intracardiac shocks: results in 33 patients

Transcription

1 THERAPY AND PREVENTION Catheter ablation of ventricular tachycardia with intracardiac shocks: results in 33 patients FRED MORADY, M.D., MELVIN M. SCHEINMAN, M.D., LORENZO A. DI CARLO, JR., M.D., JESSE C. DAVIS, M.D., JOHN M. HERRE, M.D., JERRY C. GRIFFIN, M.D., STuART A. WINSTON, D.O., MICHAEL DE BUITLEIR, M.B., CHARLES B. HANTLER, M.D., JOYCE A. WAHR, M.D., WILLIAM H. KOU, M.D., AND STEVEN D. NELSON, M. D. Downloaded from by on vember 1, 01 ABSTRACT Catheter electrical ablation of ventricular tachycardia (VT) was attempted in 33 patients who had recurrent unimorphic VT refractory to (mean + SD) antiarrhythmic drugs. Their mean age was years. Twenty-two patients had coronary artery disease, six had other types of heart disease, and five had no structural heart disease. The mean left ventricular ejection fraction was Thirty patients had only one documented morphologic type of spontaneous VT, whereas three patients had more than one. One to four shocks of 0 to 300 J each were delivered to the endocardial exit site of VT, as identified by endocardial activation mapping and pace-mapping. In each patient endocardial activation at the exit site of VT preceded the onset of the QRS complex (mean activation time msec). Pace-mapping was possible in patients, and in all but two patients the QRS complexes during VT and during pacing at the exit site of VT were very similar in at least of electrocardiographic leads. In 9 patients, shocks were delivered between an endocardial electrode (cathode) and a patch electrode on the chest wall (anode). Seven patients (including three who first received shocks using an external anode) whose VT originated in the septum received transseptal shocks between two electrodes positioned on either side of the septum. The procedure was successful in 15 patients (45%), who had no recurrence of VT either on no antiarrhythmic therapy or on the same regimen that was ineffective before ablation, over a follow-up period of months (range 5 to 35). The ablation attempt was unsuccessful in 1 patients (55%). There were no significant differences in clinical and electrophysiologic variables between patients with and without a successful outcome. Seven nonfatal complications occurred in six patients: sustained nonclinical VT immediately after the shock, ventricular fibrillation on days S and after ablation, neurologic deficits (n = ), atrioventricular block (n = ), and brachial artery thrombosis. In conclusion, catheter electrical ablation of VT has modest efficacy and is relatively safe in a selected group of patients who have predominantly one configuration of unimorphic VT. Circulation 75,. 5, 37-49, 197. THE EFFICACY of catheter ablation of ventricular tachycardia (VT) with intracardiac shocks has been highly variable, ranging from approximately 0% to 0%.`'- However, because all previous reports on the use of transcatheter shocks in the treatment of patients with VT have consisted of only a small number of patients (one to 1),1 the efficacy and morbidity of this technique have not been adequately defined. The purpose of this report is to review the results of catheter ablation of VT in 33 patients who underwent From the Division of Cardiology, Department of Intemal Medicine, University of Michigan Medical Center, and the Division of Cardiology and Cardiovascular Research Institute, University of California, San Francisco. Address for correspondence: Fred Morady, M.D., Division of Cardiology, University Hospitals, 1500 East Medical Center Drive, UH B1 F5-00, Ann Arbor, MI Received Oct. 7, 19; revision accepted Feb. 5, 197. this procedure over a 4 year period. By describing the efficacy and complications of catheter ablation of VT and the determinants of a successful outcome, we wish to better define the role of this technique in the management of patients with recurrent VT. Methods Selection of patients. Included in this report are 33 consecutive patients who underwent an attempt at catheter electrical ablation of VT at Moffitt Hospital, San Francisco, and at the University of Michigan Medical Center between September 19 and June 19. All patients had recurrent, symptomatic, sustained VT. In each patient, VT was hemodynamically stable in the electrophysiology laboratory and could be mapped. The criteria used to select patients as potential candidates for catheter ablation were not uniform during the 4 year period covered by this report. Initially, catheter ablation was attempted only in patients who were refractory to all available conventional and investigational drugs and who were not appropriate candidates for map-directed endocardial resection. As our experience with Vol. 75,. 5, May

2 MORADY et al. efficacy and risks of catheter ablation evolved, the catheter ablation technique was offered as a potential alternative to investigational antiarrhythmic drugs, cardiac electrosurgery, or an antitachycardia pacemaker. In selecting patients to undergo the catheter ablation technique there was a bias against selecting patients who had multiple configurations of spontaneous or induced VT and patients in whom a -lead electrocardiographic recording of spontaneous VT was not available. Clinical characteristics of patients (table 1). We studied 3 men and one woman, with a mean age of 5 ± 14 years ( SD, range 19 to 1). Twenty-two patients had coronary artery disease (each with a history of myocardial infarction), two patients had an idiopathic dilated cardiomyopathy, two had arrhythmogenic right ventricular dysplasia, and two had undergone successful surgical repair of tetralogy of Fallot 14 to 0 years earlier. In five patients no structural heart disease was identified by coronary angiography, a contrast ventriculogram, an echocardiogram, and exercise treadmill test. A contrast or radionu- TABLE 1 Patient characteristics clide ventriculogram demonstrated the presence of a ventricular aneurysm in 19 of the patients who had coronary artery disease. The mean left ventricular ejection fraction by contrast or radionuclide ventriculography was (range 0. to 0.9). In the subgroup of patients who had coronary artery disease, the mean left ventricular ejection fraction was The most severe symptom experienced during VT consisted of a cardiac arrest in patients, syncope in five patients, nearsyncope in patients, and palpitations and/or chest pain and dyspnea in six patients. The first episode of symptomatic VT occurred from 1 month to 0 years (mean + 4 months) before the attempt at catheter ablation. Thirteen patients had experienced more than 5 episodes of VT, whereas 0 had a history of between two and 5 episodes. The VT was incessant for 3 to 7 days before the attempt at catheter ablation in eight patients. All patients had been treated unsuccessfully (as determined either by clinical response or the response to electropharmaco- Patient Age Heart Onset of. VT Spontaneous VT. (yrs)/sex disease LVEF VT (mos) episodes Morphology CL (msec) Downloaded from by on vember 1, /M 57/M 7/M 51M 70/M 19/M 3/M 39/M 1/M 34/M 70/M 30/M 59/M 34/M 9/M 59/M 4/M 3/M 74/M 3/M /M 5/M 59/M 55/M /M 7/M 0/M 1/M /M 4/M 3/F 5/M 4/M Tet DC ne Tet ARVD ne ARVD DC ne ne ne S S 3 types RB, I LB, I LB, I Undocumented LB, I 3 types LB, I types LB, I Ind. Ind. LB, I Ind ARVD = arrhythmogenic right ventricular dysplasia; = coronary artery disease; CL = cycle length; DC = dilated cardiomyopathy; I = inferior axis; Ind. = indeterminate; LB = left bundle branch block configuration; LVEF = left ventricular ejection fraction; RB = right bundle branch block configuration; S = superior axis; Tet = tetralogy of Fallot (repaired). AIncessant VT for 3 to 7 days before ablation attempt. 3 CIRCULATION

3 Downloaded from by on vember 1, 01 logic testing) with at least two class I antiarrhythmic drugs before undergoing the catheter ablation. Nineteen of the 33 patients had failed a trial of amiodarone. Other drugs that had been unsuccessful included a fl-blocker in nine patients, verapamil in five, propafenone in three, lorcainide in two, and sotalol and encainide in one patient each. The mean number of drug trials that were either ineffective or not tolerated before catheter ablation was attempted was 3.7 ± 1. (range to 7). Other types of therapy before the attempt at catheter ablation included aneurysmectomy and endocardial resection in one patient, an automatic implantable cardioverter/defibrillator in one patient, and an antitachycardia pacemaker in two patients. Characteristics of ventricular tachycardia(table 1). Twenty-eight patients had a single documented configuration of unimorphic VT, although in many patients not every episode of VT was documented. Three patients had two or three different configurations of unimorphic VT. In one patient who had experienced a cardiac arrest in association with VT, a -lead electrocardiographic recording of the VT was not available. Electrophysiologic testing. Electrophysiologic studies were performed with patients in the fasting, unsedated state after informed consent was obtained. All antiarrhythmic drugs were discontinued at least four half-lives before the procedure in 1 patients. In 15 patients who had had a partial response to antiarrhythmic drug therapy, electrophysiologic testing and catheter ablation were performed without discontinuing drug therapy. Two. F quadripolar electrode catheters were inserted percutaneously into a femoral vein and positioned against the right ventricular apex and outflow tract or septum. For left ventricular mapping, a. F quadripolar electrode catheter was inserted through a. 7F sheath into a femoral or brachial artery and positioned within the left ventricle. The arterial pressure was monitored continuously through the sidearm of the sheath. Heparin was administered intravenously at an initial dose of 3000 U, followed by 00 U every hour. Electrocardiographic leads V,, I and III, the intracardiac electrograms, and the arterial pressure were displayed on an oscilloscope and recorded at a paper speed of 0 to 150 mm/sec with an Electronics for Medicine VR- or a Siemens Elema Minograf 7 recorder. Bipolar intracardiac electrograms were filtered from 30 to 500 Hz. Programmed stimulation was performed with a programmable stimulator (Bloom Associates, Ltd.), with stimuli twice the diastolic threshold and msec in duration. A -lead electrocardiogram was recorded whenever sustained VT that was hemodynamically stable was induced. The programmed stimulation protocol consisted of single and double extrastimuli at two basic drive cycle lengths (00 or 500, THERAPY AND PREVENTION-ARRHYTHMIA and 400 msec) and two right ventricular sites, followed by triple extrastimuli at these basic drive cycle lengths and sites. Sustained VT was defined as VT more than 30 seconds in duration or requiring direct-current countershock for termination. nsustained VT was defined as VT that was six complexes to 30 sec in duration. Induced VT was defined as "clinical" if it had the same configuration as the patient's spontaneous VT in all leads of an electrocardiogram and "nonclinical" if it had a different configuration than the patient's spontaneous episodes of VT. Mapping technique. A mapping schema described by Josephson et al.9 was used for identifying 1 areas of the right and left ventricles with biplane fluoroscopy (figure 1). Mapping of endocardial activation during VT and pace-mapping was attempted at as many of the 1 ventricular sites possible. In each patient at least sites were mapped. In four patients who were not hemodynamically stable during induced VT, 500 to 00 mg of procainamide were administered intravenously to slow the VT and allow mapping. Whenever multiple morphologies were induced, an attempt was made to map the morphology that had occurred spontaneously. The activation time at each ventricular site was defined as the interval between the first rapid deflection of the intracardiac electrogram and the earliest deflection of the QRS complex in leads, I, III, or V1 during VT. The presence or absence of fractionated potentials bridging diastole was noted. Pace-mapping was performed by pacing the right and left ventricles at a cycle length close to the VT cycle length. A - lead electrocardiogram was recorded during pacing and compared with the -lead electrocardiographic recording of VT. In seven patients pace-mapping at each ventricular site was not possible because of inability to obtain ventricular capture with pacing stimuli up to ma in strength. The grading system used to evaluate pace-maps was based on the number of electrocardiographic leads displaying QRS complexes during pacing that were very similar to the QRS complexes during VT. Pace-maps were graded as "excellent" if the QRS complexes were identical or very similar to the QRS complexes durng VT in each of the leads (figure ). Pace-maps were graded as "good" if the QRS complexes had the same bundle branch block configuration and axis as the VT and if the QRS complexes during pacing were very similar to the QRS complexes during VT in or 11 of the leads, and "poor" if the QRS complexes during pacing were very similar to the QRS complexes during VT in no more than nine of the leads (figure 3). In the initial patients in this series, the shock site was FIGURE 1. The mapping schema used to designate sites in the left ventricle and six sites in the right ventricle. Adapted from ref. 9. Vol. 75,. 5, May

4 Downloaded from by on vember 1, 01 MORADY et al. AYR) V V3 Vff a VT, CL 370 ma avly V, 1'. Pacing, CL 30 me, SIt* 1 YYVYVYAV\y\rWA,h ing 1 i!00<,~~~ 1acc Pacing, CL 30 me, site 1 1 MW FIGURE. An example of an "excellent" pace-map (in patient ), with QRS complexes during pacing at site l that are very similar to the QRS complexes during VT in all electrocardiographic leads. CL = cycle length. A, Frontal planes leads; B, precordial leads. determined primarily on the basis of the earliest endocardial activation during VT, regardless of the results of pace-mapping. In subsequent patients, the site of shock delivery was influenced by the results of pace-mapping, with shocks being delivered only to sites that had a good or excellent pace-map and at which endocardial activation preceded the onset of the QRS during VT. Catheter ablation technique. Catheter ablation was attempted with patients in the fasting, unsedated state after informed consent was obtained under an investigational protocol approved by the Human Research Committee at the University of California, San Francisco, and the University of Michigan. A. F quadripolar electrode catheter (USCI) with either a 0.5 or 1 cm interelectrode distance was positioned against the endocardium at the presumed exit site of the VT. A 1 cm patch electrode (R Corp.) was positioned on the chest in the closest possible proximity to the catheter. The distal electrode of the 40 A ' A VT, CL 370 me mapping catheter was connected to the cathodal output of a defibrillator (Life-Pak, PhysioControl, Mennen Greatbach or Air Shields), and the patch electrode was connected to the anodal sink of the defibrillator (figure 4). A different technique was used in patients whose VT originated in the interventricular septum. The distal electrode of the mapping catheter was positioned against the endocardium on the left ventricular side of the septum and connected to the cathodal output of the defibrillator. A second electrode catheter was positioned on the right ventricular side of the septum, directly opposite the left ventricular catheter. The right ventricular catheter was connected to the anodal sink of the defibrillator (figure 5). This configuration was used on the basis of a prior study in animals. ' After the patients were anesthetized with intravenous sodium thiopental or methohexital, one to four shocks of 0 to 300 J (stored energy) were delivered in synchronized fashion. The shocks had a damped sinusoidal configuration. In each patient, programmed ventricular stimulation was performed 30 to 0 min after delivery of the shocks to determine whether VT was still inducible. The patients were then monitored continuously in a coronary care unit for 1 to days, then on a telemetry ward for an additional 5 to days. Serial creatine kinase-mb fractions were measured during the first to 4 hr after the ablation attempts. One week after delivery of the shocks, an electrophysiologic study was repeated in the patients who had not had inducible VT 30 to 0 min after the shocks. Patients who had inducible VT were treated with antiarrhythmic drugs, whereas those who did not have inducible VT were discharged either without any antiarrhythmic drug therapy or on the same drug regimen that had been previously only partially effective and that had been continued throughout the hospitalization. Follow-up and evaluation of results. Follow-up data were obtained directly from the patients, either in person or by telephone. The ablation attempt was considered successful if the patient experienced no recurrence of symptomatic VT during the follow-up period, either in the absence of any antiarrhythmic drug therapy, or while being treated with the same doses of the same drugs that had been unsuccessful in controlling VT before the ablation attempt. If the patient did not have a recurrence of symptomatic VT during the follow-up period but was being treated with a drug regimen that had not been used before the ablation attempt, ablation was not considered a success because of the possibility that prevention of recurrent VT was related to the new drug regimen. Statistical analysis. Comparisons were performed with Student's t test, chi-square analysis, or Fisher's exact test. A p value <.05 was considered significant. Results Results of mapping. On the basis of endocardial activation mapping and, whenever available, pace-mapping, the exit site of VT was found to be in the left free wall in 13 patients, the interventricular septum in nine patients, and the right ventricle in 11 patients. The results of endocardial activation mapping and pacemapping in each patient are described in table. Ablation technique. In 9 patients the shocks were delivered between an electrode positioned against the endocardium (cathode) and a patch electrode on the chest wall (anode). Between one and four shocks were CIRCULATION

5 THERAPY AND PREVENTION-ARRHYTHMIA Ilr &VR VT,CL 350 me Pacing,CL 3 me sit* Pacing, CL 3 me site 1- JVVVWWA/y \ffyvnff Y~vY AAAJKVWWAA avl / avf A VT, CL 350 ms V-, V Pacing,CL 3 me site MWW 'ffwyvyy Wbvvvv Lf sec W~~~~Z~ pacing, CL 3 me site 1- FIGURE 3. An example of "poor" and "Agood" pacemaps (in patient ). A, Frontal plane leads; B, precordial leads. On the left is the VT, in the center a "poor" pace-map at site (very similar QRS complexes during VT and pacing in fewer than leads), and on the right a "good" pace-map (very similar QRS complexes during VT and pacing in or 11 leads). In this example, pacing at site 1- resulted in QRS complexes very similar to those during VT in all leads except V. CL = cycle length. V33f Downloaded from by on vember 1, 01 V4 VS B W used (mean ) per patient. The mean total energy (stored) of the shocks was 30 ± 155 J (table 3). In seven patients in whom the exit site of VT was in the interventricular septum (including three patients in whom attempts at ablation with an external patch electrode had failed) the shocks were delivered transseptally. Each patient received two shocks, and the mean total energy (stored) of the shocks was 0 ± 75 J (table 4). Short-term results. Immediately after delivery of the shocks in the patients in whom an external patch electrode was used as the anode, the clinical VT was still inducible by programmed ventricular stimulation in 11 of 9 patients. A nonclinical unimorphic VT was induced in one patient and no VT was inducible in 17 patients (table 3). ong the 17 patients in whom VT X A ^ A 't(. 1 sec. was not inducible immediately after delivery of the shocks, an electrophysiologic study 7 to days later demonstrated that the original clinical VT was inducible in three patients, a nonclinical unimorphic VT was inducible in four patients, and no VT was inducible in ten patients (table 3). Immediately after delivery of the transseptal shocks, the clinical VT was still inducible in one of the seven patients, a nonclinical unimorphic VT was inducible in one patient, and no VT was inducible in five patients (table 4). ong the six patients in whom the clinical VT was not inducible immediately after delivery of shocks, an electrophysiologic study 7 to days later demonstrated that the clinical VT was inducible in three patients, a nonclinical unimorphic VT was inducible in one patient, and no VT was inducible in two patients (table 4). Vol. 75,. 5, May

6 MORADY et al. Downloaded from by on vember 1, 01 current countershock was required to terminate the VT. Ventricular fibrillation occurred on the fifth and sixth days after delivery of 300 J in one patient (. ). He was treated transiently with amiodarone for months and subsequently had no recurrence of VT or ventricular fibrillation over 3 additional months of follow-up. Two patients (s. and 9) had transient aphasia or hemiparesis after the catheter ablation procedure, and one had a brachial artery thrombosis requiring surgical correction. One patient (. 9) had electrode Mobitz type II atrioventricular (AV) block for 3 days after delivery of 00 J transseptally and underwent FIGURE 4. A schematic representation of the technique us( ed to deliver implantation of a permanent pacemaker because of shocks between an electrode positioned against the endocardium and an impaired infranodal conduction, as reflected by infranodal block at an atrial pacing cycle length of 450 external patch electrode. LV = left ventricle. msec. Another patient (. 7) had transient :1 AV Long-term follow-up (table 5). VT recurred iin 15 pa- block for several hours after delivery of 400 J to site 13 tients during the follow-up period. The intuerval be- in the right ventricle; a permanent pacemaker was not tween delivery of the shocks and the first recuirrence of implanted in this patient and there was no recurrence of VT ranged between 1 day and months. In nine pa- AV block. tients the recurrences of VT were not fatal, whereas in In summary, 7 patients (%) had no complica- follow- tions, and six patients (1%) had seven nonfatal com- six patients death from VT occurred at a meain up interval of 9. ± 5 months. The antiar rhythmic plications. There were no instances of myocardial pertable 5. foration or pericardial tamponade. therapy used in these patients in described in Symptomatic VT did not recur in 15 pati( ents who Degree of myocardial injury. Mean peak creatine kidrugs or nase-mb fraction after delivery of the shocks was 5 were treated with either no antiarrhythmic with the same antiarrhythmic drug regimen that had ± 5 IU/liter (normal range 0 to ). been ineffective before ablation. The mean follow-up patient developed new Q waves on the electro- (range cardiogram after delivery of the shocks. New or wors- interval in these patients was 15.5 ± montlhs 5 to 35). ened congestive heart failure was not noted in any Symptomatic VT did not recur in three patients who patient. were treated with an antiarrhythmic drug regimen that had not been used before delivery of the shopcks. The mean follow-up interval in these patients was ± 0 months (range to 49). The long-term clinical results are summcarzed in table. The outcome of catheter ablation of VT was considered successful in 15 patients (45%). Comparison of patients with and without a ssuccessful outcome (table 7). Age, type of underlying heart disease, left ventricular ejection fraction, and the rnesults of mapping did not correlate with the clinical ooutcome. Patients in whom the outcome of ablation was not successful received a greater number of shocks and a greater total number of joules. There was no inducible VT 7 to days aft,er deliv- FIGURE 5. Schematic representation of the technique used to deliver ery of the shocks in 7% of patients who had a success- transseptal shocks. LV = left ventricle; RV = right ventricle. 4 ful outcome compared with 1 1% of patients without a successful outcome (p <.001). Complications (table ). One patient (. 9) developed sustained, nonclinical unimorphic VT immediately after receiving two shocks of 00 J each; direct- CIRCULATION

7 THERAPY AND PREVENTION-ARRHYTHMIA Discussion Efricacy of catheter ablation of VT. Catheter shocks delivered to the endocardium proved to be effective in preventing recurrences of VT in 45% of patients in this series. However, although the catheter ablation technique is effective in preventing VT over the long-termn in at least some patients, its success rate must be viewed with respect to the population of patients selected to undergo this procedure. It is important to note that the patients in this series were not representative of all patients who have VT refractory to pharmacologic therapy and that there was a strong bias toward selecting patients who were the most likely to respond favorably. After our initial experience suggested that success was unlikely in patients who had multiple morphologies of VT, subsequent patients were selected to undergo catheter ablation for the most part only if they had one configuration of unimorphic VT that was hemodynamically stable, allowing for adequate endocardial mapping. t included in this series were eight patients in whom the exit site of VT could not be ade- TABLE Results of programmed stimulation and mapping Clinical VT Downloaded from by on vember 1, 01 Patient Induced VT Earliest endo. nclinical. Morphology C (msec) activation (msec) Site Pace-mapA induced VT Endo NP, 500D RB,I NP ne c Good LB, 5, B Good ne 5 B Poor ne LB, I Good ne Good ne B Excellent ne 9 RB,S NP, 00 LB, I Poor ne 11 RB, Excellent ne LB, Excellent ne Good RB, I, LB, I Excellent ne Oc 1- Excellent ne 1 RB, Good types 17 B 40-4c3 Good ne 1 LB,I Good ne Excellent RB, 5, LB, I Excellent ne 1 RB, Good ne Ind Good ne 3 Ind Excellent ne Excellent ne Excellent ne RB, NP ne 7 LB, Good ne Good ne 9 B NP ne 30 LB, I Good ne Good ne 3 Ind.B NP ne NP ne -endocardial; NP -not possible due to noncapture. ASee Methods for grading system. BIncessantl spontaneous VT. cwith fractionated potentials bridging diastole. DRefers to cycle length, in msec. Vol. 75,. 5, May

8 Downloaded from by on vember 1, 01 MORADY et al. TABLE 3 Energy delivered and short-term results in patients who received shocks between internal electrode (cathode) and external patch electrode (anode) Total VT induced after ablation Patient Shock. of stored. site shocks energy (J) Immediate At 7- days Clinical NA 00 ne Clinical Clinical NA ClinicalB NA 5 00 ClinicalB NA ne ne A ne nclinical ne ne nclinical NA ne Clinical Clinical NA Clinical NA ClinicalB NA Clinical NA Clinical NA ne nclinical (NS) ne ne A Clinical NA ne ne ne nclinical ne ne ne ne Clinical NA ne Clinical ne ne ne ne ne ne ne ne ne nclinical NA not attempted; NS = nonsustained. ADivided between days. BFollowed by delivery of transseptal shocks (see table 4). 44 quately localized because of technical difficulty in mapping. Despite the bias in patient selection, the success rate was a relatively modest 45%. Nevertheless, our experience with catheter ablation of VT has indicated that this technique is of value in the management of selected patients who have recurrent VT. Many of the patients in this series who had a successful outcome had failed multiple trials of antiarrhythmic drugs (including amiodarone), were not suitable candidates for endocardial resection because of markedly depressed left ventricular function, and could not be managed successfully with an antitachycardia pacemaker; in addition, an automatic implantable cardioverter/defibrillator was not appropriate in many patients because episodes of VT occurred frequently or were incessant. Catheter ablation has resulted in control of VT for up to 35 months in these patients. Safety of catheter ablation of VT. Several complications were noted in this series, but none were fatal. In a patient who had never experienced ventricular fibrillation or a cardiac arrest before ablation, it is likely that postablation ventricular fibrillation was related to acute injury from the shocks. A second patient developed a new form of VT immediately after delivery of intracardiac shocks. Whereas malignant ventricular arrhythmias occur in a high percentage of dogs that receive endocardial shocks,' 113 our experience suggests that it is uncommon for endocardial shocks of 0 to 300 J to be proarrhythmic in man. However, because there is a risk, albeit small, of malignant arrhythmias as a complication of intracardiac shocks, we continue to monitor patients in the hospital for 7 to days after the ablation attempt. One patient in this series who received shocks to the ventricular septum developed transient complete AV block and underwent implantation of a permanent pacemaker because of residual infranodal conduction disease. AV block and bundle branch block also have been observed by other investigators in patients who received shocks to the ventricular septum." The cerebral ischemic episodes that occurred in two patients may have been caused by disruption of a thrombus in a left ventricular aneurysm or by release of a catheter-related thrombus. Also possible is cerebral ischemia caused by a gas embolus resulting from bubble formation during delivery of shocks in the left ventricle. 14 The other complications that were observed in this series consisted of transient AV block not requiring a permanent pacemaker and local complications at the site of catheter introduction into the brachial artery. Of note is that there were no instances of myocardial per- TABLE 4 Energy delivered and short-term results in patients who received transseptal shocks Pa- Site Site. Total tient of of of stored Induced VT. cathode anode shocks energy (J) Immediate At 7- days nclinical ne ne Clinical Clinical NA ne ne ne nclinical ne Clinical ne Clinical Abbreviations as in table 3. CIRCULATION

9 THERAPY AND PREVENTION-ARRHYTHMIA Downloaded from by on vember 1, 01 TABLE 5 Long-term follow-up VT during follow-up Patient AA Mos. Comparison with Patient. therapy of F/U Recurrence preablation VT status , APm Mex, AICD ne ne MexA ne, APm Prop AM, AICD Proc, Sot, Proc Proc, TocA Quinn ne ne Quin, Flec ne ne A ne A ne Verap A, ProcA Undoc. Different Different Different Death from VT Death from VT ncardiac death Death from VT Death from VT Death from VT Death from CHF Death from VT AA = antiarrhythmic; AICD = automatic implanted cardioverter/defibrillator; = amiodarone; APm = antitachycardia pacemaker; CHF = congestive heart failure; Flec = flecainide; F/U = follow-up; Mex = mexiletine; Proc = procainamide; Prop = propranolol; Quin = quinidine; Sot = sotalol; Toc = tocainide; Undoc = undocumented; Verap = verapamil. A drug regimen that was ineffective before ablation. TABLE Summary of clinical results of catheter ablation of ventricular tachycardia in 33 patients recurrence of VT 15 (45) On no antiarrhythmic drugs 9 (7) On same drug regimen ineffective before ablation (1) recurrence of VT on drug regimen not used before ablation 3 (9) Recurrent VT 15 (45) nfatal 9 (7) Fatal (1) foration or cardiac tamponade, and overall more than 0% of patients remained free of complications. *Transseptal delivery of shocks. Preliminary experiments in dogs demonstrated that shocks delivered transseptally between two catheters resulted in localized myocardial injury without perforation of the intraventricular septum.'0 This technique was applied to seven patients in the present series whose VT appeared to originate in the intraventricular septum. A successful outcome was noted in % of these patients. Of interest is that the transseptal shocks were successful in preventing further recurrences of VT in one of three Vol. 75,. 5, May

10 Downloaded from by on vember 1, 01 MORADY et al. TABLE 7 Comparison of patients with and without a successful outcome t p SuccessfulA successful value. of patients 15 1 Age (yr, mean + SD) NS Presence of 7 (47%) 13 (7%) NS Absence of heart disease 3 (0%) (11%) NS LVEF (mean ± SD) 0.3± NS Multiple VT morphologies Spontaneous 0 3 (1%) NS Induced 1 (7%) 5 (%) NS Left ventricular free-wall VT focus 7 (47%) (33%) NS Right ventricular VT focus (40%) (33%) NS Septal VT focus (13%) (33%) NS Endo. activation time at site of shock (msec) NS Presence of FBP 1 (7%) 3 (17%) NS Quality of pace-map Excellent 5/ (4%) 5/14 (3%) NS Excellent or good / (0%) /14 (%) NS Number of shocks 1.7±0..4±0.7 <.01 Total number of joules 304± <.01 inducible VT at 7- days (7%) (11%) <.001 FBP = fractionated bridging potentials; NS = not significant; other abbreviations as in tables 1 and. Aoutcome considered "successful" if no recurrence of VT after ablation, either with no drug therapy or with same drug regimen ineffective before ablation. patients who had had failed attempts at ablation with shocks delivered between an electrode position against the left ventricular aspect of the septum and an external patch electrode. Therefore transseptal shocks may at times be more effective than shocks delivered between a single electrode and an external anode in patients whose VT arises in the interventricular septum. Endocardial activation mapping. One of the criteria used for selecting patients to undergo attempts at catheter ablation was the ability to adequately localize the site of exit of VT. Therefore in all patients in this report, an endocardial site was identified at which activation during VT preceded the onset of the QRS complexes. However, the actual endocardial activation time was not helpful in predicting a successful outcome in response to ablation attempts. For example, there was no significant difference between the earliest activation time in patients with and without a successful outcome. In some patients the outcome of catheter ablation was successful although the endocardial activation time at the site of shock delivery was only msec earlier than the onset of the QRS complexes (figure ). On the other hand, the procedure was unsuccessful in some patients who had shocks delivered at 4 sites that had an activation time as early as 90 msec before the onset of the surface QRS complex (figure 7). Therefore the earliest activation time during VT is highly variable from patient to patient, and there is no absolute value for the endocardial activation time that predicts a high probability of success or failure of catheter ablation of VT. Fractionated bridging potentials during VT were noted in four patients at the presumed exit site of VT, but the success rate in these patients was only one of four. Although fractionated bridging potentials have been described as indicating proximity to the site of origin of VT,15 they were not helpful in the present series in predicting a high likelihood of success. Pace-mapping. In most patients in this series, pacemapping was used as an adjunct to endocardial activation mapping in localizing the exit site of VT.9 In general, intracardiac shocks were delivered only to sites at which pacing resulted in QRS complexes similar in configuration to the QRS complexes recorded during VT. Only two patients received shocks to sites at which the pace-maps were graded as "poor," with QRS complexes during pacing and VT being dissimilar in more than three electrocardiographic leads. The outcome was unsuccessful in both of these patients, despite the fact that endocardial activation at the shock sites preceded the onset of the QRS complexes during VT by 0 to 55 msec. Although endocardial activation at these sites occurred relatively early, the poor quality of the pace-maps may have indicated that endocardial activation occurred earlier at some other site. The success of catheter ablation of VT was independent of whether the quality of the pace-map was "excellent" or "good." Therefore concordance between pacing and VT in all leads of the electrocar- TABLE Complications of catheter ablation of VT in 33 patients. of pa- Complications tients (%) Sustained nonclinical VT immediately after shock 1 (3) Ventricular fibrillation (on days 5 and ) 1 (3) Neurologic deficit (aphasia, hemiparesis) () Transient :1 AV block 1 (3) AV block requiring pacemaker 1 (3) Brachial artery thrombosis 1 (3) Summary complications 7 () nfatal complications (1)A Fatal complications 0 AOne patient had two complications. CIRCULATION

11 THERAPY AND PREVENTION-ARRHYTHMIA FIGURE. Endocardial activation map in patient 3, showing the timing of endocardial activation at sites 14 and 1 relative to the earliest onset of the QRS complex in leads V1. I, and III. The pace-map at site 1 was excellent; however, the endocardial activation time at this site was only - msec. Earlier endocardial activation could not be demonstrated at any other site. The outcome ofcatheter ablation of VT in this patient was successful, with one shock of 150 J delivered to site 1. CL = cycle length. Downloaded from by on vember 1, 01,yj,v-!. FIGURE 7. Endocardial activation map in patient 5, showing the timing of endocardial activation at sites 14, 15, and 4- during VT. The pace-map at site 4- was considered excellent, and the endocardial activation time at this site was -90 msec. Earlier endocardial activation was not demonstrable at any other site. In spite of the excellent pace-map and identification of a site at which endocardial activation was very early relative to the QRS complex, the outcome of catheter ablation was unsuccessful in this patient, who received two shocks of 150 J each to site 4-. The patient's clinical VT was still inducible 30 min after delivery of the shocks. CL = cycle length. Vol. 75,. 5, May

12 Downloaded from by on vember 1, 01 MORADY et al. diogram is not necessary for successful ablation of VT by transcatheter shocks. However, because all but two patients had concordant QRS complexes in at least of leads, the minimum number of concordant leads required for a successful outcome is unknown. Predictors of a successful outcome. ong the selected patients in this series who underwent the catheter ablation procedure, there were no apparent predictors of a successful outcome. Although the success rate tended to be higher among patients who did not have coronary artery disease than among those who did (% vs 35%), this difference was not significant. However, it is possible that this difference would have been significant had the number of patients in each group been greater. Neither the left ventricular ejection fraction nor the exit site of VT correlated significantly with outcome. A successful outcome also did not correlate with the number of shocks or energy level used during the ablation procedure. Possible reasons for an unsuccessful outcome. The failure to ablate VT in some patients may have been related to the use of inadequate energy, especially in patients in whom the site of origin of VT may have been intramural instead of endocardial. It is likely that the site of origin of VT was not localized accurately in some patients. In addition, VT in some patients may have been caused by macroreentry1 17; in these patients, localized myocardial injury may not eliminate the reentry circuit. Comparison with prior reports. Eight prior reports included 51 patients who underwent attempts at catheter ablation with shocks up to 400 J in strength. 1 The overall success rate in these patients was approximately 30%, although in many patients the duration of follow-up was less than months. An additional 30% of patients became more responsive to antiarrhythmic drug therapy after ablation. Because of differences in patient selection, mapping technique, and ablation technique, it is difficult to compare these results with the results of the present report. Of note is that one patient died from hemodynamic collapse after receiving a single 0 J shock in the right ventricle.4 Pathologic examination revealed no evidence of myocardial perforation and the cause of death was unclear.4 There were no other deaths reported; however, the following serious complications were noted in prior reports: reversible cardiogenic shock in a patient who received four shocks of 50 to 300 J each, third-degree AV block requiring implantation of a permanent pacemaker in a patient who received six 300 J shocks to the septum, 1 a new type of VT in two patients who received two to three shocks of 50 to 300 J, and a decrease in left ventricular ejection fraction (from 0. to 0. and from 0.19 to 0.) in two patients who each received two shocks of 50 to 300 J. Limitations. Limitations in judging the efficacy of catheter ablation of VT may have led to an overestimate or underestimate of the success rate of this procedure. For example, in one patient who had frequent episodes of VT on a daily basis before ablation, there was an excellent short-term response, but the same configuration of VT recurred after months of followup. The success rate may therefore have been overestimated in the patients in this series whose follow-up has been less than months. On the other hand, the criteria used to judge longterm efficacy was rigorous and may have resulted in an underestimate of the success rate. A symptomatic recurrence of any type of VT was considered a treatment failure, even if the VT that recurred had a different configuration than the original clinical VT. In these cases it is not known whether the recurrent VT had a different site of origin than the original clinical VT or simply a different endocardial exit site. Because the occurrence of VT may be sporadic and unpredictable, an apparent decrease in the frequency of VT episodes was not considered to be a reliable end point for judging the efficacy of catheter ablation. Therefore three patients who had recurrent VT after the ablation attempt were categorized as having had an unsuccessful outcome even though the occurrence of VT was much less frequent than before the ablation attempt. It is possible that these patients did in fact have at least a partial response to the ablation attempts. Lastly, three patients who did not have symptomatic VT during follow-up were nevertheless categorized as having had an unsuccessful outcome because they were treated with an antiarrhythmic drug regimen that had not been used before ablation. It is possible that the ablation attempts contributed to the prevention of VT in these three patients. Conclusions. Catheter ablation of VT with intracardiac shocks is capable of preventing recurrences of unimorphic VT over the long term in patients both with and without coronary artery disease. However, the success rate of approximately 50% in a selected group of patients is lower than the success rate of map-directed or visually directed endocardial resection.1 In patients who have coronary artery disease, more than one configuration of VT, and a ventricular aneurysm, we favor endocardial resection over transcatheter ablation, particularly in those who would benefit from coronary artery bypass grafting. However, the results 4 CIRCULATION

13 Downloaded from by on vember 1, 01 of endocardial resection are suboptimal in patients who have markedly impaired left ventricular function or who do not have a ventricular aneurysm.'0 In these patients, the catheter ablation technique offers a relatively low-risk alternative to open-heart surgery and may be indicated if pharmacologic therapy is ineffective and if an antitachycardia pacemaker or automatic implantable cardioverter/defibrillator is not appropriate. The relatively small risk of serious complications has led us to offer the option of catheter ablation to some patients with recurrent VT of one morphology that has not necessarily been refractory to all antiarrhythmic drugs. The catheter ablation technique has been offered as a potential alternative to long-term pharmacologic therapy, particularly in patients whose VT arises in the right ventricle; in these patients, there is a smaller risk of cerebral emboli and arterial complications than in patients who require shocks in the left ventricle. The relatively high failure rate of catheter ablation is probably in large part attributable to inaccurate mapping techniques and/or inadequate tissue injury by the shocks. It is possible that the development of more precise mapping techniques and more effective techniques for selectively injuring tissue will result in more widespread application of catheter ablation for the treatment of VT. We are grateful to Hugh Sharkey, Booker Pullen, Beverly Burgie, Joan Bergeron, and Linda Abbott for their technical assistance and to Lisa Hackbarth for her secretarial assistance. References 1. Hartzler GO: Electrode catheter ablation of refractory focal ventricular tachycardia. J Coll Cardiol : 17, 193. Steinhaus D, Whitford E, Stavens C, Schneller S, McComb J, Carr J, McGovern B, Garan H, Ruskin J: Percutaneous transcatheter electrical ablation for recurrent sustained ventricular tachycardia. Circulation 70(suppl II): 11-0, 194 (abst) 3. Puech P, Gallay P, Grolleau R, Koliopoulos N: Traitement par electrofulguration endocavitaire d'une tachycardie ventriculaire recidivante par dysplasie ventriculaire droite. Arch Mal Coeur 77:, Fontaine G, Tonet JL, Frank R, Gallais Y, Farenq G, Grosgogeat THERAPY AND PREVENTION-ARRHYTHMIA Y: La fulguration endocavitaire: une nouvelle methode de traitement des troubles du rythme? Ann Cardiol Angeiol 33: 543, Downar E, Parson I, Cameron D, Waxman MB, Yao L, Easty A: Unipolar and bipolar catheter "ablation" techniques for management of ventricular tachycardia - initial experience. J Coll Cardiol 5: 47, 195 (abst). Belhassen B, Miller HI, Laniado S: Catheter ablation of incessant ventricular tachycardia refractory to external cardioversions. J Cardiol 55: 137, Klein H, Trappe HJ, Frank G, Lichtlen PR: Is fulguration of ventricular tachycardia really a new promising therapeutic approach? PACE 9: 9, 19 (abst). Belhassen B, Miller HI, Geller E, Laniado S: Transcatheter electrical shock ablation of ventricular tachycardia. J Coll Cardiol 7: 1357, Josephson ME, Waxman HL, Cain ME, Gardner MJ, Buxton AE: Ventricular activation during ventricular endocardial pacing. I1. Role of pace-mapping to localize origin of ventricular tachycardia. J Cardiol 50: 11, 19. Davis JC, Finkebeiner W, Ruder MA, DiCarlo LA Jr, Matsubara T, Chu W, Winston SA, Bharati S, Scheinman MM, Lev M: Histologic changes and arrhythmogenicity after discharge through transseptal catheter electrode. Circulation 74: 37, Lerman BB, Weiss JL, Buckley BH, Becker LB, Weisfeldt ML: Myocardial injury and induction of arrhythmia by direct current shock delivered via endocardial catheters in dogs. Circulation 9: 0, 194. Lee BI, Gottdiener JS, Fletcher RD, Rodriguez ER, Ferrans VJ: Transcatheter ablation: comparison between laser photoablation and electrode shock ablation in the dog. Circulation 71: 579, Kempf FC, Falcone RA, lozzo RV, Josephson ME: Anatomic and hemodynamic effects of catheter-delivered ablation energies in the ventricle. J Cardiol 5: 373, Bardy GH, Coltorti F, Ivey TD, Alferness C, Rackson M, Hansen K, Stewart R, Greene HL: Some factors mediating bubble formation with catheter-mediated defibrillator pulses. Circulation 73: 55, Josephson ME, Wit AL: Fractionated electrical activity and continuous electrical activity: fact or artifact? Circulation 70: 59, Mason JW, Stinson EB, Oyer PE, Winkle RA, Hunt S, Anderson KP, Derby GC: The mechanisms of ventricular tachycardia in humans determined by intraoperative recording of the electrical activation sequence. Int J Cardiol : 13, Miller JM, Harken AH, Hargrove WC, Josephson ME: Patterns of endocardial activation during sustained ventricular tachycardia. J Coll Cardiol : 0, Miller JM, Kienzle MG, Harken AH, Jospehson ME: Subendocardial resection for ventricular tachycardia: predictors or surgical success. Circulation 70:, Krafchek J, Lawrie GM, Roberts R, Magro SH, Wyndham CRC: Surgical ablation of ventricular tachycardia: improved results with a map-directed regional approach. Circulation 73: 39, Garan H, Nguyen K, McGovern B, Buckley M, Ruskin JN: Perioperative and long-term results after electrophysiologically directed ventricular surgery for recurrent ventricular tachycardia. J Coll Cardiol : 01, Kehoe RF, Zheutlin T, Finkelmeier B, Steinman R, Loeb J, Michaelis L, Moran J: Visually-directed endocardial resection for ventricular arrhythmia: long-term outcome and functional status. J Coll Cardiol 5: 497, 195 (abst) Vol. 75,. 5, May