Download The Lateral Tunnel Fontan Procedure for Hypoplastic Left Heart

Pediatr Cardiol (2007) 28:426–432
DOI 10.1007/s00246-007-9002-5
The Lateral Tunnel Fontan Procedure for Hypoplastic Left Heart
Syndrome: Results of 100 Consecutive Patients
J. C. Hirsch Æ R. G. Ohye Æ E. J. Devaney Æ
C. S. Goldberg Æ E. L. Bove
Published online: 3 August 2007
Ó Springer Science+Business Media, LLC 2007
Abstract The Fontan procedure for hypoplastic left heart
syndrome (HLHS) is well established. Multiple surgical
techniques including extracardiac conduits and autologous
tissue connections have been developed. We reviewed the
results of 100 consecutive patients undergoing the lateral
tunnel modification of the Fontan procedure at the University of Michigan. A cross-sectional retrospective study
was performed for 100 consecutive patients identified in
the University of Michigan Congenital Heart Surgery
database with the diagnosis of HLHS. All patients had
undergone a lateral tunnel Fontan procedure between June
2000 and August 2004. The medical record was reviewed
to assess patient, procedural, and morphologic determinants of outcome. Hospital survival was 97% and intermediate-term survival was 96% with a median follow-up
time of 34 months. Preoperative mean pulmonary artery
pressure, right ventricular end diastolic pressure, aortic
cross-clamp time, and tricuspid valve regurgitation were
not associated with late right ventricular function or survival. Three patients required takedown of the lateral tunnel Fontan in the early postoperative period. A positive
association was found between protein-losing enteropathy
and prolonged (>2 weeks) postoperative pleural drainage
(p = 0.035). No patient required cardiac transplantation or
J. C. Hirsch (&) R. G. Ohye E. J. Devaney E. L. Bove
Section of Cardiac Surgery, Division of Pediatric Cardiovascular
Surgery, University of Michigan School of Medicine, 1500 E.
Medical Center Drive, Mott F7380, Ann Arbor
MI 48109-0223, USA
e-mail: [email protected]
C. S. Goldberg
Section of Pediatric Cardiology, University of Michigan School
of Medicine, 1500 E. Medical Center Drive, Mott F7380,
Ann Arbor, MI 48109-0223, USA
123
late intervention on the Fontan pathway. At the time of
follow-up, 100% of patients were New York Heart Association class I or II and 90% were in normal sinus rhythm.
The lateral tunnel Fontan procedure for HLHS can be
performed with acceptable early and intermediate-term
risk. There was a low prevalence of late rhythm disturbances and other complications. Protein-losing enteropathy
and prolonged pleural drainage were associated.
Staged palliation for the surgical management of single
ventricle congenital heart defects is widely accepted, but
with variable techniques employed at each individual stage
[2, 5]. However, the overall benefits and risks of these
modifications have not been rigorously tested. The choice
of technique is often based on surgeon/center preference.
At the University of Michigan, more than 600 Fontan
procedures have been performed between 1992 and 2005.
For patients with suitable anatomy, we continue to prefer
the lateral tunnel technique.
Our choice of the lateral tunnel Fontan is partially based
on our preference for the hemi-Fontan over the bidirectional Glenn as the second-stage palliation. The hemiFontan creates a large cavopulmonary connection that
facilitates augmentation of the branch pulmonary arteries
and simplifies the completion of the inferior vena caval
pathway at the time of the Fontan (Fig. 1). Although the
hemi-Fontan is a technically more difficult procedure, the
benefits are worth the increased effort at the hemi-Fontan
stage when a longer operation is better tolerated in the
postoperative phase than following the more hemodynamically demanding recovery period after the Fontan. In a
prior publication from our institution, this method of staged
Pediatr Cardiol (2007) 28:426–432
427
Fig. 1 The anastomosis
between the cavoatrial junction
and the central pulmonary
arteries is shown on the left. The
final reconstruction allows only
superior vena cava return to
enter the lungs while retaining
the connection with the right
atrium to simplify the
subsequent lateral tunnel Fontan
operation. Reprinted with
permission from Bove and
Mosca [5]
Fig. 2 Three-dimensional model of typical hemi-Fontan procedure
based on anatomic data from MR scans and angiocardiograms.
Reprinted with permission from Bove et al. [3]
repair resulted in a survival rate of 94% for patients
undergoing a hemi-Fontan followed by a lateral tunnel
Fontan [8].
In addition, computational fluid dynamic modeling of
cavopulmonary connections for hypoplastic left heart
syndrome (HLHS) has demonstrated the optimal caval
offset achieved by the lateral tunnel Fontan with the
lowest power loss compared to traditional total cavopulmonary connections. The energy-efficient flow within the
cavopulmonary connection of the hemi-Fontan (Fig. 2)
followed by the lateral tunnel Fontan (Fig. 3) procedure is
due to the streamlined geometry of the anastomotic
design. This study demonstrated the important hemodynamic differences between the hemi-Fontan and bidirectional Glenn, which only become apparent following
completion of the Fontan procedure. The ‘‘optimal’’ lateral offset of the bidirectional Glenn combined with the
lateral tunnel or extracardiac Fontan directs the superior
vena cava (SVC) flow preferentially into the right pulmonary artery, whereas the larger volume inferior vena
Fig. 3 Model of completion lateral tunnel Fontan as performed after
hemi-Fontan procedure by removal of the intraatrial patch. Reprinted
with permission from Bove et al. [3]
cava (IVC) flow is directed to the smaller left lung. In
contrast, the blood flow into the branch pulmonary
arteries is not only more efficient with less energy losses
but also more uniformly distributed to each lung with the
hemi-Fontan/lateral tunnel Fontan approach [3]. This
increased performance appears to be the result of the
divergent shape in both the SVC and the IVC anastomoses and the anteroposterior offset of the two caval
anastomoses in relation to the pulmonary arteries [2].
Since 1992, 94% of single ventricle patients at the
University of Michigan have received a lateral tunnel
Fontan as the third-stage palliation. To improve our
understanding of the intermediate-term outcomes for the
lateral tunnel Fontan, we performed this retrospective study
123
428
Pediatr Cardiol (2007) 28:426–432
Fig. 4 A polytetrafluoroethylene patch is fashioned to
channel inferior vena caval
return to the previously
constructed anastomosis with
the pulmonary arteries. The
tunnel is completed by suturing
the patch within the right
atriotomy closure
to evaluate outcomes in a single diagnostic group (HLHS)
undergoing the lateral tunnel Fontan procedure at the
University of Michigan.
Materials and Methods
Data Collection
We performed a cross-sectional retrospective study on 100
consecutive patients identified through the University of
Michigan Congenital Heart Surgery database with HLHS
and a lateral tunnel Fontan procedure performed between
June 2000 and August 2004. Only patients with classic
HLHS were included, as defined by right ventricle-dependent systemic circulation and systemic outflow tract
obstruction. We obtained preoperative, operative, and
hospital course data on each patient through a review of
paper and electronic medical records.
Preoperative data collected were patient demographics,
cardiac catheterization information on both mean pulmonary
artery (PA) pressures and right ventricular end diastolic
pressure (RVEDP), and echocardiographic results of right
ventricular function and tricuspid valve regurgitation. We
converted right ventricular function and tricuspid valve
regurgitation to ordinal variables. For right ventricular
function, 0 = normal, 1 = mildly depressed, 2 = moderately
depressed, and 3 = severely depressed. For tricuspid valve
regurgitation, 0 = none or trace, 1 = mild, 2 = moderate,
and 3 = severe. Operative data included age and weight at
repair, cardiopulmonary bypass (CPB) and aortic crossclamp (XC) times, and concurrent procedures. Hospital
course data encompassed postoperative arrhythmias, chest
tube drainage >2 weeks, additional procedures including
takedown of the Fontan, discharge right ventricular function
123
and tricuspid valve regurgitation based on echocardiographic
data, discharge rhythm, and length of stay. Follow-up data
obtained from our pediatric cardiology clinic notes or from
communication with the referring cardiologist were length of
follow-up, the presence of a fenestration, New York Heart
Association (NYHA) classification, most recent echocardiographic data on right ventricular function and tricuspid
valve regurgitation, current rhythm, interval interventions or
operations, current medications, and any subsequent complications. Institutional review board approval was obtained
prior to the initiation of the study.
Surgical Technique
All 100 patients had undergone a prior hemi-Fontan procedure according to techniques described previously [2].
The lateral tunnel procedure was performed during continuous cardiopulmonary bypass at 28–32°C. The IVC
pathway was constructed with a patch of 0.6-mm Gore-Tex
(W. L. Gore & Associates, Flagstaff, AZ, USA). The patch
was sutured around the internal orifice of the IVC and
continued above the right pulmonary veins until the previous patch placed at the hemi-Fontan procedure was
encountered. The previous patch was excised in its entirety.
A 2.8- or 3-mm punch was utilized to construct a fenestration. The Fontan was completed by closing the atriotomy
incorporating the anterior half of the Gore-Tex patch
(Fig. 4) [6]. Mediastinal and bilateral pleural drains were
placed in the operating room in all patients.
Statistical Analysis
Data were recorded on a Microsoft Excel spreadsheet
(Microsoft, Redmond, WA, USA) and transferred to the
Pediatr Cardiol (2007) 28:426–432
429
Table 1 Patient demographic, preoperative, and intraoperative data
Table 2 Concomitant surgical procedures
Pre- and intraoperative parametric data, mean ± SD
Procedure
n
Age (months)
24 ± 6
Tricuspid valve repair
15
Weight (kg)
11.1 ± 1.5
Pulmonary artery augmentation
2
Mean PA pressure (mmHg)
13 ± 3
Epicardial pacemaker placementa
2
RVEDP (mmHg)
8±3
a
CPB time (min)
65 ± 19
XC time (min)
28 ± 9
One for preoperative complete heart block and one for preoperative
sick sinus syndrome
Preoperative nonparametric data (n = 100), %
Table 3 Postoperative morbidity
RV systolic function
Normal
79
Mild dysfunction
18
Moderate dysfunction
Severe dysfunction
3
0
Tricuspid valve regurgitation
None/trace
41
Mild
37
Moderate
16
Severe
6
Complication No. (n = 100)
Postoperative arrhythmia
19
Prolonged pleural drainage (>2 weeks)
16
Seizure
5
Diaphragm plication
4
Thoracic duct ligation
4
Fontan takedown
3
Stroke
2
Epicardial pacemaker placement
2
CPB, cardiopulmonary bypass; PA, pulmonary artery; RV, right
ventricle; RVEDP, right ventricular end diastolic pressure; SD,
standard deviation; XC, cross-clamp
ECMO
1
SAS statistical package (SAS Institute, Cary, NC, USA) for
analysis. Normal data are expressed as mean ± standard
deviation and non-normal data as medians and ranges.
Differences in dichotomous variables were compared by
using chi-square or Fisher’s exact test. Wilcoxon rank sum
tests were performed for comparisons involving ordinal
variables.
concomitant procedure (n = 15). When patients were subcategorized into those who did and those who did not
receive a tricuspid valve repair, we observed no significant
difference in tricuspid valve regurgitation at the time of
discharge or follow-up, nor was there a significant difference in long-term RV function or survival.
ECMO, extracorporeal membrane oxygenation
Postoperative Data
Results
Preoperative and Intraoperative Data
A summary of the preoperative and intraoperative data is
presented in Table 1. We defined a high RVEDP as >9
mmHg and high PA pressure as 15 mmHg representing
the upper quartiles. Longer XC time was defined as any
time greater than the median of 25 minutes. High RVEDP
and high PA pressure did not correlate with late RV
function. Longer XC times demonstrated a trend toward
decreased late RV function (p = 0.07), but this was not
statistically significant. There was also a trend toward
lower preoperative PA pressure ( 15 mmHg) having an
association with postoperative survival (p = 0.054). Preoperative RV function did not predict postoperative arrhythmias. Table 2 summarizes the operative procedures
performed in combination with the lateral tunnel Fontan
procedure. Tricuspid valve repair was the most common
The postoperative complications are listed in Table 3.
Major complications include arrhythmias, prolonged
pleural drainage, and seizures. Table 4 contains information on RV function, tricuspid valve regurgitation, and
rhythm both at the time of discharge and at follow-up.
Although 19% of patients had an arrhythmia postoperatively (complete heart block, junctional rhythm, junctional
ectopic tachycardia, atrial flutter, supraventricular tachycardia, and sick sinus syndrome), 95% of patients were
discharged home in normal sinus rhythm. Two patients
required placement of a permanent pacemaker in the
postoperative period. One of these patients developed
complete heart block following a tricuspid valvuloplasty
and the other patient had sick sinus syndrome, which was
poorly tolerated. We dichotomized the length of time for
pleural drainage into patients draining for <2 weeks or >2
weeks. Sixteen percent of patients drained for >2 weeks.
Prolonged chest tube drainage was significantly associated
123
430
Pediatr Cardiol (2007) 28:426–432
Table 4 Discharge and follow-up data
Table 5 Late complications and interventions
Discharge (%)
Complication No. (n = 81)
Follow-up (%)
PLE
RV systolic function
5
Normal
70
84
Late death
4
Mild dysfunction
28
14
Device fenestration closure
3
Moderate dysfunction
2
2
Arrhythmia
2
Severe dysfunction
0
0
Fenestration creation
1
Tricuspid valve regurgitation
Recurrent effusions
1
None/trace
48
52
Stroke
1
Mild
37
35
Seizure
1
Moderate
15
13
Tricuspid valve replacement
1
0
0
Severe
Rhythm
Normal sinus rhythm
95
90
Junctional
2
3
Sick sinus syndrome
0
5
Paced
3
3
RV, right ventricle
with the development of protein-losing enteropathy (PLE)
(p = 0.035). Prolonged chest tube drainage and PLE had no
association with the late presence of a fenestration. Three
patients required takedown to a hemi-Fontan in the postoperative period. The median length of stay in the hospital
was 10 days (range, 3–121).
The overall hospital survival was 97%. One patient
developed refractory supraventricular tachycardia requiring
emergent placement on extracorporeal membrane oxygenation for support. This patient had a significant anoxic
brain injury and the family chose to withdraw support.
Another patient with sepsis and low cardiac output developed persistent seizure activity and progressed to brain
death. The third patient underwent multiple interventions
on the left pulmonary artery including a stent. Following
the Fontan, the patient’s left pulmonary artery thrombosed
and the patient was converted to a hemi-Fontan but still
progressed to multisystem organ failure and support was
withdrawn.
Follow-Up Data
Of the 100 patients who underwent a lateral tunnel Fontan
procedure during the study period, 84% of the hospital
survivors were available for follow-up. The median followup time was 34 months (range, 1–72). All patients were in
NYHA class I or II. Table 4 contains information regarding
RV function, tricuspid valve regurgitation, and rhythm at
the time of follow-up. Table 5 outlines late complications
in patients available for follow-up. We could confirm the
status of the fenestration in 55 patients and found patency
123
PLE, protein-losing enteropathy
in 62%. One patient required creation of a fenestration for
high Fontan pathway pressure and PLE. Three patients
underwent fenestration closure with an Amplatzer device
(AGA Medical, Plymouth, MN, USA). All of the patients
are taking aspirin 40.5 mg per day except for 4 patients
who were on coumadin. The indications for coumadin
included left pulmonary artery stent, mechanical tricuspid
valve, superior vena caval thrombus, and left ventricle
thrombus. We noted 4 late deaths between 2 and 46 months
following the lateral tunnel Fontan procedure. No patients
required transplantation, Fontan conversion, or intervention
on their Fontan pathway.
Discussion
Surgical and medical management of patients with a functional single ventricle has been significantly refined since
1971, when Fontan and Baudet first described total cavopulmonary anastomosis for the palliation of tricuspid atresia
[11]. After de Leval et al. as well as Jonas and Castaneda
introduced the lateral tunnel Fontan procedure, the longterm survival rate for patients with HLHS improved, as did
their freedom from late arrhythmias [7, 13, 19]. This study
demonstrates excellent hospital and intermediate-term survival in the highest risk single ventricle population.
A number of studies have demonstrated that elevated pulmonary artery pressures and resistance, pulmonary artery
distortion, impaired ventricular function, significant tricuspid valve regurgitation, and prolonged cardiopulmonary
bypass and aortic cross-clamp times adversely influence
survival [4, 14, 15]. Recent improvements in operative
management have neutralized many of these risk factors. In
a prior report from our group, survival was adversely
affected by elevated preoperative PA pressures and prolonged CPB times [16]. In this study, which examined a
similar patient population, decreased survival was associated with PA pressures 15 mmHg (p = 0.054), but this
Pediatr Cardiol (2007) 28:426–432
trend did not reach statistical significance. In the current
study, we were unable to demonstrate an association
between CPB and XC time with survival or late RV function. These findings may be attributable to modifications in
our surgical technique demonstrated in an earlier study [4].
Given the small sample size and overall excellent survival
and functional status, our study may not have adequate
power to determine these relationships.
We previously demonstrated the importance of concomitant tricuspid valve repair for improved survival and preserved RV function in patients with significant tricuspid
valve regurgitation [18]. This study again demonstrated that
tricuspid valvuloplasty could produce successful and durable
outcomes. We continue to incorporate tricuspid valvuloplasty as part of the lateral tunnel Fontan procedure when
warranted; 15% of patients underwent a concomitant repair
in this study. Improvements in tricuspid valve regurgitation
and RV function at the time of discharge and late follow-up
were durable in this population. Furthermore, the addition of
the valvuloplasty, despite longer cardiopulmonary bypass
times, did not negatively affect survival.
As the operative success for single ventricle patients
improves and the survivors of the procedure age, concern
regarding late complications has increased. Late risks
include arrhythmias and PLE. Early postoperative arrhythmias, atriopulmonary connections, and significant
atrioventricular valve regurgitation have been previously
shown to be associated with late arrhythmias [1, 9]. In our
cohort of patients, none of these could be shown to be
associated with late arrhythmias. However, 90% of our
patients had a normal sinus rhythm at the time of followup, and only two patients had late dysrrhythmias. Proponents of extracardiac conduit Fontan procedures believe
there may be a decreased incidence of atrial arrhythmias
with this technique due to fewer atrial suture lines and the
avoidance of exposure of the atrium to elevated pressures
[17]. Although this was not demonstrated in the current
study, follow-up remains short.
The pathogenesis of PLE remains poorly understood.
Feldt et al. [10] found multiple factors associated with the
late development of PLE in Fontan patients, including
ventricular anatomy other than dominant left ventricle,
elevated preoperative end diastolic pressure, prolonged
CPB time, and lengthened hospital stay (>15 days). These
authors did not specifically evaluate prolonged postoperative pleural effusions, but they speculated that there might
be an association. In this study, we demonstrated an
association between PLE and pleural drainage >2 weeks.
However, no association was found between PLE and high
RVEDP or prolonged CPB time. This may be due to the
relatively short CPB times in this study. Furthermore, the
moderate length of follow-up in this study may not be
adequate to fully evaluate the risk of developing PLE.
431
In conclusion, we continue to prefer the lateral tunnel
Fontan for the third-stage palliation in functional single
ventricle patients with suitable anatomy. Based on
computational fluid dynamic modeling, the hemi-Fontan
procedure followed by the lateral tunnel Fontan procedure minimizes energy losses and optimizes the distribution of IVC blood flow into both lungs [3]. In a cohort
of patients from our institution, neurodevelopmental
outcomes for preschool and early school-age children
following the Fontan operation were found not to be
statistically different from the general population, even in
patients with HLHS [12]. The use of autologous tissue
for the majority of the reconstruction may allow for
growth potential and reduce the theoretical risk of
thromboembolism. With the exception of a specific
indication for coumadin, the patients in this cohort were
managed on low-dose aspirin alone, without any documented cases of thromboembolism.
There are limitations to this study that could be
addressed in future studies. The length of follow-up may be
insufficient to detect the incidence of late complications of
the Fontan procedure, including PLE, arrhythmias, and late
thrombus formation. The results represent the experience
of a single institution. Multicenter studies comparing extracardiac and lateral tunnel Fontan procedures for specific
diagnoses such as HLHS will help elucidate differences
between these procedures in a larger cohort and eliminate
single center bias.
References
1. Alphonso N, Baghai M, Sundar P, et al. (2005) Intermediateterm outcome following the Fontan operation: a survival,
functional, and risk-factor analysis. Eur J Cardiothorac Surg
28:529–535
2. Bove EL (1998) Current status of staged reconstruction for
hypoplastic left heart syndrome. Pediatr Cardiol 19:308–315
3. Bove EL, de Leval MR, Migliavacca F, et al. (2003) Computational fluid dynamics in the evaluation of hemodynamic performance of cavopulmonary connections after the Norwood
procedure for hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 126:1040–1047
4. Bove EL, Lloyd TR (1996) Staged reconstruction for hypoplastic left heart syndrome, contemporary results. Ann Surg
224:387–394
5. Bove EL, Mosca RS (1996) Surgical repair of the hypoplastic left
heart syndrome. Prog Pedeatr Cardiol 5:23–35
6. Bove EL, Ohye RG, Devaney EJ (2004) Hypoplastic left heart
syndrome: conventional surgical management. Semin Thorac
Cardiovasc Surg Pediatr Card Surg Annu 7:3–10
7. de Leval MR, Kilner P, Gewillig M, Bull C (1988) Total cavopulmonary connection: a logical alternative to atriopulmonary
connection for complex Fontan operations. Experimental studies
and early clinical experience. J Thorac Cardiovasc Surg 96:682–
695
8. Douglas WI, Goldberg CS, Mosca RS, et al. (1999) Hemi-Fontan
procedure for hypoplastic left heart syndrome: outcome and
suitability for Fontan. Ann Thorac Surg 68:1361–1367
123
432
9. Durongpisitkul K, Porter CJ, Cetta F, et al. (1998) Predictors of
early- and late-onset supraventricular tachyarrhythmias after
Fontan operation. Circulation 98:1099–1107
10. Feldt RH, Driscoll DJ, Offord KP, et al. (1996) Protein-losing
enteropathy after the Fontan operation. J Thorac Cardiovasc Surg
112:672–680
11. Fontan F, Baudet E (1971) Surgical repair of tricuspid atresia.
Thorax 26:240–248
12. Goldberg CS, Schwartz EM, Brunberg JA, et al. (2000) Neurodevelopmental outcome of patients after the Fontan operation: a
comparison between children with hypoplastic left heart syndrome
and other functional single ventricle lesions. J Pediatr 137:646–652
13. Jonas RA, Castaneda AR (1988) Modified Fontan procedure:
atrial baffle and systemic venous to pulmonary artery anastomotic
techniques. J Card Surg 3:91–96
14. Kaulitz R, Ziemer G, Luchmer I, et al. (1996) Modified Fontan
operation in functionally univentricular hearts: preoperative risk
factors and intermediate results. J Thorac Cardiovasc Surg
112:658–669
123
Pediatr Cardiol (2007) 28:426–432
15. Mayer JE Jr, Bridges N, Lock K, et al. (1992) Factors associated
with marked reduction in mortality for Fontan operations in
patients with single ventricle. J Thorac Cardiovasc Surg 103:444–
452
16. Mosca RS, Kulik TJ, Goldberg CS, et al. (2000) Early results of
the Fontan procedure in one hundred consecutive patients with
hypoplastic left heart syndrome. J Thorac Cardiovasc Surg
119:1110–1118
17. Nurnberg JN, Ovroutski S, Alexi-Meskishvili V, et al. (2004)
New onset arrhythmias after the extracardiac conduit Fontan
operation compared with the intra-atrial lateral tunnel procedure: early and midterm results. Ann Thorac Surg 78:1979–
1988
18. Ohye RG, Gomez CA, Goldberg CS, et al. (2004) Tricuspid valve
repair in hypoplastic left heart syndrome. J Thorac Cardiovasc
Surg 127:465–472
19. Stamm C, Friehs I, Mayer JE Jr, et al. (2001) Long-term results of
the lateral tunnel Fontan operation. J Thorac Cardiovasc Surg
122:1219–1228