Methods

The collaborative regional hospital network consists of 7 different hospitals (catchment area of approximately 400,000) that regularly transfer patients to the interventional electrophysiology center. In this network, patients with ES admitted to the collaborating hospitals were included if catheter ablation had been scheduled after contacting the interventional electrophysiology center from December 2007 to December 2009. ES was defined as the occurrence of ≥3 episodes of ventricular arrhythmia separated by >5 minutes during a 24-hour period not due to ischemia. The patients were transferred to the interventional center as soon as they had been stabilized for transportation. After admission, all patients were continuously monitored using 12-lead electrocardiographic monitoring in the intensive care unit to document any clinical arrhythmia. Myocardial ischemia was ruled out as the mechanism for electrical instability. For rhythm stabilization, if recurrent ventricular arrhythmia occurred, conventional strategies using a specified protocol were used as applicable: (1) intensified β-blocker therapy including an intravenous bolus (n = 32); (2) adequate sedation using midazolam (n = 18); (3) additional or intensified antiarrhythmic medication using class I antiarrhythmics and/or amiodarone (n = 24); (4) general anesthesia (n = 5); (5) increased V-pace (n = 4); and/or (6) hemodynamic left ventricular support (n = 3). Cardiogenic shock was defined as prolonged hypotension (>3 hours, <70 mm Hg systolic) under intravenous pressor agents or hemodynamic device support such as an intra-aortic balloon pump or extracorporeal membrane oxygenation.

All patients or family members provided informed consent. The ablation procedure was usually scheduled for the next working day after admission (<24 hours) or on the day or night after admission if no temporary rhythm stabilization was achieved (acute ablation). The patients were studied using an electroanatomic 3-dimensional mapping system (CARTO XP, CARTO-3, Biosense Webster, Diamond Bar, California, and NaVX, St. Jude Medical, St. Paul, Minnesota) under deep sedation using disoprivan or midazolam. The invasive electrophysiologic study was performed using a recently published standardized protocol and pacing from multiple sites from the right and left ventricle (programmed stimulation and/or burst stimulation). In brief, during sinus rhythm, a left ventricular substrate map (bipolar voltage map) was constructed using a 3.5 cooled-tip ablation catheter (NaviStar Thermo, Cool, EZ-Steer NAV Thermo, Cool, Biosense Webster). The zones of scar (bipolar voltage ≤0.5 mV), and zones of damaged myocardium within the scar border zone (bipolar voltage of 0.5 to 1.5 mV) were identified. Ablation strategies were then based on the individual clinical ventricular arrhythmia and documented structural ventricular abnormalities (scar areas and underlying structural heart disease).

The clinical ventricular arrhythmias were defined as either any spontaneously occurring electrocardiographically documented ventricular arrhythmia, or, if only ICD-Holter data were accessible (without electrocardiographic documentation), induced ventricular arrhythmia with a cycle length comparable to ICD-documented cycle length.

In patients with ischemic cardiomyopathy (ICM) and monomorphic VT ablation was directed towards the border zone of the documented scar region. The site of ablation was chosen according to the congruence of local pace map and the clinical arrhythmia electrocardiogram. Regional (or circumferential) scar encircling ablation was performed, as previously presented ( Figure 1 ). In patients with nonischemic cardiomyopathy (NICM), scar regions were identified, and linear ablation was directed to the interconnect scar and other electrically inert tissue areas (e.g., mitral valve annulus, other scar region, aortic valve annulus; Figure 2 ). The location of the ablation was again chosen according to the congruence between clinical arrhythmia and the local pace map electrocardiogram. In patients with documented ventricular fibrillation (VF), scar mapping was performed, and focal ablation was directed toward areas presenting with Purkinje-like potentials within the scar border zone (producing ventricular premature beats with electrocardiographic morphology comparable to the clinical inductors of VF). If patients had ongoing hemodynamically stable ventricular tachycardia (VT), activation mapping was performed, and ablation was directed toward critical components of the VT. The ablation strategy was individually adapted to the structural abnormalities identified during substrate mapping. In cases in which no ablation success during ablation in the left ventricle was achieved, an epicardial percutaneous ablation approach was used when the electrocardiographic findings of the clinical arrhythmia was susceptible of an epicardial origin ( Figure 3 ). In cases in which clinical ventricular arrhythmia on the electrocardiogram was suspicious of a right ventricular origin (n = 3), mapping and ablation was also directed toward the right ventricle.

(A) Electrocardiogram of clinical arrhythmia and (B) substrate map ( red indicates bipolar voltage <0.5 mV indicating scar and purple , bipolar voltage >1.5 mV indicating normal myocardium) in patient with ES and posterior myocardial infarction. Ablation strategy involved circumferential ablation (red dots) within scar border zone of posterior infarct scar owing to hemodynamic instability and occurrence of 4 different VT morphologies, all related to posterior scar.

(A) Fluoroscopic image and (B) electroanatomic substrate map of patient with NICM. (A) Ablation directed toward epicardial anterior left ventricular surface ( ^# position of ablation catheter) after percutaneous epicardial puncture. Before ablation, additional selective coronary angiography was performed. (B) Endocardial (endo) and epicardial (epi) bipolar voltage map and documentation of scar region at basal aspect of left anterior wall (red). Multiple ablations (red dots) within scar border zone were constituted.

(A,B) Holter-electrocardiographic recordings, (C) intraprocedural electrocardiogram and intracardiac electrocardiograms and (D) electroanatomic substrate map in patient with recurrent episodes of ventricular fibrillation late after posterior and septal myocardial infarction. (A) Multiple episodes of sustained ventricular arrhythmia and (B) triggering ventricular premature beat with comparable morphology in 2 episodes (*). (C) Ventricular premature beat in this case did not trigger ventricular fibrillation (*) with shorter QRS duration than intrinsic beats and Purkinje-like potential preceding local electrocardiogram in ablation catheter electrodes 1 and 2 (ABL 1/2). (D) Substrate map indicating posteroseptal scarring and mapping of specific left ventricular conduction system (green dots). Ablation directed toward earliest Purkinje-like potential in septal region (red dots) , and ablation eliminated ventricular premature beats and ventricular fibrillation.

Patients with mechanical devices for left ventricular support were studied in sinus rhythm using a transseptal (intra-aortic balloon pump) or transaortic (extracorporal membrane oxygenation) approach during ongoing device therapy.

During ablation, all inducible ventricular arrhythmias were targeted, but the procedure was stopped if other ventricular arrhythmias after ablation of the clinical needed direct current-shock defibrillation >3 times. The extent of the ablation was left to the discretion of the operator. In all patients, programmed ventricular stimulation to identify procedural success was performed after ablation using ≥2 right ventricular and 2 left ventricular pacing sites (2 pacing cycle length with ≤3 extrastimuli and burst pacing up to a cycle length of 250 ms) if no further arrhythmia was induced. Complete success was considered an inability to induce any sustained ventricular arrhythmia at the end of the procedure. Partial success was defined as the noninducibility of all clinical VTs but the induction of ≥1 nonclinical VT. Failure was considered present if ≥1 clinical VT was inducible at the end of the procedure.

During follow-up, the patients’ antiarrhythmic medication was unchanged from their chronic medication before the onset of ES, except for those patients with ablation failure (for whom the medication was intensified, including treatment with amiodarone plus mexiletine plus β blockade; Table 1 ).

All patients underwent ICD implantation (5 patients with biventricular pacing devices) and were followed up after 1, 3, and 6 months, and every 6 months thereafter. Follow-up included ICD-Holter interrogation and documentation of any ventricular arrhythmia episode triggering ICD treatment. Only appropriate ICD therapies to terminate ventricular arrhythmia were considered. The primary end point was recurrence of sustained ventricular arrhythmia during the follow-up period. The interval to the first appropriate ICD therapy was documented. Any recurrence of ES was documented. Overall rhythm success was defined as no sustained ventricular arrhythmia documented by the ICD or on an electrocardiogram after the first ablation procedure. Cumulated rhythm success was considered success after the initial hospital stay, including multiple ablation sessions.

All procedure-related adverse events were monitored, and death and the cause of death were documented during follow-up.

The patient characteristics are described as the mean ± SD for quantitative data and the number (percentage) for categorical variables. Follow-up is given as the median (range). The recurrence of ventricular arrhythmia episodes during follow-up is given as the percentage of the overall patients treated. The Kaplan-Meier technique was used to calculate the cumulative event rates for survival, freedom from ventricular arrhythmia recurrence (overall rhythm success), and freedom from ventricular arrhythmia recurrence after hospital discharge (cumulated rhythm success). Statistical analyses were performed using the Statistical Package for Social Sciences statistical software, version 18.0 (SPSS, Chicago, Illinois).