| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2004;78:1979-1988
© 2004 The Society of Thoracic Surgeons
a Departments of Congenital Heart Disease/Pediatric Cardiology, Berlin, Germany
b Cardiothoracic and Vascular Surgery, German Heart Institute Berlin, Berlin, Germany
Accepted for publication February 6, 2004.
* Address reprint requests to Dr Nürnberg, Klinikum Links der Weser, Abteilung für Angeborene Herzfehler/Kinderkardiologie, Senator-Wessling-Strasse 1, 28277 Bremen, Germany
jan-hendrik.nuernberg{at}klinikum-bremen-ldw.de
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentLimitations of the StudyConclusionsAcknowledgmentsReferences |
|---|
METHODS: Seventy-four consecutive patients received either an LTFO (n = 29, 5 recordings in 1992 to 9 recordings in 1997) or an ECFO (n = 45, 11 recordings in 1995 to 5 recordings in 2001). The rhythm was documented preoperatively and postoperatively with standard electrocardiogram (ECG) recording and ECG monitoring. During follow-up all patients had 2–8 (median 3) standard ECG recordings per year. Additionally 45 patients (65%) had a Holter ECG at least once a year.
RESULTS: Median follow-up post-ECFO was 4.4 years (1.6–7.2) and post-LTFO it was 7.9 years (5.4–11.1). There were 5 early deaths (3 LTFO, 2 ECFO) and 1 late death (LTFO) (total mortality 8%). Sinus rhythm persisted in 37 ECFO patients (86%) as compared with 13 LTFO patients (50%) (p < 0.001). The incidence of new onset supraventricular tachyarrhythmias (SVTs) post-ECFO compared with LTFO was lower: 5 patients (11%) versus 11 patients (38%) early postoperatively (p < 0.001) and none versus 7 patients (27%) during follow-up (p < 0.001), respectively. Early postoperatively 10 LTFO patients (34%) and another 3 patients during follow-up required permanent pacemaker implantation due to bradyarrhythmias, but none of the ECFO patients required this.
CONCLUSIONS: Our data suggest that ECFO decreases the incidence of postoperative new onset arrhythmias during early and midterm follow-up compared with LTFO.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentLimitations of the StudyConclusionsAcknowledgmentsReferences |
|---|
|
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentLimitations of the StudyConclusionsAcknowledgmentsReferences |
|---|
|
|
ECFO patients undergo more surgical procedures before the final Fontan operation than LTFO patients (1.9 vs 1.3 operations per patient, respectively). Twenty-five ECFO patients and 12 LTFO patients had previously received an aortopulmonary shunt. The bidirectional cavopulmonary shunt (BCPS) was performed in 36 ECFO and 8 LTFO patients, respectively. Five LTFO patients underwent previous hemi-Fontan operation (ie, establishment of the right atrial to right pulmonary artery connection during the BCPS procedure and patch closure of the superior caval vein orifice). A single-stage Fontan procedure was performed in 5 LTFO and 4 ECFO patients, respectively. These data are displayed in Table 4.
|
All LTFOs were performed through a median sternotomy during low-to-moderate hypothermic cardiopulmonary bypass (CPB) with crystalloid cardioplegia. The lateral tunnel was established using a Gore-Tex polytetrafluoroethylene vascular patch that was fixed with straight intraatrial suture lines between the orifices of the inferior and superior caval vein and in some cases immediately adjacent to the crista terminalis. In 17 out of 29 patients (59%) an incision across the superior cavoatrial junction was created to enlarge the right atrial-to-pulmonary artery anastomosis. Routinely a fenestration with a diameter of 4–5 mm was punched directly into the patch.
Placement of the extracardiac conduit was achieved during normothermic CPB without cardiac arrest (33 patients) whereas during the introduction of the procedure the initial 7 patients experienced moderate hypothermia and cardioplegia, as did 5 other later patients because of additional intracardiac procedures. A tubular Gore-Tex vascular graft (diameter 16–24 mm) was interposed between the inferior caval vein and the right pulmonary artery. This surgical technique is described in detail elsewhere [8]. A direct fenestration (n = 11) or a small polytetrafluoroethylene (PTFE) tube with a diameter of 3–5 mm (n = 6) was placed between the conduit and the right atrial free wall in 17 patients (38%). The operation related data are presented in Table 5.
|
Definition of Arrhythmias
In comparison with normal sinus rhythm (NSR) at the age-dependent heart rate (see previously published data) [9], normofrequent atrial rhythm, junctional rhythm at a physiologic rate including atrioventricular (AV) dissociation, and retrograde or absent P-waves were defined as nonsinus rhythm (non-SR). Documented atrial flutter, atrial fibrillation, and AV junctional tachycardias were classified as SVTs. Sinus bradycardia and slow junctional rhythm were categorized as bradyarrhythmias (BAs) and separated from bradycardia because of complete heart block.
Statistical Analysis
The data were analyzed with the software SPSS for Windows 9.0 (SPSS Inc., Chicago, IL). As the groups do not fulfill the criteria for a gaussian distribution descriptive statistics are expressed as median (range). The nonparametric Mann–Whitney U test (including separate variances) and the 
2 test were used to compare variables between the two groups as appropriate. Independent factors associated with postoperative arrhythmias were sought using multiple logistic regression. Risk factors for postoperative new onset BAs or early and late SVTs identified by univariate analysis were selected for the multivariate analysis if there was no sizable direct correlation between these factors. The Cox proportional hazard regression analysis was applied to analyze the risk of late arrhythmias because of the differing follow-up periods. Differences were considered statistically significant if the p value was less than 0.05.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentLimitations of the StudyConclusionsAcknowledgmentsReferences |
|---|
Surgical Procedure
The duration of CPB throughout the Fontan operation and the aortic cross-clamp time (n = 12) was shorter in the ECFO group (p < 0.001, Table 5).
Early and Late Mortality
There were 2 early deaths (within 30 days after the operation) in the ECFO group and 3 early deaths in the LTFO group. Another adult patient with double inlet left ventricle died 5 years post-LTFO at the age of 22 years. This patient had a favorable morphologic result (no pulmonary artery stenoses or distortion) as well as favorable hemodynamic conditions (sinus rhythm, left atrial pressure 6 mm Hg, and mean pulmonary artery pressure 8 mm Hg during cardiac catheterization 3 years postoperatively). A first period of atrial flutter with 2:1 AV conduction developed 6 months before his death. He was treated successfully with electrical countershock after digoxin and captopril medication had been initiated. Either SVTs with rapid conduction or ventricular tachycardias are most likely the reason for his sudden death.
Arrhythmias
PREOPERATIVE
Although ECFO patients previously underwent more surgical procedures (Table 4) no difference in the preoperative rhythm was noted. NSR was present in 42 out of 45 ECFO patients (93%) and 27 out of 29 LTFO patients (93%) (Fig 1). Three ECFO patients experienced non-SR preoperatively (1 patient with atrial rhythm in heterotaxy syndrome, 1 patient with atrial rhythm from the coronary sinus region, and 1 patient with AV dissociation). The 2 remaining LTFO patients presented normal sinus node function but had permanent dual chamber pacemakers due to complete heart block after resection of a restrictive foramen bulboventriculare.
|
|
|
|
The development of transient early postoperative SVT before discharge was present in a considerably higher percentage of LTFO patients than ECFO patients (11 out of 29 patients [38%] vs 5 out of 45 patients [11%], p = 0.005) (Table 6). The majority of these early postoperative SVTs were of short duration and could be treated successfully with antiarrhythmic medication alone or if necessary by electrical countershock after both ECFO and LTFO. Only 1 37-year-old ECFO patient was receiving antiarrhythmic treatment (amiodarone) on discharge because of atrial flutter as compared with 5 LTFO patients who required postdischarge antiarrhythmic medication with either amiodarone, sotalol, or verapamil due to SVT. The majority of the early new onset postoperative arrhythmias post-ECFO were transient whereas in the LTFO group they persisted until after discharge.
MEDIUM-TERM FOLLOW-UP
When preoperatively present sinus rhythm persisted in 37 ECFO patients (86%) as compared with 13 LTFO patients (50%) (p < 0.001). Postoperative non-SR developed in three ECFO patients, however they maintained a normal heart rate after discharge. New onset arrhythmias did not develop in any of the 43 ECFO patients. The amiodarone treatment in the only patient with postoperative atrial flutter could be discontinued 6 months after discharge without recurrence of SVT until the last follow-up.
Out of 26 surviving LTFO patients, new onset SVT developed in 7 patients (24%) 1–4.5 years postoperatively (Fig 3). Recurrent atrial flutter required permanent antiarrhythmic medication and repeated cardioversion. Hemodynamically considerable bradyarrhythmia 2.3, 3.5, and 3.7 years after surgery, respectively (Fig 2), with the need for permanent pacing developed in three additional patients out of 10 with postoperatively exhibited non-SR, tachyarrhythmia, or bradyarrhythmia.
The overall 5-year freedom from new onset postoperative SVT and BA was 100% post-ECFO versus 73% for BA and 62% for SVT post-LTFO (Kaplan–Meier analysis, log rank with p = 0.0008 and p = 0.0001, respectively) (Figs 2, 3). Therefore the development of arrhythmias in patients with LTFO during follow-up occurred more frequently than in patients with ECFO (Figs 1–3; Table 6). Permanent pacemaker insertion due to early or late postoperative BAs was not necessary in any of the ECFO patients but was required in 7 out of 26 LTFO patients (27%) excluding the 2 patients with preoperatively implanted permanent pacemakers.
Arrhythmia Predictors
The application of multiple logistic regression analysis identified LTFO as well as the right ventricle in systemic position as notable independent risk factors for the development of new onset early postoperative tachyarrhythmias (p = 0.006 and p = 0.002, respectively). The incision across the cavoatrial junction during the LTFO was the sole substantial parameter for the actualization of postoperative BAs (p = 0.001). Taking the Cox proportional hazards model into consideration a loss of NSR on discharge (including normofrequent non-SR and tachyarrhythmic and bradyarrhythmic syndromes) and a previously performed atrioseptectomy were identified as considerable independent predictors for the development of late SVTs (p = 0.013 and p = 0.09, respectively). The right ventricle in systemic position also demonstrated a high tendency to be significant (p = 0.055). None of the other parameters tested yielded a level of significance. The results are summarized in Tables 7 and 8.
|
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentLimitations of the StudyConclusionsAcknowledgmentsReferences |
|---|
There is no doubt that the postoperative arrhythmias in Fontan patients are of multifactorial genesis. Obtaining NSR with AV synchrony (or normofrequent supraventricular rhythm in the presence of atrial isomerism/heterotaxy syndromes) and the avoidance of considerable bradycardia [9, 11, 13, 14] are possibly the most important factors regarding the prevention of tachyarrhythmias. Extensive suture lines followed by scar tissue formation in the right atrium, direct injury of the sinus node or its arterial supply during a previous hemi-Fontan operation, injury of the crista terminalis, large atrial wall incisions at the time of intracardiac procedures, and multiple-stage operations may predispose patients to atrial arrhythmias even after the latest modification of the Fontan operation [13, 15–18]. Experimental and clinical electroanatomic mapping procedures in patients superseding Fontan and Mustard operations have illustrated that the crista terminalis and right atrial scars can serve as nonconductive barriers that predispose a patient, in particular, to intraatrial reentrant tachyarrhythmias [19–21]. Gandhi and associates demonstrated that avoidance of suture lines near the crista terminalis reduces the incidence of atrial flutter after the modified Fontan operation [16], but the presence of extensive right atrial trabeculations may require shallow suture lines along the crista terminalis to avoid baffle leaks during LTFO [6]. Experience at the Mayo Clinic illustrates that patients having the baffle sewn to the crista terminalis experienced nearly twice the incidence of arrhythmias as compared with those having the baffle sewn anteriorly and well removed from the crista terminalis. The incidence of SND is higher in patients in whom incisions were carried across the cavoatrial junction [22, 23].
Another important factor is the persisting pressure load of the right atrial free wall even if it is only the portion within the lateral tunnel that conditions obvious myocardial hypertrophy [24] which itself, together with the surgical scars and anatomical barriers, contributes to the arrhythmogenic substrate. In contrast to LTFO use of the extracardiac conduit eliminates this additional predisposing factor although patients with ECFO are not free from scarred areas within the right atrium after previous operations. Nevertheless ECFO patients may be susceptible to tachyarrhythmias in the case of elevated pressure in the systemic atrium if the ventricular loading conditions are altered because of AV-valve incompetence (particularly in tricuspid or AV-canal morphology) or residual obstruction (minor restriction of the foramen bulboventriculare, a hypoplastic aortic arch, or a remaining noncompliant region after coarctation repair) that are more remarkable during increased physical activity of the growing child and will promote ventricular hypertrophy with resulting changes of the diastolic properties [25].
SND is, alongside SVTs, one of the dominant arrhythmias superseding the Fontan operation. Preoperative SND has been observed in as many as 7% of Fontan candidates [13]. Intracardiac electrophysiological studies revealed abnormal sinus node function and delayed intraatrial conduction in about 60% of patients after modified Fontan operations [26]. A higher risk of developing postoperative arrhythmias, especially bradyarrhythmia, due to an abnormal sinus node function [4, 27] is evident in patients with left atrial isomerism and heterotaxy. The age of a patient at the time of the Fontan operation seems to play a role in the frequency of atrial arrhythmias which occur less frequently in children undergoing Fontan completion at a younger age [4, 28] so that the duration of atrial and ventricular volume overload with resulting myocardial changes is shorter. Although our two patient groups differ considerably in their age at operation (Table 2) this occurrence could not be determined as an independent risk factor (Tables 7 and 8).
Stamm and associates observed freedom from new SVTs to be 96% and 91% at 5 and 10 years, respectively, in their series of LTFO patients, but BAs developed in one-third of their patients during the same time period. Preoperative and early postoperative BAs proved to be the most potent predictors of late BAs [11].
Analyzing 100 patients post-ECFO, Tokunaga and associates demonstrated that short-term to midterm follow-up illustrated that the event-free rate of the ECFO group was substantially superior compared with patients treated by LTFO [29]. Shirai and associates [30] observed a lower incidence of arrhythmias in a series of 16 patients post-ECFO compared with the data presented in the literature for post-LTFO, but at follow-up their patients exhibited a high incidence of asymptomatic SND detected by noninvasive testing. Petrossian and associates [2] also observed a low incidence of arrhythmias in their series of 51 patients post-ECFO at early and midterm median follow-up of 1.9 years. In a small series of 14 patients post-ECFO, sinus rhythm persisted during follow-up for 6–54 months [31].
There is still controversy about the role of interruption of the sinus node artery and incision of the superior cavoatrial junction during the hemi-Fontan operation and LTFO in the development of SND early or late postoperatively. It has been postulated that the nearer one operates to the sinus node, the more one expects to witness SND [13, 17]. Temporary loss of sinus rhythm in patients after TCPC was also noted by Kavy and associates [5] and was attributed to surgical compromise of the sinus node area. Cohen and associates identified SND in 30% of their patients after a mean follow-up time of 3.5 ± 1.7 years succeeding various modifications of the Fontan operation. The incidence was increased after 4 years of follow-up. Patients with atrial flutter were substantially at risk of SND. The authors concluded that SND might be, in part, due to the routine use of the intermediate hemi-Fontan operation because surgical incisions could injure the sinus node artery [13]. However later on the same authors reported a similar incidence of early SND post-ECFO and post-LTFO and concluded that avoidance of surgery near the sinus node has no effect on the development of early SND, however they do not present follow-up data [9]. In our opinion this conclusion seems to be anatomically and physiologically incorrect. The course of the sinus node artery is remarkably variable and barely predictable. Twenty years ago Anderson and associates stated, "It therefore behoves the surgeon to treat the entire superior cavoatrial junction with the utmost respect if he is to avoid damage to the sinus node and postoperative arrhythmic sequelae" [32]. The observations after repair of the superior sinus venosus type atrial septal defect (SND in 10% of patients vs 0.3% in patients with secundum type atrial septal defect) [33] as well as after modifications of the atrial switch operation for d-transposition of the great arteries [34] support the view that preservation of the superior cavoatrial junction with the sinus node and its artery is mandatory.
The incidence of new onset early arrhythmias in our patients was considerably higher post-LTFO than post-ECFO. Only 50% of LTFO patients versus 86% of ECFO patients remained in sinus rhythm at the end of the follow-up period. Contrary to Stamm and associates [11] we were able to distinguish the incision across the superior cavoatrial junction as the sole independent predictor of postoperative BAs. Multivariate analysis identified LTFO and the presence of a systemic right ventricle as single and independent predictors for the development of tachyarrhythmias early postoperatively. Irrespective of the surgical approach used to establish the Fontan circulation, our data indicate that the development of late tachyarrhythmias seems to be dependent upon non-SR on discharge and a preceding atrial septectomy. What is astonishing is that the latter was performed predominantly in our ECFO group (Table 4) with no newly occurring SVTs during follow-up so far. However patient age or previous surgery rate could not be identified as independent risk factors for the development of arrhythmias after the final Fontan operation.
Our findings are very similar to data presented by Azakai and associates [6]. In their series the LTFO group exhibited a higher incidence of postoperative SND and supraventricular tachyarrhythmia and LTFO was the only independent predictor of early and midterm postoperative atrial arrhythmias. Dilawar and associates recently reported SND in 21% of patients post-LTFO and in 59% of patients post-ECFO. They explain their conflicting results with different follow-up times (twice as long in the ECFO group). In addition the two Fontan procedures were performed by three different surgeons using different approaches with a high percentage of right atrial appendage to superior caval vein orifice incisions during the staging superior cavopulmonary connection [7].
Methodological Aspects
SURGICAL PROCEDURE
The two patient cohorts in this study are homogenous with respect to the surgical procedure. The majority of all operations (90%) were performed by one surgeon. Only extracardiac Fontan patients as already reported [8, 35] and those with an intraatrial lateral tunnel were included. Patients with other total TCPCs (ie, a cavopulmonary anastomosis with atrial septal defect closure in tricuspid atresia or direct atriopulmonary connections) were withdrawn. With respect to the underlying major cardiac defects both groups were considerably inhomogeneous—tricuspid atresia and double inlet left ventricle being predominant. A direct comparison between the diagnoses of the groups did not yield statistical significance (Table 3).
ARRHYTHMIAS
Their classification was oriented toward those arrhythmias that can be identified and differentiated by standard 12-lead surface ECG including prolonged registration and Holter ECG recordings although, in particular, the detection of atrial dysrhythmias is limited with Holter techniques because of the tiny atrial signals. Invasive electrophysiological investigations or even electroanatomical mapping procedures used to identify and define complex intraatrial reentrant tachycardias, to test AV conduction properties, or to test sinus node recovery were not performed routinely. In permanently paced patients the telemetric data was reviewed and offered additional information, however because standard pacemakers are not currently capable of storing intracardiac ECG tracings in the case of atrial high-frequency episodes, the value of this source is limited even if bipolar leads with little interference were implanted. Possibly one has to assume that any noninvasive arrhythmia detection study is handicapped because of a pronounced "gray zone" of undetected or incorrectly detected subclinical arrhythmias. Nevertheless the proportion of Holter records in our study (LTFO 88%, ECFO 53%) that were directly analyzed in our institution is substantially higher than in other studies comparing the two Fontan procedures (Table 1) including the appreciably populated LTFO study by Stamm and associates (20% Holter ECGs) [11].
|
Limitations of the Study |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentLimitations of the StudyConclusionsAcknowledgmentsReferences |
|---|
|
Conclusions |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentLimitations of the StudyConclusionsAcknowledgmentsReferences |
|---|
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentLimitations of the StudyConclusionsAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentLimitations of the StudyConclusionsAcknowledgmentsReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  C. Mavroudis, B. J. Deal, C. L. Backer, R. D. Stewart, W. H. Franklin, S. Tsao, K. M. Ward, and R. A. DeFreitas 111 Fontan Conversions with Arrhythmia Surgery: Surgical Lessons and Outcomes Ann. Thorac. Surg., November 1, 2007; 84(5): 1457 - 1466. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Kanter The Fontan Right Atrium--In Context Circulation, April 3, 2007; 115(13): 1698 - 1700. [Full Text] [PDF]
|
|
|
|
|
|
P. Khairy, N. Poirier, and L.-A. Mercier Univentricular Heart Circulation, February 13, 2007; 115(6): 800 - 812. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. B. Meyer, G. Zamora, G. Wernovsky, R. F. Ittenbach, P. R. Gallagher, S. Tabbutt, P. J. Gruber, S. C. Nicolson, J. W. Gaynor, and T. L. Spray Outcomes of the Fontan Procedure Using Cardiopulmonary Bypass with Aortic Cross-Clamping Ann. Thorac. Surg., November 1, 2006; 82(5): 1611 - 1620. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L.S. Morales, D. J. Dibardino, B. E. Braud, A. L. Fenrich, J. S. Heinle, W. K. Vaughn, E. D. McKenzie, and C. D. Fraser Jr Salvaging the Failing Fontan: Lateral Tunnel Versus Extracardiac Conduit Ann. Thorac. Surg., October 1, 2005; 80(4): 1445 - 1452. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. P. Graham Jr The Year in Congenital Heart Disease J. Am. Coll. Cardiol., June 7, 2005; 45(11): 1887 - 1899. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
