Author + information
- Received November 30, 2017
- Revision received January 22, 2018
- Accepted January 25, 2018
- Published online June 18, 2018.
- Jorge Romero, MDa,b,
- William G. Stevenson, MDa,c,
- Akira Fujii, MDa,
- Sunil Kapur, MDa,
- Samuel H. Baldinger, MDa,d,
- Nishaki K. Mehta, MDa,
- Roy M. John, MDa,c,
- Gregory F. Michaud, MDa,c,
- Laurence M. Epstein, MDe,
- Bruce A. Koplan, MDa,
- Usha B. Tedrow, MD, MSca and
- Saurabh Kumar, BSc(Med)/MBBS, PhDa,f,∗ ()
- aCardiovascular Division, Brigham and Women’s Hospital, Department of Medicine, Harvard Medical School, Boston, Massachusetts
- bDepartment of Medicine (Cardiology), Montefiore-Einstein Center for Heart and Vascular Care, New York, New York
- cArrhythmia and Electrophysiology Program, Department of Medicine, Vanderbilt University Medical Center, Nashville, Tennessee
- dDepartment of Cardiology, Bern University Hospital and University of Bern, Bern, Switzerland
- eZucker School of Medicine, Hofstra/Northwell, Northwell Health, Manhassett, New York
- fDepartment of Cardiology, Westmead Hospital, Sydney, Australia
- ↵∗Address for correspondence:
Dr. Saurabh Kumar, Department of Cardiology, Westmead Hospital, Hawkesbury Road, Sydney, New South Wales 2145, Australia.
Objectives This study sought to examine the relationship between the number of oral antiarrhythmic drug (AAD) failures before referral for ventricular tachycardia (VT) ablation and subsequent clinical outcomes.
Background Failure of AADs prompts referral for VT ablation.
Methods Consecutive patients (n = 669) with sustained VT who were referred for a first-time ablation were divided into 2 groups according to the number of oral Class 1 or 3 AAD failures before referral: single-drug failure (≤1 AAD; n = 256) or multidrug failure (>1 AADs; n = 413). Outcomes were stratified according to underlying disease type (no structural heart disease [SHD] [n = 87]; ischemic cardiomyopathy [ICM] [n = 368]; and ischemic cardiomyopathy [NICM] [n = 214]) and reported at a mean follow-up of 35 ± 46 months.
Results Patients with multidrug failure, compared with patients with single-drug failure, had more advanced SHD and required more extensive ablation to control arrhythmia. Multidrug failure, compared with single-drug failure, was associated with lower ventricular arrhythmia–free survival in ICM (46 ± 4% vs. 58 ± 6%; p = 0.03) and NICM (26 ± 5% vs. 49 ± 6%; p = 0.008), but not in the absence of SHD (71 ± 8% vs. 85 ± 7%; p = 0.10). Overall survival was lower in multidrug failure versus single-drug failure groups in patients with ICM (71 ± 3% vs. 84 ± 4%; p = 0.03) and NICM (70 ± 5% vs. 88 ± 4%; p < 0.001). Multidrug failure was independently associated with a higher risk of ventricular arrhythmia recurrence (hazard ratio: 1.6; p = 0.01) and mortality in NICM (hazard ratio: 2.6; p = 0.008), but not in ICM.
Conclusions Patients with SHD and failure of multiple oral AADs before VT ablation referral have more advanced heart disease and worse clinical outcomes following ablation, especially in NICM.
- antiarrhythmic drugs
- catheter ablation
- ischemic cardiomyopathy
- nonischemic cardiomyopathy
- ventricular tachycardia
Recurrent defibrillator shocks for ventricular arrhythmia (VA) lead to poorer quality-of-life metrics and increased mortality rates (1,2). Although antiarrhythmic drugs (AADs) can reduce VA recurrences, adjunctive percutaneous catheter ablation is even more efficacious in reducing VA burden (3,4). Referral for catheter ablation is usually prompted when AADs fail to control VA, an approach supported by recommendations from a recent consensus statement on management of VA (5). However, the relationship between the number of AAD failures before VT ablation referral and subsequent clinical outcomes is unknown. Failure of multiple drugs before referral for ablation may represent 1 of the following: 1) a group of patients with more advanced heart disease or comorbidities whose clinical threshold for referral for catheter ablation may be higher; 2) delayed referral for ablation as a result of the physician’s and/or patient’s preference; or 3) a combination of these factors. In this study, we examined the impact of the number of AAD failures before referral for VT ablation on outcomes after ablation among 3 distinct subgroups of patients: 1) absence of structural heart disease (SHD); 2) ischemic cardiomyopathy (ICM); and 3) nonischemic cardiomyopathy (NICM).
This was a retrospective single-center study comprising 669 consecutive patients who presented for their first percutaneous catheter ablation of sustained monomorphic VT between 1999 and 2014 at the Brigham and Women’s Hospital in Boston, Massachusetts. All patients gave written informed consent for the procedure. The Partners’ Institutional Review Board approved this retrospective analysis, conforming to the ethical guidelines of 1975 Declaration of Helsinki.
There were 87 patients with no SHD, 368 patients with ICM, and 214 patients with NICM. Drug failure was defined as the drug’s inability to control VT (breakthrough episodes) with the clinician prompting escalation of AAD (including addition of more than 1 AAD).
To explore the association between the number of drug failures and outcomes, we classified patients within each disease subtype into the number of failed orally administered Class 1 or 3 AADs before referral for catheter ablation. Patients were divided into 2 groups: 1) failure of no AAD or a single AAD to control VT (termed “single-drug failure” for this study; n = 256); or 2) failure of >1 AADs to control VT (“multidrug failure”; n = 413).
Use of beta-blockers, nondihydropyridine calcium-channel blockers, and/or use of intravenous Class 1 or 3 AADs as a temporizing measure, was not considered in the final count of number of drug failures.
All patients underwent echocardiography and/or magnetic resonance imaging to screen for the presence of SHD and to define ventricular function. The primary basis for the distinction between ICM and NICM was the presence of coronary artery disease confirmed with coronary angiography. NICM was defined according to the criteria of the European Society of Cardiology Working Group for Myocardial and Pericardial Diseases (6). Both left ventricular (LV) and right ventricular ICMs were included. Excluded from this study were the following: 1) patients with premature ventricular contractions or ventricular fibrillation initiated by premature ventricular contractions as the procedural indication; and 2) patients with any prior catheter ablation for VT.
Mapping and ablation
Our approach to percutaneous endocardial and epicardial mapping and ablation has been described previously (7). Briefly, programmed ventricular stimulation was performed with 1 to 3 extrastimuli delivered after a drive train of 600 ms and 400 ms from 2 right ventricular sites. This was repeated from at least 1 LV site, if VT was noninducible from right ventricular stimulation in patients undergoing LV mapping. Isoproterenol or epinephrine was used at the discretion of the operator. The stimulation protocol was repeated after ablation unless it was thought that it would jeopardize the patient’s hemodynamic stability. The morphology of the induced VTs were compared with those of spontaneously occurring VTs before ablation when available. Sustained monomorphic VT was defined as continuous VT for ≥30 s or a single VT that required an intervention for termination (cardioversion, pacing) (8).
We defined “spontaneous VT” as any inducible VT with an identical 12-lead electrocardiographic morphology and rate (within 20 ms) compared with a clinical VT occurring spontaneously. If 12-lead electrocardiograms of the presenting, spontaneous VT were not available before ablation, the rate cutoff and intracardiac electrogram morphology from the implantable cardioverter-defibrillator (ICD) tracings were compared with the inducible VTs (9). “Undocumented VTs” were defined as inducible sustained monomorphic VTs that did not match the clinical VT occurring spontaneously (i.e., nonclinical VT).
For VT associated with SHD, substrate mapping was performed with a particular focus on the scar region. Mapping was performed with an ablation catheter or, when available, a multispline catheter (PentaRay, Biosense Webster, Diamond Bar, California) and the CARTO electroanatomic mapping system (Biosense Webster). Areas of low voltage (<1.5 mV) (10) and electrically unexcitable scar were identified (11). As previously described, ablation was guided by substrate mapping during sinus rhythm and by limited assessment with activation and entrainment mapping during VT (7). Ablation targets were presumptive channels and exits within low-voltage areas identified from a paced QRS interval morphology yielding a ≥10/12 pace-map match to the VT QRS interval morphology abnormal fractionated potentials, double potentials, or late potentials during sinus or paced rhythm at sites where pacing captured, particularly with stimulus–QRS intervals of >40 ms. If no low-voltage area was identified, ablation was attempted at the likely exit region identified as sites with pre-systolic electrograms during VT or where pace mapping yielded a 12/12 pace-map match to the targeted VT QRS interval morphology. If the procedure was hemodynamically tolerated, VT was induced again, and activation or entrainment mapping was performed. If the procedure was not tolerated, VT was terminated with radiofrequency ablation, burst pacing or cardioversion, and further ablation guided largely by substrate mapping.
Radiofrequency ablation was delivered with an irrigated catheter (ThermoCool, ThermoCool SF, Biosense Webster) at a power of 30 to 50 W targeting an impedance drop of 10 to 18 ohms. Before the approval of ThermoCool irrigated catheters in the United States (∼2006), a nonirrigated 4-mm-tip ablation catheter, a nonirrigated 8-mm-tip ablation catheter, or an internally irrigated ablation catheter (Boston Scientific, Marlborough, Massachusetts) was used. Applications were repeated at target areas until they were rendered electrically unexcitable with unipolar pacing at 10 mA at 2-ms pulse width (11). Epicardial mapping was performed using the percutaneous approach if VT was suspected to be of epicardial origin, generally after endocardial ablation failed to abolish inducible VT. 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.
The approach to the ablation of idiopathic VT relied on a combination of assessment of putative origin of VT on the basis of 12-lead electrocardiographic morphology activation mapping during VT, assessment of pre-potentials, and entrainment mapping (if possible) for fascicular VT and/or pace mapping when the VT was not reliably sustained or hemodynamically tolerated (for papillary muscle, LV summit, or right ventricular outflow tract VTs). Voltage mapping and entrainment mapping were performed to exclude scar-mediated re-entry.
Antiarrhythmic drug and ICD management post ablation
In patients with SHD, AADs prescribed long term were either continued at the same dose or a reduced dose or were discontinued as recommended by the treating physician. AADs initiated recently to control multiple VTs or incessant VT were discontinued. In patients with no SHD, AADs were generally discontinued at the time of ablation or after the first follow-up visit at 3 months if the procedure was deemed acutely successful on the basis of noninducibility of the clinical VT and the absence of recurrence in follow-up, but it was ultimately at the discretion of the referring physician. Details of defibrillator programming post ablation are provided in the Online Methods section.
Acute procedural outcomes were reported as follows:
1. Complete success: defined as noninducibility of any VT, either “spontaneous” or “undocumented”;
2. Partial success: defined as abolition of at least 1 “spontaneous” VT, but other “spontaneous” or “undocumented” VTs remained inducible;
3. Failure: persistent inducibility of “spontaneous” VT.
Major complications were reported; definitions are shown in the Online Methods section.
In follow-up, outcomes reported were as follows:
1. VA-free survival: defined as absence of any sustained VT or ventricular fibrillation recurrence;
2. Survival free of cardiac transplantation;
3. Overall survival.
Follow-up was obtained from review of records of all hospital and outpatient clinic visits and from referring cardiologists, electrophysiologists, and primary care physicians. In patients with an ICD, all device interrogations were analyzed. The National Social Security Death Index was searched for mortality information. Follow-up time was defined from the date of first ablation to the time of death or last clinical contact. Procedural outcomes were reported at mean follow-up after a single procedure.
The Statistical Package for the Social Sciences for Windows (SPSS release 23, IBM, Armonk, New York) was used for analysis. Continuous variables were expressed as mean ± SD if normally distributed; median and interquartile range 25% to 75% (Q25 to Q75) or full ranges were used if the data were clearly skewed. Where normal distribution was not present, log transformation of the raw values was performed to meet the assumption of homogeneity of variance. Where applicable, paired sample Student’s t-tests were performed using raw values (if normally distributed) or log-transformed values (if not normally distributed). Acute procedural success and complications were compared as categorical variables with the Fisher exact test. Overall survival, survival free of VA, and transplant-free survival were estimated using the Kaplan-Meier procedure and were reported at mean follow-up. Cox proportional hazard models were created to determine factors associated with VA recurrence and all-cause mortality, specific to heart disease type. Models predicting mortality used VA recurrence as a time-dependent covariable. Hazard ratios (HRs) and 95% confidence intervals (CIs) were used to express risk of VA recurrence and mortality. A 2-tailed p value <0.05 was considered statistically significant.
Single-drug and multidrug failures were present, respectively, in 55 and 32 patients in the absence of SHD, in 107 and 261 patients with ICM, and in 94 and 120 patients with NICM. Age and sex were not significantly different between patients with the single-drug and multidrug failures within disease groups (Table 1). In both ICM and NICM patient groups, patients with multidrug failure had greater impairment in LV function and a higher proportion of patients with a New York Heart Association functional heart failure class ≥II and a higher incidence of failure of amiodarone, sotalol, and mexiletine to control VA. Use of concurrent beta-blockers did not significantly differ between the groups (Table 1). Baseline characteristics were similar in the patients with multidrug versus single-drug failure in the absence of SHD.
Sustained monomorphic VT was the procedural indication for all patients. A history of VT storm or incessant VT was more common with multidrug failure compared with single-drug failure in patients with all 3 types of heart disease (Table 2). In patients with ICM and NICM (but not patients with no SHD), those with multidrug failure had a greater number of inducible VTs and a higher incidence of slower VTs (Table 2). There was no significant difference in the incidence of epicardial ablation with single-drug versus multidrug failure in patients with any type of heart disease. Radiofrequency ablation time was significantly longer in the multidrug failure group compared with the single-drug failure group in patients with ICM and NICM, but not in patients without SHD (Table 2). Procedure time was longer in the multidrug failure group compared with the single-drug failure group only in patients with NICM.
Acute ablation outcomes
There were no differences in acute procedural complete success, partial success, or failure in the multidrug versus single-drug failure groups in the absence of SHD, ICM, or NICM (Table 2). Major complications were numerically higher in the multidrug versus single-drug failure groups in all groups, but not significantly so (Table 2).
Ventricular arrhythmia recurrences at mean follow-up
Mean follow-up from the first ablation was 35 ± 46 months (median 14 months). In patients with no SHD, there was no significant difference in VA-free survival between the patients with multidrug failure and those with single-drug failure (71 ± 8% vs. 85% ± 7%, respectively; p = 0.1 at mean follow-up of 35 months).
In ICM, VA-free survival was worse in patients with multidrug failure versus those with single-drug failure (46 ± 4% vs. 58 ± 6%; p = 0.03) (Figure 1A). In NICM, VA-free survival was substantially worse in patients with multidrug versus single-drug failure at mean follow-up (26 ± 5% vs. 49 ± 6%; p = 0.008 at mean follow-up) (Figure 1B). Analysis of Kaplan-Meier curves showed an early separation that persisted over follow-up (Figures 1A and 1B).
In multivariable analysis, multidrug failure was associated with an increased risk of VA recurrence in patients with NICM (HR: 1.6; 95% CI: 1.1 to 2.4; p = 0.01), but not in patients without SHD (Online Tables 1 and 2). In patients with ICM, multidrug failure was associated with an increased risk of VA recurrence in univariable analysis (HR: 1.5; 95% CI: 1.02 to 2.2; p = 0.04), but not in multivariable analysis (Online Table 3).
Overall survival and survival free of cardiac transplantation
There were no deaths in patients with no SHD. At mean follow-up of 35 ± 46 months, overall survival was worse in multidrug failure groups versus single-drug failure groups with ICM (71 ± 3% vs. 84 ± 4%; p = 0.03) (Figure 2A). In patients with NICM, those with multidrug failure also had worse survival compared with those with single-drug failure (70 ± 5% vs. 88 ± 4%; p < 0.001) (Figure 2B)
No cardiac transplantations were performed in patients without SHD. Transplant-free survival was similar in multidrug failure versus single-drug failure groups with ICM at mean follow-up (96 ± 2% vs. 100%; p = 0.90). Transplant-free survival was significantly worse in multidrug failure versus single-drug failure groups at mean follow-up (80 ± 5% vs. 96 ± 3%; p = 0.01).
Predictors of mortality
Mortality was used as the endpoint for Cox regression analysis (1-survival). In patients with NICM, multidrug failure was associated with increased mortality on multivariable analysis (HR: 2.6; 95% CI: 1.2 to 5.4; p = 0.008) (Online Table 4). In patients with ICM, multidrug failure was associated with mortality in univariable analysis (HR: 1.6; 95% CI: 1.04 to 2.5; p = 0.03), but not in multivariable analysis (Online Table 5). Amiodarone failure, however, was independently associated with mortality in ICM (HR: 1.9; 95% CI: 1.1 to 3.2; p = 0.01).
Outcomes with classification into 3 tiers of drug failures
Further subanalysis was performed according to failure of no AAD or a single AAD (≤1 AAD), 2 to 3 AADs, or >3 AADs (Table 3). At mean follow-up of 35 ± 46 months, VA-free survival was significantly worse among patients with NICM with an increasing number of drug failures. Overall survival was also worse in patients with NICM with an increasing number of drug failures (Table 3).
Among patients with ICM, VA-free survival was worse with failure of 2 to 3 AADs compared with ≤1 AAD (Table 3). Overall survival was worse with an increasing number of drug failures in patients with ICM (Table 3).
Impact of amiodarone failure
Amiodarone failed in 408 patients with SHD, and 307 of these patients had multidrug failures. In patients with ICM, those with a history of any prior amiodarone failure had worse VA-free survival when compared with patients with nonamiodarone drug failure (44 ± 4% vs. 70 ± 6%; p = 0.005) and worse overall survival (70 ± 3% vs. 92 ± 4%; p = 0.003). Amiodarone failure (compared with failure of other AADs) was associated with an increased risk of VA (HR: 1.6; 95% CI: 0.99 to 2.6; p = 0.06) and mortality (HR: 1.9; 95% CI: 1.1 to 3.2; p = 0.01) on multivariable analysis in patients with ICM (Online Tables 3 and 5).
In patients with NICM (with either single-drug or multidrug failure), those with a history of any prior amiodarone failure had similar VA-free survival (35 ± 5% vs. 36 ± 7%; p = 0.6), but worse overall survival (69 ± 5% vs. 93 ± 3%; p < 0.001). Amiodarone failure (compared with failure of other AADs) was not an independent predictor of VA recurrence or mortality on multivariable analysis (Online Tables 2 and 4).
When examining only those patients with single-drug failure before ablation referral, the lower VA-free survival for those with amiodarone failure did not reach statistical significance (ICM: 46 ± 8% vs. 60 ± 16%; p = 0.2; NICM: 50 ± 11% vs. 49 ± 11%; p = 0.95). However, overall survival was worse in patients with amiodarone failure when compared with patients with nonamiodarone drug failure (ICM: 76 ± 6% vs. 94 ± 5%, p = 0.02; NICM: 77 ± 8% vs. 96 ± 4%; p = 0.08).
In this study, we examined the relationship between the number of oral AAD failures before referral and subsequent outcomes after percutaneous catheter ablation of VT. We followed a large group of 669 patients undergoing their first VT ablation for a mean of 35 months and examined outcomes stratified by underlying heart disease (no SHD, ICM, and NICM). The pertinent findings of this study were as follows:
1. Patients with NICM-related VT who were referred for ablation after failure of multiple, orally administered Class 1 or 3 AADs had worse heart disease, more inducible VTs, and substantially worse outcomes of arrhythmia recurrence, heart transplantation, and mortality compared with patients referred for ablation following no drug failure or single-drug failure. The differences in outcomes were even more compelling with increasing numbers of drug failures (≤1 vs. 2 to 3 vs. failure of >3 AADs) before referral.
2. Although multidrug failure was associated with worse VA-free survival and overall survival among patients with ICM, the association was weaker and was primarily influenced by worse outcomes in patients with amiodarone failure versus failure of other AADs.
3. Multidrug failure was not associated with worse outcomes in patients without SHD.
In ICM, the association with better outcomes for referral before multiple-drug failure was also evident, but this was predominantly influenced by worse outcomes in patients with failed amiodarone treatment. This finding is consistent with those reports in the VANISH (Ventricular Tachycardia Ablation versus Escalation of Antiarrhythmic Drugs) study, where any amiodarone failure was associated with worse outcomes in ICM (4). In NICM, the differences in outcomes (VA recurrence, transplantation, mortality) between multidrug failure and no drug failure or single-drug failure were most pronounced, a finding that was anticipated but not previously reported. In contrast to patients with SHD, multidrug failure did not confer any difference in outcomes in the absence of SHD.
Our study suggests that the worse clinical outcomes in patients with multidrug failure versus single-drug failure may be attributable to the former group’s having more advanced heart disease with likely more complex VT substrate that was more difficult to treat compared with patients referred after no drug failure or single-drug failure, thus leading to inevitably poorer clinical outcomes after ablation. One other possibility is that patients with multidrug failure represented “late” referrals for ablation, whereas patients with no drug failure or single-drug failure represented “early” referrals for ablation. Given that data on timing between VT onset and referral for ablation were not available, we could not differentiate between late referral per se and more advanced heart disease as the predominant factor driving the worse outcomes. However, it is plausible that patients who had failures of multiple orally administrated AADs were more likely to have undergone ablation at a later time point in their disease process compared with patients who had only a single-drug failure or no AAD failure. A study where patients are well matched for clinical characteristics, randomized to early vs. late ablation, is needed to confirm or refute our findings.
Two prior studies specifically evaluated the relationship between timing of referral and catheter ablation for monomorphic VT in patients with SHD (12,13). In both studies, the time elapsed from defibrillator therapy to ablation was used to define the metric of early vs. late ablation. Frankel et al. (13) studied 98 patients with either ICM or NICM and compared outcomes of patients referred “late” (defined as those with 2 or more episodes of VT, with the first and most recent episode separated by at least 1 month) versus “early” (everybody else outside of this definition) for VT ablation. In that study, similar to ours, a greater proportion of patients in the late referral group vs. the early referral group had failures of amiodarone (62.9% vs. 47.2%; p = 0.05) or other nonamiodarone AADs (62.9% vs. 50%; p = 0.3). In contrast to our study, however, the mean ejection fraction and the incidence of VT storm were not higher in the late referral group compared with the early referral group. Also similar to our findings, Frankel et al. (13) reported better VT-free survival at 1 year with early compared with late referral, although the sample size was not large enough to observe a mortality difference between the groups.
Dinov et al. (12) reported outcomes of 300 patients with either ICM or NICM who were classified as “early” (within 30 days), “delayed” (1 month to 1 year), or “very late” (>1 year) after the first documented VT. Similar to our study, the very late referral group had worse ejection fractions, a higher incidence of New York Heart Association functional heart failure class >II, a higher burden of VT, a higher incidence of VT storm, a numerically higher incidence of failed Class 1 or 3 AADs, more inducible VTs, and longer ablation and procedure times compared with the early ablation group. Also similar to our study, the very late ablation group had a higher incidence of VT recurrence (64.5%) compared with the delayed (62%) and early ablation (37%) groups (p < 0.001). Outcomes were not stratified according to type of SHD, and the sample size was not large enough to observe a mortality difference among the groups (12).
Although the definition of “early” versus “late” referral varied between the aforementioned studies and ours, the numerically higher incidence of failed Class 1 or 3 AADs, worse heart disease, and worse outcomes in patients with delayed referral suggest that the patients captured within each of these groups, regardless of method of classification, are indeed likely to be similar. The incremental knowledge provided by our study is a 2-fold larger sample size compared with the large study by Dinov et al. (12) and stratification of outcomes into ICM and NICM, which are 2 disease processes known to have markedly different natural histories and outcomes.
Prior studies used the time elapsed from ICD shock (either from the first episode of VT or between consecutive VT episodes) as the metric for classification of early versus late referral. VT recurrences however, can be highly variable in pattern and clustering, with some patients experiencing a single shock followed by a long quiescence, whereas others can experience episodes with greater frequency (14,15). Thus, timing from VT event to referral for ablation for dichotomizing early versus late ablation does not take into account the clustering, frequency, and severity VT events or the intervening complexities of adjunctive AAD intervention initiated by treating physicians when faced with patients with recurrent VT. However, using the number of failed AADs for defining early versus late referral is also fraught with assumptions. There may have been other potential clinical reasons that may not have been captured in this study that would have led clinicians to refer patients either earlier or with delay, independent of the actual number of AADs used. These reasons may be related to severity and distribution of ventricular scarring, VT rate, specific subtype of cardiac disease, presence of absence of heart failure, tolerability of the usual doses of AADs, and other comorbidities. Thus, multidrug failure may indeed be a marker for disease progression over time, and a number of relevant factors may lead to escalating drug therapy before ablation referral. Whether early referral to ablative therapy (vs. late referral) leads to improvement in clinical outcomes after ablation warrants further prospective study. Further study is needed to differentiate whether multidrug failure represented delayed referral for ablation or an inherently sicker population with inevitably worse outcomes. Nevertheless, the results of our study have important clinical implications in that patients with multidrug failure who presenting for VT ablation may require more intensive pre-operative optimization and post-procedural monitoring and care to compensate for potentially worse outcomes after ablation. Furthermore, the magnitude of difference between multidrug and single-drug failures could serve to power future prospective trials.
The study is retrospective. Unrecognized clinical factors may have influenced whether patients were referred before failure of single or multiple drugs. This report is from a high-volume center for VT ablation, and referral biases are present. It is unclear whether the same results are generalizable to less experienced centers. Consequently, results may be skewed to the sickest cohort of patients with ICM and NICM. Patients were also drawn from a 15-year study period; hence differences in technology and VT ablation technique may have had an impact on procedural outcome. A large sample size accumulated over years would be necessary to explore outcome differences. Appropriately, this is a large series comparing multidrug with single-drug failure before referral for catheter ablation with has a long duration of follow-up; it suggests a relationship between the number of drug failures before referral for ablation and death after catheter ablation. The lack of a significance difference with early versus delayed ablation in the no-SHD group is likely the result of small numbers. Further prospective study is needed in this group.
The reasons for worse outcomes associated with multidrug failure cannot be definitely established from our study. Our study did not have a control arm of patients who did not undergo ablation. It is more likely that patients whose condition remains stable with 1 AAD would not be referred for VT ablation, whereas patients with failure of multiple AADs are more likely to be referred for VT ablation. However, the question of VT ablation referral versus continuation or escalation of AAD use has been recently addressed (4). Nevertheless, our study has an inevitable selection bias such that patients with a long history of disease and VT will also be more likely to experience ablation failure as well. The possibility of drug proarrhythmia is also acknowledged, although distinction from true drug failure can be challenging. We did not have data on drug dosing or drug changes after ablation, which may have had an effect on outcomes. Multidrug failure may be a marker for worse heart disease, as was present with extensive arrhythmia substrate, consistent with the greater number of inducible VTs. Exposure to drug toxicities could conceivably play a role. Whether earlier ablation would improve outcomes cannot be inferred from our data.
Patients with SHD who underwent de novo catheter ablation for VT after failure of multiple AADs had worse heart disease, more extensive arrhythmia substrate and procedural characteristics, and worse clinical outcomes of VT recurrence and death when compared with patients who had no AAD failures or a single-AAD failure before referral. Differences in clinical outcomes were more pronounced in NICM. Further study is needed to differentiate whether multidrug failure represented delayed referral for ablation or an inherently sicker group of patients with inevitably worse outcomes.
COMPETENCY IN MEDICAL KNOWLEDGE: Patients with SND who are referred for catheter ablation for VT after failure of >1 oral AAD have worse heart disease, more complex VT substrate, and worse clinical outcomes of arrhythmia recurrence and mortality compared with patients referred after failure of no AAD or a single AAD drug for the control of VT.
TRANSLATIONAL OUTLOOK 1: Whether early referral for catheter ablation, before the failure of multiple oral AADs fail to control VT will result in better outcomes compared with late ablation (after multidrug failure) needs to be studied in a prospective randomized trial.
TRANSLATIONAL OUTLOOK 2: Further work is needed to create a standard definition of what is considered an early or late referral for VT ablation.
Dr. Kumar is a recipient of the Neil Hamilton Fairley Overseas Research scholarship co-funded by the National Health and Medical Research Council and the National Heart Foundation of Australia; and the recipient of the Bushell Traveling Fellowship funded by the Royal Australasian College of Physicians. Dr. Tedrow has received consulting fees or honoraria from Boston Scientific and St. Jude Medical; has received research funding from Biosense Webster and St. Jude Medical; and is on the Fellowship Course faculty for St. Jude Medical, Medtronic, Biosense Webster, and Boston Scientific. Dr. John has received consulting fees or honoraria from St. Jude Medical; and receives lecture honoraria from Biosense Webster and Abbott. Dr. Michaud has received consulting fees or honoraria from Boston Scientific, Medtronic, and St. Jude Medical; has received research funding from Boston Scientific and Biosense Webster; and has a received teaching honoraria from Biotronik. Dr. Epstein is a consultant with Spectranetics, Medtronic, and Abbott; and has received speaker fees from Spectranetics, Medtronic, and Abbott. Dr. Koplan has received research grants from Biosense Webster and St. Jude Medical. Dr. Stevenson is co-holder of a patent for needle ablation that is consigned to Brigham and Women’s Hospital; and has received speaker honoraria from Boston Scientific and Abbott Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Francis Marchlinski, MD, served as Guest Editor for this paper.
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
- antiarrhythmic drug
- confidence interval
- hazard ratio
- implantable cardioverter-defibrillator
- ischemic cardiomyopathy
- left ventricular
- nonischemic cardiomyopathy
- structural heart disease
- ventricular arrhythmia
- ventricular tachycardia
- Received November 30, 2017.
- Revision received January 22, 2018.
- Accepted January 25, 2018.
- Santangeli P.,
- Muser D.,
- Maeda S.,
- et al.
- Sapp J.L.,
- Wells G.A.,
- Parkash R.,
- et al.
- Baldinger S.H.,
- Nagashima K.,
- Kumar S.,
- et al.
- Aliot E.M.,
- Stevenson W.G.,
- Almendral-Garrote J.M.,
- et al.
- Yoshida K.,
- Liu T.Y.,
- Scott C.,
- et al.
- Marchlinski F.E.,
- Callans D.J.,
- Gottlieb C.D.,
- Zado E.
- Soejima K.,
- Stevenson W.G.,
- Maisel W.H.,
- Sapp J.L.,
- Epstein L.M.
- Dinov B.,
- Arya A.,
- Bertagnolli L.,
- et al.
- Dorian P.,
- Borggrefe M.,
- Al-Khalidi H.R.,
- et al.