Author + information
- Received October 31, 2001
- Revision received March 21, 2002
- Accepted April 5, 2002
- Published online July 3, 2002.
- Cevher Ozcan, MD*,
- Arshad Jahangir, MD*,
- Paul A Friedman, MD, FACC*,
- David L Hayes, MD, FACC*,
- Thomas M Munger, MD, FACC*,
- Robert F Rea, MD, FACC*,
- Margaret A Lloyd, MD, FACC*,
- Douglas L Packer, MD, FACC*,
- David O Hodge, MS†,
- Bernard J Gersh, MB, ChB, DPhil, FACC*,
- Stephen C Hammill, MD, FACC* and
- Win-Kuang Shen, MD, FACC*,* ()
- ↵*Reprint requests and correspondence:
Dr. Win-Kuang Shen, Division of Cardiovascular Diseases and Internal Medicine, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905, USA.
Objectives We evaluated the incidence and predictors of sudden death after atrioventricular (AV) node ablation and pacemaker implantation.
Background Sudden death may occur after radiofrequency catheter ablation of the AV node and pacemaker implantation in patients with atrial fibrillation (AF). Whether it is related to the procedure or to pre-existing heart disease remains unclear.
Methods All patients who had radiofrequency catheter ablation of the AV node and pacemaker implantation for rate control of medically refractory AF were identified retrospectively and observed prospectively. All patients with sudden death after ablation were identified. The relationship between the procedure and sudden death was defined on the basis of the time between the two as “likely,” “possibly” or “unlikely.”
Results Of 334 consecutive patients with AF who underwent AV node ablation, nine had sudden death after the ablation. Four patients (1.2%) had sudden death likely related to the procedure: in 3 patients, arrest occurred within 48 h after the procedure; in one patient, arrest occurred four days after the procedure. In three other patients (0.9%), sudden death was possibly related to the procedure because the event occurred within three months afterward. The remaining two deaths were unrelated to the procedure. Diabetes, New York Heart Association functional class (≥II), preprocedure ventricular arrhythmia, mitral or aortic stenosis, aortic regurgitation and chronic obstructive pulmonary disease were independent predictors for sudden death.
Conclusions Sudden death likely or possibly related to catheter ablation occurred in 7 of 334 patients (2.1%). Risk of sudden death is highest within two days after the procedure.
Catheter ablation of the atrioventricular (AV) node by radiofrequency energy and permanent pacemaker implantation is an alternative nonpharmacologic treatment for controlling the ventricular rate in symptomatic patients with atrial fibrillation (AF) for whom drug therapy has failed (1–4). Although AV node ablation does not eliminate AF, several studies have reported a reduction of symptoms and improvement in the quality of life, exercise tolerance and left ventricular function (4–8). Moreover, we recently demonstrated that treatment with AV node ablation and permanent pacemaker implantation does not have a negative effect on long-term overall survival when compared with pharmacologic treatment (9).
Although AV node ablation and permanent pacemaker implantation is considered effective and safe treatment for patients who have AF, life-threatening complications, such as ventricular arrhythmia and sudden death, have been reported to occur after the procedure (10–15). Previous studies showed that the incidence of these complications ranged from 3.1% to 6.7% (7,12,13). The annual rate of sudden death has been estimated to range from 1.9% to 3.7% (13,15). Because of the high prevalence of pre-existing cardiovascular disease (CVD) and the likely presence of an associated arrhythmogenic substrate in patients undergoing AV node ablation and pacemaker implantation for AF, it has been difficult to distinguish cardiac arrest secondary to pre-existing underlying disease versus procedure-related cardiac arrest.
In this study, we evaluated the incidence and predictors of sudden death after AV node ablation and pacemaker implantation in patients with AF that was refractory to drug therapy. Individual events were defined to identify the following: 1) sudden death likely related to the procedure; 2) sudden death possibly related to the procedure; 3) sudden death unlikely related to the procedure; and 4) predictors of sudden death.
Patients included in this study were all those with a history of AF who had radiofrequency AV node ablation and pacemaker implantation from July 1990 through December 1998 at Mayo Clinic (Rochester, Minnesota). Patients with indications for AV node ablation include symptomatic patients with paroxysmal or chronic AF refractory to pharmacologic therapy for maintaining sinus rhythm or ventricular rate control. Patients who had direct current (DC) AV node ablation were excluded from the study.
Sudden death was defined as death that was witnessed and the patient died within 1 h of the onset of symptoms or death that occurred within 1 h from when the patient was last seen to when the body was found. The death was defined as being “likely,” “possibly” or “unlikely” related to the procedure. If sudden death occurred within 48 h after the procedure or occurred at any time after the procedure in the absence of any CVD, the cause was defined as “likely” related to the procedure. If sudden death occurred between two days and three months after the procedure in the presence of CVD, the cause was defined as “possibly” related to the procedure. If sudden death occurred later than three months after the procedure in the presence of CVD, the cause was defined as “unlikely” related to the procedure.
Data were collected from a centralized system that provided complete records of AV node ablation and pacemaker implantation in patients observed at Mayo Clinic. These records provided a detailed history and diagnosis for all outpatient encounters, including emergency room visits, house and nursing home calls, as well as data recorded for inpatient care, death certificates and autopsies.
The follow-up period began at the time of the ablation procedure and ended in January 1999 or at the time of death. Patients were observed in the pacemaker clinic at three-month intervals for the first year. Thereafter, a yearly survey was conducted. Causes of death were determined from review of hospital records, death certificates and phone calls to the local physician or family members. Follow-up was complete for all patients.
AV node ablation and pacemaker implantation
Radiofrequency ablation of the AV node was performed by standard techniques in the electrophysiology laboratory (9). Complete AV block was achieved in all patients. The pacemaker was implanted the next day or immediately after ablation (since 1994). Seven patients (2%) required a left-sided approach to achieve AV block, and 24 patients (7%) required a second or third procedure because of recurrent AV conduction after the first attempt. A rate-responsive ventricular pacemaker (VVIR) was implanted when the patient had AF at the time of the procedure and when cardioversion to restore and maintain sinus rhythm was no longer attempted. A dual-chamber, rate-adaptive pacemaker (DDDR) was implanted when the patient had a history of paroxysmal AF and had sinus rhythm at the time of the procedure or sinus rhythm was restored by cardioversion. Before 1997, the lower pacing rate was programmed at 60 beats/min for temporary or permanent pacemakers after AV block. Since 1997, the lower pacing rate was set at 90 beats/min immediately after ablation and was gradually decreased by 10 beats/min monthly to 60 beats/min.
Data are presented as percentages or as means ± SD. A p value <0.05 was considered significant. Univariate and multivariate associations of baseline variables with sudden death were assessed using logistic regression analysis (16). The following variables were considered potential risk factors: demographic features (age and gender), clinical history (syncope, angina, congestive heart failure [CHF] and New York Heart Association [NYHA] functional class), heart disease (ischemic heart disease, cardiomyopathy and valvular heart disease) and associated clinical conditions (diabetes mellitus [DM], chronic obstructive pulmonary disease [COPD], cerebral vascular disease, hypertension and cancer). Significant univariate factors are presented as point estimates and 95% confidence intervals of the risk ratios. The multivariate model for risk factor analysis was not constructed because of the low number of events.
A total of 334 patients with AF (175 men and 159 women) who underwent AV node ablation and pacemaker implantation at Mayo Clinic from 1990 through 1998 were included in this study (Table 1). The mean age of patients in the study group was 68 ± 11 years (range, 32 to 95 years) at the time of ablation and pacing. Patients were observed for a mean of 36 ± 26 months (range, 3 days to 99 months; median, 33 months).
Classification of sudden death
Nine patients (six women and three men; mean age, 68 ± 7 years; range, 56 to 78 years) had sudden death after the ablation (Table 2). According to our definitions, four patients had sudden death likely related to the procedure because the event occurred within 48 h of the procedure in three cases, and one patient had sudden death four days after the procedure in the absence of any CVD. Sudden death in three patients was possibly related to the procedure because the event occurred between two days and three months after the procedure and in the presence of CVD. Another two patients died suddenly at 27 and 43 months after the procedure, and both had underlying structural heart disease. Therefore, we considered their deaths as unlikely related to the procedure.
Sudden death likely or possibly related to ablation
Of the 9 patients who had sudden death, 7 (2.1%; 5 women and 2 men; mean age, 67 ± 8 years; range, 56 to 78 years) had arrests that were likely or possibly related to the procedure (timing of the event after the procedure: mean, 10 ± 9 days; range, 1 to 35 days). The patients’ clinical characteristics are summarized in Tables 1 and 2. Baseline characteristics of these seven patients were not significantly different from those of the overall study cohorts except for the history of ventricular rhythm disturbances and COPD. Ventricular rhythm disturbances include previously documented premature ventricular complexes, nonsustained ventricular tachycardia and accelerated idioventricular rhythm.
Six of these seven patients had structural heart disease. Ventricular arrhythmia was present in four of the seven patients before the ablation (two with frequent premature ventricular complexes, i.e., >10/h; two with nonsustained ventricular tachycardia). The circumstances of the sudden deaths and the clinical outcomes are shown in Figure 1. Six of the seven patients had AF at the time of the ablation. Among those six patients, the mean heart rate was 112 ± 17 beats/min (range, 90 to 131 beats/min) before the ablation and 39 ± 25 beats/min (range, 0 to 75 beats/min) after ablation. An escape rhythm was documented in six of the seven patients immediately after AV node ablation. The Q-Tc interval of the escape rhythm after ablation (mean, 440 ± 65 ms) was not significantly prolonged compared with the interval before ablation (mean, 438 ± 55 ms) (p = 0.96). All except one patient (pacing rate, 90 beats/min) had the lower pacing rate programmed at 60 beats/min immediately after ablation. The initiating rhythm at the time of cardiac arrest was documented in five patients: ventricular fibrillation in four patients and polymorphic ventricular tachycardia in one patient. The rhythm was not documented in the remaining two patients. Two patients died at the time of cardiac arrest; five patients were resuscitated acutely, but three of them died within two months after arrest. The two survivors were alive at 35 and 39 months of follow-up after the cardiac arrest.
Predictors for sudden death
The following were independent predictors for sudden death in patients with AF after AV node ablation and permanent pacing: DM, NYHA functional class (≥II), preablation ventricular arrhythmia, mitral stenosis, aortic stenosis, aortic regurgitation and COPD (Table 3). The following were not independent predictors for sudden death in this population: age, gender, cardiac disease, cardiac surgery, hypertension, antiarrhythmic drug therapy, duration of AF, diagnosis of paroxysmal or chronic AF, low left ventricular ejection fraction, ablation time, Q-Tc interval, bundle branch block, cycle length or duration of AF and pacing mode.
Sudden death unlikely related to ablation
During the long-term follow-up, two patients died suddenly and unexpectedly several months after the procedure. Both patients had ischemic heart disease. One was a 69-year-old man with coronary artery disease, cardiomyopathy, hypertension and low left ventricular ejection fraction (0.30) who died 43 months after the ablation. The second patient was a 75-year-old woman with coronary artery disease and hypertension who died 27 months after the ablation. Pacemaker malfunction was not detected at the last follow-up examination before death. Thus, these patients were considered as having sudden death that was unlikely related to the ablation procedure.
Within the total study cohort, 74 patients had died at the time of latest follow-up. A cardiac cause of death was documented in 47 patients (64%). The most common cause of death was CHF, which occurred in 24 patients (32%).
In this clinical follow-up study of radiofrequency catheter ablation of the AV node and permanent pacemaker implantation in patients with medically refractory AF, the major findings include the following: 1) sudden death likely related to the procedure occurred in four patients (1.2%); 2) sudden death possibly related to the procedure was documented in three patients (0.9%); 3) the highest risk of sudden death was within 48 h after the ablation; and 4) DM, NYHA functional class (≥II), ventricular arrhythmia, mitral stenosis, aortic stenosis, aortic regurgitation and COPD were independent predictors for early sudden death. Although the overall incidence of sudden death related to the procedure is low, our data support the recommendation that in-hospital monitoring for a minimum of two days after the ablation and pacemaker implantation should be considered for patients with predictors of increased risk.
Sudden death after AV node ablation and pacemaker implantation
In a nonrandomized prospective, international, multicenter study, sudden death occurred in 3.8% of patients within the first year after DC catheter ablation of the AV junction (17). In another study, one of 17 patients who underwent radiofrequency ablation died suddenly two months after the procedure (18). A nonrandomized prospective study compared 54 patients treated with radiofrequency ablation with 49 patients treated with DC ablation (13). The incidence of sudden death was 2% in the DC group and 3.7% in the radiofrequency group. Kay et al. (7)recently reported a 3.2% rate of sudden death (five of 156 patients) in a one-year follow-up in a prospective nonrandomized study. Moreover, Darpo et al. (15)retrospectively analyzed the incidence of sudden death after radiofrequency ablation and reported a 1.9% annual incidence of sudden death. A multicenter study of 585 patients who underwent DC or radiofrequency ablation (19)and a meta-analysis of 1,181 patients who underwent radiofrequency ablation (20)reported risks of 1.04% and 2.0% annual sudden death mortality rates, respectively. However, the combined incidence rate of sudden death and malignant ventricular arrhythmias was high, ranging from 6.0% to 6.7% (13,14). In most of these studies, the cause-and-effect relationship between the procedure and sudden death was difficult to assess because of the various times between the cardiac arrest and the procedure and the high prevalence of underlying cardiac disease in these populations.
In our study, we observed that the overall incidence of sudden death after radiofrequency AV node ablation and permanent pacemaker implantation was 2.7%, and the incidence of cardiac arrest likely related to the procedure was 1.2%. The 1.2% incidence is similar to results reported by Darpo et al. (15)and lower than results from other studies. Although whether the procedure is a cause of sudden death cannot be precisely determined, the temporal relationship between the procedure and the event, as strictly defined in our study, and the detailed pursuit to determine the presence or absence of comorbid CVDs render the cause-and-effect relation likely. Other potential explanations for different results among the studies include differences in patient characteristics, methods, sizes and time frames.
Mechanism of sudden death
It has been conjectured that the smaller lesion associated with radiofrequency ablation compared with that associated with DC ablation may lead to a reduction in the risk of sudden death after AV node ablation. In patients who underwent DC AV node ablation, the reported incidence of sudden cardiac death was 2% to 3.8% (13,17). The difference between the incidence of sudden death in our study and the incidence of sudden death from DC ablation is not large.
The exact underlying mechanisms of unexpected sudden death remain elusive. Potential mechanisms include pacemaker failure in the absence of an escape rhythm (7,21), ventricular arrhythmias associated with coexisting heart disease (22–24)or exacerbated repolarization abnormalities secondary to an abrupt change in heart rate (14,25–27). In our study, pacemaker function had been normal at last follow-up in all patients with sudden death; however, postmortem examination of the pacemaker characteristics was not available. Although repolarization abnormalities mediated by bradycardia have been a suspected mechanism of sudden death after AV node ablation, the complexity and potential multiple mechanisms of arrhythmogenesis after AV node ablation are highlighted by the reports of one patient with cardiac arrest despite a higher pacing rate at 90 beats/min and one sudden death in a patient without underlying heart disease (14,25). The patient with sudden death despite a pacing rate of 90 beats/min also had ischemic heart disease, CHF and poor left ventricular ejection fraction. The absolute Q-Tc interval changes after ablation were not significantly different in six of the seven patients with an escape rhythm. Nevertheless, such observation does not preclude a rate-dependent repolarization abnormality as a potential arrhythmogenic mechanism as we now begin to understand the complex interactions among dispersion of refractoriness, ionic modulations and autonomic modulations (26,27).
Predictors for sudden death
Risk factors for sudden death after AV node ablation and pacemaker implantation have not been thoroughly evaluated in published works. In our study patients, DM, NYHA functional class (≥II), ventricular arrhythmia, valvular heart disease and COPD were independent predictors for sudden death likely or possibly related to the procedure. Previous studies reported that most patients who had cardiac arrest had ischemic heart disease, rheumatic heart disease or COPD. Because of the high prevalence of ischemic heart disease, CHF and low ejection fraction in our study patients, these factors were not independent predictors of early sudden death. The observation that DM may be an independent predictor for sudden death is intriguing. Repolarization prolongation related to oral antihyperglycemic agents (e.g., adenosine triphosphate-sensitive potassium channel blockers and sulfonylurea drugs) has been reported (28). Among the nine patients who experienced sudden death, five had DM, and four were receiving an oral antihyperglycemic agent.
This study should be interpreted in light of the limitations imposed by the retrospective nature of the study design. The multivariate model was used in our study to minimize the effect of baseline differences. In this study, therapy with ablation and pacemaker was selected for patients with AF refractory to drug therapy. This selection was not random. However, inclusion of consecutive patients minimizes other selection biases. Our study does not compare the risk of sudden death among the various treatments for maintaining control of ventricular rate or sinus rhythm in patients with AF. The power of the statistical analysis is limited by the low event rate.
This study provides the incidence of sudden death in patients with drug-refractory AF after radiofrequency AV node ablation and permanent pacing in a large group of consecutive patients. The incidence of sudden death likely or possibly related to the procedure is 2.1% (seven patients), and three of the seven events occurred within 48 h of the procedure in patients with identifiable risks. Physicians and patients should be aware of such risk when considering this therapeutic option. The final decision is reached by weighing the benefit-risk ratio, while recognizing that therapy with AV node ablation and pacemaker implantation is highly effective in controlling symptoms in patients refractory to medical therapy (1–8). This procedure does not decrease long-term survival when compared with medical therapy (9). Longer in-hospital monitoring and observation (i.e., two to three days) should be considered in patients with DM, COPD, significant ventricular arrhythmia or valvular heart disease.
- atrial fibrillation
- congestive heart failure
- chronic obstructive pulmonary disease
- cardiovascular disease
- direct current
- diabetes mellitus
- New York Heart Association
- Received October 31, 2001.
- Revision received March 21, 2002.
- Accepted April 5, 2002.
- American College of Cardiology Foundation
- Prystowsky E.N.,
- Benson D.W. Jr..,
- Fuster V.,
- et al.
- Brignole M.
- Marshall H.J.,
- Harris Z.I.,
- Griffith M.J.,
- Gammage M.D.
- Kay G.N.,
- Ellenbogen K.A.,
- Giudici M.,
- et al.
- Huang S.K.,
- Bharati S.,
- Graham A.R.,
- Lev M.,
- Marcus F.I.,
- Odell R.C.
- Langberg J.J.,
- Chin M.C.,
- Rosenqvist M.,
- et al.
- Olgin J.E.,
- Scheinman M.M.
- Walker S.H.,
- Duncan D.B.
- Evans G.T. Jr..,
- Scheinman M.M.,
- Bardy G.,
- et al.
- Jackman W.M.,
- Wang X.Z.,
- Friday K.J.,
- et al.
- Wood M.A.,
- Brown-Mahoney C.,
- Kay G.N.,
- Ellenbogen K.A.
- Mehta D.,
- Curwin J.,
- Gomes J.A.,
- Fuster V.
- Haider A.W.,
- Larson M.G.,
- Benjamin E.J.,
- Levy D.
- Garratt K.N.,
- Brady P.A.,
- Hassinger N.L.,
- Grill D.E.,
- Terzic A.,
- Holmes D.R. Jr..