Author + information
- Received February 20, 2018
- Revision received March 19, 2018
- Accepted March 21, 2018
- Published online July 16, 2018.
- Vidal Essebag, MD, PhDa,b,∗ (, )
- Jacqueline Joza, MDa,
- Pablo B. Nery, MDc,
- Steve Doucette, MScd,
- Isabelle Nault, MDe,
- Lena Rivard, MDf,
- Lorne Gula, MDg,
- Marc Deyell, MDh,
- Jean-Marc Raymond, MDi,
- Chris Lane, MDj and
- John L. Sapp, MDk
- aMcGill University Health Centre Research Institute, Montreal, Canada
- bHôpital Sacré-Coeur de Montréal, Montreal, Canada
- cResearch Methods Unit, University of Ottawa Heart Institute, Ottawa, Canada
- dDepartment of Community Health and Epidemiology, Dalhousie University, Halifax, Canada
- eInstitute Universitaire de Cardiologie et de Pneumologie de Quebec, Quebec, Canada
- fMontreal Heart Institute, Montreal, Canada
- gWestern University, London, Canada
- hSt. Paul’s Hospital, Vancouver, Canada
- iCentre Hospitalier de L’Universite de Montreal, Montreal, Canada
- jRoyal Jubilee Hospital, Victoria, Canada
- kDepartment of Medicine, QEII Health Sciences Centre and Dalhousie University, Halifax, Canada
- ↵∗Address for correspondence:
Dr. Vidal Essebag, Cardiac Electrophysiology, McGill University Health Centre, 1650 Cedar Avenue, Room E5-200, Montreal QC H3G 1A4, Canada.
Objectives This study sought to evaluate the predictive value of noninducibility on long-term outcomes.
Background The traditional endpoint for catheter ablation of ventricular tachycardia (VT) is noninducibility of VT by programmed stimulation; however, the definition of inducibility remains variable and its prognostic value limited by nonstandardized periprocedural antiarrhythmic drug therapy and implantable cardioverter-defibrillator programming in prior observational studies. The VANISH trial randomized patients with prior myocardial infarction and VT to ablation (with an endpoint of noninducibility of VT ≥300 ms after ablation) versus antiarrhythmic drug escalation.
Methods Patients enrolled in the VANISH study randomized to catheter ablation were included. The relationship between post-ablation inducibility and the primary composite endpoint (death, VT storm >30 days, or appropriate implantable cardioverter-defibrillator shock >30 days) was assessed using a time-to-event analysis, adjusting for other clinical and procedural characteristics.
Results A total of 129 patients from the ablation arm were included in the primary analysis, of which 51 were noninducible post-ablation compared with 78 who had inducible VT or in whom inducibility testing was not performed. There were no significant baseline characteristic or procedural differences except for increased implantable cardioverter-defibrillator shocks before randomization in the noninducible group. In multivariate analysis, inducibility significantly increased the risk of death, appropriate shock, or VT storm after 30 days (HR: 1.87; p = 0.017).
Conclusions Inducibility of any VT post-ablation was associated with an increased risk of the composite endpoint in the VANISH trial. A randomized trial is required to confirm whether more aggressive ablation targeting faster induced VTs (<300 ms) can improve outcomes.
Radiofrequency catheter ablation has emerged as an important adjunctive therapy for patients with ventricular tachycardia (VT) and prior myocardial infarction (MI) (1). The determination of acute procedural success following catheter ablation for VT has relied on noninducibility by programmed electrical stimulation (PES) (2). In patients with clinically presenting sustained monomorphic VT, the likelihood of inducing the same type of VT during electrophysiology study is 93%, suggesting that this should be a useful marker of procedural success. However, the usefulness of noninducibility as an acute endpoint relies not just on the reproducible induction of the clinical VT, but of any future clinically relevant VT (3). Furthermore, the temporal evolution of ablation lesions including lesion expansion (healthy myocyte loss) or regression (resolution of edema), short-term alterations in autonomic tone and loading conditions during the ablation procedure, and the long-term evolution of underlying coronary artery disease substrate have challenged the usefulness of inducibility as an endpoint (2).
Despite these considerations, the utility of inducibility has been assessed by the rate of spontaneous recurrences during long-term follow-up (4–6). Complete elimination of VT inducibility after ablation has been associated with reduced VT recurrence and cardiac mortality in observational studies (4,7), whereas others failed to confirm these results (8–10). The correlation of outcome to endpoints in these trials is derived from early reports that may not be applicable to contemporary patients undergoing VT ablation. Procedural techniques and strategies have evolved and a greater diversity of patients are now undergoing ablation (11). Moreover, current knowledge regarding the predictive value of noninducibility is limited to observational studies where ablation techniques, antiarrhythmic regimens, and implantable cardioverter-defibrillator (ICD) programming have been nonuniform.
The VANISH trial randomized patients with ischemic heart disease and VT to catheter ablation (with endpoint of noninducibility of VT ≥300 ms after ablation) versus antiarrhythmic drug (AAD) escalation (12). A significantly lower rate of the composite primary outcome of death, VT storm, or appropriate ICD shocks was found in patients undergoing catheter ablation. We sought to investigate the relationship between acute inducibility results following catheter ablation and long-term outcomes in patients with VT following MI.
VANISH population and procedure
Details of the VANISH study have been described previously (12). Briefly, 259 patients with ischemic cardiomyopathy and an ICD who had VT despite AAD therapy were randomized to either catheter ablation with continuation of baseline antiarrhythmic medications or to escalation of AAD therapy. Qualifying VT episodes occurred within 6 months of enrollment and were defined as any 1 of the following: 3 or more episodes of VT treated with antitachycardia pacing, of which at least 1 episode was symptomatic; 1 or more appropriate ICD shocks; 3 or more episodes of VT within 24 h; or sustained VT at a rate below the programmed detection rate of the ICD. Patients were followed for a mean of 27.9 ± 17.1 months.
This VANISH substudy population included all patients who underwent a catheter ablation procedure for VT (Figure 1). As per protocol, VT induction with PES from 2 ventricular sites at 2 drive cycle lengths (CLs) with up to triple ventricular extrastimuli, coupled not closer than 180 ms, was performed at the start of the procedure. If the clinical VT was not induced, isoproterenol could be administered at a dose adjusted to achieve a 30% increase in baseline heart rate. All clinical VTs and all induced VTs with a CL ≥300 ms were targeted for ablation. Although inducible VTs with CL <300 ms were not specifically targeted for ablation, there was no restriction on additional substrate ablation. If sustained hemodynamically tolerated VT was induced, activation and/or entrainment mapping with ablation was performed. If hemodynamically nontolerated VT was induced, a substrate ablation approach that could include pace-mapping was performed. If the clinical arrhythmia was VT with a CL <300 ms, or if the clinical VT was not inducible, then a purely substrate-based approach was used. The procedural endpoint post-ablation was no inducible VT with CL ≥300 ms with repeat induction testing post-ablation. Patients remained on unchanged AAD therapy unless contraindicated. ICDs were programmed according to a standardized protocol, and there was blinded adjudication of endpoints.
Substudy inducibility definition
For this substudy, noninducibility post-ablation was defined as no monomorphic VT (of any CL) induced with PES. This included any inducible clinical monomorphic VT, VT with CL ≥300 ms, or monomorphic VT <300 ms. Inducible polymorphic VT or ventricular fibrillation was not included in the inducible group.
The primary outcome was defined as a composite of death, VT storm (3 or more documented episodes of VT within 24 h), or appropriate ICD shock after a 30-day treatment period. The secondary outcome was defined as a composite of appropriate ICD shock at any time, VT storm at any time, and sustained VT below ICD detection limit at any time.
Patients were stratified according to noninducibility post-ablation, where inducible post-ablation and inducibility not performed were treated together as a distinct group. Baseline demographics and procedural characteristics of patients undergoing ablation were summarized as means and standard deviation and frequencies with percentage. Comparisons between inducibility categories were made using the Student t test or Fisher exact test as appropriate. Univariate associations between baseline characteristics and noninducibility were assessed using logistic regression and summarized as odds ratios and 95% confidence intervals (CIs). All variables showing an association at a significance level lower than 0.2 on univariate analysis were selected for an automated variable selection method entered into a final multivariable model of predictors of inducibility of any VT post-ablation (in the 129 patients randomized to VT ablation). Variables were chosen using a forward selection technique with criteria for entry set at a significance level of 0.05. Sensitivity analyses were performed excluding patients in whom inducibility post-ablation was not performed, and a separate analysis including 48 crossover patients from the AAD arm who underwent VT ablation. Analyses of long-term prognostic value of noninducibility were limited to the 129 patients randomized to ablation (given that management and follow-up of patients who went on to catheter ablation subsequently from the AAD arm was not standardized). Time-to-event analyses were used to assess the relationship of inducibility of any VT on the primary outcome (composite endpoint of death, VT storm, or appropriate ICD shock) and the secondary outcome (composite of appropriate ICD shock, VT storm, and sustained VT below ICD detection limit). Survival rates were summarized between inducibility groups using Kaplan-Meier product-limit estimates. Univariate and multivariate associations between baseline demographics on both outcomes were assessed using an identical selection technique, with inducibility of any VT forced in the final multivariate model. A separate sensitivity analysis with inducibility post-ablation redefined as any clinical VT or VT ≥300 ms (as opposed to any inducible VT) was also performed. Effect sizes were shown as hazard ratios (HR) and 95% CIs. All analyses were performed in SAS version 9.4 (SAS Institute, Cary, North Carolina).
Substudy patient population
The 129 patients who underwent catheter ablation as randomized in the VANISH trial were included in the primary analysis. Of these, 39.5% (51 of 129) patients were rendered noninducible for any VT, 41.9% (54 of 129) remained inducible with VT of any CL, and 18.6% (24 of 129) did not have final induction testing performed post-ablation. Baseline characteristics comparing noninducible patients with those who were either inducible or did not have final induction testing are shown in Table 1 (Online Table 1 displays characteristics for all 3 groups). No significant differences in baseline characteristics were noted except for a lower number of ICD shocks within the prior 3 months in the noninducible arm. No significant differences in procedural characteristics were present between the 2 groups as shown in Table 2 (procedural characteristics for all 3 groups are shown in Online Table 2). An additional 48 patients from the AAD arm underwent ablation totaling 177 patients (secondary sensitivity analysis). Of these, 37.3% (66 of 177) were noninducible, 42.9% (76 of 177) were inducible, and 19.8% (35 of 177) did not have PES performed post-ablation. This substudy did not include repeat ablation procedures in either arm.
Predictors of inducibility post-ablation
There were no baseline characteristics that were shown to be significantly predictive of whether VT was inducible post-ablation among the 129 patients in the ablation arm (Online Table 3). In a sensitivity analysis that excluded patients who did not have inducibility testing performed post-ablation, no baseline characteristics were found to be significantly predictive (Online Table 4). Similar findings were noted in a secondary sensitivity analysis, which included the 48 patients from the AAD arm who subsequently underwent VT ablation and excluded patients who did not have inducibility testing performed (Online Table 5).
Long-term prognostic value of noninducibility
In multivariate analysis including the 129 patients in the VT ablation arm, inducibility of VT post-ablation was associated with an increase in the primary outcome of death, appropriate shock, or VT storm after 30 days (HR: 1.87; 95% CI: 1.12 to 3.11; p = 0.017) (Table 3). The presence of diabetes mellitus was associated with a decrease in the primary outcome (HR: 0.50; 95% CI: 0.28 to 0.88; p = 0.017). Inducibility was also associated with an increase in the secondary composite endpoint of appropriate ICD shock at any time, VT storm at any time, and sustained VT below the ICD detection limit (HR: 2.63; 95% CI: 1.36 to 5.10; p = 0.004). The presence of diabetes mellitus was associated with a decrease in the secondary outcome (HR: 0.45; 95% CI: 0.23 to 0.89; p = 0.02) as was beta-blocker use (HR: 0.21; 95% CI: 0.09 to 0.52; p = 0.007). In a multivariate sensitivity analysis excluding patients for whom inducibility testing post-ablation was not performed, inducibility post-ablation remained predictive of primary and secondary outcomes (HR: 2.30; 95% CI: 1.31 to 4.03; p = 0.04; and HR: 2.19; 95% CI: 1.18 to 4.68), respectively.
VT CL criteria to define inducibility
The previously described primary analysis found that inducibility defined as any inducible VT (excluding polymorphic VT/ventricular fibrillation) significantly predicted primary and secondary outcomes (Table 3, Figures 2A and 2B). In a sensitivity analysis defining inducibility as any clinical VT or VT ≥300 ms, inducibility did not significantly predict outcomes (HR: 1.32; 95% CI: 0.82 to 2.12, p = 0.25; and HR: 1.46; 95% CI: 0.83 to 2.57, p = 0.19) for the primary and secondary outcomes, respectively (Table 3, Figures 2C and 2D). The Kaplan-Meier curve in Online Figure 1 demonstrates the primary outcome based on presence of noninducibility and testing not performed.
This analysis confirms that noninducibility of any VT after catheter ablation for patients with prior MI is significantly associated with a decreased risk of the primary outcome of a composite endpoint of death, appropriate shock, or VT storm after 30 days in addition to the secondary outcome of a composite of appropriate ICD shock at any time, VT storm at any time, and sustained VT below the ICD detection limit. Although the VANISH design allowed a procedural endpoint if only VT <300 ms was inducible, our analysis suggests that even these faster VTs may predict worse outcomes. These results are particularly important given that the VANISH trial was a randomized controlled trial with standardized ICD programming and AAD use.
Although we clearly found that patients rendered noninducible with VT ablation had a better prognosis, we were unable to identify any baseline clinical characteristics that were predictive of noninducibility in this patient population. Other studies have demonstrated the impact of procedural success on all-cause mortality (13). Patients with left ventricular ejection fraction <35% and those with New York Heart Association functional class III/IV were found to have the most benefit from acute procedural success defined as VT noninducibility and elimination of local abnormal ventricular activities (13). Future research with a larger sample size may help eventually determine predictive characteristics of patients for whom ablation is most likely to render VT noninducible and may be most likely to benefit from ablation therapy.
The ideal endpoint after catheter ablation for VT remains uncertain; however, VT noninducibility with PES has been consistent as a result of its practical application and presumed fundamental qualities. However, the lack of a consistent definition, variations in stimulation protocols, and inconsistent AAD use and ICD programming have cast doubt on its use as a procedural endpoint (2,14). Expert consensus guidelines define endpoints for ablation in patients with post-MI VT as: 1) noninducibility of the clinical VT; 2) modification of the induced VT CL; and 3) noninducibility of any VT (2). The basis for these guidelines is not well supported, because results from prior nonrandomized studies (4,6,8–10,15–20) have reported inconsistent results. A recent meta-analysis demonstrated a lower mortality risk in patients whose VT is rendered noninducible at procedure completion, particularly in those where elimination of all inducible VTs were performed (21); however, this meta-analysis analyzed cohort studies and yielded unadjusted estimates of risk. Other important studies of catheter ablation have been shown to reduce arrhythmia recurrence in patients with VT but were not powered to test whether a significant reduction of ICD therapies after ablation improves survival (8–10).
Of the larger trials demonstrating increased survival with noninducibility of the clinical VT as an endpoint, Sauer et al. (4) describe a total of 208 patients who underwent 327 VT ablations from 1999 to 2005. A long-term survival benefit was noted in patients who did not have any inducible clinical VT at the end of the ablation procedure (HR: for 3-year all-cause mortality 0.42; 95% CI: 0.21 to 0.85; p = 0.02). Notably, VT ablation performed after the year 2003 (with the implementation of electroanatomic mapping and irrigated radiofrequency technologies) was demonstrated to be more effective than previous in achieving a satisfactory long-term outcome (4). Tokuda et al. (15) in 2013 similarly found an association between persistence of inducible clinical VT and increased VT recurrence and mortality in 518 patients undergoing ablation. Other smaller studies have demonstrated consistent results (9,22).
Several trials used a defined ablation endpoint of any inducible VT. Della Bella et al. (17) performed a prospective analysis of 528 patients with nonischemic or ischemic VT who underwent 634 VT ablation procedures. Complete success (defined as noninducibility of any VT) was associated with a decreased incidence of combined cardiac death (8% vs. 18.7%; log-rank p < 0.001; HR: 0.539; p = 0.038), as compared with the presence of any inducible VT post-ablation or baseline noninducibility. Dinov et al. (6) demonstrated that the probability for VT recurrence in patients with a partially successful ablation (elimination of the clinical VT only) was almost 2-times higher as compared with those with complete elimination of any clinical or nonclinical stable monomorphic VT (HR: 1.9; 95% CI: 1.004 to 3.58; p = 0.048). Yokokawa et al. (20) retrospectively analyzed data from 1,064 patients who underwent ablation for post-infarction VT. Noninducibility of any VT was independently associated with lower mortality (adjusted HR: 0.65; 95% CI: 0.53 to 0.79; p < 0.001) (20). More recently, Fujii et al. (23) found that in 83 patients with ischemic cardiomyopathy that were rendered noninducible post-ablation, inducible nonsustained VT was independently associated with VT recurrence (HR: 3.66; 95% CI: 1.3 to 11.1).
Perhaps alternative endpoints after VT ablation should be sought. Komatsu et al. (13) analyzed 195 patients with ischemic cardiomyopathy and nonischemic cardiomyopathy who underwent substrate ablation targeting elimination of local abnormal ventricular activities. Success was defined as achievement of both identified local abnormal ventricular activities elimination and VT noninducibility. All-cause mortality was significantly lower in patients with acute procedural success (15% vs. 41%; log-rank p < 0.001) (13). Other acute endpoints, such as ablation of uniform premature ventricular complex morphologies that were documented to trigger spontaneous episodes of VT, remain to be clarified.
The VANISH ablation protocol included mapping and ablation approaches dependent on the hemodynamic stability and CL of the induced VTs. Although a substrate ablation protocol targeted late potentials within scar, there may have been interoperator differences in approaches to substrate ablation. Furthermore, both the hemodynamic stability of VT and the ability to induce VT may be influenced by patient and procedural factors including autonomic tone and anesthesia. Our study allowed the use of sedation or general anesthesia according to operator preference.
Inducibility of any VT post-ablation was associated with an increased risk of the composite outcome in the VANISH trial. Our analysis suggests that inducibility of nonclinical fast VTs (CL <300 ms) may predict worse outcomes; achievement of noninducibility of any VT as an acute procedural endpoint should be considered. A randomized trial targeting faster induced VTs is required to determine whether a more aggressive ablation strategy can improve outcomes.
COMPETENCY IN MEDICAL KNOWLEDGE: Ablation of VT in patients with ischemic heart disease has been shown to be superior to antiarrhythmic drug escalation following recurrent arrhythmia. Noninducibility of VT on programmed electrical stimulation has been routinely used as an endpoint for termination of ablation despite the absence of evidence in prior randomized studies.
TRANSLATIONAL OUTLOOK 1: Noninducibility of any VT post-VT ablation in patients with prior myocardial infarction is useful as an endpoint for termination.
TRANSLATIONAL OUTLOOK 2: Inducibility of VTs post-VT ablation that are faster than the clinical tachycardia may be a marker for a worse prognosis.
This study was supported by an operating grant from the Canadian Institutes of Health Research, with additional funding from St. Jude Medical and Biosense Webster, and a Clinical Research Scholar Award to Dr. Essebag from Fonds de recherché du Quebec-Santé. Canadian Arrhythmic Network (CANet) Investigators (to Drs. Essebag, Joza, Nery, Nault, Rivard, Gula, Deyell, Lane, Sapp). Dr. Essebag has received honoraria from Biosense Webster Inc., Boston Scientific, St. Jude Medical, Abbott, and Medtronic Inc. Dr. Rivard has received consulting fees from Biosense Webster. Dr. Deyell has received research funding from Biosense Webster; and honoraria from Biosense Webster and Abbott. Dr. Raymond has received lecture fees from Biosense Webster. Dr. Lane has received honoraria from Medtronic, Biosense Webster, Forest Laboratories Canada, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb/Pfizer, and Servier. Dr. Sapp has received research funding from Biosense Webster, Abbott, St. Jude Medical, and Philips; is a consultant for Biosense Webster; and received speaker honoraria from Abbott and Medtronic Inc. 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
- antiarrhythmic drug
- confidence interval
- cycle length
- hazard ratio
- implantable cardioverter-defibrillator
- myocardial infarction
- programmed electrical stimulation
- ventricular tachycardia
- Received February 20, 2018.
- Revision received March 19, 2018.
- Accepted March 21, 2018.
- 2018 American College of Cardiology Foundation
- Proietti R.,
- Joza J.,
- Essebag V.
- Aliot E.M.,
- Stevenson W.G.,
- Almendral-Garrote J.M.,
- et al.
- Dinov B.,
- Fiedler L.,
- Schonbauer R.,
- et al.
- Carbucicchio C.,
- Santamaria M.,
- Trevisi N.,
- et al.
- Stevenson W.G.,
- Wilber D.J.,
- Natale A.,
- et al.
- Calkins H.,
- Epstein A.,
- Packer D.,
- et al.
- Proietti R.,
- Roux J.F.,
- Essebag V.
- Sapp J.L.,
- Wells G.A.,
- Parkash R.,
- et al.
- Komatsu Y.,
- Maury P.,
- Sacher F.,
- et al.
- Priori S.G.,
- Blomstrom-Lundqvist C.,
- Mazzanti A.,
- et al.
- Tokuda M.,
- Kojodjojo P.,
- Tung S.,
- et al.
- Della Bella P.,
- Baratto F.,
- Tsiachris D.,
- et al.
- Yokokawa M.,
- Desjardins B.,
- Crawford T.,
- Good E.,
- Morady F.,
- Bogun F.
- Reddy V.Y.,
- Neuzil P.,
- Taborsky M.,
- Ruskin J.N.
- Yokokawa M.,
- Kim H.M.,
- Baser K.,
- et al.
- Ghanbari H.,
- Baser K.,
- Yokokawa M.,
- et al.
- Fukunaga M.,
- Goya M.,
- Hiroshima K.,
- et al.
- Fujii A.,
- Nagashima K.,
- Kumar S.,
- et al.