| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a Division of Cardiovascular Thoracic Surgery, Children’s Memorial Hospital, Northwestern University, Chicago Illinois
b Division of Cardiology, Children’s Memorial Hospital, Northwestern University, Chicago Illinois
c Department of Surgery, Feinberg School of Medicine, Northwestern University, Chicago Illinois
d Department of Pediatrics, Feinberg School of Medicine, Northwestern University, Chicago Illinois
Accepted for publication June 26, 2007.
* Address correspondence to Dr Mavroudis, Division of Cardiovascular Thoracic Surgery, M/C #22, Children’s Memorial Hospital, 2300 Children’s Plaza, Chicago, IL 60614. (Email: cmavroudis{at}childrensmemorial.org).
Presented at the Forty-third Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 29–31, 2007.
| Pediatric cardiac surgery:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionFootnotesAcknowledgmentsReferences |
|---|
Methods: Since 1994, 111 patients (mean age 22.5 ± 7.9 years) underwent Fontan conversion with arrhythmia surgery. The series was divided into three time periods: (1) 1994 to 1996 (initial isthmus ablation, n = 9, group I); (2) 1996 to 2003 (early modified right atrial maze and Cox-maze III, n = 51, group II); and (3) 2003 to 2006 (recent modifications of the modified right atrial maze and left atrial Cox-maze III for both atrial fibrillation and left atrial reentry tachycardia, n = 51, group III).
Results: There were one early (0.9%) and six late deaths (5.4%); six patients required cardiac transplantation (5.4%). Two late deaths occurred after transplantation. Renal failure requiring dialysis occurred in four patients (3.6%). Mean hospital stay was 13.7 ± 12.1 days. Mean cross-clamp time was 70.8 ± 41.6 minutes. Four risk factors for death or transplantation were identified: presence of a right or ambiguous ventricle, preoperative protein-losing enteropathy, preoperative moderate-to-severe atrioventricular valve regurgitation, and long (>239 minutes) cardiopulmonary bypass time. In intergroup comparisons (groups I and II versus group III), three trends were noted: increased incidence of concomitant surgical repairs (p = 0.03), older patients (p = 0.01), and increased incidence of left atrial reentry tachycardia and atrial fibrillation (p = 0.04). Late recurrence of atrial tachycardia ensued in 15 of 111 (13.5%); 8 of 51 in group II (15.7%) and 4 of 51 in group III (7.8) (p = 0.3).
Conclusions: Fontan conversion with arrhythmia surgery is safe and efficacious. Based on improved results and evolving surgical techniques, selection criteria can be more clearly defined.
Atriopulmonary to total cavopulmonary artery extracardiac Fontan conversion with arrhythmia surgery and pacemaker therapy has been established for Fontan patients with troublesome arrhythmias, pathway obstructions, valvar dysfunction, and decreasing functional status [1–8]. These problems are not unlike the constellation of residual and acquired lesions that complicate the clinical course and long-term survival of adults with repaired complex congenital heart disease [9–11]. Fontan conversion with arrhythmia surgery therefore is an operation with inherent and considerable variation, which is highly dependent on the fundamental diagnosis, the original atriopulmonary connections, the type of arrhythmia, and the severity of the associated lesions.
We have sought to update and report our experience as warranted by technique modification, changing clinical patterns, and outcome analysis [1–8]. The purpose of this paper is to highlight the surgical solutions that were applied to this complex and changing group of patients and to analyze the results in an attempt to identify risk factors for poor outcomes.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionFootnotesAcknowledgmentsReferences |
|---|
Patient Population
The series was divided into three time periods based on the evolution of our arrhythmia surgery techniques. Group I consists of the first nine patients in our series in whom we used isthmus cryoablation for arrhythmia control. The next 102 consecutive cases had variations of the modified right atrial maze for atrial reentry tachycardia and the Cox-maze III procedure for atrial fibrillation and, more recently, left atrial reentry tachycardia. We subdivided this population temporally into group II, 51 patients operated between December 1996 and January 2003, and group III, 51 patients operated between May 2003 and November 2006, in order to document any changing patterns of presentation and evolving therapeutics. The preoperative diagnoses and types of arrhythmias are noted in Table 1. Clinical characteristics are noted in Table 2.
|
|
Surgical Techniques
We have previously reported our operative modifications to facilitate arrhythmia surgery in light of differing anatomic substrates [8]. Operative preferences include repair of atrioventricular valves preferentially over valve replacement and atrial reconstructive techniques to ensure unobstructed pathways. With complex connections made at the initial Fontan operation, the need to tailor our approach to the individual patient evolved (Table 3), and required adaptation of our techniques to allow optimal access to anatomic landmarks. For instance, prior partition of the atrium as seen with double-inlet ventricles or patch occlusion of the diminutive right-sided atrioventricular valve required patch takedown in order to access the coronary sinus and valve annulus to perform the maze procedure.
|
|
|
Pulmonary Venous Obstruction, Residual Shunts, and Anomalous Systemic Venous Return
Pulmonary venous obstruction was often caused by an enlarging right atrium with posterior encroachment on the right-sided pulmonary veins, or by coronary sinus dilatation with posterior encroachment on the left-sided pulmonary veins. Right atrial wall reduction and atrial septal defect creation addressed right-sided obstruction, and unroofing of the coronary sinus relieved left-sided pulmonary venous obstruction [8].
Residual ventricular septal defects were closed as indicated to optimize the extracardiac Fontan circulation. Anomalous systemic venous return (eg, in heterotaxy syndrome patients), was treated by incorporation into the extracardiac venous circuit pathway by direct anastomosis when possible.
Cardiac Transplantation Considerations
The prospect of future cardiac transplantation in these patients led us to modify our reconstructive techniques to eliminate use of homograft material to avoid the problems of increased panel-reactive antibody levels [13]. Between 1996 and 1998, we used aortic homograft material to create Fontan connections in only three patients; polytetrafluoroethylene (PTFE) tube grafts were used in all other extracardiac Fontan connections. Use of homograft materials in concomitant surgical procedures has been limited. In group II, six of 13 patients had pulmonary artery reconnection using aortic homograft, which decreased to two of eight patients in group III. One patient had right and main pulmonary artery patch arterioplasty using autograft material harvested from excess atrial tissue.
Statistical Analysis
Comparison of variables among groups I, II, and III was made by 
2 analysis for categoric variables and by analysis of variance for continuous variables. Because of the temporal differences among groups, time to untoward events rather than number of events was compared using statistical survival methods. Association of patient variables with death or transplantation was determined by multivariate Cox proportional hazards modeling. The variables included in the initial model are listed in the Appendix and the final model was created by stepwise backward elimination of nonsignificant variables. Freedom from death, death or transplantation, or recurrence of atrial tachycardia (Figs 2; 3)
was calculated by the Kaplan-Meier product-limit methodology. Statistics were performed with StatView (SAS Institute, Inc, Cary, NC).
|
|
|
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionFootnotesAcknowledgmentsReferences |
|---|
Perioperative Complications
Unwanted cavitary entry during resternotomy occurred in three patients and inadvertent entry into the left internal mammary artery occurred in one patient, all resulting in femoral cannulation. Elective femoral cannulation was performed in eight patients because of gigantic and centrally located right atria. At administration of cardioplegia, left main coronary artery entry was noted in one patient as a consequence of atrioventricular groove dissection; this was repaired with suture and a PTFE patch. One patient underwent reexploration for postoperative bleeding. Renal failure requiring dialysis occurred in four patients (3.6%); one died 11 months postoperatively after refusing chronic dialysis therapy, two underwent cardiac transplantation (one survived), and the remaining patient recovered normal renal function. No patient had a clinically apparent neurologic event. Mediastinitis occurred in two patients (1.8%): one required sternal debridement and another was treated with vacuum-assisted closure. One patient with preoperative chronic hepatic dysfunction, ascites, and decreased ventricular function required chronic intermittent peritoneal drainage for significant ascites and is living 16 months postoperatively. She is a potential candidate for liver transplantation despite favorable Fontan hemodynamics; however, because of Fontan physiology, a preceding cardiac transplant is likely necessary for hepatic graft survival. Two other patients, ages 2.6 years (hypoplastic left heart syndrome) and 38.8 years (tricuspid atresia) with mild preoperative ventricular dysfunction developed compensated protein-losing enteropathy 3 and 2.5 years after Fontan conversion, respectively.
Mortality and Cardiac Transplantation
There was one early (0.9%) and six late deaths (5.4%). Six patients (5.4%) required cardiac transplantation one week and 6, 8, 10, 11, and 33 months after Fontan conversion with arrhythmia surgery. Two of the six late deaths occurred 4 and 24 days after cardiac transplantation (ten months and seven months, respectively, after Fontan conversion with arrhythmia surgery). Among the four other late deaths, a 41-year-old patient died five months postoperatively of ventricular failure with severe atrioventricular valve regurgitation while being evaluated for cardiac transplantation at another institution. A 27-year-old patient died suddenly two years postoperatively; at autopsy the right coronary artery was thrombosed. A 21-year-old patient died suddenly 2.5 years postoperatively after sedation administration during a psychiatric hospitalization. A 27-year-old patient died 11 months postoperatively after refusing further dialysis. Overall freedom from death or cardiac transplantation is noted in Figure 2. Four preoperative and operative risk factors were identified by Cox proportional hazards modeling as significant risk factors for death or transplantation: presence of a right or ambiguous ventricle, preoperative protein-losing enteropathy, moderate-to-severe atrioventricular valve regurgitation, and long (>239 minutes) cardiopulmonary bypass time (Table 5). In contrast to our earlier analysis [5], older age at surgery, severe ventricular dysfunction, and long ischemic time (>100 minutes) were no longer risk factors.
|
Arrhythmia Recurrence
All patients had postoperative prophylactic antitachycardia medical therapy. For patients who had a modified right atrial maze, beta receptor antagonists were started several days postoperatively after a noninvasive electrophysiologic study was performed. For patients undergoing a Cox-maze III procedure amiodarone was started in the operating room after pacemaker placement. Antiarrhythmic medications were withdrawn three months postoperatively; patients were then closely monitored for recurrence. Late recurrence of atrial tachycardia (>30 days after surgery) occurred in 15 of 111 (13.5%). The twelve surviving, nontransplanted patients who experienced arrhythmia recurrence are on antiarrhythmic medications. The recurrence rate in group I was three of nine (33%). Of the 102 patients in groups II and III, atrial arrhythmias recurred in 12 (11.9%). In the group II patients, eight of 51 (15.7%) had atrial tachycardia recurrence, and in group III, four of 51 (7.8%) had atrial tachycardia recurrence (p = 0.3, group II versus group III). Freedom from arrhythmia recurrence is shown in Figure 3.
Among 33 patients with right atrial tachycardia undergoing modified right atrial maze, the recurrence rate was 9% (3 of 33) with an average follow-up time of 6 ± 2 years (range, 1 to 10 years). After Cox-maze III procedure for atrial fibrillation, there was no recurrence of atrial fibrillation, although seven of these patients (7 of 53, 13%) experienced atrial tachycardia with an average follow-up time of 3 ± 2.5 years (range, 1 month to 9 years). After Cox-maze III for left atrial reentry tachycardia, two patients experienced recurrence of atrial tachycardia (2 of 12, 17%) with an average follow up time of 1.5 ± 1 years (range, 6 months to 5 years).
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionFootnotesAcknowledgmentsReferences |
|---|
Atrioventricular valve repair remains a challenge due to abnormal valve morphology in addition to regurgitation or stenosis [16, 17]. Our experience is small; however, it appears that the Alfieri valvuloplasty technique has recently resulted in improved and sustained valvar function in patients with mitral valves. Tricuspid valve repair has not been as successful in our experience; valve replacement should be considered in these patients with preserved right ventricular function and tricuspid regurgitation, reserving cardiac transplantation for those patients with severe ventricular dysfunction. Patients with regurgitant common atrioventricular valves also present with individualized problems because no commercially available annular rings are suitable and because there is no large experience with different types of repairs. More creative annulo-valvuloplasty techniques may need to be developed to treat these patients if these valves are to be preserved.
The development of atrial tachycardia in atriopulmonary Fontan patients is likely multifactorial, including atrial incisions-suture load (similar to findings in patients with Mustard and Senning operations [18]), dilatation, fibrosis, hypertrophy, atrioventricular valve regurgitation, and fundamental diagnosis. The more recent lateral tunnel and extracardiac Fontan modifications [19, 20] are accruing enough follow-up time to indicate that the lateral tunnel [21], more than the extracardiac [22] Fontan patients, are experiencing atrial arrhythmias, consistent with the increased suture lines necessary for the lateral tunnel procedure. Comprehensive follow-up of all Fontan patients will be necessary to detect and analyze the problems that may occur with the Fontan modifications being performed today.
There has been a shift in presenting arrhythmias from predominantly right atrial reentry tachycardia to predominantly atrial fibrillation. We also identified several patients with left-sided atrial reentry tachycardia circuits due to left atrial suture lines and dilatation. The determinants of atrial fibrillation are not entirely clear but this change in arrhythmia referral pattern may reflect the increased number of prior ablation procedures that delay surgical referral [23]. This is a disturbing trend because atrial fibrillation causes greater functional disability, results in thrombus formation, and its pharmacological treatment has significant associated negative side effects. Furthermore, the surgical treatment of atrial fibrillation in association with Fontan conversion requires a longer cross-clamp time. Consideration should be given to earlier referral for those patients with refractory arrhythmias to reverse this concerning trend.
The arrhythmia operation has undergone a metamorphosis to a modified right-sided maze procedure for right atrial reentry tachycardia, and to the left atrial Cox-maze procedure for atrial fibrillation and left atrial reentry tachycardia. The modified right-sided maze procedure is more variable because of different anatomical substrates; however, the principles of ablating the areas of slow conduction remain. When the tricuspid valve and the coronary sinus are present, many circuits surrounding the coronary sinus are possible. We therefore ablate the areas between the coronary sinus and the atrial septal defect, the coronary sinus and the inferior vena cava, as well as the inferior vena cava and the tricuspid annulus. The other lesions involving the crista terminalis and the base of the atrial appendage remain the same.
Our adaptation of the Cox-maze III procedure for atrial fibrillation was very effective for control of atrial fibrillation. The Cox-maze III procedure was also applied to patients with left atrial reentry tachycardia. However, several patients had postoperative organized atrial tachycardia, possibly because all of the arrhythmia circuits were not adequately addressed by the Cox-maze III. Intraoperative mapping in one of these patients revealed continued left atrial tachycardia with entry occurring across the dome of the left atrium. This led us to modify the Cox-maze III procedure to the left atrial Cox-maze procedure, as described in Figure 1, which is used for both atrial fibrillation and for left atrial reentry tachycardia with the expectation that postoperative atrial tachycardia will be controlled.
Special lesions are placed for heterotaxy, bilateral venae cavae, and individualized reconstructions, which may extend into the left atrium and cause additional areas of slow conduction and contribute to unusual tachycardia circuits, which may involve the dome of the left atrium. Additional ablative lesions will be necessary to connect the left-sided vena cava to the confluence of pulmonary veins in the case of heterotaxy syndrome. In the case of left atrial suture lines, resection or ablation of the area of slow conduction will be necessary. Previous atrial patches used to partition the atrium or isolate the diminutive right-sided atrioventricular valve must be removed in order to expose anatomic landmarks and perform the modified right-sided maze procedure [8].
Fontan conversion is but one treatment modality for the failed Fontan circulation [24]. The incidence of cardiac transplantation in this series was 5.4% and we were able to identify specific risk factors for this outcome: presence of a right or ambiguous systemic ventricle, preoperative protein-losing enteropathy, moderate-to-severe atrioventricular valve insufficiency, and prolonged cardiopulmonary bypass time. Of note, advancing age and severe ventricular dysfunction were no longer predictors of postoperative mortality or transplantation. It is our experience that severe ventricular dysfunction attributed to chronic persistent arrhythmia, the negative inotropic effects of antiarrhythmic medications, or anatomic pathway obstruction may be improved with surgical intervention. Indications for cardiac transplantation in lieu of Fontan conversion are severe ventricular dysfunction not attributable to identified hemodynamic or arrhythmia causes and protein-losing enteropathy. Risk factors such as ambiguous or right ventricle and severe atrioventricular valve insufficiency are potentially neutralized with better selection criteria and improved valve repair techniques as we have shown in the most recent cohort of patients. Just as importantly, Fontan conversion does not preclude future cardiac transplantation. Recognition of the potential for future cardiac transplantation led us to modify our reconstructive procedures to avoid the use of allograft material, which has been shown to increase panel-reactive antibody levels and interfere with optimal immunosuppression [13].
This analysis has several limitations in that it is a retrospective review of evolving surgical techniques on a changing, more complex patient population. In addition, any analysis of risk factors comparing the three groups faces a time bias: patients in group III have a mean follow-up period of 18 months, versus 11 years for group I patients and 6.25 years for group II.
This series represents a window onto long-term outcomes of patients with atriopulmonary Fontan connections. Despite the introduction of lateral tunnel and extracardiac connection techniques, it is likely that the problems we encountered will arise with some regularity in the future. Fontan patients should be treated in the same manner as other patients with repaired complex heart disease with a high expectation of success, assuming selection criteria are met and proper operative techniques are employed. While cardiac transplantation will continue to be an important mode of therapy for the failed Fontan, Fontan conversion with arrhythmia surgery continues to be an efficacious procedure for patients with increasingly complex anatomic lesions, and atrial fibrillation or left atrial tachycardia.
|
Appendix |
|---|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionFootnotesAcknowledgmentsReferences |
|---|
I and others in the congenital heart surgery community would like to acknowledge Dr Mavroudis’ team for the education of colleagues in the United States and abroad about the failing Fontan circulation and their evolving surgical techniques. They have been gracious hosts to many of us in Chicago.
The results in this manuscript relate to simple endpoints: early and late mortality, recurrence of arrhythmia, and the need for cardiac transplantation. The most important contribution from their experience is the thorough understanding of different atrial tachyarrhythmias that occur in the failing Fontan circulation and their methods of ablation, which have evolved over time. They have learned by experience over a 12-year period and have altered their ablation sets accordingly, the result of which is a profound reduction in arrhythmia recurrence. They have gravitated toward a complete left-sided and right-sided maze procedure performed with cryoablation with lesion sets that are almost identical to the Cox-maze III, which was described by Dr Cox almost two decades ago.
The importance and the need to address both atria at the time of operation cannot be overemphasized. Although we have not routinely performed intraoperative mapping, we currently perform biatrial Cox-maze III cryoablation procedures in these patients. This issue leads to the first of my four questions. Did you perform a multivariate analysis to determine predictors of recurrent atrial tachyarrhythmias?
While early mortality is exceptionally low in your hands for this complex group of patients, the late events, death, and the need for transplantation when it did occur, did so relatively early, within a year after conversion. To their credit, the authors have included many high-risk patients, those with ventricular dysfunction, significant atrioventricular valve regurgitation, and protein-losing enteropathy. Indeed, one wonders whether patients like this benefit from conversion or whether they should go directly to transplantation. In our experience of 70 Fontan conversions, we have also had low early mortality and have found that late outcome is optimized when there is low ventricular end-diastolic pressure and when the expected systolic and diastolic ventricular function are close to normal after conversion.
None of your patients received a fenestration at the time of conversion. It has been our experience that some risk factors for the Fontan circulation are almost always present at the time of conversion. Consequently, we use the fenestration frequently to help facilitate the early postoperative course with the intent to close it in the cardiac cath lab late postoperatively if the clinical situation permits. My second question—could you comment on the role of the fenestration in the setting of conversion?
In your manuscript you have addressed only the endpoints of death, arrhythmia recurrence, and the need for transplant. Other endpoints critical to this patient population should be discussed. My third question, could you comment on the impact of conversion in patients with protein-losing enteropathy, especially in light of the fact that none of your patients received fenestration? Results of transplantation in patients with protein-losing enteropathy is high risk and the PLE does not always resolve, making transplantation not always a viable alternative. I wish to congratulate Dr Mavroudis and will leave you with one final question. Is there a role for prophylactic conversion prior to the onset of atrial arrhythmias? Thank you.
DR MAVROUDIS: Thank you, Dr Dearani, for your kind words. We are aware of your excellent results from you and your colleagues at the Mayo Clinic. You asked about the predictors of recurrent atrial tachycardia, and whether we performed a multivariate analysis to identify any causes. We did not do a multivariate analysis using that as an endpoint. Our recurrence rate is relatively low; however, we did note a significant recurrence rate in our group I patients who had simple isthmus cryoablation. We can therefore say that isthmus cryoablation is a significant predictor of recurrent atrial tachycardia. Recurrent tachycardia in our group II and group III patients is likely multifactorial and would require a large number of patients to identify any risk factors.
You asked about the role of fenestration. We didn’t use fenestration in any of these patients. Our rationale for not using fenestration is that these patients generally come to us with oxygen saturations in the 90% to 95% range. We worry about postoperative desaturation, which may impact unfavorably on ventricular function due to myocardial oxygen demand issues. Whether a fenestration in this setting of borderline ventricular function can optimize the postoperative course remains untested. So we really can’t answer your question because we have never used fenestration in this setting, and our results so far have supported this policy. We look to you and your group to assess this important strategy, which may cause us to alter our approach.
We have encountered protein-losing enteropathy in three of our operated patients, and by and large the results have not been good. None of the patients were relieved of their protein-losing enteropathy after the Fontan conversion. One died after cardiac transplantation due to enteric complications, another one went on to transplantation and did well, and the third is still living but has not been cured of protein-losing enteropathy. So we are prone to suggest that this operation is not an answer to protein-losing enteropathy and that these patients ought to be referred for cardiac transplantation. The difficulty with this approach is that there might be some patients with significant venous pathway obstruction who might benefit from Fontan conversion, especially if the conversion results in significant reduction in central venous pressure. At the present time, however, we don’t have enough experience with this type of patient to make any definitive recommendations. Thank you for your thoughtful questions and conscientious review of our manuscript.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionFootnotesAcknowledgmentsReferences |
|---|
|
Footnotes |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionFootnotesAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionFootnotesAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. L. Backer The Fontan Procedure Our Odyssey Continues. J. Am. Coll. Cardiol., July 8, 2008; 52(2): 114 - 116. [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 |
