You are on page 1of 6

Modified Fontan Without Use of Cardiopulmonary

Bypass
Vincent K. H. Tam, MD, Bruce E. Miller, MD, and Kathy Murphy, MSN
Section of Cardiothoracic Surgery and Department of Anesthesiology, Emory University School of Medicine; and Sibley Heart
Center, Egleston Children’s Hospital, Atlanta, Georgia

Background. Direct cavopulmonary connection using in either group. Early postoperative hemodynamics ap-
an extracardiac conduit has a number of theoretical pear to be significantly improved in the No CPB group.
advantages in the staged management of children with Transfusion of cryoprecipitate and platelets was signifi-
single ventricular congenital heart defects. With appro- cantly less in the group without the use of cardiopulmo-
priate planning, completion Fontan using an extracardiac nary bypass (p ⴝ 0.026, p < 0.001, respectively). Review of
connection may be accomplished without the use of the most recent 12 patients also demonstrated a substan-
cardiopulmonary bypass. tially shorter extubation time and intensive care unit
Methods. From January 1995 to October 1997, 32 con- stay. The length of hospital stay was significantly shorter
secutive patients underwent completion Fontan using an (p ⴝ 0.036).
extracardiac cavopulmonary connection. Twenty-one of Conclusions. Completion Fontan without the use of
these patients had completion Fontan without the use of cardiopulmonary bypass results in improved immediate
cardiopulmonary bypass (No CPB group). Their postop- postoperative hemodynamics, and decreased use of
erative outcome was retrospectively compared with a blood and blood products. The most recent group ap-
second group of 11 patients who underwent completion pears to demonstrate a more rapid recovery time and
Fontan with an extracardiac conduit with the use of shorter hospital stay (p ⴝ 0.036).
cardiopulmonary bypass. (Ann Thorac Surg 1999;68:1698 –704)
Results. There was no operative or hospital mortality © 1999 by The Society of Thoracic Surgeons

S ince the original description of right heart bypass by


Fontan and Baudet in 1971 [1] as treatment for
children with tricuspid atresia, many modifications have
cardiopulmonary bypass. The early gratifying result
stimulated us to plan for and develop techniques to
accomplish the extracardiac Fontan without the use of
been suggested. De Leval and colleagues [2] proposed cardiopulmonary bypass. This report describes the first
the concept of direct cavopulmonary connection as a way 21 consecutive patients and compares their early out-
to minimize kinetic energy loss in the Fontan circuit. come with 11 other consecutive patients who underwent
Interest on the use of an extracardiac inferior vena cava an extracardiac Fontan with the use of cardiopulmonary
to pulmonary artery conduit has recently been rekindled bypass during the same time period.
by Marcelleti and colleagues [3, 4].
The extracardiac modified Fontan has a number of
Material and Methods
theoretical advantages. Because the conduit is con-
structed outside the heart, the operation may be per- Patient Population
formed with the patient supported on cardiopulmonary From January 1995 to October 1997, 32 consecutive pa-
bypass, without arresting the heart. In selected patients, tients between the ages of 18 months and 5 years,
the extracardiac Fontan may be accomplished without weighing between 8.6 to 19 kg, underwent the modified
the use of cardiopulmonary bypass. In January 1995, an Fontan operation with an extracardiac inferior vena cava
18-month-old 11-kg child with hypoplastic left heart to pulmonary artery connection (Table 1). Twenty-one
syndrome successfully underwent placement of an extra- patients underwent surgery without the use of cardiopul-
cardiac inferior vena cava to pulmonary artery conduit monary bypass (No CPB group), while 11 patients had
without the use of cardiopulmonary bypass. This patient surgery with the use of cardiopulmonary bypass (CPB
had bilateral bidirectional Glenn shunts placed at the age group). All patients except 1 had previously undergone
of 6 months, and the presence of dual sources of blood placement of a superior vena cava to pulmonary artery
flow to the lungs allowed the construction of an inferior anastomosis. One patient in each group had bilateral
vena cava to pulmonary artery conduit without the use of bidirectional Glenn shunts. Cardiac catherization was

Presented at the Thirty-fourth Annual Meeting of The Society of Thoracic


This article has been selected for the open discussion
Surgeons, New Orleans, LA, Jan 26 –28, 1998. forum on the STS Web site:
Address reprint requests to Dr Tam, Emory Clinic, 1365 Clifton Rd, Suite http://www.sts.org/section/atsdiscussion/
A2236, Atlanta, GA 30322; e-mail: vtam01@emory.edu.

© 1999 by The Society of Thoracic Surgeons 0003-4975/99/$20.00


Published by Elsevier Science Inc PII S0003-4975(99)01067-X
Ann Thorac Surg TAM ET AL 1699
1999;68:1698 –704 FONTAN WITHOUT CARDIOPULMONARY BYPASS

Table 1. Patient Characteristics


Heart
Defects
Age (mo) Weight (kg) HLHS TA Others

No CPB 24.2 ⫾ 5.1 10.8 ⫾ 1.1 14 4 3


CPB 26.8 ⫾ 14.6 12.1 ⫾ 2.8 4 1 6
p 0.57 0.17

CPB ⫽ cardiopulmonary bypass; HLHS ⫽ hypoplastic left heart


syndrome; TA ⫽ tricuspid atresia.

performed before surgery in all patients. The calculated


pulmonary vascular resistance for the No CPB group was
2.0 ⫾ 0.5, while the calculated pulmonary vascular resis-
tance for the CPB group was 2.2 ⫾ 0.9. One patient in the
No CPB group has a calculated pulmonary vascular
resistance greater than 3.0 Woods unit, while 2 of the CPB
group had similarly elevated calculated pulmonary vas-
cular resistance. These differences were not statistically
significant. Additional features that may confer increased
risk for the completion Fontan operation are listed for
both groups in Table 2.
It has been our practice to address all potential ana-
tomical issues at the bidirectional Glenn stage, such that Fig 1. Right branch pulmonary artery has been isolated and a longi-
tudinal arteriotomy made. A vascular clamp is used to direct supe-
at completion Fontan, connection between inferior vena
rior vena cava blood flow to the left lung only.
cava and pulmonary artery only needs to be established.
Of these 32 consecutive patients, 26 patients needed
inferior vena cava (IVC) to pulmonary artery connections
only, and were candidates for completion Fontan without are heparinized initially with a 400-U/kg dose of heparin.
the use of cardiopulmonary bypass. Six patients required A vascular clamp is used to isolate the right branch
pulmonary artery reconstruction and were not felt to be pulmonary artery, diverting superior vena cava blood
candidates. Of the 26 patients for whom no cardiopulmo- flow to one lung only (Fig 1). The extracardiac conduit to
nary bypass was planned, in 5 patients, cardiopulmonary pulmonary artery anastomosis is then performed using a
bypass was used because of difficulty with placement of continuous polypropylene suture technique. Once this
the IVC cannula. anastomosis is completed, superior vena cava blood flow
is reestablished to both lungs (Fig 2). Next, an inferior
Surgical Technique vena cava to atrial shunt is constructed using two right-
In all patients, abnormal systemic arterial to pulmonary angle metal-tip venous cannulas (DLP, Grand Rapids,
arterial collateral vessels are divided as much as possible. MI). This inferior vena cava to atrial shunt is first placed
In the cardiopulmonary bypass group, standard bicaval in the inferior vena cava, and then the atrium, to provide
venous and ascending aortic cannulation is used. In both for an alternative pathway for inferior vena cava blood
groups, right and left branch pulmonary arteries are flow to the atrium (Fig 3). A vascular clamp is then placed
dissected completely to the hilar branches. All patients across the inferior portion of the right atrium, cephalad to
the inferior vena cava junction. The inferior vena cava to
extracardiac conduit anastomosis is performed (Fig 4).
Table 2. Risk Factors for Completion Fontan
The conduit is then deaired and blood flow from the
No CPB AVVR inferior vena cava is established to the branch pulmonary
Decreased systemic RV function arteries. The inferior vena cava to atrial shunt is then
Diffusely small branch PA, AVVR removed (Fig 5). All patients had a 4-mm fenestration
LPA stent, SVT, renal failure, pleural effusion created between the IVC conduit and the pulmonary
after Glenn venous atrium (Fig 6). Because of the concern with the
SVT, pacemaker, AVVR, decreased ventricular lack of growth of the extracardiac conduit, typically a 20-
function
to 22-mm conduit is used.
CPB AVVR
No previous Glenn
Analysis
The early postoperative course for these two groups of
AVVR ⫽ moderate atrioventricular valve regurgitation; CPB ⫽ car- patients were compared. Variables analyzed include in-
diopulmonary bypass; LPA ⫽ left pulmonary artery; PA ⫽ pulmo-
nary artery; RV ⫽ right ventricle; SVT ⫽ supraventricular tachy- traoperative and early postoperative hemodynamics,
cardia requiring transcatheter ablation. transfusion of blood and blood products in the first 24
1700 TAM ET AL Ann Thorac Surg
FONTAN WITHOUT CARDIOPULMONARY BYPASS 1999;68:1698 –704

Fig 3. The inferior vena cava is divided from the right atrium and
the cardiac end oversewn. The inferior vena cava to the right atrial
shunt allows venous blood to continue to return to the right atrium.

different. However, when the last 12 patients in the No


CPB group were compared with the CPB group, the
Fig 2. The extracardiac conduit to pulmonary artery anastomosis length of ventilator hours, intensive care unit stay, and
has been completed. Superior vena cava blood flow is now reestab-
lished to both right and left lungs. An inferior vena cava to right
atrial shunt has been placed.

hours of mechanical ventilation, length of intensive care


unit stay, and length of hospital stay. Categorical vari-
ables were compared using Fisher’s exact test, while
continuous variables were compared using Student’s t
test. All results for the continuous variables are reported
as mean ⫾ standard deviation. Results were considered
significant if the p value was less than 0.05.

Results
There was no operative or hospital mortality in either
group. Patients in the No CPB group have significantly
lower pulmonary venous atrial pressures at the end of
surgery, 6 hours, and 12 hours postoperatively (Table 3).
By 24 hours, these differences disappeared (Fig 7). Al-
though proportionately more patients were treated with
milrinone in the CPB group, this difference was not
statistically significant. Cryoprecipitate and platelet
transfusions were significantly more common in the CPB
group (Table 4). Although there was a substantial differ-
ence in the use of more packed red blood cells for the
CPB group, this difference was not statistically signifi-
cant. Time to extubation, length of intensive care unit Fig 4. The inferior vena cava to extracardiac conduit anastomosis is
stay, and length of hospital stay were not statistically performed.
Ann Thorac Surg TAM ET AL 1701
1999;68:1698 –704 FONTAN WITHOUT CARDIOPULMONARY BYPASS

Table 3. Perioperative Hemodynamics


CVP Atrial Pressure
(mm Hg) (mm Hg)

End of surgery
No CPB 12.7 ⫾ 2.6 4.9 ⫾ 1.4
CPB 14.6 ⫾ 2.4 6.6 ⫾ 2.1
p 0.053 0.040
6 Hours postop
No CPB 11.2 ⫾ 3.5 4.2 ⫾ 2.7
CPB 12.6 ⫾ 2.3 6.5 ⫾ 2.1
p 0.22 0.016
12 Hours postop
No CPB 10.9 ⫾ 3.4 4.6 ⫾ 2.8
CPB 13.2 ⫾ 3.1 7.3 ⫾ 3.0
p 0.086 0.031
24 Hours postop
No CPB 13.5 ⫾ 5.0 6.7 ⫾ 4.4
CPB 13.5 ⫾ 2.2 6.7 ⫾ 2.4
p 1.00 1.00

CPB ⫽ cardiopulmonary bypass; CVP ⫽ central venous pressure.

tions. One patient developed refractory ventricular


tachycardia and fibrillation secondary to digoxin toxicity
approximately 18 hours after admission to the intensive
care unit. A loading dose of digoxin had been given
Fig 5. Completion Fontan using an extracardiac conduit has been intravenously over a short period of time, 16 to 18 hours
accomplished without the use of cardiopulmonary bypass. A fenes- postoperatively because of the development of supraven-
tration may be placed between the extracardiac conduit and the
tricular tachycardia (even though this patient was on
atrium.
maintenance digoxin preoperatively). Because this was a
witnessed event in the intensive care unit, cardiopulmo-
hospital stay were all substantially different, with the nary resuscitation was begun immediately and the pa-
length of hospital stay achieving statistical significance. tient brought emergently to the operating room. The
Four patients in the No CPB group suffered complica- child was quickly placed on cardiopulmonary bypass.
Despite being on CPB, multiple attempts at defibrillation
were unsuccessful. This rhythm was resistant to multiple

Fig 6. Angiogram demonstrating direct inferior vena cava to main


pulmonary artery connection, accomplished without the use of car-
diopulmonary bypass. Because of the slight rotation of the patient,
the branch pulmonary artery confluence is superimposed on the
Glenn anastomosis, which had been placed in the very distal right Fig 7. Pulmonary venous atrial pressures at the end of surgery, 6,
branch pulmonary artery. 12, and 24 hours postoperatively. * p ⬍ 0.05.
1702 TAM ET AL Ann Thorac Surg
FONTAN WITHOUT CARDIOPULMONARY BYPASS 1999;68:1698 –704

Table 4. Number of Patients Who Received Perioperative


Transfusion (24 Hours)
PRBC FFP Cryo Platelets

No CPB 17/20 (369 ⫾ 227 mL) 7/20 2/20 4/20


CPB 10/10 (563 ⫾ 271 mL) 3/10 5/10 9/10
p 0.532 1.00 0.026 ⬍ 0.001

CPB ⫽ cardiopulmonary bypass; Cryo ⫽ cryoprecipitate; FFP ⫽


fresh-frozen plasma; PRBC ⫽ packed red blood cells.

phamacologic medications. Dilantin briefly interrupted


the ventricular tachycardia. Digoxin binding antibody,
however, terminated the ventricular tachycardia upon
administration. The digoxin level before binding anti-
body, while the patient was on cardiopulmonary bypass,
was well above 5.0 ng/mL. He subsequently developed
multiple complications, including renal failure, empy- Fig 9. Construction of the Glenn anastomosis toward the left branch
ema, and remained in the intensive care unit for a period pulmonary artery in the same patient with hypoplastic left heart
of 50 days. Fortunately, he has recovered since with no syndrome.
discernable neurologic deficit. One patient suffered a
perioperative stroke with the development of left upper
extremity paresis. The etiology of this insult remains clinical outcomes. Staging the Fontan operation with the
unclear. Two patients developed superficial wound in- bidirectional Glenn anastomosis, and the deliberate cre-
fections, requiring treatment with antibiotics only. In the ation of a residual right-to-left atrial shunt, have un-
CPB group, 1 patient developed a severe tracheitis and doubtedly contributed significantly to the improved early
required continuous sedation, paralysis and prolonged outcome of these patients. The investigations of De Leval
endotracheal intubation for approximately 13 days. Per- and associates [2] and Sharma and associates [5] have
sistent pleural effusions occurring after 2 weeks, requir- added further guidance in optimizing the geometry of
ing catheter drainage, occurred in 4 patients in the No the cavopulmonary connections. The use of an extracar-
CPB group and in 5 patients in the CPB group. This diac inferior vena cava to pulmonary artery connection
difference was not statistically significant. has a number of theoretical advantages. It avoids exten-
sive atrial suture lines, exposure of the atrium to higher
venous pressures, and theoretically better preservation
Comments of kinetic energy in the Fontan circuit. In addition,
Surgical management of children with single ventricular because the connection is created outside the heart,
physiology has continued to evolve. Many institutions aortic cross-clamping could be entirely avoided. These
have adopted various approaches with improved early theoretical advantages have prompted the application of
the extracardiac cavopulmonary connection at our
institution.
The outcome after completion Fontan operations may
be further optimized with appropriate planning. Metic-
ulous attention is given to achieving appropriate growth
and development of branch pulmonary arteries with
avoidance of any significant pulmonary artery stenosis
(Fig 8). In the newborn period, excessive pulmonary
blood flow is avoided. A bidirectional Glenn anastomosis
is constructed in early to mid infancy. We have not
utilized the hemi-Fontan approach. The superior vena

Table 5. Hematology, Coagulation (Postoperative Day 1,


3:00 a.m.)
Hgb Hct PT PTT

No CPB 13.3 ⫾ 1.9 39.1 ⫾ 5.6 15.7 ⫾ 3.4 47.6 ⫾ 19.2


CPB 14.5 ⫾ 1.9 42.9 ⫾ 6.1 15.4 ⫾ 2.1 42.9 ⫾ 10.7
p 0.10 0.11 0.79 0.40
Fig 8. Optimal growth and development of the right and left branch CPB ⫽ cardiopulmonary bypass; Hgb ⫽ hemoglobin (gm/100 ml);
pulmonary arteries in an infant 5 months after the Norwood opera- Hct ⫽ hematocrit; PT ⫽ prothrombin time (in seconds); PTT ⫽
tion for hypoplastic left heart syndrome. partial thromboplastin time (in seconds).
Ann Thorac Surg TAM ET AL 1703
1999;68:1698 –704 FONTAN WITHOUT CARDIOPULMONARY BYPASS

cava (SVC) anastomosis is either diverted toward the Table 6. Postoperative Course Comparing the 12 Most
right, or more typically toward the left branch pulmonary Recent Patients in the No CPB Group
artery (Fig 9). This tends to encourage appropriate Ventilator ICU Stay Hospital Stay
growth of the left branch pulmonary artery, simplifying (h) (h) (days)
the later completion Fontan procedure. Having the SVC
No CPB 8.2 ⫾ 5.3 51.1 ⫾ 28.4 7.0 ⫾ 3.8
anastomosis in the very proximal right branch pulmo-
CPB 50.1 ⫾ 95.7 93.9 ⫾ 108.9 10.8 ⫾ 4.3
nary artery, occasionally into the proximal left branch p 0.20 0.23 0.036
pulmonary artery, easily allows construction of the extra-
cardiac Fontan without the use of cardiopulmonary by- CPB ⫽ cardiopulmonary bypass; ICU ⫽ intensive care unit.
pass. Need for simultaneous extensive pulmonary artery
reconstruction, atrioventricular valve repair, and other
Conclusions
procedures would exclude the use of this technique. Review of our early experience with the construction of
Theoretically, we avoid the inflammatory sequelae asso- extracardiac cavopulmonary connection without the use
ciated with the use of cardiopulmonary bypass. The of cardiopulmonary bypass has been encouraging. Early
postoperative hemodynamics appear to be improved
results in this preliminary group indeed suggest that the
with less frequent use of blood and blood products.
immediate hemodynamics are improved compared with
Analysis of the most recent group has also suggested a
the CPB group. The use of blood and blood products are
quicker recovery and shorter hospital stay (Tables 5, 6).
more limited. With further refinement, shorter immedi-
Hopefully, with further refinement, this would translate
ate recovery time and shorter hospital stays are achiev- into not only improved short-term but perhaps long-term
able. However, because blood is exposed to the foreign outcome for our patients as well.
material of the IVC to right atrial shunt, we can not
assume the usual inflammatory response to cardiopul-
We acknowledge the support and advise by Drs Robert A.
monary bypass is entirely avoided. A prospective ran- Guyton, Willis H. Williams, Kirk R. Kanter, James M. Bailey,
domized trial with measurement of inflammatory medi- Steve Toscone, Nina Guzetta, Shiva Sharma, David Jones, and
ators may indeed provide interesting results. the physicians of the Children’s Heart Center, Atlanta, Georgia.
The obvious single disadvantage of the extracardiac We are also grateful for the expert assistance by Vertis Walker in
the preparation of this manuscript.
Fontan is the lack of growth of the conduit. To minimize
the likelihood for replacement of the extracardiac con-
duit, we have utilized a conduit of 20 to 22 mm in References
diameter. Whether the IVC will grow to compensate for
1. Fontan F, Baudet E. Surgical repair of tricuspid atresia.
the fixed length of the conduit remains to be seen. In Thorax 1971;26:240– 8.
terms of conduit material, we have generally preferred a 2. De Leval M, Kilner P, Gewillig M, Bull C. Total cavopulmo-
valveless ascending aortic homograft conduit. Typically, nary connection: a logical alternative to atrial pulmonary
connection for complex Fontan operations. J Thorac Cardio-
the homograft is divided at or above the level of the vasc Surg 1988;96:682–95.
sinotubular ridge. Valve tissue and associated ventricular 3. Marcelletti C, Corno A, Giannico S, Marino B. Inferior vena
muscle are excised. Because of the posterior location of cava-pulmonary artery extracardiac conduit. J Thoracic Car-
the IVC and the more anterior location of the right diovasc Surg 1990;100:228–32.
4. Black MD, van Son JAM, Haas GS. Extracardiac Fontan
pulmonary artery, with the potential for right upper operation with adjustable communication. Ann Thorac Surg
pulmonary vein compression, the arch of the homograft 1995;60:716– 8.
provides the perfect geometry for this extracardiac con- 5. Sharma S, Goudy S, Walker P, et al. In-vitro flow experiments
for the determination of the optimal geometry of the total
duit. Since the homograft provides a nonthrombogenic
cavopulmonary connection for surgical repair of children
surface, we have not routinely used anticoagulation, with functional single ventricle. J Am Coll Cardiol 1996;27:
except for once daily aspirin. 1264–9.

DISCUSSION
DR EDWARD L. BOVE (Ann Arbor, MI): It was not clear to me cation. Obviously, if a child needed to have a completion Fontan
in your paper just what preoperative indications you used for operation with, for instance, AV valve repair, then obviously that
patients undergoing a Fontan without the use of bypass. Can would not be possible. In the kids who are appropriately
you identify those patients before surgery and did you have prepared with well-developed branch pulmonary arteries with a
patients that had to be converted to bypass even though the Glenn anastomosis, doing the operation without bypass is quite
original plan was to do it without the pump. feasible. Early in our experience there were a few patients that
were converted to bypass because we had difficulty placing the
DR TAM: What we have tried to do is to address any and all IVC cannula distal enough without significant blood loss for us
surgical issues at the Glenn stage so that when these children to do the operation.
came back for the Fontan operation, generally speaking, we tend
not to have difficulty with branch pulmonary artery stenosis or DR J. TERRANCE DAVIS (Columbus, OH): Could you give us
other unresolved surgical issues, eg, restrictive atrial communi- a little more information about what kind of conduit you use and

You might also like