Author + information
- Received September 13, 2018
- Revision received October 11, 2018
- Accepted October 12, 2018
- Published online January 21, 2019.
- Tomofumi Nakamura, MD, PhDa,b,
- Ryohsuke Narui, MD, PhDa,
- Qi Zheng, MDb,
- Hirad Yarmohammadi, MD, MPHb,
- Usha B. Tedrow, MD, MSb,
- Bruce A. Koplan, MD, MPHb,
- Gregory F. Michaud, MDa,
- William G. Stevenson, MDa and
- Roy M. John, MD, PhDa,∗ ()
- aCardiovascular Division, Department of Medicine, Vanderbilt University Medical Center, Nashville, Tennessee
- bCardiovascular Division, Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts
- ↵∗Address for correspondence:
Dr. Roy M. John, Vanderbilt Heart and Vascular Institute, 1215 21st Avenue South, Nashville, Tennessee 37232.
Objectives This study sought to evaluate the incidence and significance of atrioventricular (AV) block associated with ventricular arrhythmia (VA) ablation.
Background Attempted ablation of VAs that arise from the septum carries a risk of AV block.
Methods Data from 1,418 patients who had catheter ablation for drug-refractory VAs were evaluated. Two analyses were conducted. The first analysis assessed the patient and procedure characteristics associated with ablation-induced AV block. The second analysis investigated outcome differences between patients with and without AV block. For the second analysis, patients with AV block (Group I) were compared with a 1:2 propensity score–matched control group (Group II) and with patients with pre-existing AV block before ablation (Group III).
Results Twenty-one (1.6%) patients developed AV block. In multivariable analysis, nonischemic cardiomyopathy (odds ratio: 3.33; 95% confidence interval: 1.32 to 8.40; p = 0.011) and transcoronary ethanol ablation (odds ratio: 46.50; 95% confidence interval: 14.10 to 153.00; p < 0.001) were independently associated with AV block. Subsequent to the AV block, 9 patients were upgraded from an implantable cardioverter-defibrillator to cardiac resynchronization therapy with defibrillator (CRT-D), 2 had de novo CRT-D implantation, 5 had pre-existing CRT-D, and 5 had pacing without CRT. VAs recurred in 33% of patients in Group I, 17% in Group II (log-rank p = 0.842), and 35% in Group III (p = 0.636). The composite outcome of heart failure hospitalization, heart transplantation, or death occurred in 29% of patients in Group I, 17% in Group II (p = 0.723), and 45% in Group III (p = 0.303).
Conclusions Complete AV block occurs in fewer than 2% of patients undergoing VA ablation and does not appear to be associated with the worse outcome of heart failure hospitalization, heart transplantation, or death.
Catheter ablation is an important therapeutic option for ventricular arrhythmias (VAs), especially those that are life threatening or refractory to medical therapy. The anatomic location of the arrhythmogenic substrate varies based on the etiology. Whereas ischemic scars are distributed primarily in the subendocardium of infarcts, the substrate in nonischemic cardiomyopathies (NICMs) is more variable, often involving the base of the heart in the perivalvular area or interventricular septum. Ablation of arrhythmia substrate in the septum has the potential for conduction system injury (1,2). In addition, VAs may originate from the His-Purkinje system and may require creation of fascicular block for treatment (3,4). The majority of patients with VAs will have implantable cardioverter-defibrillators (ICDs) in place, but creation of a heart block often requires revision of existing pacing systems to provide atrioventricular (AV) or intraventricular synchrony, with potential additional complications (5). As the implementation of cardiac resynchronization therapy (CRT) is not beneficial in all patients with heart block, creation of complete AV block may have detrimental effects on ventricular function and heart failure symptoms in some patients. The influence of AV block on outcomes in patients undergoing ablation for VA is not well defined. The aims of the present study were to: 1) define the incidence of AV block in patients undergoing ablation for VA; 2) identify predictors for occurrence of AV block; and 3) assess outcomes of patients with AV block compared with a matched cohort without AV block and a cohort with pre-existing AV block unrelated to the ablation.
Between 2007 and 2017, a total of 1,418 patients underwent their first catheter ablation for recurrent episodes of ventricular tachycardia (VT) or premature ventricular contractions at our institution. Coronary angiography or stress testing incorporating imaging excluded ischemia as a trigger for VAs. The distinction between ischemic cardiomyopathy and NICM was based primarily on the presence of relevant coronary artery disease confirmed with coronary angiography or history of prior myocardial infarction. NICM was defined by the presence of persistent left ventricular (LV) systolic impairment in the absence of coronary artery disease, but excluding valvular heart disease, congenital heart disease, arrhythmogenic right ventricular cardiomyopathy, and cardiac sarcoidosis, which were evaluated independently. Each patient gave written informed consent for the procedure. Data collection was conducted in accordance with the protocols approved by the Human Research Committees of Brigham and Women’s Hospital.
Electrophysiological study and ablation
Procedure details have been described previously elsewhere (6,7). In brief, multipolar electrode catheters were positioned in the right ventricular (RV) apex and the His bundle region via the femoral veins. LV mapping was performed from a retrograde aortic or transatrial septal approach. Electroanatomical mapping and ablation was performed using a 3.5-mm-tip open-irrigated catheter (ThermoCool, ThermoCool SmartTouch, ThermoCool SF, or ThermoCool SmartTouch SF, Biosense Webster, Diamond Bar, California), a 4-mm-tip nonirrigated catheter (NaviStar, Biosense Webster), or a multipronged multipolar catheter (Pentaray, Biosense Webster) using the CARTO mapping system (CARTO3 or CARTO XP, Biosense Webster). In the electroanatomic mapping system bipolar electrograms were high-pass filtered at 20 to 30 Hz and low-pass filtered at 400 Hz. Bipolar electrograms were also band pass filtered from 30 to 500 Hz and digitally recorded along with a 12-lead surface ECG utilizing the CardioLab EP system (GE Healthcare, Buckinghamshire, United Kingdom). Programmed electrical stimulation with 1 to 4 extrastimuli of a minimum-coupling interval of 180 ms, applied following 2 basic drives (typically 600 ms and then 400 ms), and burst pacing from 2 ventricular sites were used to attempt to induce arrhythmias. Infusions or boluses of isoproterenol or epinephrine were used to attempt to induce VAs in selected patients at the discretion of the operator.
Epicardial mapping was performed if endocardial ablation failed and epicardial arrhythmia origin was suspected. Percutaneous subxiphoid pericardial access was obtained as previously described (8). Coronary angiography was performed before epicardial ablation to avoid coronary injury; high output pacing was also performed to avoid ablation in close proximity to the phrenic nerve.
Radiofrequency (RF) energy was delivered from irrigated catheters at a power of 25 to 50 W with irrigation rates of 7 to 30 ml/min targeting an impedance drop of 10 to 20 Ω. Applications at target regions were usually repeated until unipolar pacing at 10 mA at 2-ms stimulus strength failed to capture.
Selected patients who had arrhythmias refractory to standard ablation techniques underwent additional ablation strategies, including the use of dual-site unipolar RF ablation (9), transcoronary ethanol ablation (10), and use of an investigational intramural needle ablation (11).
Patients who had developed complete AV block that persisted at least at to hospital discharge were included in the AV block group.
Data collection and follow-up
Data were collected from a centralized database containing records of all patients treated and followed at Brigham and Women’s Hospital and its associated Partners Healthcare sites. These records include emergency department visits, outpatient clinic visits, and data recorded during inpatient care as well as follow-up progress notes from referring physicians monitoring out-of-area patients. Routine interrogation of ICDs was obtained at 3-month intervals or within 24 h following a clinic event. The study primary endpoint was recurrence of a VA that required intervention such as appropriate ICD shock or antitachycardia pacing delivery, external cardioversion, or antiarrhythmic drug administration for termination, or resulted in repeat ablation after the hospital discharge. The second endpoint was a composite of heart failure hospitalization, heart transplantation, and death.
A Shapiro-Wilk test was performed for continuous variables for the evaluation of distribution normality. The variables with normal distribution are expressed as mean ± SD and an unpaired Student’s t-test was used for comparison. The variables with skewed distribution are expressed as median (25th and 75th percentile) and Mann-Whitney U test was used to compare. Categorical variables are presented as count and percentage and Fisher exact test was performed for comparison. Two analyses were conducted in the present study. The first analysis was to identify the baseline patient characteristics and procedures to predict the incidence of complete AV block associated with ablation. The second analysis investigated the differences in outcomes between patients with and without complete AV block. For the first analysis, a logistic regression analysis was performed to investigate the risks of procedure related block. The covariates with a p value < 0.05 in the univariate analysis were selected for testing in the multivariable analysis. This analysis was conducted within the entire cohort but 70 patients with pre-existing complete AV block were excluded. For the second analysis, patients with procedural AV block (Group I) were compared with a matched cohort without AV block (Group II) and a group with pre-existing AV block (Group III) with the aim to evaluate the impact of AV block on the clinical outcome. To account for differences in baseline characteristics between the presence and absence of complete AV block, propensity score stratification methods were utilized for selecting the components of Group II. The patients with AV block were matched in a 1:2 ratio to a comparator group who did not have AV block with ablation. Groups were matched according to a propensity score generated using the following covariates: sex, age, body mass index, arrhythmia type, underlying heart disease (i.e., no structural heart disease, ischemic heart disease, and NICM), New York Heart Association functional class, implanted device (i.e., ICD and cardiac resynchronization therapy with defibrillator [CRT-D]), LV ejection fraction, and history of atrial fibrillation. A 1:2 matching was then performed based on this propensity score. Covariate balances between the groups were checked using tests noted previously and standardized differences, in which imbalance was defined as an absolute value >0.2. Survival curves were created using the Kaplan-Meier method and compared with the log-rank test. A p value <0.05 was considered significant. All statistical analyses were performed utilizing R software version 3.3.3 (R Foundation for Statistical Computing, Vienna, Austria).
Twenty-one (1.6%) patients developed persistent complete AV block during ablation for VAs. The individual patient characteristics and results summary are shown in Table 1. Most of these patients had evidence of pre-existing intraventricular conduction abnormalities at the baseline (90%). LV abnormalities were most common, with 8 (38%) patients demonstrating complete left BBB and 5 (24%) patients with anterior or posterior fascicular block. Right BBB was observed in 7 (33%) patients and first-degree AV block was observed in 6 (29%) patients. During ablation, His and left bundle potential was observed at the ablation site of AV block in only 1 patient each, although the possibility of creating AV block was a consideration in most of the cases based on catheter position. Table 2 shows the comparison of baseline characteristics between the patients with and without complete AV block after ablation and Table 3 summarizes the result of univariate and multivariable logistic regression analysis. On univariate analysis, NICM (odds ratio [OR]: 4.10; 95% confidence interval [CI]: 1.57 to 11.10; p = 0.002) and performance of transcoronary ethanol ablation (OR: 57.20; 95% CI: 14.90 to 209.40; p < 0.001) were predictors for risk of AV block. These remained independent risks in multivariable regression analysis (OR: 3.33; 95% CI: 1.32 to 8.40; p = 0.011; OR: 46.50; 95% CI: 14.10 to 153.00; p < 0.001, respectively).
Of the 21 patients with complete AV block, 5 patients had a previously implanted CRT-D and received no change in pacing system. Of the 12 patients with single- or dual-chamber ICDs, 9 underwent an upgrade to CRT-D. One patient with a dual-chamber pacemaker did not have a change in pacing system. For 3 patients who did not have an implanted pacing system, de novo CRT-D implantation was performed for the 2 patients with impaired LV function, with a dual-chamber ICD for the 1 patient with preserved LV systolic function. The individual post-procedure outcomes for patients with AV block (Group I) are shown in Table 1 and comparison to the propensity score–matched cohort (Group II) is shown in Table 4. There were no differences with regard to baseline characteristic between the groups, except for pre-existing conduction abnormalities and ablation site. In Group II, 25 (60%) patients had any conduction abnormality (p = 0.018 vs. Group I): left BBB in 10, axis deviation in 11, right BBB in 4, and first-degree AV block in 6. The target ablation site was significantly different (p < 0.001). In Group I, all 21 patients had a septal target site ablated, with RF in 20 patients and ethanol ablation of a septal branch of the left anterior descending artery in 1 patient. In contrast, in Group II, the septum was targeted for ablation in 22 of 42 patients (RF ablation in 21 patients and ethanol ablation in 1 patient).
A comparison of Group I with the patients with pre-existing AV block (Group III) is also shown in Table 4. The patients in Group III had worse New York Heart Association functional class (p = 0.009) and LV ejection fraction (p = 0.004) at baseline compared with Group I.
Recurrent VA was observed in 7 (33%) patients in Group I, 7 (17%) in Group II, and 25 (35%) in Group III (Figure 1). Five of 7 patients in Group I underwent a second ablation and the target arrhythmia was arising from the site same as the prior procedure (basal septum) in all cases. The incidence of recurrent of VA was not different between groups (p = 0.842 for Group I vs. Group II and p = 0.636 for Group I vs. Group III). Three patients in Group I experienced heart failure hospitalization during follow-up, and 2 of them had severely impaired LV function at the baseline. One patient who did not undergo upgrade of a pre-existing ICD to CRT-D subsequently developed severe heart failure that required placement of a LV assist device, and subsequently died of subarachnoid hemorrhage 1,958 days after ablation (Patient #14 in Table 1). None of the patients who underwent successful upgrade to a CRT-D developed progression of heart failure symptoms during the follow-up period. Four patients in Group I died after a median of 553 days of follow-up. Two patients died of heart failure (Patients #14 and #15 in Table 1) and 2 patients died of uncertain cause (Patients #3 and #4 in Table 1). The composite outcome of heart failure hospitalization, heart transplantation, and death was not significantly different among groups; 6 (29%) patients in Group I, 7 (17%) in Group II (p = 0.723 vs. Group I), and 32 (45%) in Group III (p = 0.303 vs. Group I) (Figure 2). No device infections or malfunctions requiring revision were observed during the follow-up period.
To our knowledge, a systematic analysis of AV block and its consequences following ablation for VAs has not been previously conducted. The main findings of the present study are: 1) conduction system injury causing AV block occurs in <2% of patients; 2) ablation of VA substrates in NICM entails a higher risk for conduction system injury, especially when transcoronary ethanol ablation is employed; and 3) the occurrence of complete AV block did not appear to influence VA recurrence, although there was trend toward a higher VT recurrence in patients with heart block. Outcomes from heart failure status did not appear to be different between the groups.
Higher prevalence of conduction system involvement in NICM versus ICM
Septal involvement with inflammation and scar is common in NICM. In an autopsy study of idiopathic dilated cardiomyopathy hearts, visible scar involved the ventricular septum in 13% and it was associated with higher incidence of baseline left BBB (12). The basal septum and periaortic area is a common region for scar-related VTs in NICM (13,14). In the present study, 57% of the patients with AV block had NICM.
In the patients with ischemic heart disease, the territory of coronary artery occlusion defines the location of the arrhythmogenic substrate. The blood supply to the anterior ventricular septum is mainly from the left anterior descending artery, but any of 3 main branches can potentially contribute to the basal septum. Of our 5 cases with ischemic heart disease who developed complete AV block, the infarct artery was the left anterior descending artery in 2 patients, left circumflex artery in 1, right coronary artery in 1, and uncertain due to 3-vessel disease in 1 patient.
Ablation modality and risk for injury to conduction system
VTs originating from the basal septum can be difficult to ablate, likely due to an origin from deep within the septum. When conventional ablation fails, more aggressive strategies are sometimes tried, and not surprisingly, these are associated with a risk of AV block. Transcoronary ethanol ablation was associated with a risk of AV block, which is expected when the basal septum is targeted. The incidence of AV block during septal ethanol ablation for VT in our patients has been previously found to be 38.5% (10). Bipolar RF (which was not available in this study) and simultaneous unipolar RF ablation also have a risk of causing heart block in this area. More recently, we have also been using an infusion needle ablation catheter (11). All of these emerging techniques risk AV block when ablation is performed at the basal septum.
Role of early implementation of CRT pacing in patients with LV dysfunction and AV block
The composite outcome of heart failure hospitalization, heart transplant, and death was not different between the groups. Of the 21 patients with AV block, 16 patients had CRT-D at the time of hospital discharge that was implanted either before or after ablation. The remaining 5 patients received RV pacing. Four of these had preserved LV function and none reached a composite endpoint. One patient who did not undergo upgrade despite reduced LV ejection fraction developed severe heart failure (Patient #16 in Table 1). When AV block occurs, it seems prudent to implement biventricular pacing, especially for the patients with reduced LV ejection fraction. The benefit of CRT pacing in patients with heart block and pre-existing systolic LV dysfunction has been demonstrated in prior studies (15–17).
Although the ablation data are collected prospectively, this study was a retrospective analysis of our database. These patients are largely referred to our tertiary center and undoubtedly are a selected group. The implementation of resynchronization and its programming was left to the treating physician. Catheter-induced BBB are not uncommon and usually resolve. Although of interest, information was not adequately captured to allow retrospective review. In Group II, there were 6 patients who were noted to have new left BBB (n = 4) or right BBB (n = 2) present at the end of procedure, persistence of block after hospital discharge were not evaluated. The limited number of patients with AV block and outcomes rendered point estimates are imprecise and also may have led to type II error, thereby accounting for lack of significance. Further study with a larger patient population is warranted to confirm these findings.
Conduction system injury causing persistent AV block occurs in fewer than 2% of patients undergoing VA ablation. Although there is a trend toward more arrhythmia recurrences in patients with AV block, this did not reach statistical significance. If appropriate pacing therapy is implemented including CRT when indicated, conduction system injury is not associated with worse outcomes (combined endpoint of heart failure and mortality) when compared with patients who do not have AV block with ablation. For patients with drug-refractory VA, the risk of heart block from ablation may be adequately offset if VAs are controlled.
COMPETENCY IN MEDICAL KNOWLEDGE: Conduction system injury causing persistent AV block occurs in fewer than 2% of patients undergoing VA ablation and is not associated with worse outcomes of heart failure and mortality, if appropriate pacing therapy is provided. For patients with drug-refractory VA, the risk of heart block from ablation may be adequately offset if VAs are controlled.
TRANSLATIONAL OUTLOOK: Further studies are needed to address the influence of persistent bundle branch block in patients, with impaired LV systolic function are also thought to be affected. Studies to evaluate outcomes during the long-term follow-up period are also needed.
Drs. Nakamura and Narui have received a scholarship from the Japanese Heart Rhythm Society, Tokyo Japan. Dr. Tedrow has received speaker honoraria from Abbott Medical, Biosense Webster, Medtronic, and Boston Scientific. Dr. Michaud has received speaker honoraria from Boston Scientific, Biosense Webster, Abbott, Pfizer, Medtronic, and Biotronik. Dr. Stevenson has received speaker honoraria from Abbott Medical, St. Jude Medical, Boston Scientific, and Medtronic; and is also co-holder of a patent for needle ablation that is consigned to Brigham and Women’s Hospital. Dr. John has received speaker honoraria from Biosense Webster, Abbott, and Medtronic. All other 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
- bundle branch block
- confidence interval
- cardiac resynchronization therapy
- cardiac resynchronization therapy with defibrillator
- implantable cardioverter-defibrillator
- left ventricular
- nonischemic cardiomyopathy
- odds ratio
- ventricular arrhythmia
- ventricular tachycardia
- Received September 13, 2018.
- Revision received October 11, 2018.
- Accepted October 12, 2018.
- 2019 American College of Cardiology Foundation
- Yoshida K.,
- Yokokawa M.,
- Desjardins B.,
- et al.
- Sasaki T.,
- Miller C.F.,
- Hansford R.,
- et al.
- Liu Y.,
- Fang Z.,
- Yang B.,
- et al.
- Zhan X.,
- Liang Y.,
- Xue Y.,
- et al.
- Poole J.E.,
- Gleva M.J.,
- Mela T.,
- et al.
- Kumar S.,
- Fujii A.,
- Kapur S.,
- et al.
- Nagashima K.,
- Choi E.K.,
- Tedrow U.B.,
- et al.
- Kumar S.,
- Bazaz R.,
- Barbhaiya C.R.,
- et al.
- Iyer V.,
- Gambhir A.,
- Desai S.P.,
- Garan H.,
- Whang W.
- Tokuda M.,
- Sobieszczyk P.,
- Eisenhauer A.C.,
- et al.
- Sapp J.L.,
- Beeckler C.,
- Pike R.,
- et al.
- Oloriz T.,
- Silberbauer J.,
- Maccabelli G.,
- et al.
- Piers S.R.,
- Tao Q.,
- van Huls van Taxis C.F.,
- Schalij M.J.,
- van der Geest R.J.,
- Zeppenfeld K.
- Brignole M.,
- Botto G.L.,
- Mont L.,
- et al.