Author + information
- Received March 11, 1996
- Revision received April 16, 1997
- Accepted July 2, 1997
- Published online November 1, 1997.
- David A Harrison, MDA,1,1,
- Louise Harris, MB, FACCA,* (, )
- Samuel C Siu, MD, FACCA,
- Cynthia J MacLoghlinA,
- Michael S Connelly, MBBSA,
- Gary D Webb, MD, FACCA,
- Eugene Downar, MD, FACCA,
- Peter R McLaughlin, MD, FACCA and
- William G Williams, MD, FACCA
- ↵*Dr. Louise Harris, The Toronto Hospital, 150 Gerrard Street W, 3 Gerrard PMCC3-562, Toronto, Ontario, M5G 2C4, Canada.
Objectives. We sought to determine the features associated with sustained monoform ventricular tachycardia (VT) in adult patients late after repair of tetralogy of Fallot (TOF) and to review their management.
Background. Patients with repair of TOF are at risk for sudden death. Risk factors for ventricular arrhythmia have been identified from patients with ventricular ectopic beats because of the low prevalence of sustained VT.
Methods. From a retrospective chart review of patients assessed between January 1990 and December 1994, 18 adult patients with VT were identified and compared with 192 with repaired TOF free of sustained arrhythmia.
Results. There was no significant difference in age at repair, age at follow-up or operative history. Patients with VT had frequent ventricular ectopic beats (6 of 9 vs. 21 of 101), low cardiac index ([mean ± SD] 2.4 ± 0.4 vs. 3.0 ± 0.8) and more structural abnormalities of the right ventricle (outflow tract aneurysms and pulmonary or tricuspid regurgitation) than control patients. Electrophysiologic map-guided operation was performed in 10 of 14 patients who required reoperation. VT has reoccurred in three of these patients. Four patients did not undergo operation (three received amiodarone; one underwent defibrillator implantation). Two patients with VT also had severe heart failure and died.
Conclusions. Most patients with VT late after repair of TOF have outflow tract aneurysms or pulmonary regurgitation, or both. These patients have a greater frequency of ventricular ectopic beats than arrhythmia-free patients after repair of TOF. A combined approach of correcting significant structural abnormalities (pulmonary valve replacement or right ventricular aneurysmectomy, or both) with intraoperative electrophysiologic-guided ablation may reduce the potential risk of deterioration in ventricular function and enable arrhythmia management to be optimized.
Follow-up studies, now available for >25 years after operation [1–4], show that patients with tetralogy of Fallot (TOF) have an excellent life expectancy. When late mortality occurs, it is often sudden and attributed to ventricular arrhythmia [2, 5–7]. Although ventricular arrhythmias are believed to be a common mechanism of death in this population [8, 9], the prevalence of sustained ventricular tachycardia (VT) is low; therefore, risk factors and prognosis for patients with VT have usually been extrapolated from patients with frequent ventricular extrasystoles or brief, nonsustained salvos of VT [10–16].
Potential risk factors for ventricular arrhythmia that have been identified include older age at intracardiac repair [11, 16], increased right ventricular systolic pressure , moderate or severe pulmonary regurgitation , ventricular dysfunction and increased cardiopulmonary bypass time . Optimal management of the patient with sustained VT has yet to be determined.
A group of 18 patients with previous repair of TOF and documented sustained monoform VT were reviewed to determine whether risk factors associated with this arrhythmia could be identified and to report the long-term results of management of these patients.
Patients with VT were identified by a retrospective chart review performed in all patients with repair of TOF assessed at the Toronto Congenital Cardiac Centre for Adults (TCCCA) between January 1990 and December 1994. (This center follows up patients with congenital heart disease who are >18 years old; their demographic and clinical data are prospectively collected in a computerized database.) Patients presented with either 1) sustained symptomatic VT documented by a 12 lead-electrocardiogram (ECG), requiring either medical therapy or electrical cardioversion to terminate the arrhythmia, or 2) palpitations usually associated with syncope or near syncope. These patients were subsequently found to have sustained monoform VT at electrophysiologic testing. Sustained VT is defined as VT >30 s in duration. A control, arrhythmia-free group (n = 192) was derived from the 236 patients with no history of sustained VT who were followed up at the TCCCA between January 1990 and December 1994. Forty-four patients were excluded because of treatment for atrial arrhythmias (n = 17), ventricular extrasystole (n = 11) or a remote and poorly documented tachycardia of uncertain etiology (n = 16).
Surgical history, current therapy and electrophysiologic and hemodynamic data were obtained from review of health records. Angiographic data was interpreted by two observers (D.H., P.M.) who had no knowledge of the patients’ clinical status.
Standard 24-h ambulatory ECG monitoring is obtained approximately every 3 years for patients followed up at our center as part of a prospective follow-up protocol using either Cardiodata PR3 or Oxford MR4 monitor systems. The tapes were analyzed using the Cardiodata Mark 4 system, Oxford Excel or Marquette Electronics SXP Analysis System. Underlying rhythm, frequency of atrial and ventricular extrasystoles, presence of ventricular couplets and duration of nonsustained VT (>3 beats) were determined.
In those patients who underwent electrophysiologic study (17 of 18), induction of VT was performed using a standard protocol of pacing from the right ventricular apex at a cycle length of 600 and 400 ms, followed by three extrasystoles. If unsuccessful, this protocol was repeated from a midseptal site, and burst pacing to a maximum of 250 ms was also attempted.
The intraoperative electrophysiologic mapping procedure has been described in detail elsewhere [17, 18]. At the time of operation, a customized right ventricular balloon electrode array for recording endocardial activation was introduced through the right atrium into the body of the right ventricle and right ventricular outflow tract, with the patient on normothermic cardiopulmonary bypass. A sock was then positioned over the surface of the heart for epicardial recording. This enabled the simultaneous recording of endocardial and epicardial activation from 224 sites using a previously described multiplexed recording system . Programmed stimulation using a similar protocol to that outlined was used intraoperatively to initiate VT.
1.3 Hemodynamic Measurements
Sixteen of the 18 patients with VT underwent right and left heart catheterization as part of their arrhythmia assessment. Postrepair measurement of pulmonary artery pressures, right ventricular outflow tract gradient and cardiac output were available for comparison with a subgroup of the control patients who had previously consented to postrepair evaluation. From this group an age-matched hemodynamic comparison group was formed: Each patient with VT was matched with an “arrhythmia-free” patient for both age at intracardiac repair (within 1 year) and age at follow-up cardiac catheterization.
1.4 Angiographic Measurements
In patients presenting with VT, cardiac angiography was obtained to assess pulmonary and tricuspid valve competence and right ventricular morphology and function. In describing right ventricular morphology, we defined an aneurysmas a discrete area of the outflow tract with diastolic deformity and systolic filling; diffuse dilation of the outflow tract was not included in the definition. Because angiography was not performed routinely in patients without arrhythmias, a comparison group was derived from a random sample of patients undergoing angiography for hemodynamic indications (congestive heart failure, before pulmonary valve replacement).
1.5 Data Analysis
Results are presented as mean value ± SD, and 95% confidence intervals are provided where appropriate. Patients with VT and control patients without VT were compared using a two-tailed Student ttest. Hemodynamic data were analyzed using the Wilcoxon rank sum test. Categoric data were examined using chi-square or Fisher exact tests, where appropriate.
A total of 18 patients with VT (14 male, 4 female; 20 to 62 years old) were identified and compared with 192 “arrhythmia-free” patients (108 male, 84 female; 18 to 68 years old). There were no significant differences in mean age at intracardiac repair (11.2 ± 11.8 years, 95% confidence interval [CI] 5.3 to 17.1 vs. 11.5 ± 10.5 years, 95% CI 10.0 to 13.0, respectively) or age at last follow-up visit (33.2 ± 10.9 years, 95% CI 27.8 to 38.6 vs. 30.7 ± 10.4 years, 95% CI 29.2 to 32.2 years, respectively). Palliative shunt operations had been performed in 11 of 18 patients with VT (Blalock-Taussig in 9, Waterston in 2; 95% CI 0.36 to 0.83) and in 118 of 192 control patients (Blalock-Taussig in 105, Waterston, Potts in 5; 95% CI 0.55 to 0.68).
2.2 Electrophysiologic Findings
Monoform VT was documented electrocardiographically in 15 of the 18 patients before referral for electrophysiologic study. The primary symptom at presentation was resuscitated cardiac arrest (n = 1), syncope (n = 3), presyncope (n = 10) and palpitation alone (n = 1). The remaining three patients had experienced recurrent palpitations, with frequent multiform ventricular extrasystoles on 24-h ambulatory ECG monitoring, and were subsequently found to have inducible sustained VT at electrophysiologic study. For the complete group of 18 patients, the mean rate of VT was 230 beats/min (range 188 to 300), with a left bundle branch block configuration in 75% of patients.
Electrophysiologic study was obtained in 17 of the 18 patients: The remaining patient with documented VT presented with severe mitral regurgitation subsequent to bacterial endocarditis and was referred for urgent operation. Using the protocol described, 14 of 17 patients who had electrophysiologic study had inducible VT at electrophysiologic testing.
Twenty-four hour ambulatory ECGs obtained without antiarrhythmic therapy and before presentation with VT were available for 9 of the 18 patients. Frequent ventricular extrasystoles (defined as >30/min in any hour or nonsustained VT [>3 beats] or both) were present in six of these patients (67%, 95% CI 30% to 93%). In comparison, 21 of 101 arrhythmia-free patients (21%, 95% CI 13% to 29%) had frequent extrasystoles (p < 0.01).
2.3 Hemodynamic Characteristics
Right heart catheterization with an age-matched comparison group was available for 15 of 18 patients with VT (Table 1). Cardiac index was significantly lower in the group presenting with VT. Cardiac output was measured by the Fick technique in half the patients and by the thermodilution technique in the remainder. It is recognized that the latter technique may be less reliable in the presence of pulmonary or tricuspid regurgitation. There was no difference in either right ventricular or pulmonary artery systolic or diastolic pressures. There was a nonsignificant trend toward lower pulmonary artery diastolic pressure and to greater pulmonary artery pulse pressure in patients with VT.
Angiograms were available for review in 14 patients with VT. Moderate or severe tricuspid regurgitation was present in 3 patients, moderate or severe pulmonary regurgitation in 11 and a right ventricular outflow tract aneurysm in 11. At least one of these abnormalities was present in 12 (85%) of 14 patients. The comparison group included 14 patients without VT who underwent cardiac angiography for other clinical indications (heart failure, preoperative valvular assessment): 7 of these patients (50%) had no significant abnormality, and none had moderate or severe tricuspid regurgitation. Moderate or severe pulmonary regurgitation was present in five patients and right ventricular outflow tract aneurysm in six.
Medical management of VT involved the use of two or more different antiarrhythmic medications in 16 of the 18 patients. The medications used included amiodarone (13 patients), mexiletine (9 patients), quinidine (7 patients), sotalol (7 patients), procainamide (4 patients) and propafenone (4 patients).
Operation subsequent to the onset of VT was performed in 14 patients. Pulmonary valve repair (n = 1) or replacement (n = 10) was undertaken in 11 of the patients. In addition, 9 of these 11 patients had right ventricular outflow tract reconstruction with aneurysmectomy. VT could be induced in eight patients undergoing pulmonary valve reoperation (valve replacement in seven, valve repair in one), all of whom underwent cryoablation. The infundibular septum at its juncture with the ventricular septal defect (VSD) patch, parietal band or right ventricular free wall was the predominant region identified by intraoperative electrophysiologic mapping as the site of origin of VT.
Two patients underwent cryoablation without pulmonary valve surgery: resection of an outflow tract aneurysm with intraoperative cryoablation in one, cryoablation alone to a reentry circuit mapped to the site of a surgical pledget in the other.
One patient had mitral valve replacement subsequent to bacterial endocarditis. No electrophysiologic assessment was obtained in this patient intraoperatively, and this patient continued to experience problems with left ventricular dysfunction and prosthetic valve failure and died of heart failure 9 months after operation (8 years after presentation with the first episode of VT).
Four patients did not undergo operation subsequent to the onset of VT. One patient diagnosed at our center but subsequently managed elsewhere underwent placement of an implantable cardioverter-defibrillator (ICD) as primary management without surgical assessment. The other three patients were treated with amiodarone: One patient had a large residual VSD, moderate right ventricular dysfunction, moderate pulmonary regurgitation and severe tricuspid regurgitation and was treated for heart failure but died while awaiting surgical closure of his VSD; a second patient was awaiting operation at the time of this writing; a third patient did not have a structural indication for operation and has remained arrhythmia free 1 year after commencing amiodarone therapy.
As previously described, two patients died late after onset of VT. Both patients were in New York Heart Association functional class IV before their deaths. One of these patients died suddenly at home while receiving aggressive diuresis for exacerbation of heart failure. Sudden death occurred in 2 of 192 patients in the control group: 1 had known coronary artery disease and moderate left ventricular dysfunction, and the second patient had biventricular dilation present at autopsy.
Operation was an integral part of the management for 14 of the 18 patients with VT. Follow-up data are available for the surgical patients at 4.0 ± 3.5 years since operation. All but one surgical patient had assessment for cryoablation intraoperatively (Fig. 1). In three of these patients, VT could not be induced at operation. One of these three patients has had a clinical recurrence of monoform VT and is now receiving amiodarone therapy. Two patients have remained free of recurrence: One has stopped therapy, and VT is no longer inducible at electrophysiologic study; the other is maintained with the same therapy as that received preoperatively.
Cryoablation was performed in the 10 patients with VT inducible intraoperatively. Eight patients who underwent cryoablation have been free of sustained monoform VT; two have had recurrence (one now has an ICD; the other received amiodarone). Two patients had brief, nonsustained asymptomatic salvos of VT on routine monitoring: Mexiletine therapy was commenced in one and amiodarone in the other, and both have since remained arrhythmia free.
Monoform VT was identified in 18 of 254 patients late after repair of TOF. Factors found in the present study to be associated with VT include frequent ventricular extrasystoles and low cardiac index. Important abnormalities of the right ventricle (nonsignificant trends for each of moderate to severe pulmonary regurgitation, right ventricular outflow tract aneurysms and moderate to severe tricuspid regurgitation) were also more common in patients with VT.
3.1 Ventricular Ectopic Beats
Although ambulatory monitoring data were not available for all patients, the group with VT nonetheless exhibited a significantly higher frequency of ventricular extrasystoles. The latter have previously been identified as a significant risk factor for sudden death in TOF [6, 9]and correlate with ability to induce VT at electrophysiologic study . Risk factors for ventricular ectopic activity (Lown grade ≥2) after repair of TOF have been reported to include older age at repair [11, 16], moderate or severe pulmonary regurgitation , increased right ventricular systolic pressure , ventricular dysfunction and longer cardiopulmonary bypass time .
3.2 Influence of Age
Compared with the arrhythmia-free cohort, patients with VT in our study were of comparable age at intracardiac repair and at follow-up and had had a similar surgical history before intracardiac repair. However, many underwent operation at an older age than that considered current practice, with the majority of patients now undergoing repair in infancy. A significant effect of increasing age at time of definitive repair on ventricular arrhythmias has been previously reported [11, 16]. Older average age at time of initial repair was also observed in our control group and may reflect the nature of the population followed up at the TCCCA: patients are ≥18 years old at initial referral, and many underwent operation in an earlier surgical era, when corrective operation was unusual at <5 years of age.
3.3 Pulmonary Insufficiency
Twelve (86%) of 14 patients with VT who underwent angiography had important abnormalities of the right heart: moderate or severe tricuspid or pulmonary regurgitation or right ventricular outflow tract aneurysms. Our observation that valvular abnormalities are more often present in patients with VT is consistent with the previous observations of Zahka et al. . In their prospective study of valvular function in 59 patients after repair of TOF, the presence of at least moderate pulmonary regurgitation correlated with the presence of ventricular couplets or nonsustained VT on ambulatory monitoring. Similarly, the presence of these arrhythmias correlated with any severity of tricuspid regurgitation. As in our study, Zahka et al. found no relation between ventricular arrhythmias and right ventricular outflow tract gradient. The mechanism by which pulmonary regurgitation might facilitate the initiation of VT is not certain. Zahka et al. observed a correlation between the severity of pulmonary regurgitation and diastolic right ventricular area: Right ventricular enlargement with myocardial stretch would in turn provide an appropriate substrate for ventricular arrhythmias. The importance of right ventricular volume overload as a potential risk factor for VT has recently been addressed by Gatzoulis et al. , who found that patients with restrictive right ventricular physiology by Doppler assessment after repair were protected from the effects of pulmonary regurgitation and late ventricular arrhythmias, whereas patients with VT exhibited nonrestrictive physiology, right ventricular enlargement and QRS prolongation.
3.4 Outflow Tract Aneurysms
The coexistence of significant pulmonary regurgitation with aneurysmal dilation of the right ventricular outflow tract may provide a second explanation for the prevalence of pulmonary regurgitation in patients with VT. Of the 14 patients with ventricular arrhythmias who underwent angiography, 79% had outflow tract aneurysms. Our comparison group with 43% prevalence of aneurysm had undergone angiography for clinical indications and as such are unlikely to represent a normal population of patients after repair of TOF. Earlier reports of aneurysm formation after repair of TOF have suggested an incidence of 4% to 6% during long-term follow-up [20, 21].
Aneurysmal dilation of the right ventricular outflow tract appears to provide a suitable substrate for ventricular arrhythmias in these patients. Histopathologic studies [22, 23]in this population have confirmed the presence of isolated bundles of myocardial cells, surrounded by extensive fibrosis in the margin of the right ventricular outflow tract patch/right ventricular free wall junction. These findings are not unlike the changes observed in the border zone of the aneurysm in ischemic heart disease, a zone widely accepted as providing the substrate for VT in ischemic heart disease [24, 25]. We and others [26, 27]previously confirmed the mechanism of VT in TOF to be due to reentry. Premature stimuli during electrophysiologic mapping can evoke transient block and critical delays in activation, facilitating the initiation of the reentrant circuit. These abnormalities of conduction are presumably the functional manifestations of the pathologic findings previously described. The more recent study by Misaki et al. attempted to correlate similar abnormalities of conduction in patients with VT after repair of TOF with the histologic findings described above. Although some of the histologic changes described may represent the natural history of this congenital abnormality, there is considerable evidence that most of these changes are the consequence of the initial surgical repair. This finding is supported by observations of Dietl et al. , who compared the outcome of the transatrial versus the transventricular approach for initial repair of TOF. Although the comparison was of a consecutive rather than concurrent series of patients, the results suggested a significant reduction in ventricular arrhythmias using the transatrial approach with preservation of pulmonary valve and right ventricular function. Dietl et al. concluded that a right ventriculotomy scar significantly increased the risk of ventricular arrhythmias.
3.5 Follow Up
The majority of our patients (14 of 18) underwent operation subsequent to the development of VT; in most instances for severe pulmonary regurgitation and right ventricular outflow tract aneurysm resection. Suitability for intraoperative electrophysiologic mapping was assessed in all but one of the surgical patients, and cryoablation was performed in 10. With postsurgical follow-up of 4.0 ± 3.5 years, there has been arrhythmia free survival in 10 of the surgical group.
3.6 Limitations of the Study
This was a retrospective study in a patient cohort from a tertiary referral center. As such, we made no attempt to comment on the incidence of sustained VT in the adult population with repair of TOF. However, the majority of patients are followed up serially through the TCCFA, which serves as the continuity center for the Hospital for Sick Children, where most of the definitive repair of TOF for our region is performed.
Although the ambulatory ECG data are consistent with the observation of others [9, 10, 16], the hemodynamic/angiographic comparison group may not be reflective of the average patient with repair of TOF during follow-up. The latter is biased toward the patient being investigated for symptoms—usually congestive heart failure or preoperative valve assessment.
The management described was individualized. This was a retrospective study and not part of a randomized, controlled trial. In addition, because of the small sample size, the study may not have sufficient power to detect clinically relevant differences.
3.7 Conclusions and Clinical Applications
Two patients have died in the group with VT. Both patients had severe symptomatic heart failure. We believe that correction of the underlying structural problems may be as important in the long-term outcome of these patients as management of their potentially life-threatening arrhythmias. Indeed, it is possible that surgical repair alone may have an important moderating effect on the ventricular arrhythmias. Although treatment of the arrhythmia may be possible through the use of antiarrhythmic medication or an ICD, such an approach may not be ideal in the long term for all patients. Eighty percent of patients undergoing intraoperative map-guided operation in the present study have remained free of sustained VT during long-term follow-up. Patients with repair of TOF who present with sustained VT should be assessed for surgical intervention. A combined approach of correcting significant structural abnormalities in conjunction with intraoperative electrophysiologically guided ablation enables arrhythmia management to be optimized and may reduce the potential risk of deterioration in ventricular function.
We gratefully acknowledge Cyrus Salimi for assistance with data acquisition.
↵1 Dr. Harrison was a research fellow of the Heart and Stroke Foundation of Canada at the time this study was undertaken.
☆ This work was supported by a term grant from the Heart and Stroke Foundation of Ontario, Ottawa.
- confidence interval
- electrocardiogram, electrocardiographic
- implantable cardioverter-defibrillator
- ventricular septal defect
- ventricular tachycardia
- Toronto Congenital Cardiac Centre for Adults
- tetralogy of Fallot
- Received March 11, 1996.
- Revision received April 16, 1997.
- Accepted July 2, 1997.
- The American College of Cardiology
- Garson A,
- Nihill MR,
- McNamara DG,
- Cooley DA
- Gillette PC,
- Yeoman MD,
- Mullins CE,
- McNamara DG
- Dunnigan A,
- Pritzker MR,
- Benditt DG,
- Benson DW Jr.
- Garson A,
- Randall DC,
- Gillette PC,
- et al.
- Deanfield JE,
- McKenna WJ,
- Presbitero P,
- England D,
- Graham GR,
- Hallidie-Smith K
- Sullivan ID,
- Presbitero P,
- Gooch VM,
- Aruta E,
- Deanfield JE
- Zahka KG,
- Horneffer PJ,
- Rowe SA,
- et al.
- Webb Kavey R,
- Thomas FD,
- Byrum CJ,
- Blackman MS,
- Sondheimer HM,
- Bove EL
- Downar E,
- Parson I,
- Mickleborough LL,
- Cameron DA,
- Yao LC,
- Waxman MB
- Gatzoulis MA,
- Till JA,
- Somerville J,
- Redington AN
- Misaki T,
- Tsubota M,
- Watanabe G,
- et al.
- Horowitz LN,
- Spear JF,
- Moore EN
- Downar E,
- Harris L,
- Kimber S,
- et al.
- Dietl CA,
- Cazzaniga ME,
- Dubner SJ,
- Perez-Balino NA,
- Torres AR,
- Favaloro RG