Author + information
- Received March 27, 2012
- Revision received July 11, 2012
- Accepted July 16, 2012
- Published online January 8, 2013.
- Miki Yokokawa, MD,
- Benoit Desjardins, MD, PhD,
- Thomas Crawford, MD,
- Eric Good, DO,
- Fred Morady, MD and
- Frank Bogun, MD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Frank Bogun, Cardiovascular Center, SPC 5853, University of Michigan, 1500 E. Medical Center Drive, Ann Arbor, Michigan 48109-5853
Objectives The purpose of this study was to assess the determinants of ventricular tachycardia (VT) recurrence in patients who underwent VT ablation for post-infarction VT.
Background The factors that predict recurrence of VT after catheter ablation in patients with prior infarctions are not well described.
Methods Catheter ablation was performed in 98 consecutive patients (88 males [90%]; mean age 67 ± 10 years; ejection fraction 27 ± 13%) with post-infarction VT. Electrograms from the implantable cardioverter-defibrillator were analyzed, and VTs were classified as clinical, nonclinical, or new clinical.
Results A total of 725 VTs were induced during the ablation procedure. All VTs were targeted. In 76 patients, 105 clinical VTs were inducible. Critical sites were identified with entrainment mapping and pace-mapping (≥10 of 12 matching leads) for 75 of 105 clinical VTs (71%) and for 278 of 620 nonclinical VTs (45%). Post-ablation, the clinical VT was not inducible in any patient, and all VTs were rendered noninducible in 63% of the patients. Over a mean follow-up period of 35 ± 23 months, 65 of 98 patients (66%) had no recurrent VTs and 33 (34%) had VT recurrence. A new VT occurred in 26 of 33 patients (79%), and a prior clinical VT recurred in 7 patients (21%). Patients with recurrent VT had a larger scar area as assessed by electroanatomic mapping compared with patients without recurrent VTs (93 ± 40 cm2 vs. 69 ± 30 cm2; p = 0.002). In patients with repeat procedures, the majority of inducible VTs for which a critical area could be identified were at a distance of 6 ± 3 mm to the prior ablation lesions.
Conclusions Patients with recurrent VTs have a larger scar as assessed by electroanatomic mapping. Most recurrent VTs were new, and the majority of these VTs were mapped to the vicinity of prior ablation lesions in patients with repeat procedures.
Although ablation for post-infarction ventricular tachycardia (VT) is effective in reducing frequent implantable cardioverter-defibrillator (ICD) discharges, recurrent VTs still occur in many patients depending on the duration of follow-up. The reasons for recurrence are unclear because most of these arrhythmias are only documented by an ICD. ICD electrograms have been shown to have distinct morphological characteristics enabling different VT morphologies to be distinguished (1). The purpose of the current study was to assess the reasons for VT recurrence based on comparisons of ICD electrograms recorded during and after the ablation procedure.
Ninety-eight patients (88 [90%] male; mean age 67 ± 10 years; ejection fraction 27 ± 13%) with post-infarction VT and an implanted ICD were referred for catheter ablation of VT between January 2004 and April 2011 (Table 1). All patients had a history of myocardial infarction (anterior in 32 [33%], inferior in 66 [67%]). Before ablation, all patients had failed to respond to antiarrhythmic drug treatment. Sixty-one patients (62%) were being treated with amiodarone, and 15 (25%) of the 61 were treated with a combination of amiodarone and mexiletine. Fourteen patients (14%) were treated with sotalol, 3 (3%) with mexiletine, 3 (3%) with dofetilide, and 1 (1%) with verapamil at the time of ablation.
The indications for VT ablation were ICD discharges in 84 patients (86%), recurrent palpitations in 26 (27%), and incessant VT in 13 (13%). Ten patients (10%) had undergone an unsuccessful ablation procedure before referral. Before the ablation procedure, a stress test and/or cardiac catheterization was performed. Nineteen patients had a nuclear stress test, 67 patients had a cardiac catheterization, and 5 patients had both a nuclear stress test and cardiac catheterization. Two patients underwent an exercise stress test. In 5 patients with VT storm and renal insufficiency, neither a stress test nor a catheterization was performed. In 11 patients, a revascularization procedure was performed before the procedure.
Electrophysiology study, mapping, and ablation
The study protocol was approved by the institutional review board at the University of Michigan. After written informed consent was obtained, a multipolar electrode catheter was inserted into a femoral vein and was positioned in the right ventricular apex. A 7-F multipolar catheter was placed at the His bundle position. Programmed ventricular stimulation was performed from 2 right ventricular sites using up to 4 extrastimuli (2). The protocol was continued until either refractoriness or a coupling interval of 200 ms was reached with a minimum of 3 extrastimuli or if ≥5 shocks were delivered to terminate VT. For all induced VTs, ICD recordings were performed with near- and far-field recordings, and these were compared with the previously documented VTs from the ICD. If none of the clinical VTs was induced, programmed stimulation was repeated with isoproterenol at 2 μg/min if there was no inducible ischemia present before the ablation procedure.
An electroanatomic mapping system (CARTO, Biosense Webster, Inc., Diamond Bar, California) was used in all patients, with an 8-F mapping/ablation catheter that had a 3.5-mm irrigated tip electrode and a 2-mm ring electrode separated by 1 mm (Thermocool, Biosense Webster, Inc.). In 7 of 98 patients, a 4-mm tip catheter was used; in 2 patients, an 8-mm tip catheter was used; and in 3 patients, both a 4-mm and an 8-mm tip catheter were used. In the remaining 86 patients, a 3.5-mm irrigated-tip catheter was used. Intracardiac electrograms were filtered at 50 to 500 Hz. The intracardiac electrograms and leads V1, I, II, and III were displayed on an oscilloscope and recorded at a speed of 100 mm/s. Systemic heparinization with an activated clotting time ≥250 s was maintained throughout the procedure.
Left ventricular access was obtained by using a retrograde aortic approach and a trans-septal approach in 3 patients. A left ventricular endocardial voltage map with a mean of 429 ± 184 points was constructed during sinus rhythm in 81 patients and during a paced rhythm in 17 patients. Pace-mapping was performed at low voltage, which was defined as a bipolar voltage <1.5 mV. A total of 347 ± 169 low-voltage sites were sampled per patient. Capture of at least 2 consecutive beats to allow for adequate analysis for pace-mapping was achieved at 56 ± 23 sites per patient. Pace-mapping was not performed if the catheter was close to a site (≤3 mm) where pace-mapping was previously performed.
Bipolar pace-mapping was performed at 10 mA and a pulse width of 2 ms. If no capture occurred, the pacing output was increased to 20 mA. Post-ablation stimulation was repeated at the same pulse strength, and ablation was repeated if capture was still present.
Radiofrequency energy was delivered at a power of 35 to 50 W targeting an impedance drop of 10 ohms using a 3.5-mm open-irrigated tip catheter (Thermocool, Biosense Webster). For an 8-mm tip catheter (n = 5), a power of 70 W was used, with a target temperature of 60 C°. For 4-mm nonirrigated tip catheters (n = 7), a power of 50 W with a temperature of 60°C was used.
All inducible VTs were targeted for ablation. Radiofrequency ablation was performed at critical sites of the VT re-entry circuit. A critical site was defined as a site showing concealed entrainment with matching stimulus-QRS and electrogram-QRS intervals (3) or where VT terminated during pacing without global capture (4). Entrainment mapping was performed within the low-voltage area. Sites with pre-systolic or discrete systolic activity were sought (5–7). Entrainment mapping was performed at these sites, and radiofrequency energy was delivered if concealed entrainment was demonstrated or if there was VT termination by a subthreshold pacing stimulus. If delivery of radiofrequency energy failed to terminate VT within 30 to 60 s, the catheter was moved to an alternative site.
If a VT was not hemodynamically tolerated, a site was considered critical if there was a matching pace-map. A QRS-morphology match was required in at least 10 of 12 leads. Once a site with a matching pace-map was identified, the site was labeled and radiofrequency energy was delivered. After assessing for capture post-ablation, the catheter was moved to a different site where pace-mapping was again performed. If a hemodynamically tolerated VT was reproducibly inducible, this VT was targeted with entrainment mapping initially. This was followed by pace-mapping throughout the scar.
Radiofrequency energy was delivered for 60 to 120 s and/or until the capture threshold post-ablation was >10 to 20 mA. When there were matching pace-maps, radiofrequency ablation was performed in the entire area in which the match was present. If there was a prior inferior wall myocardial infarction, a linear ablation lesion was created to connect the infarct scar with the mitral annulus. Otherwise, no linear lesions were created.
If no clinical VTs were inducible, radiofrequency energy was delivered at sites with isolated potentials and fractionated electrograms. The ablation endpoint was noncapture with pacing stimuli, similar to ablation sites identified by pace-mapping.
After ablation, the entire stimulation protocol was repeated at 2 right ventricular sites. If VTs remained inducible, further mapping and ablation were performed until a critical site was no longer identifiable.
Classification of VTs
The ICD electrograms of VTs that were documented before the VT ablation procedure were classified as clinical VTs. Induced VTs were correlated with the real-time morphology of the ICD electrograms to assess whether a VT was clinical (Fig. 1). This was done by visual comparison as previously described (1). Induced VTs that were not previously documented by the ICD were defined as previously undocumented nonclinical VTs. VTs that were documented during follow-up were compared with both clinical VTs and the VTs induced during the ablation procedure. If a new VT morphology was detected during follow-up, this was defined as a new clinical VT.
Two observers independently compared the ICD electrograms before, during, and after ablation. Discrepancies were resolved by consensus. The kappa value was 0.88 for agreement that a particular real-time tracing from the ICD programmer during the VT induction matched or did not match with a spontaneous VT. Both observers were blinded in the post hoc analysis.
In 68 of 98 patients, both near- and far-field ICD electrograms were available for analysis during the initial ablation procedure. In 30 patients only, either a near-field electrogram or a far-field electrogram was available for the clinical VT. In 21 of these 30 patients, comparison of the electrograms revealed that a single induced VT matched with the stored ICD electrogram of a clinical VT. In 3 patients, the ICD-stored electrograms of the clinical VTs were similar to 2 of the induced VTs. In 6 patients, the ICD-stored electrograms of the clinical VTs did not match any of the real-time ICD recordings of the induced VT. Therefore, in 27 of 30 patients (90%), all induced VTs could be classified as clinical or nonclinical.
In patients with recurrent VTs and repeat procedures, the location of critical sites of inducible VTs were analyzed with respect to the prior ablation procedure. Specifically, we assessed whether:
1. Critical sites of new VTs were located adjacent to a prior area of ablation. In addition to the topographic correlation of the electroanatomic maps (Fig. 2), to precisely correlate both procedures, we determined whether these noncapturing sites were new and due to ablation at the initial procedure or existed before the initial ablation procedure. Pace-mapping or entrainment mapping was performed to identify critical sites (Fig. 3) and to correlate with the location of prior ablation lesions (Figs. 2 and 3). The distance of new sites with noncapture to sites critical for new VTs was determined. The critical sites were defined as being adjacent to prior ablation lesions if they were not separated by points where capture with a different morphology occurred. In patients without new areas of noncapture, prior ablation points from previous procedures were identified by aligning both electroanatomic maps using similar fiducial points (aortic cusp, mitral annulus, and left ventricular apex). The coordinates of the ablation points of the initial procedure were labeled and displayed in a customized MATLAB program. After aligning the fiducial points of both maps, the distance between prior ablation points and sites with matching pace-maps were determined.
2. Critical sites of induced VTs were located in scar areas that were not thought to participate in ventricular arrhythmias based on the initial mapping procedure. The closest distance of ablation sites from the previous procedure and new critical sites was measured as outlined here.
Before discharge, 2 patients required repeat ablation procedures. All patients were seen every 3 to 6 months in an ICD device clinic. If a patient previously was receiving amiodarone, this medication was continued at the same dose. Amiodarone was discontinued if there were adverse effects (9 of 61 patients [15%]). The dose was reduced from 400 to 200 mg/day if patients had no adverse effects and no arrhythmias 3 to 6 months after the procedure. Amiodarone eventually was discontinued in 14 patients in whom no VT was inducible at the end of the procedure. Twelve patients (12%) were treated with sotalol, 7 (7%) with mexiletine, 2 (2%) with dofetilide, and 2 (2%) with verapamil.
Continuous variables are expressed as mean ± SD and were compared by using the Student t test. Discrete variables were compared by using the Fisher exact test or by chi-square analysis, as appropriate. A p value <0.05 was considered significant.
Recordings of VTs before ablation
A total of 331 VTs were documented by using ICD electrograms before the ablation procedure (3 ± 4 VTs per patient). In 37 of 98 patients (38%), there was only 1 documented VT based on ICD electrogram analysis. Multiple VTs were documented in the remaining patients. The clinical VTs had a mean cycle length of 356 ± 74 ms; when induced in the electrophysiology laboratory, their mean cycle length was 366 ± 100 ms (p = 0.11). However, 145 induced nonclinical VTs had VT cycle lengths (±30 ms) similar to those of the clinical VTs. A mean of 2 ± 2 VTs per patient with similar VT cycle lengths (±30 ms) compared with the clinical VT were induced.
Inducible VTs during the ablation procedure
A total of 725 VTs were induced by programmed stimulation, and 147 VTs (20%) were reproducibly inducible (Table 2). A total of 105 of 725 induced VTs (14%) were clinically documented before ablation, and 86% of the induced VTs were previously not documented (6 ± 5 VTs per patient). In 76 patients (78%), at least 1 clinical VT was induced (1 ± 1 per patient). In 11 patients, programmed stimulation was repeated during isoproterenol infusion, and a clinical VT was induced in 7 additional patients. In the remaining patients, demonstration of incomplete revascularization precluded the use of isoproterenol. In 22 patients (22%), a clinical VT was not inducible. There were no significant differences in clinical characteristics between patients with and without inducible clinical VT. In these patients, sites with isolated potentials were also targeted. There was no difference in outcome whether this approach was performed compared with the approach in which clinical VTs were targeted with pace-mapping (recurrences occurred in 9 of 22 vs. 24 of 76 patients; p = 0.41).
Mapping and ablation
Critical endocardial sites were identified for 75 of 105 clinical VTs (71%) and for 278 of 620 nonclinical VTs (45%), and radiofrequency energy was applied at these locations. In 16 of 105 clinical VTs (15%) and 26 of 620 nonclinical VTs (4%), critical sites were identified by entrainment mapping. Nonglobal capture occurred in 4 clinical and 1 nonclinical VT. Pace-mapping was used to identify critical sites of 55 of 105 clinical VTs and 251 of 620 nonclinical VTs.
After ablation, programmed stimulation was repeated in 95 of 98 patients (97%). In 60 of 95 patients (63%), no VT was inducible post-ablation. The clinical VTs that were induced before ablation were no longer inducible in all patients in whom programmed ventricular stimulation was repeated at the end of the procedure. Fifty-four of 725 VTs (7%) were inducible, and 671 VTs (93%) were no longer inducible. No critical endocardial sites could be identified for the 54 inducible VTs. In the patients in whom VT was inducible post-ablation, VT was inducible with triple extrastimuli in 24 patients, double extrastimuli in 9, and a single extrastimulus in 2. The manner of VT induction did not correlate with recurrence of VT (7 of 24 [29%] vs. 4 of 9 [44%] vs. 1 of 2 [50%], respectively; p = 0.63). The cycle length of the inducible VT did not correlate with VT recurrence (358 ± 96 ms vs. 332 ± 57 ms; p = 0.25). Furthermore, there was no significant difference in scar area between patients with and without inducible VT at the end of the procedure (83 ± 33 cm2 vs. 73 ± 37 cm2; p = 0.23).
For 307 of 725 VTs (42%), no critical endocardial sites were identified, yet these VTs were no longer inducible post-ablation.
Follow-up after the initial procedure
The mean follow-up period after the procedure was 35 ± 23 months (1 to 89 months) (Table 3). No patients were lost to follow-up. Sixty-five of 98 patients (66%) did not have VT recurrence. In 16 of 33 patients, a new VT resulted in VT recurrence; in 7 patients, only a prior clinical VT recurred. In 10 of 33 patients, both a prior clinical VT and a new VT occurred. Patients with recurrent VTs had a larger endocardial scar area as assessed by electroanatomic mapping. Whether a VT was inducible at the end of the ablation procedure did not predict recurrence.
In the 33 patients with recurrent VTs, ICD electrograms showed 89 different VTs recurring after ablation (Table 4). Fifteen of 89 VT events (17%) were due to VTs that were previously recorded by the ICD (prior clinical VT). The remaining 74 VTs were not previously documented. With respect to the prevalence of the recurrent VTs, a median of 77% were due to new VTs, and the remaining 23% were due to the previously documented VTs.
Repeat mapping procedures
In 14 patients (14%), a repeat procedure was performed (Figs. 2 to 4⇓). Two of the 14 patients had recurrence of only a prior clinical VT, and 6 of the patients had recurrence of a new VT. In 6 of 14 patients, both a prior clinical and a new VT recurred. A total of 49 different VTs were documented before the repeated ablation procedure (4 ± 3 VTs per patient; 8 of these VTs were previously documented, and 41 were new VTs). During the repeat procedures, 78 VTs (6 ± 3 VTs per patient) were induced. These included 8 previously documented clinical VTs in 8 patients. New clinical VTs that were not documented before or during the initial ablation procedure were inducible at the repeat procedure in all 14 patients. Critical sites were identified for 47 of 78 inducible VTs (60%). The majority of the critical sites that were identified were for new VTs (85%) that were not documented before or during the initial ablation procedure. The majority of the critical sites were adjacent to previous ablation lesions (53%), and the second largest group (32%) was localized to areas that were previously not thought to participate in the arrhythmias (based on pace-mapping data in the initial procedure). The mean distance between sites critical for new VTs and noncapturing sites arising in areas of prior ablation was 6 ± 3 mm, and it was 11 ± 7 mm for new VTs arising from areas without prior ablation (p = 0.0004). Critical sites for prior clinical VTs accounted for the remaining 15% of critical areas that were identified in these patients. These VTs were previously either not inducible or were targeted by pace-mapping or entrainment mapping.
After the procedure, no VT could be induced in 11 patients. After the last procedure, 75 of 98 patients had no additional VT recurrences, resulting in a total success rate of 77%.
In 1 patient with recurrent VTs, an epicardial procedure was performed, and the recurrent clinical VT was successfully ablated. This was the only patient in this series in whom an epicardial procedure was performed.
Periprocedural complications consisted of pericardial tamponade that responded to percutaneous drainage (n = 1), stroke secondary to thromboembolism (n = 1), heart failure (n = 1), groin hematoma requiring blood transfusion and surgery (n = 1), and hematuria (n = 1). Two patients had recurrent VT storm and required another ablation procedure before discharge.
The total follow-up period after the first procedure was 39 ± 23 months (3 to 89 months); after the last procedure, the follow-up was 34 ± 18 months (7 to 56 months). Twelve patients died during follow-up. One patient died secondary to worsening heart failure, 1 patient died suddenly with recurring ventricular arrhythmias, and 10 patients died of noncardiac causes.
There was no significant difference in mortality between patients with and without recurrence of VTs (6 of 33 [18%] vs. 6 of 65 [9%]; p = 0.20). Amiodarone was discontinued in 23 patients. There was no difference in VT recurrence regardless of whether amiodarone was discontinued (p = 0.77).
VT recurred in approximately one third of the patients in this series. The main reason for VT recurrence was the occurrence of new VTs that were most often mapped to sites adjacent to scar from prior ablation lesions or areas that were initially not thought to contain VT circuits. Patients with VT recurrence had more extensive scar than patients without recurrent VT.
VT recurrence after ablation has been described in up to 57% of patients (8) with prior myocardial infarctions. Previous efforts to analyze the reasons for VT recurrence have been mainly limited by the inability to adequately characterize the recurring VT. These attempts were confined to comparisons of VT cycle lengths, which are subject to inaccuracy due to cycle length variation secondary to antiarrhythmic treatment or due to different VTs with similar cycle lengths in a given patient. The morphology of ICD electrograms has been used successfully to identify clinical VTs (1) and is well suited to analyze reasons for recurrent VTs.
The main reason for VT recurrence in this study was the occurrence of new VTs. The majority of VTs that recurred were different from the initially induced and targeted VTs. These circuits are either new circuits or modifications of pre-existing circuits. A clear distinction between these 2 possibilities is impossible because all of the circuits were not mapped. It is intriguing, however, that critical areas were found adjacent to clusters of ablation lesions deployed during the initial ablation procedure that resulted in the lack of local capture during the repeat procedure. Nonexcitable post-infarction scar has been described to form part of a fixed border of the re-entry circuit (9). It is therefore possible that some of the ablation lesions from prior procedures formed the borders for some of the recurring new VTs. Complete linear lesions are difficult to achieve in the ventricle, yet complete linear lesions may be one way to reduce VT recurrence post-ablation. Real-time intraprocedural magnetic resonance imaging may help to assure that ablation lesions are complete.
In patients with repeat procedures, about one-third of the mapped VTs were located in areas that were not thought to be involved in the arrhythmogenic substrate during the initial procedure and were therefore not targeted for ablation. Lack of inducibility most likely accounted for the inability to identify these areas participating in the arrhythmogenic substrate. It is therefore not surprising that in patients with VT recurrence, the infarct scars as measured by using electroanatomic mapping were larger.
Clinical VTs that were thought to be effectively ablated during the initial procedure sometimes recurred. In 2 patients with repeat procedures, the VTs recurred despite the VT terminating during radiofrequency ablation at a site with concealed entrainment. It is important to note that despite the inability to induce a particular VT at the end of a procedure, these VTs can still recur.
Left-sided programmed stimulation was not performed in all patients in whom the clinical VT could not be induced. In addition, programmed stimulation with isoproterenol was not performed in all patients due to incomplete revascularization. This might have reduced the rate of noninducible patients and may have resulted in a lower recurrence rate.
Because only a near- or a far-field ICD electrogram of the clinical VT was available in one-third of patients for the initial procedure, a few clinical VTs might not have been correctly identified. However, all patients had their ICDs reprogrammed to record both near- and far-field electrograms after the initial procedure.
The procedure was guided by entrainment mapping and pace-mapping. Interestingly, a critical area was identified by using these mapping techniques in only about 50% of the inducible VTs. This finding suggests that many VT circuits had epicardial or intramural components or were composed of functional elements. Other potential reasons for the lower yield of pacing sites matching the induced VTs include the possibility that sites with matching pace-maps within the scar were simply missed and that the delivery of radiofrequency energy at sites of matching pace-maps altered the scar substrate. A complex 3-dimensional arrhythmogenic substrate with shared circuits (10) might explain why an epicardial procedure was necessary in only 1 of the 98 patients.
Although pace-mapping only was performed at low-voltage sites, there were many more low-voltage sites than pace-mapping sites. This discrepancy was due to catheter instability and dislodgement before pacing, inconsistent capture, and exclusion of sites ≤3 mm from prior pace-mapping sites.
Although there was no difference in mortality in this small study, it was not powered to detect a difference in mortality between patients with and without recurrence.
Dr. Bogun has received a grant from the Leducq Foundation. Dr. Desjardins was supported by National Institutes of Health grant 7K23EB006481. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- implantable cardioverter-defibrillator
- ventricular tachycardia
- Received March 27, 2012.
- Revision received July 11, 2012.
- Accepted July 16, 2012.
- American College of Cardiology Foundation
- Yoshida K.,
- Liu T.Y.,
- Scott C.,
- et al.
- Hummel D.,
- Strickberger S.,
- Daoud E.,
- et al.
- Bogun F.,
- Bahu M.,
- Knight B.,
- et al.
- Stevenson W.,
- Khan H.,
- Sager P.,
- et al.
- Soejima K.,
- Stevenson W.G.,
- Maisel W.H.,
- Sapp J.L.,
- Epstein L.M.