bipolar EAM in whom septal ablation was performed, abnormal, multicomponent, fractionated bipolar EGMs with normal voltage amplitude by standard criteria in the regions of interest were targeted during ablation.
In the group of patients with unipolar septal voltage abnormalities on EAM but normal DE-CMR (n = 9), the septum was targeted for ablation in 4 (44.4%). In this group, clinical VT was localized to the septum based on either best pace map or entrainment mapping to the septum. In these 4 patients, the mechanism of VT was focal or triggered in 1 and re-entrant in 3 (bipolar or unipolar voltage abnormalities seen in all 3 of these patients; 1 had bipolar and unipolar voltage abnormalities in a region of noncompacted tissue in the apical septum with no evidence of scar on DE-CMR, whereas the other 2 had more basal bipolar or unipolar septal voltage abnormalities in the absence of DE-CMR septal scar. Figure 3 shows a representative case.
Our study demonstrates that the presence of septal substrate, as indicated by bipolar or unipolar voltage abnormality, is present in many (83%) patients with NILVCM and VT undergoing EAM and catheter ablation. The voltage abnormality is uncommonly isolated to the septum (7% for unipolar and 21% for bipolar). Similar to the free-wall involvement, the process appears to extend from the base toward the apex with less involvement at the apical segments. The presence of only unipolar voltage abnormality recorded from the septal endocardium strongly suggests the presence of midwall abnormalities and correlates with the presence of midmyocardial abnormalities seen on CMRI. Furthermore, we demonstrated an optimal cutoff of 4.8 mV for unipolar EGM abnormality that corresponds to CMR imaging–delayed enhancement, which defined dense scar.
Characterization of the arrhythmogenic substrate in patients with NILVCM and VT continues to evolve. As opposed to ischemic cardiomyopathy, which has a more defined distribution, regions of scar in NILVCM have been more difficult to characterize (12). Recent studies using electroanatomical voltage mapping and CMR imaging have suggested that the scar pattern in patients with NILVCM can be classified into 2 major subtypes: anteroseptal or inferolateral (13,14). Necropsy studies showed the presence of both free-wall and septal scar, with visible scar involving the septum in 14% of patients (15). Initial descriptions of endocardial substrate abnormalities in patients with NILVCM from our group showed a high prevalence of basal perivalvular scar, with the overwhelming majority (88%) of inducible VT originating from this area (1). This was further expanded in a study by Cano et al. (3), in which patients with NILVCM and suspected epicardial VT undergoing endocardial and epicardial mapping showed areas of confluent low voltage involving the basal lateral LV, present in 82% of the study population. The results of the current study further support these findings; in this cohort, the majority of patients with septal scar on EAM had basal involvement (96%).
The importance of septal substrate was initially described by Haqqani et al. (6), in which 71% of patients had low-voltage areas involving the basal septum with isolated septal-voltage abnormalities being uncommon (11.6%). Our study further expands upon this previously published work, with 89% of this cohort having septal unipolar EGM abnormalities and confirming that only a small subset of patients with NILVCM and VT have isolated septal substrate (7% on unipolar EAM), with the majority having both free-wall and septal involvement (42% and 82% for bipolar and unipolar EAM, respectively).
Unipolar EGMs have a larger field of view than endocardial bipolar EGMs, helping to identify abnormalities of the midmyocardium and epicardium, and we have previously described the use of unipolar mapping to identify VT substrate (4,6). Our study underscores the importance of unipolar mapping to identify midmyocardial septal substrate in patients with NILVCM and VT. In this cohort, unipolar mapping identified scar in more than one third of patients not identified by bipolar mapping alone. Furthermore, 30% of patients who had the septum targeted for ablation showed normal septal bipolar EAMs but abnormal unipolar EAMs.
It has been shown that DE-CMR is fairly reliable for identifying the location of an anatomically defined scar and is helpful in identifying an intramural substrate very common with basal septal scar associated with NICM. However, we believe that MRI using standard late gadolinium enhancement imaging techniques requires a certain depth of scar to be recognized consistently. It has been shown that layered scar on the epicardium, confirmed by epicardial bipolar voltage mapping, can be missed with MRI, and a mismatch between MRI and voltage map finding at regions of origin of VT has been previously reported (16). The current study shows that unipolar mapping is a more sensitive method to define this substrate. The majority of patients with septal substrate defined on DE-CMR were also identified with unipolar mapping (89%). However, a small subset in this cohort with normal DE-CMR had substrate defined with unipolar mapping that was also targeted for ablation. We believe that unipolar EGMs may be a more sensitive (albeit less specific) tool for identifying a possible substrate abnormality associated with VT. The 8.3-mV voltage cutoff for defining abnormal LV unipolar voltage has now been validated with 2 separate “normal” patient cohorts and represents a statistical definition of an abnormal unipolar signal (4,17). On the septum, the presence of unipolar abnormalities should make one suspicious of an intramural substrate, even in the absence of bipolar endocardial abnormalities. Importantly, as the unipolar EGMs appear to become mildly abnormal with any degree of replacement fibrosis, the mere presence of unipolar abnormalities does not specifically indicate the presence of a substrate worth targeting for VT ablation (17,18). Furthermore, it is important to recognize that, in the setting of hypertrophy, epicardial layered scar confirmed with bipolar EGMs may not be present using the standard voltage cutoff (<8.3 mV) for endocardial unipolar EGMs, and a higher cutoff with adjustment of the voltage slider bar may be required (19). In the current cohort, adjusting unipolar cutoffs down to 4.8 mV correlated best with CMR imaging–defined septal scar, helping to further delineate the dense midseptal scar frequently seen in patients with NILVCM. A higher cutoff (8.3 mV) may allow for identification of patchy, heterogeneous substrate for VT.
From a practical point of view, the challenge—when dealing with VT with suspected septal origin in NICM—is identifying a high likelihood of midseptal scar in the absence of specific activation and entrainment mapping information identifying a site of origin consistent with a VT isthmus site. In these cases, data obtained from pre-procedural MRIs as well as data gleaned from unipolar EGMs and pacing, which suggest the presence of midmyocardial scar consistent with a VT substrate, can be used as complementary and confirmatory information before ablation. The presence of midmyocardial late gadolinium enhancement, the presence of more dramatic unipolar abnormalities (voltage <5.0 mV), and the presence of significant transseptal conduction delay (>40 mS) when pacing the right side of the septum and recording from the left (20) all support the presence of midmyocardial scar. These findings can guide the clinical decision to attempt aggressive septal ablation if there is persistent VT, QRS morphology during VT, and pacing consistent with a septal origin (6).
Finally, until we are able to detail the VT circuits and the precise location of VT substrates in NICM, determining whether data obtained from imaging versus EGMs should represent the gold standard for localizing the VT substrate will be subject of additional study and ongoing debate.
This study was a nonrandomized retrospective analysis of a cohort of patients with NICM undergoing VT ablation, with a subset undergoing DE-CMR. As detailed voltage mapping of the RV septum was not performed in the majority of patients before septal ablation, we are unable to determine whether unipolar voltage cutoffs used for the LV septum (8.3 mV) are applicable when mapping from the RV side of the septum. Prospective evaluation of unipolar and bipolar voltages of the RV septum in patients with and without structural abnormalities is necessary to determine optimal voltage cutoffs for defining septal VT substrate when recording from the RV. Also, contact force-sensing catheters were not widely available at the time of ablation, and all ablations were performed with noncontact force-sensing ablation catheters. Although we attempted to visualize the catheter tip on intracardiac echocardiography to assure contact during mapping, it is possible that inadequate contact may have reduced the amplitude of recorded voltage during mapping. Furthermore, EAM was performed in all patients using point-by-point mapping, and it is possible that the use of multipolar small-electrode catheters may further enhance the ability to detect voltage abnormalities. Further prospective studies involving the comparative use of DE-CMR and EAM for substrate identification in NILVCM, with a concentration on outcomes, may be appropriate.
Septal substrate in patients with NILVCM and VT is common (83%) with only midmyocardial involvement in 37%. The substrate is consistently greater at the basal rather than apical septum. Identification of midseptal VT substrate is facilitated by DE-CMR. Near-uniform unipolar voltage abnormality is present when DE-CMR is abnormal. A unipolar cutoff of 4.8 mV on the LV septum correlates with DE-CMR–defined scar. A small but important subset of patients with normal DE-CMR will have septal low voltage on EAM, and VT eliminated targeting the defined abnormal voltage regions on the septum.
COMPETENCY IN MEDICAL KNOWLEDGE: Septal substrate is common in patients with NICM and VT. Although delayed-enhancement CMR imaging can identify septal VT substrate in most cases, electroanatomic bipolar and unipolar voltage mapping may also identify septal VT substrate in patients with normal MRIs. A unipolar voltage cutoff of 4.8 mV on the LV septum correlates with scar defined on CMR imaging.
TRANSLATIONAL OUTLOOK: Larger series are necessary to confirm the correlation between septal voltage abnormalities seen with electroanatomic mapping and scar on CMR imaging in patients with NICM. Future studies should examine voltage cutoffs to define scar both from the RV side of the septum and when mapping with multielectrode mapping catheters.
↵∗ Dr. Liang and Dr. D’Souza contributed equally to this work and are joint first authors.
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
- delayed gadolinium-enhanced cardiac magnetic resonance
- electroanatomic mapping
- implantable cardioverter defibrillator
- left ventricular
- magnetic resonance imaging
- nonischemic cardiomyopathy
- nonischemic left ventricular cardiomyopathy
- right ventricular
- ventricular tachycardia
- Received March 5, 2018.
- Revision received April 16, 2018.
- Accepted April 19, 2018.
- 2018 American College of Cardiology Foundation
- Hsia H.H.,
- Callans D.J.,
- Marchlinski F.E.
- Marchlinski F.E.
- Cano O.,
- Hutchinson M.,
- Lin D.,
- et al.
- Hutchinson M.D.,
- Gerstenfeld E.P.,
- Desjardins B.,
- et al.
- McCrohon J.A.,
- Moon J.C.,
- Prasad S.K.,
- et al.
- Soejima K.,
- Stevenson W.G.,
- Sapp J.L.,
- Selwyn A.P.,
- Couper G.,
- Epstein L.M.
- De Cobelli F.,
- Pieroni M.,
- Esposito A.,
- et al.
- Desjardins B.,
- Yokokawa M.,
- Good E.,
- et al.
- Marchlinski F.E.,
- Callans D.J.,
- Gottlieb C.D.,
- Zado E.
- Callans D.J.,
- Ren J.F.,
- Michele J.,
- Marchlinski F.E.,
- Dillon S.M.
- Piers S.R.,
- Tao Q.,
- van Huls van Taxis C.F.,
- Schalij M.J.,
- van der Geest R.J.,
- Zeppenfeld K.
- Oloriz T.,
- Silberbauer J.,
- Maccabelli G.,
- et al.
- Betensky B.P.,
- Dong W.,
- D'Souza B.A.,
- Zado E.S.,
- Han Y.,
- Marchlinski F.E.
- Campos B.,
- Jauregui M.E.,
- Park K.M.,
- et al.
- Santangeli P.,
- Marchlinski F.E.
- Betensky B.P.,
- Kapa S.,
- Desjardins B.,
- et al.