Outcomes of Fontan conversion for failing Fontan circulation: mid-term results

Abstract

OBJECTIVES

We investigated the results of a revision of a previous Fontan connection to total cavopulmonary connection (TCPC) in patients with failing Fontan circulation.

METHODS

From July 1998 to April 2013, 21 patients who had failing Fontan circulation underwent revision of the previous Fontan operation to TCPC. The median age at TCPC conversion was 17.9 years (range, 4.6–38.1 years) and the median interval between initial Fontan operation and TCPC was 13.8 years (range, 2.1–25.4 years). There were 37 indications for Fontan revision in 21 patients. The indications were huge right atrium (n = 15), atrial arrhythmia (n = 8), intra-atrial thrombi (n = 6), protein-losing enteropathy (PLE) (n = 3) and more than mild atrioventricular valve regurgitation (n = 5). The previous Fontan operation was revised to extracardiac conduit replacement (n = 20) and intra-atrial lateral tunnel (n = 1). Concomitant surgery for atrial arrhythmia was performed in 8 patients. Fenestration was performed in 7 patients. The median follow-up duration was 7.1 years (range, 0.3–13.4 years).

RESULTS

There were no operative deaths and two late deaths occurred 7.9 and 8.1 years after operation. Actuarial 5- and 10-year survival rates were 92.3 and 83.1%, respectively. Postoperative complications included bleeding (n = 3), deep sternal infection (n = 1) and prolonged pleural effusion for more than 2 weeks (n = 5). During follow-up, atrial arrhythmia recurred in 6 patients, PLE recurred in 2 patients and pleural effusion recurred in 2 patients. All patients were classified as New York Heart Association Class I (n = 15) or Class II (n = 4).

CONCLUSIONS

Fontan conversion to TCPC in patients with failing Fontan circulation can be performed with low risk of morbidity and mortality. The procedure confers better quality of life and is functional for patients with failed Fontan circulation.

INTRODUCTION

Fontan first presented this new surgical treatment in 1971 [1]. Since that time, functional single ventricle has been routinely treated with Fontan operation. However, atriopulmonary Fontan operation was associated with right atrial enlargement, increased systemic venous pressure, atrial arrhythmias, formation of atrial thrombi and pulmonary embolism [2, 3]. To solve these complications, Fontan conversion and thrombectomy with concomitant arrhythmia surgery were performed for patients who had dilated right atrium, large thrombus in the right atrium and atrial tachyarrhythmia after Fontan completion [4–7]. Conversion to total cavopulmonary connection (TCPC) is expected to result in better haemodynamic and functional status. Thus, we investigated the results of revision of the previous Fontan connection to TCPC for patients with failing Fontan circulation.

MATERIALS AND METHODS

Patients

From July 1998 to April 2013, 21 patients underwent conversion of previous Fontan connection to intra-atrial or extracardiac TCPC at Yonsei Cardiovascular Center. We retrospectively reviewed the patients' medical records. There were 14 males (66.7%) with the median age of 17.9 years (range, 4.6–38.1 years). The median interval from the first operation was 13.8 years (range, 2.1–25.4 years). The initial type of surgery was atriopulmonary Fontan operation in 16 patients, right atrial appendage to the infundibulum of right ventricle connection in 2 patients and lateral tunnel Fontan operation in 3 patients. The types of congenital heart disease leading to Fontan palliation were tricuspid atresia in 10 patients, double outlet right ventricle in 5 patients, corrected transposition of great arteries in 2 patients and other anomalies in 4 patients. There were 37 causes for Fontan revision in 21 patients. The causes of reoperation were right atrial enlargement in 15 patients, atrial arrhythmia in 8 patients, right atrial thrombus in 6 patients, protein-losing enteropathy (PLE) in 3 patients and more than mild atrioventricular valve regurgitation in 5 patients. Patients' characteristics are summarized in Table 1.

Table 1:

Patient characteristics

Age at revision	Median age of 17.9 years (range, 4.6–38.1 years)
Interval from original Fontan operation	Median 13.8 years (range, 2.1–25.4 years)
Gender, n (%)	Male 14 (66.7), female 7 (33.3)
Original diagnoses, n (%)	TA: 10 (47.6), DORV: 5 (23.8), ccTGA: 2 (9.5) PA with IVS: 1 (4.8), Heterotaxy: 2 (9.5), TGA, VSD, PS: 1 (4.8)
Heart rhythm
NSR, n (%)	10 (47.6)
A fib., n (%)	5 (23.8)
A flutter, n (%)	3 (14.3)
Atrial tachycardia, n (%)	2 (9.5)
CAVB with PPM, n (%)	1 (4.8)
Causes of revision
Dilated RA, n (%)	15 (71.4)
Arrhythmia, n (%)	8 (38.1)
RA thrombi, n (%)	6 (28.6)
PLE, n (%)	3 (14.3)
AVVR, n (%)	5 (23.8)
Fenestration, n (%)	7 (33.3)
Arrhythmia surgery, n (%)	8 (38.1)

TA: tricuspid atresia; DORV: double outlet right ventricle; ccTGA: congenitally corrected transposition of great arteries; PA with IVS: pulmonary atresia with intact ventricular septum; TGA: transposition of great arteries; VSD: ventricular septal defect; PS: pulmonary stenosis; NSR: normal sinus rhythm; A fib.: atrial fibrillation; A flutter: atrial flutter; CAVB: complete atrioventricular block; PPM: permanent pacemaker; RA: right atrium; PLE: protein-losing enteropathy; AVVR: atrioventricular valve regurgitation.

Operation

The operation was performed on cardiopulmonary bypass under moderate hypothermia (30°C). The previous atriopulmonary, intra-atrial lateral tunnel and atrioventricular connection were removed and revised to extracardiac conduit (n = 20) and intra-atrial lateral tunnel (n = 1) Fontan operation was performed. The extracardiac conduit Fontan was performed by placing polytetrafluoroethylene tube grafts (median size 20 mm, range 16–24 mm) in 18 patients and autologous pericardial flap in 2 patients. Intra-atrial lateral tunnel was constructed with atrial flap. Fenestration between the conduit and the atrium (median size 5 mm) was made in seven cases, including 3 patients who had PLE.

The indication for making fenestration was PLE or increased pulmonary artery pressure (PAP) >20 mmHg. Arrhythmia surgery was performed in 8 patients. Of these 8 patients, right-sided maze procedures (one radiofrequency ablation and two cryoablations) were performed in 3 patients, isthmus ablations (all cryoablations) were performed in 3 patients, lateral tunnel ablation with radiofrequency was performed in 1 patient and biatrial approach, similar to the Cox-Maze III procedure with cryoablation, was performed in 1 patient. Epicardial DDD pacemaker (n = 2) or lead (n = 1) implantation was performed simultaneously in 3 recent patients to treat bradyarrhythmia. Concomitant atrioventricular valve repair was performed in 3 patients. The median cardiopulmonary bypass time was 161 min (range, 82–343 min) and the aortic cross-clamp time was 76 min (range, 36–189 min). This study was approved by the Institutional Review Board at Yonsei University Health System with a waiver to obtain consent based on this retrospective study.

Statistical analysis

Continuous variables are presented as the mean ± standard deviation or median and range depending on the normality of their distribution. Continuous variables were compared using paired t-test. Survival analyses were performed using the Kaplan–Meier actuarial method. This analysis was performed with SPSS software version 14.0 (SPSS, Inc., Chicago, IL, USA). A value of P < 0.05 was considered statistically significant.

RESULTS

Operative and early postoperative outcomes

There was no operative death. Prolonged pleural effusion for more than 2 weeks developed in 7 patients. Postoperative bleeding that required re-exploration occurred in 3 patients. Deep mediastinal infection developed in 1 patient and was successfully treated by debridement and mediastinal irrigation without long-term sequelae. The median postoperative hospital stay was 25 days (range, 6–57 days). Eight patients remained in the hospital for more than 30 days due to prolonged plural effusion in 7 patients and postoperative deep sternal infection in 1 patient.

Late outcomes

Follow-up was complete in all patients. The median follow-up period was 8 years (range, 6 months–13.8 years). Two late deaths occurred 7.9 and 8.1 years after the operation due to PLE. The actuarial survival rate was 92.3 and 83.1% at 5 and 10 years, respectively. During follow-up, atrial arrhythmia recurred in 6 patients, PLE recurred in 2 patients and pleural effusion developed in 2 patients. There were 3 patients with PLE. The PLE symptoms improved immediately after the operation, but recurred within 2 months after the operation in 2 patients, who died later as mentioned above. Another patient's symptoms improved, but PLE was not cured completely (stool alpha-1 trypsin level remained above the normal range). Follow-up haemodynamic study was performed in 14 patients more than 1 year after the operation. PAPs decreased, while ejection fraction (EF) increased, which was not statistically significant (P > 0.05). However, ventricular end diastolic pressures (VEDPs) significantly increased after the operation (P < 0.05) (Table 2). The patients' quality of life was assessed according to the New York Heart Association (NYHA) classification; 78.9% of patients (15/19) were in NYHA class I and 21.1% of patients (4/19) were in NYHA class II after Fontan conversion.

Table 2:

Haemodynamic data of patients after conversion to TCPC (n = 14)

	Preoperative	Postoperative	P-value
PAP (mmHg)	17.8 ± 3.8	15.4 ± 4.2	0.085
VEDP (mmHg)	9.6 ± 4.5	12.8 ± 3.7	<0.001
Ejection fraction (%)	55.9 ± 11.3	61.8 ± 7.3	0.145

PAP: pulmonary artery pressure; VEDP: ventricular end diastolic pressure; TCPC: total cavopulmonary connection.

Arrhythmia

Five patients were in atrial fibrillation before the conversion to TCPC. Of these 5 patients, 4 patients underwent arrhythmia surgery. Patient 2 underwent radiofrequency ablation around the previous lateral tunnel suture site due to paroxysmal atrial fibrillation. After the operation, his rhythm changed into junctional rhythm or atrioventricular block. Therefore, the patient underwent epicardial DDD pacemaker implantation. Patient 3 underwent isthmus cryoablation and obtained normal sinus rhythm; however, the patient had recurrence of paroxysmal atrial tachycardia 5 months after surgery and required antiarrhythmic medication (amiodarone). Patient 4 underwent a biatrial Maze procedure with cryoablation similar to the Cox-Maze III procedure and simultaneous implantation of epicardial DDD pacemaker; however, atrial fibrillation recurred 2 months after surgery and the pacemaker mode was changed to VVI. Patient 5 underwent a right-sided Maze procedure with cryoablation and obtained junctional rhythm; however, the patient had recurrence of paroxysmal atrial fibrillation 4 months after surgery. Patient 1, who did not undergo arrhythmia surgery recovered normal sinus rhythm after surgery, which has continued to persist for 13.7 years. Our management for atrial fibrillation has gradually progressed from no surgery or isthmus ablation to a biatrial Maze procedure. The unsatisfactory results of atrial fibrillation management were probably due to our insufficient degree of understanding of arrhythmia surgery.

Three patients exhibited atrial flutter. Patient 6 underwent a right-sided Maze procedure with radiofrequency and obtained normal sinus rhythm, which has continued to persist for 1.4 years. Patient 7 underwent isthmus cryoablation and obtained normal sinus rhythm; however, the patient experienced recurrence of paroxysmal atrial flutter 11 months after surgery and required antiarrhythmic medication (verapamil). Patient 8 underwent a right-sided Maze procedure with cryoablation and simultaneous implantation of epicardial DDD pacemaker due to preoperative recurrent supra ventricular arrhythmia and has maintained sinus rhythm for 1 year. Patient 9, who had non-sustained atrial tachycardia, underwent isthmus cryoablation and obtained sinus rhythm, which has continued to persist for 9.4 years. The results of arrhythmia surgery are summarized in Table 3.

Table 3:

Results of arrhythmia surgery (n = 8)

Patient number	Arrhythmia	Year	Surgery	Energy source	PPM insertion	Rhythm (postop.)	Last rhythm	Antiarrhythmic medication	Atrial wall reduction
1	A fib.	2000	No	(−)	(−)	NSR	NSR	(−)	(+)
2	A fib.	2006	Lateral tunnel ablation	RF	(+) Postop. VVI	Junctional or AV block	AV block	(−)	(−)
3	A fib	2005	Isthmus ablation	Cryo	(−)	NSR	Paroxysmal tachycardia	Amiodarone	(−)
4	A fib.	2012	Biatrial maze	Cryo	(+) Intraop. DDD	NSR with DDD pacing	A fib. recurrence, VVI pacing	(−)	(+)
5	A fib.	2013	Right side maze	Cryo	(−)	Junctional rhythm	Paroxysmal A fib.	(−)	(+)
6	A flutter	2004	Isthmus ablation	Cryo	(−)	NSR	Paroxysmal A flutter	Verapamil	(+)
7	A flutter	2012	Right side maze	RF	(−) Intraop. lead insertion	NSR	NSR	(−)	(+)
8	A flutter	2012	Right side maze	Cryo	(+) Intraop. DDD	NSR with DDD pacing	NSR with DDD pacing	(−)	(+)
9	Atrial tachycardia	2004	Isthmus ablation	Cryo	(−)	NSR	NSR	(−)	(−)

A fib.: atrial fibrillation; NSR: normal sinus rhythm; PPM: permanent pacemaker; RF: radiofrequency; AV: atrioventricular.

DISCUSSION

Our results demonstrate that conversion of a failing Fontan connection to a TCPC can be accomplished safely and successfully, with resultant improvement in functional and haemodynamic outcomes. The common problems of the previous Fontan connection in our patient population were both poor functional ability and atrial dysrhythmia. There were no operative deaths during the procedures and only two late deaths, which were caused by previous PLE. Conversion to TCPC did not seem to affect PLE. Of 3 patients with PLE, 2 patients died due to PLE recurrence, and the third patient still has functional impairment, and is not cured. Excluding the two late deaths, all other patients were evaluated as NYHA class I or II, and their haemodynamic data showed improving results of PAP and EF although these data did not reach statistical significance. The haemodynamic data displayed significantly increased VEDPs. We assume that this results from the possibility of increased preload after conversion to TCPC. Although VEDPs increased, other parameters showed satisfactory results and patients had good function; therefore, it might not be a concern. To evaluate functional improvement, we can only show functional class and haemodynamic data. However, we recently included routine cardiopulmonary exercise tests, and, therefore, will obtain adequate data in the future.

The development of atrial tachycardia in atriopulmonary Fontan patients is a considerable problem. It develops from many factors including atrial incisions-suture load, dilatation, fibrosis, hypertrophy, atrioventricular valve regurgitation and fundamental diagnosis. There has been a shift in presenting arrhythmias from predominantly right atrial re-entry tachycardia to predominantly atrial fibrillation. atrial fibrillation causes greater functional disability, results in thrombus formation, and its pharmacological treatment is associated with significant adverse side effects. Surgical treatment of atrial fibrillation in association with Fontan conversion requires long cross-clamp time. The arrhythmia operation has undergone changes to a modified right-sided Maze procedure for right atrial re-entry tachycardia, and to a left atrial Cox-Maze procedure for atrial fibrillation and left atrial re-entry tachycardia. Conversion to the extracardiac type of Fontan with concomitant arrhythmia surgery has achieved effective arrhythmia control in the long term.

Mavroudis et al. [8] reported that this strategy had an incidence of late death need for transplantation of 5.4% and late recurrence of arrhythmia of 15.7%. Our experiences with eight arrhythmic surgeries resulted in seven right-sided ablations. Of these 7 cases, 3 patients had recurrent tachycardia and 2 patients had atrial fibrillation. Another patient, who underwent biatrial Maze procedure and concomitant permanent pacemaker (PPM) (DDD) insertion had recurrent atrial fibrillation We assume that unsatisfactory outcomes of arrhythmia surgery resulted from our insufficient degree of understanding of arrhythmia surgery and our passive approach in the past. During the last 2 years, we attempted to perform aggressive arrhythmia surgery including intraoperative PPM insertion. Prophylactic conversion regardless of symptoms, arrhythmias or abnormal haemodynamics, is under consideration in the present era; however, we should determine the optimal timing of conversion. Recent reports [9, 10] indicate that two simple electrocardiographic markers, namely the P wave duration and dispersion, are informative when considering the conversion procedure before the onset of sustained tachyarrhythmia.

Limitations

The limitations of the present study include its retrospective nature and the small number of patients. Evaluation of functional status was subjective using functional class.

CONCLUSIONS

Fontan conversion to TCPC in patients with failing Fontan circulation can be performed with low risk of morbidity and mortality, and improve haemodynamic and functional conditions. The conversion does not appear to have a positive effect on PLE. More aggressive arrhythmic surgery is required for atrial fibrillation.

FUNDING

This study was supported by a faculty research grant of Yonsei University College of Medicine for 6-2014-0053.

Conflict of interest: none declared.