Inappropriate implantable cardioverter–defibrillator (ICD) therapies can lead to significant adverse events and increased mortality. These therapies are often the result of supraventricular tachycardias (SVTs). The objective of this study was to evaluate the incidence of SVT leading to inappropriate shocks in a large cohort of patients with ICDs and assess the efficacy of radiofrequency ablation (RFA) in decreasing these therapies. Patients with ICDs and recurrent SVTs were identified. A cohort of patients with ICD therapies subsequently underwent electrophysiologic study and RFA. Eighty-four patients (13%) were found to have SVT leading to 122 inappropriate ICD shocks and 130 episodes of antitachycardia pacing therapies. Median time to SVT onset after ICD implantation was 269 days. Electrophysiologic studies were performed in 30 patients. Successful RFA was performed for atrial tachycardia, atrial flutter, or atrioventricular nodal reentrant tachycardia in 22 patients. Ninety-five percent of patients who underwent successful SVT ablation had no further inappropriate ICD therapies compared to 63% of patients in whom ablation was not performed during a mean follow-up of 20.7 ± 11.9 months. In conclusion, SVT is responsible for a significant number of inappropriate ICD therapies. RFA is an effective strategy to substantially decrease subsequent inappropriate ICD therapies.
The life-saving clinical benefits of implantable cardioverter–defibrillator (ICD) therapy are well described. Although effective treatment of life-threatening ventricular arrhythmias is the goal, approximately 1 of 4 patients receive inappropriate ICD therapies because of misclassification of supraventricular arrhythmias by the device. Substantial adverse events from these inappropriate shocks include pain, psychological trauma, proarrhythmia, and a twofold increase in mortality. Although many inappropriate therapies are attributed to atrial fibrillation (AF) or sinus tachycardia, non-AF supraventricular tachycardia (SVT) appears to account for many inappropriate shocks. SVT in patients with heart failure also appears to be malignant and is an independent predictor of adverse events including mortality. We assessed the incidence of non-AF SVT in a large cohort of ICD recipients and evaluated the efficacy of catheter-based radiofrequency ablation (RFA) on subsequent inappropriate therapies and clinical outcome.
Methods
All patients who underwent insertion of an ICD with or without biventricular pacing at Albert Einstein Medical Center in Philadelphia, Pennsylvania from July 2005 to July 2009 and received follow-up at this institution’s device clinic were included in this retrospective study.
Demographic and clinical characteristics including age, gender, co-morbidities, severity and cause of ventricular dysfunction, medications, implant characteristics, time to first onset of SVT relative to date of implantation, and basic device parameters were ascertained from review of the medical record. Data from transthoracic echocardiograms before and after device insertion were recorded and included left ventricular ejection fraction, atrial size, and presence of valvular disease. Baseline programmed parameters from the implanted cardiac devices including the use of SVT discriminators were recorded.
Intracardiac electrographic recordings of tachyarrhythmia episodes in patients who previously received ICD or cardiac resynchronization therapy with defibrillator (CRT-D) devices were evaluated and characterized by 2 electrophysiologists. Episodes of AF were excluded from this analysis. Patients with documented preimplantation SVT were excluded from analysis. Tachyarrhythmias were determined to be SVTs if atrial electrographic frequency exceeded ventricular electrographic frequency or if variations in atrial cycle length resulted in variation in ventricular cycle length. SVT was also identified if therapies (antitachycardia pacing, cardioversion, or defibrillation) were delivered with evidence of persistence of atrial activity with atrioventricular dissociation. SVT was classified as a long RP interval if the RP/RR ratio was ≥0.5; for an SVT with cycle length >400 ms, the RP interval was ≥180 ms; and a short RP interval was defined as an RP/RR ratio <0.5 or an RP interval <180 ms. The presence of variable or <1:1 atrioventricular conduction was noted. Atrial and ventricular cycle lengths of the tachyarrhythmia were measured.
Patients with recurrent, inappropriate ICD therapies (shocks or antitachycardic pacing) for non-AF sustained SVT despite medications and programmed SVT discriminators, worsening heart failure symptoms, and/or decreasing left ventricular ejection fraction thought to be related to tachycardia-induced cardiomyopathy after ICD implantation were offered the option to undergo electrophysiologic study and possible RFA of the culprit arrhythmia. During these procedures, a standardized induction protocol of incremental pacing and extrastimuli from the atrium and ventricle were performed on and off isoproterenol 0.5 to 20 μg/min as tolerated clinically. All induced atrial arrhythmias were recorded. Standardized maneuvers were performed to identify the mechanism of SVT, i.e., atrioventricular nodal reentrant tachycardia, atrioventricular reentrant tachycardia, atrial tachycardia, or atrial flutter. After the diagnosis RFA was performed, as possible, of the induced arrhythmia(s). Electroanatomical mapping was used as necessary.
In addition, data including inducibility of arrhythmias were collected from the subgroup of patients who underwent electrophysiologic testing for any reason before device implantation.
Patients were followed routinely for subsequent SVT burden including inappropriate therapies. After RFA of induced arrhythmias, no programming changes were made to patients’ ICDs that would have affected SVT discrimination or inappropriate therapies.
Descriptive statistics were reported as mean ± SD for continuous variables, which were normally distributed. Group comparisons of continuous variables were evaluated using 2-tailed t tests. Group comparisons of categorical variables were evaluated using chi-square testing. A 2-tailed p value <0.05 was considered statistically significant. Statistical analyses were performed using SPSS 14.0 (SPSS, Inc., Chicago, Illinois).
Results
In total 660 patients underwent ICD implantation during the 5-year study period. Eighty-four patients (13%) had episodes of SVT. Demographic information and pertinent medical history are presented in Table 1 . SVT occurred in 33 patients with CRT-D (11.8%) and in 51 patients with ICD (13.4%). Most patients had structural heart disease. Most patients in the study group had significantly decreased left ventricular systolic function before ICD implantation, with a mean left ventricular ejection fraction of 28.8 ± 16.8%. All but 11 of the study group patients had a left ventricular ejection fraction ≤45%.
| Age (years) | 63.8 ± 16.5 |
| Men | 58 (69%) |
| Medical history | |
| Hypertension | 45 (54%) |
| Diabetes mellitus | 21 (25%) |
| Hyperlipidemia | 16 (19%) |
| Coronary artery disease | 35 (42%) |
| Ischemic cardiomyopathy | 25 (30%) |
| Nonischemic cardiomyopathy | 36 (43%) |
| Chronic kidney disease | 18 (21%) |
| Pulmonary hypertension | 20 (24%) |
| Chronic obstructive pulmonary disease | 13 (15%) |
| Sleep apnea | 4 (5%) |
| Stroke | 4 (5%) |
| Hypertrophic cardiomyopathy | 2 (2%) |
| Channelopathy | 2 (2%) |
| Left ventricular dyssynchrony | 27 (32%) |
| Preimplantation echocardiographic data | |
| Ejection fraction (%) | 28.8 ± 16.8 |
| Left atrial enlargement | 49 (58%) |
| Mean left atrial dimension (cm) | 4.3 ± 0.6 |
| Right atrial enlargement | 25 (30%) |
| Tricuspid regurgitation | 54 (64%) |
| Medications | |
| β Blockers | 81 (97%) |
| Angiotensin-converting enzyme inhibitors | 66 (78%) |
| Angiotensin receptor blockers | 8 (10%) |
| Hydralazine | 16 (19%) |
| Nitrates | 33 (39%) |
| Diuretics | 66 (78%) |
| Digoxin | 15 (18%) |
| Calcium channel blockers | 9 (11%) |
| Antiarrhythmic medications (excluding above) | 12 (14%) |
| Amiodarone | 11 (13%) |
| Sotalol | 1 (1%) |
| Indications for implantable cardioverter–defibrillator implantation | |
| Ventricular tachycardia | 6 (7%) |
| Syncope, low ejection fraction | 4 (5%) |
| Primary prevention of sudden cardiac death | 70 (83%) |
| Brugada syndrome | 2 (2%) |
| Hypertrophic cardiomyopathy | 2 (2%) |
ICDs were programmed based on physician preference to back-up VVI pacing at 40 to 50 beats/min for single-chamber devices or to DDD or DDDR 60 to 120 beats/min, with long atrioventricular delays to minimize right ventricular pacing, for dual-chamber devices. CRT-D devices were programmed to foster biventricular pacing. Percent right atrial pacing ranged from 0% to 35%, with a mean of 5.3%. Percent time the right ventricle was paced ranged from 0% to 99%, with a mean of 58%, although this mean decreased to 25% in those patients with standard single- or dual-chambered ICDs. Percent time that the left ventricle was paced in patients with CRT-D devices ranged from 91% to 99%, with a mean of 96.8%.
Time to onset of SVT relative to the date of device implantation averaged 407.9 days, with a range of 2 to 1,866 days. Median time of onset of SVT from implantation was 269 days. SVT appeared <10 days after implantation in 3 patients (3.6%). Based on analysis of stored atrial and ventricular electrograms from the SVT episodes, the 84 patients had 57 long RP, 8 short RP, and 19 variable RP tachycardias.
All available SVT discriminators were activated in all patients. Sustained rate duration cutoffs were set to nominal levels. Based on implanting physician preference, rate cutoffs for ventricular tachycardia in our ICD patients were 165 beats/min (cycle length 364 ms) or 180 beats/min (cycle length 333 ms) and rate cutoffs for the ventricular fibrillation zone were 190 beats/min (cycle length 316 ms) or 220 beats/min (cycle length 286 ms). Appropriate SVT discrimination avoided device therapies in 56 of 84 patients (66%). In 25 of the 28 patients (89%) who received ICD therapies for SVT, their clinically observed ventricular rates during SVT decreased in their programmed ventricular tachycardia zone, whereas in 3 of the 28 patients (11%) their SVT ventricular rate decreased in their programmed ventricular fibrillation zone.
Twenty-six of the 28 patients (93%) receiving therapies for SVT had an average of 4.2 shocks per patient (range 1 to 25), and 2 of the 28 patients (7%) had frequent antitachycardia right ventricular pacing therapy with 6 and 9 pace outs, respectively. In addition, the 26 patients receiving inappropriate ICD shocks for SVT also received an average of 4.5 antitachycardia pacing therapies (range 1 to 24). In most therapy attempts, antitachycardia pacing was generally unsuccessful as would be expected for focal atrial tachycardia and atrial flutter, and low and high energy shocks were frequently unsuccessful leading to multiple shock therapies being applied for each SVT event.
Ten of the 28 patients (36%) who received inappropriate ICD therapies for SVT also had episodes of sustained ventricular tachycardia. In 8 of these 10 patients (80%) the cycle length of ventricular tachycardia was within 40 ms of the ventricular cycle length of the SVT causing inappropriate shocks, and in 2 of the 8 patients (25%) the rate of ventricular tachycardia was slower than the rate of the clinical SVT.
Diagnostic electrophysiologic studies were performed in 30 of the 84 patients (36%) who had sustained SVT episodes. Electrophysiologic testing was performed in 28 patients because of recurrent ICD discharges caused by SVT and in 2 patients because of worsening left ventricular systolic function thought to be caused by tachycardia-induced cardiomyopathy. During routine electrophysiologic study, 26 patients (86%) were found to have 38 inducible SVTs ( Table 2 ). Four patients did not have inducible arrhythmias. One patient with inducible SVT could not be characterized because of repeated termination but was suspected of having macroreentrant typical atrial flutter. Mechanisms of all 38 induced arrhythmias are listed in Table 3 .
| Noninducible patients during electrophysiologic testing | 4 |
| Inducible patients during electrophysiologic testing | 26 |
| Focal atrial tachycardia | 10 |
| Right atrial flutter | 7 |
| Right atrial and left atrial flutter | 1 |
| Right atrial flutter and atrial tachycardia | 6 |
| Right atrial flutter/atrioventricular nodal reentrant tachycardia/atrial tachycardia | 1 |
| Unknown | 1 |
| Atrial tachycardia (n = 20) | |
| Right atrium | |
| Superior crista terminalis | 8 |
| Parahisian | 1 |
| Coronary sinus ostium | |
| Inferolateral tricuspid annulus | 1 |
| Right atrial appendage | 1 |
| Left atrium | |
| Left superior pulmonary vein | 1 |
| Right superior pulmonary vein | 1 |
| Roof | 1 |
| Uncharacterized | 2 |
| Atrial flutter (n = 17) | |
| Typical cavotricuspid isthmus dependent | 14 |
| Atypical right atrial flutter | 2 |
| Left atrial flutter | 1 |
| Typical atrioventricular nodal reentrant tachycardia (n = 1) |
Two patients had 2 different SVTs recognized sequentially after ICD implantation and underwent a second electrophysiologic study. One patient had isthmus-dependent atrial flutter 19 days after ICD implantation and a focal right atrial tachycardia 77 days after ICD implantation. The second patient had an isthmus-dependent atrial flutter 149 days after ICD implantation and then a focal right atrial tachycardia 521 days after ICD implantation.
RFA was performed in 22 of 26 patients (85%) who had 31 inducible sustained SVTs during electrophysiologic testing ( Figure 1 ) . Four patients did not undergo ablation because 2 patients had focal atrial tachycardia that repeatedly degenerated into AF preventing mapping, 1 patient had rapid atrial tachycardia with hemodynamic compromise, and 1 patient had atypical atrial flutter that repeatedly terminated but eventually was noninducible preventing further characterization. Of the 31 SVTs in patients who underwent ablation, 28 (93%) were successfully ablated. One example of a successfully ablated arrhythmia is shown in Figure 2 . One left atrial tachycardia was not ablated at the patient’s request to avoid a transseptal puncture. One focal atrial tachycardia repeatedly terminated preventing mapping. One left atrial flutter repeatedly terminated and became noninducible. There were no periprocedural complications associated with electrophysiologic testing or ablation in these patients.
