Definitive operation for pulmonary atresia with intact ventricular septum

J

THoRAc CARDIOVASC SURG

1987;93:454-64

Definitive operation for pulmonary atresia with intact ventricular septum Results in twenty patients Definitive operation was performed on 20 patients (aged 25 to 178 months) with pulmonary atresia and intact ventricular septum. All patients bad one or more prior preliminary palliative procedures. Right ventricular outflow tract reconstruction with atrial septal defect closure and shunt removal was done on 10 patients. Tricuspid annular circumference was at least 70% of normal in seven patients and between 55 % and 70 % in three patients. Two patients died during hospitalization. The eight surviving patients were asymptomatic 3 to 145 months after operation. The modifiedFontan operation was performed on 10 patients. None of these patients had a tricuspid annular circumference greater than 70 % of normal; the circumference was less than 55 % in nine patients and between 55 % and 70 % in one patient. One patient died during hospitalization and one died later. FoUow-up 6 to 48 months after operation showed that six patients were in the New York Heart Association Class I and two were in Class Il, Mter effective preliminary palliation of pulmonary atresia with intact ventricular septum, definitive operation can be done with an operative risk of 15 % (three of 20 patients) and exceUent late results. Right ventricular outflow tract reconstruction can be done as a complete repair for patients who have adequate tricuspid annular size. The modified Fontan operation is the only option for definitive repair when the tricuspid anulus is severely hypoplastic.

Ernerio T. Alboliras, M.D., Paul R. Julsrud, M.D., Gordon K. Danielson, M.D., Francisco J. Puga, M.D., Hartzell V. Schaff, M.D., Dwight C. McGoon, M.D., Donald J. Hagler, M.D., William D. Edwards, M.D., and David J. Driscoll, M.D., Rochester, Minn.

h e reported experience with definitive operation for pulmonary atresia and intact ventricular septum is limited.!" In the past, many infants with this rare condition did not survive long enough to have a definitive operation.v" 7 However, with prostaglandin therapy to maintain postnatal patency of the ductus arteriosus" and improved operative and postoperative techniques," 10 more infants are surviving infancy and becoming candidates for correction or definitive palliation. From the Sections of Pediatric Cardiology, Thoracic and Cardiovascular Surgery, and Medical Pathology, and the Department of Diagnostic Radiology, Mayo Clinic and Mayo Foundation, Rochester, Minn. Read in part at the Second World Congress of Paediatric Cardiology, New York, June 2-6, 1985. Received for publication Feb. 27, 1986. Accepted for publication April 8, 1986. Address for reprints: David J. Driscoll, M.D., Section of Pediatric Cardiology, Mayo Clinic, 200 First St. S.W., Rochester, Minn. 55905.

454

Creating continuity between the main pulmonary artery and the right ventricular outflow tract with closure of the atrial septal defect and the preexisting shunts (definitive correction) has been advocated for patients who have a detectable right ventricular infundibulum." The Fontan operation 1I, 12 (definitive palliation) is an alternative to this operative technique, particularly for patients with an incompletely developed right ventricle or a severely diminutive tricuspid valve apparatus.v- 13 The purpose of this study was to determine the operative and long-term results of definitive operation for pulmonary atresia with intact ventricular septum and to define guidelines for selecting the appropriate type of operation. Methods

We reviewed the records of all 51 patients with pulmonary atresia and intact ventricular septum who were evaluated at the Mayo Clinic from 1969 through 1984. Twenty of these patients had definitive operations

Volume 93 Number 3 March 1987

Pulmonary atresia with intact ventricular septum

and are the subjects of this report. Definitive operation is any procedure that is designed to completely separate the pulmonary and systemic circulations. Cardiac catheterization data, angiocardiograms, twodimensional echocardiograms, and operative reports were studied to evaluate the anatomy of the right ventricle before definitive operation. The tripartite classification of Goor and Lillehei," which was advocated by Bull and colleagues," was utilized to assess the right ventricular morphology. With this approach, the presence or absence of the inlet, trabecular, and infundibular portions of the right ventricular cavity was determined. The inlet portion is defined as the part of the ventricle incorporating the tricuspid valve apparatus. The trabecular portion is the heavily trabeculated part distal to the insertion of the papillary muscle toward the right ventricular apex. The infundibular portion is the outlet that leads to the atretic pulmonary valve. Right ventricles are classified into one of three categories as proposed by Bull and colleagues 15: (1) those with three portions present, (2) those in which the trabecular portion of the right ventricle is obliterated by hypertrophied muscle and is effectively absent, and (3) those in which both the trabecular and infundibular portions are obliterated by hypertrophied muscles, so that only the inlet portion is present. Listed in order of acceptance, the diameter of the tricuspid anulus was determined by (1) intraoperative measurement (four patients), (2) angiographic measurement (10 patients), and (3) echocardiographic measurement (six patients). For patients who had right ventriculography, the diameter of the tricuspid valve anulus was measured from end-diastolic frames with the greater dimension of either the anteroposterior or the lateral view. The actual diameter of the anulus was calculated with the known diameter of the catheter to obtain a magnification factor. Two-dimensional echocardiographic measurements were used only if intraoperative measurement or a right ventriculogram was unavailable. The circumference was calculated with the assumption that the anulus was a circle, using the formula: C=1TD

in which C is the circumference and D is the measured diameter. There were no significant differences in angiographic and echocardiographic measurements in four of the 20 patients when both studies were available (two-tailed t test, p = 0.18). The annular dimensions obtained for living patients in our study were compared with the autopsy data for normal patients reported by Rowlatt, Rimoldi, and Lev 16 after determining a postmortem shrinkage factor. This was accomplished by correlating angiographic and post-

455

160

E g

140 120

Q)

o

c:

... E ... '0

100

0

40

Q) Q)

80 /-

::J

()

'61 c:

-c

/-

~

60

•

20 OK;....JL......J.--L......L....J-....L......L....L..-..L.-L-J-.J........L.....J

o

20

40

60

80

100 120 140

Autopsy circumference (mm)

Fig. 1. Comparison of angiographic (angio) measurement and autopsy measurement of tricuspid annular circumference of eight patients who had a right ventriculogram and did not have tricuspid valve or right ventricular hypoplasia and who died within the past 2 years. Angiographic circumference Autopsy (mm) + 0.1193 circumference (mm) = ---'---------'-------1.0407 r = 0.98 (p < 0.001). Standard error = 0.104. Solid line. Line of identity. Broken line. Line of regression.

mortem tricuspid valve annular measurements for eight cases chosen at random (irrespective of the type of congenital heart defect), in which right ventriculography and autopsy had been done within the past 2 years and there was no tricuspid valve or right ventricular hypoplasia (Fig. 1). We determined the following with linear regression analysis: Postmortem circumference

Angiographic circumference + 0.1193

= -------'---:..:.:...:----'------

1.0407

The correlation coefficient between the postmortem and the angiographic circumference was 0.98 (p < 0.(01). With this formula antemortem measurements of tricuspid valve anulus taken from right ventriculograms (or operative or echocardiographic measurements, as indicated) were corrected to postmortem measurements and compared with the normal mean of Rowlatt, Rimoldi, and LeV. 16 The normal mean was obtained with the formula from data by Rowlatt and colleagues": Normal mean (mm) = [-16.1170 + 5.9802 log (surface area in squared meters X 10,000)] X 10

The Journal of Thoracic and Cardiovascular Surgery

Alboliras et al.

4 56

Table I. Tricuspid valve circumference related to right ventricular morphology and type of definitive operation Right ventricular morphology Tricuspid anulus (% normal)

All three portions present

>70 55-70 <55 Total

7 4 4 15

I

Inlet and outlet present 0 0 0 0

I

Inlet only present 0 0 5 5

Definitive operation RVOTR 7

I

Fontan

3

0 I

0 10

9 10

Legend: RVOTR. Right ventricular outflow tract reconstruction.

Table II. Prior operative procedures related to type of definitive repair Procedure Waterston shunt Aorta-MPA shunt Aorta-LPA shunt Blalock-Taussig shunt Balloon septostomy Blalock-Hanlon septectomy T ransventricular pulmonary valvotomy Transpulmonary valvotomy RVOTR (ASD not closed) PDA ligation Total

RVOTR

9 0

Fontan 5 2

0

I

0

5 5

Total 14 2 I 5 5

0 I

0

I

4

2

6

2

I 2

I

3 3 2

24

42

I I 18

Legend: RVOTR, Right ventricular outflow tract reconstruction. MPA. Main pulmonary artery. LPA. Left pulmonary artery. ASD, Atrial septal dcfccr PDA. Patent ductus arteriosus.

Previous investigators-" have determined that the right ventricular volume in pulmonary atresia with intact ventricular septum is proportional to the size of the tricuspid valve anulus. Because of the known limitations of determination of right ventricular volume in this malformation.!'" volumetric data were not analyzed. All associated malformations and prior palliative procedures were noted. The type of definitive repair was correlated with the morphologic and hemodynamic data. After hospital discharge, the patients were evaluated through outpatient follow-up at the Mayo Clinic or correspondence with the patients and referring physicians.

Results There were 14 male and six female patients. The mean age at definitive operation was 78.2 months (range, 25 to 178 months). Of the 20 patients, 15 had a tripartite right ventricle, and five had no trabecular or infundibular portion. No patient had a visible right

ventricle-coronary artery fistula. Ten patients had repair with right ventricular outflow tract reconstruction, and 10 had a modified Fontan procedure as definitive palliation (Table I). Prior palliative procedures. A total of 42 operative procedures had been done before definitive repair (Table 11). All patients had one or more prior palliative operations. Initial systemic-pulmonary artery shunt operations were performed at other institutions in 18 of the 20 patients. Opening the right ventricular outflow tract, either by pulmonary valvotomy (transpuimonary or transventricular) or by patch reconstruction without closure of the atrial septal defect, had been performed as a palliative procedure for nine of the 15 patients who had a tripartite right ventricle. When measured before definitive correction, the tricuspid valve anulus of patients who had this palliative procedure averaged 70.8% of normal (range, 47.3% to 104.0%). This was not significantly different from the mean tricuspid annular circumference of the six patients who had a tripartite right ventricle and who did not have an opening created between the right ventricular outflow tract and the main pulmonary artery (mean, 74.6%; range, 38.5% to 99.1%). At the time of definitive operation, all patients who had had prior pulmonary valvotomy required either further outflow tract reconstruction (with or without conduit) or the Fontan operation. Serial studies were available for two of the eight patients who had prior pulmonary valvotomy or outflow tract reconstruction without atrial septal defect closure. In both, tricuspid circumference increased more than expected, relative to change of body surface area. In one patient the tricuspid annular circumference increased from 52.7% to 67.6% of normal during a period of 3 years, and the patient eventually had right ventricular outflow tract reconstruction with atrial septal defect and shunt closure. In the other patient the tricuspid annular circumference increased from 30.7% to 47.3% of normal during a period of 3 years, but this increase in circum-

Volume 93 Number 3

Pulmonary atresia with intact ventricular septum

March 1987

457

Table m. Hemodynamic data of 10 patients who had right ventricular outflow tract reconstruction with ASD and shunt closure Systolic pressures (mm Hg) Preoperative

Definitive correction Case

1*

Age (mo)

I

Prior operation

RV

Waterston Waterston, PDA ligation, pulmonary valvotomy, RVOTR (open ASD and shunt) Pulmonary valvotomy, Waterston Waterston Pulmonary valvotomy Waterston Balloon septostomy, Waterston Waterston, pulmonary valvotomy Waterston, pulmonary valvotomy Waterston, Blalock-Hanlon, pulmonary valvotomy

Technique

2

55 142

Patch Patch

3

39

Patch

4 5 6 7

54 52 120 96

Patch Patch Hancock conduit Hancock conduit

8

52

Hancock conduit

9

83

Hancock conduit

10*

32

Hancock conduit, TVR

Postoperative

I

PA

RV

95 40

25

75 30

106

33

30

119 93 86 192

31 14

22 50 40 60

20 30 40 50

100

21

30

28

160

15

30

25

40

40

50

50

I

PA 30

Legend: ASD,Atrial septal defect. RY, Rightventricle. PA, Pulmonary artery. PDA, Patentductusarteriosus. RYOTR, Rightventricular outflow tract reconstruction. TYR, Tricuspid valve replacement. 'Death during hospitalization.

Table IV. Tricuspid annular circumference in 10 patients who had right ventricular outflow tract reconstruction Normal mean

circumference (mm)

Pat hoconverted circumference (mm) *

(mmtt'

Percent!

61.8 59.7 54.4 58.6 69.0 79.0 64.0 56.6 42.4 39.3

59.5 57.5 52.4 56.4 66.4 76.0 61.6 54.5 40.9 37.9

64.3 85.1 63.0 65.2 68.4 76.7 70.6 52.4 68.8 60.0

92.5 67.6 83.2 86.5 97.1 99.1 87.3 104.0 59.4 63.2

Surface area (m 2)

Angiofechofoperative

1 2

0.59

3

0.56 0.61 0.69 0.95 0.75 0.48 0.70 0.50

Case

4 5 6

7 8 9 10

1.31

Angiojechojopcrative circumference + 0.11993 'Pathology-converted circumference = - " - - - ' - - ' - - - = - - - - - - - - - - - 1.0407

tNormal mean (rnrn) = [-16.1170 + 5.9802 log (surface area in 01' X 10,000)] tMean, 84.0% of normal.

ference was judged to be inadequate, and a Fontan operation was performed. Right ventricular outflow tract reconstruction. Right ventricular outflow tract reconstruction with closure of the atrial septal defect and all existing palliative shunts was performed on 10 patients (Table

X 10;

reported by Rowlatt, Rimoldi, and Lev."

III). The mean age at operation was 72.5 months (range, 32 to 142 months). In one patient the atrial septal defect was closed 6 years after right ventricular outflow tract reconstruction. All 10 patients had a tripartite right ventricle. The tricuspid valve circumference averaged 84.0% of normal (range, 59.4% to

The Journal of

4 58

Alboliras et al.

Thoracic and Cardiovascular Surgery

Fig. 2. Methods used for right ventricular outflow tract reconstruction. Left. Portions of hypertrophied infundibulum, atretic pulmonary valve, and part of main pulmonary artery are resected. Middle. Valveless outflow patch (Dacron, homograft dura, or pericardium) is onlayed over the resected parts. Right. Second option: valved conduit between outlet and pulmonary artery.

104.0%) (Table IV). Two patients had Ebstein's malformation of the tricuspid valve, one of whom also had a double-chambered right ventricle. Valveless outflow patch. In five of the 10 patients, outflow tract patch reconstruction was accomplished without inserting a valve. Portions of the hypertrophied infundibulum and the obstructive pulmonary valve were resected (Fig. 2). A pericardial patch was used in three patients, a Dacron patch in one patient, and a homograft dural patch in one patient. Valved conduit. In five of the 10 patients a Hancock porcine-valved conduit was used to reconstruct the right ventricular outflow tract (Fig. 2). Two patients had severe tricuspid insufficiency: one received a porcine valve replacement and the other an annuloplasty. In the two patients with Ebstein's malformation, the tricuspid valve was competent so that nothing was done to the valve. Outcome. There were two patients who died during hospitalization among the 10 patients with right ventricular outflow tract reconstruction (mortality rate, 20%). The patient with Ebstein's malformation and doublechambered right ventricle (Case 1) died during the operation. The patient's tricuspid anulus was 92.5% of normal. Another patient (Case 10) died from intractable heart failure on the seventh postoperative day. This patient also had a tricuspid valve prosthesis, and the tricuspid anulus was 63.2% of normal and was the second smallest in this group of 10 patients. Preoperative and postoperative right ventricular pressures were available from either cardiac catheterization or intraoperative measurements (Table III). The two deaths during hospitalization were of patients with

persistent right ventricular hypertension. The patient who had right ventricular outflow tract reconstruction before atrial septal defect closure (Case 2) had a preoperative right ventricular systolicpressure of 40 mm Hg, which decreased to 30 mm Hg after the final operation. The seven other surviving patients had a mean preoperative right ventricular systolic pressure of 122.3 mm Hg (range, 86 to 192 mm Hg). The postoperative right ventricular systolic pressure of these seven patients averaged 37.4 mm Hg (range, 22 to 60 mm Hg), with an average decrease in right ventricular pressure of 84.8 mm Hg (range of decrease, 43 to 132 mm Hg). Postoperative pulmonary artery pressure was available in eight of the 10 patients. The mean pressure gradient across the right ventricular outflow tract was 4.9 mm Hg (range, 0 to 20 mm Hg), with an average pulmonary arterial systolic pressure of 34.1 mm Hg (range, 20 to 50 mm Hg). Follow-up. The eight surviving patients have had follow-up for a mean of 65 months (range, 3 to 145 months). There were no late deaths. Four patients are on digitalis therapy. All eight patients are in New York Heart Association (NYHA) Fundational Class I. Fontan operation. A modified Fontan operation was performed on 10 patients (Tables I and V). The mean age at operation was 83.8 months (range, 25 to 178 months). In five of the 10 patients, analysis of right ventricular morphology revealed only the inlet portion present. No patient had only the inlet and the infundibulum present. In the other fivepatients all three portions of the right ventricle (inlet, trabecular, and infundibular) were present, but these structures and the tricuspid valve were believed to be too small for right ventricular

Volume 93 Number 3

Pulmonary atresia with intact ventricular septum

March 1987

459

Table V. Hemodynamic data of 10 patients who had modified Fontan operation Pressures (/11m Hg)

Case

Age at operation (mo)

Physiologic correction + concomitant procedures (shunt closure excluded)

12

144 89

13

72

14

52

15

103

16

178

17*

40

18

82

RA-RPA (tricuspid valve excised) RA-MPA, modified Glenn

19

25

RA-RV-MPA

20

53

RA-RV-MPA

II

RA-MPA RA-RV-MPA, modified Glenn RA-RPA, patch enlargement, LPA stenosis RA-RPA (tricuspid valve patched) RA-RV-MPA, modified Glenn RA-RPA

Preop. Mean PA

Postop. mean RA

107 45

17 19

16 22

125

13

10

170

12

16

101

19

21

100

10

20

80

8

18

9

19

22

10

10

25

20

17

Prior operation

Systemic RV

Waterston, Blalock-Taussig Waterston, pulmonary valvotomy, RVOTR (open ASD and shunt) Balloon septostomy, Blalock-Taussig Balloon septostomy, Waterston Balloon septostomy, Waterston Balloon septostomy, Blalock-Taussig Aorta-MPA shunt Waterston, Blalock-Taussig, aorta-LPA shunt Balloon septostomy, aorta-MPA shunt, pulmonary valvotomy Pulmonary valvotomy, PDA ligation, Blalock-Taussig, RVOTR (open ASD and shunt)

I

Legend: RV. Right ventricle. PA, Pulmonary artery. RA, Right atrium. MPA, Main pulmonary artery. RVOTR, Right ventricular outtlow tract reconstruction. ASD. Atrial septal defect. RPA, Right pulmonary artery. LPA. Left pulmonary artery. PDA, Patent ductus arteriosus. 'Death during hospitalization.

Table VI. Tricuspid annular circumference in 10 patients who had the Fontan operation Case

Surface area (m 2)

Angiolechofoperative circumference (mm)

Pathoconverted circumference (mm)*

Normal mean (mm) *

Percenii

II 12 13 14 15 16 17 18 19 20

0.92 0.72 0.76 0.68 0.87 1.49 0.55 0.75 0.45 0.75

34.7 37.7 28.3 26.7 18.8 47.1 28.3 29.8 37.7 34.6

33.5 36.3 27.3 25.8 18.2 45.4 27.3 28.7 36.3 33.4

75.9 69.5 70.9 68.0 74.4 88.4 62.5 70.6 57.3 70.6

44.1 52.2 38.5 37.9 24.5 51.4 43.7 40.7 63.4 47.3

'See Table IV. t Mcan, 44.4% of normal.

outflow tract reconstruction and closure of atrial septal defect and prior palliative shunts. In two of these five patients (Cases 12 and 20), right ventricular outflow tract reconstruction without closure of the atrial septal defect had been done as a palliative procedure. However, because right ventricular and tricuspid annular

growth was insufficient, a Fontan operation was performed-2 years in one case and 3 years in the other-after palliative outflow tract reconstruction. The circumference of the tricuspid valve of the 10 patients who had the Fontan operation averaged 44.4% of normal (range, 24.5% to 63.4%) (Table VI). The

The Journal of Thoracic and Cardiovascular Surgery

4 6 0 Alboliras et al.

Fig. 3. Methods used to establish continuity between right atrium and pulmonary arteries. Left, Right atrial appendageright ventricle-main pulmonary artery connection with anterior patch onlay. Right, Direct anastomosis of right atrium to right pulmonary artery, with pericardial patch augmentation.

circumferences of the tricuspid valve of the two patients who had a prior right ventricular outflow tract reconstruction were 52.2% and 47.3% of normal before definitive correction. Right atrial-pulmonary arterial connection. Six patients had a direct right atrial-pulmonary arterial connection (Fig. 3). In three of these patients the right. atrial appendage was anastomosed to the right pulmonary artery, with pericardial patch augmentation; the tricuspid valve was left intact in two patients and was patched in one patient. In two patients the right atrial appendage was anastomosed to the main pulmonary artery, with anteriorly located pericardial patch augmentation and the tricuspid valve left intact. Only in one patient was the tricuspid valve excised and continuity of the resultant common atrioventricular cavity and the pulmonary artery established by anastomosing the right atrial appendage to the right pulmonary artery. One of these three modifications for establishing right atrialto-pulmonary arterial connection was performed in four of the five patients who had only the inlet present and in two of the five patients who had a tripartite right ventricle. Right atrial-right ventricular-main pulmonary arterial connection. Right atrial-right ventricular-main pulmonary arterial connection was performed in four patients (Fig. 3). In these patients the right atrial appendage was used as a flap that was anastomosed to the edges of a ventriculotomy, which extended cranially onto the main pulmonary artery. A generous pericardial or dura mater graft was used to construct the sides and

roof of the conduit. These modifications were used in three of the five patients with a tripartite right ventricle and in one of the five patients with only the inlet present. Complicating conditions. Four of the 10 patients who had the Fontan operation had severe stenosis of a pulmonary artery from a prior shunt. In three of the four patients, a modified Glenn anastomosis," which consisted of a side-to-end connection between the superior vena cava and the right pulmonary artery, was also performed with the Fontan operation; the right pulmonary artery was transected at the stenotic site. The junction of the superior vena cava and right atrium was left patent. One of the four patients had a left pulmonary artery stenosis from a prior Blalock-Taussig shunt; this site was enlarged with a pericardial patch. Hospital course. There was a 3-year-old patient who, while hospitalized, died of low cardiac output 7 days after the operation (mortality rate, 10%). One patient had a stroke, and another developed the postpericardiotomy syndrome. Follow-up. The mean period of follow-up was 18 months (range, 6 to 48 months). There was one late death caused by arrhythmia; the patient died 39 months after the operation. The patient had the sick sinus syndrome and developed atrial flutter at the time of death. Of the remaining eight survivors, six are in NYHA Functional Class I, and two are in Class II. Five are taking cardiac medications (two patients, digoxin alone; three patients, digoxin, captopril, and furosemide). Two patients had further operations: ligation of an unrecognized patent ductus arteriosus in one patient and enlargement of the superior vena cava-right pulmonary artery connection with removal of thrombus in the other. Discussion Pulmonary atresia with intact ventricular septum is a difficult management problem. In an early report, there were only three survivors among 123 patients treated surgically." However, advances in medical and surgical care and the understanding of the natural history and hemodynamics of this lesion have resulted in improved survival rates from palliative and corrective opera-

tions."6. 19.20

Palliative operation. Because of hypoxemia from diminished pulmonary blood flow and restrictive atrial septal defect, one or more palliative procedures may be necessary early in life to permit survival to an appropriate age for definitive repair.'? When possible, establishing continuity between the right ventricular cavity and

Volume 93 Number 3 March 1987

the main pulmonary artery is desirable for those neonates with an identifiable right ventricular outflow tract. 5,7,10,17 Continuity can be achieved by transventricular valvotomy, transpulmonary valvotomy, or infundibular resection and outflow tract patching. These procedures decrease right ventricular hypertension, cause regression of myocardial hypertrophy, enhance right ventricular compliance, restore normal coronary flow patterns, and promote right ventricular cavity growth." The increase in right ventricular size and tricuspid annular circumference was documented in two of our patients. Ideally, opening the right ventricular outflowtract will therefore increase the possibility of the patient achieving a suitably sized right ventricular chamber for eventual definitive repair. Definitive operation. There are two options for definitivetreatment.':' If the size of the right ventricle is adequate, the preferred option is correction by right ventricular outflow tract reconstruction and closure of the atrial septal defect and existing left-to-right shunts. However, for patients with an incompletely developed right ventricle, especially those with a severely diminutive tricuspid valve apparatus, definitive palliation with the Fontan operation I I has been advocated.v- 12, 13 Role of tricuspid valve annular size. The tripartite approach to right ventricular morphology'<" is helpful in selecting the type of initial palliative procedure. Patients who have an infundibular cavity may be suitable candidates for pulmonary valvotomy. However, the approach to definitive repair is based primarily on the actual size of the tricuspid anulus and the right ventricular cavity. Tricuspid valve size is a better index of the adequacy of the right side of the heart for repair because of the limitations in determining right ventricular volumes in this condition.!" 18 since the volume of the right ventricle in pulmonary atresia with intact ventricular septum is directly proportional to the tricuspid valve circumference" and since the size of the right ventricle can be enlarged by a right ventricular onlay patch. Because the available nomograms of tricuspid annular dimensions are from autopsy data, Bull and colleagues" used antemortem (angiographic) and postmortem measurements of tricuspid valves in patients with pulmonary atresia and intact ventricular septum and determined a shrinkage factor of 1.43:1. To derive their angiographic normal values, they increased the autopsy normal data of Rowlatt, Rimoldi, and Lev" by this factor. Because the data from Rowlatt and associates were taken from normal hearts, we believe that comparing angiographic and postmortem tricuspid annular dimensions of patients who have a nonhypoplastic tricuspid valve anulus and right ventricle, as was done in

Pulmonary atresia with intact ventricular septum

46 I

our study, reflects a more appropriate shrinkage factor. This approach avoids the uncertainty of the effect on the tricuspid annular dimension of the shrinkage of the heavily trabeculated, muscle-bound right ventricle in pulmonary atresia with intact ventricular septum. The nature of shrinkage also may be different in patients whose tricuspid valves are dysplastic. Furthermore, any minor shrinkage in absolute dimension of the hypoplastic tricuspid valve may actually be major shrinkage if related to the actual size of the valve. Because of these differences in deriving the shrinkage factor, our normal mean tricuspid annular circumference is smaller than that reported by Bull and colleagues." By our derivation, a tricuspid valve circumference that is 70% of normal or greater (after pathology conversion) is indicative of adequate right ventricular size for right ventricular outflow tract reconstruction. This percentage coincides with the limits of the lower two standard deviations for normal of Rowlatt, Rimoldi, and LeV. 16 Seven of our 20 patients had tricuspid valve circumference greater than 70%; all had three portions of the right ventricle present, and all had right ventricular outflow tract reconstruction. With the data of Bull and associates," we found that only three of our 20 patients had tricuspid valve diameters above the lower 99% confidence limit. We consider an annular circumference of between 55% and 70% as borderline. Four of our 20 patients had circumferences in this category. All four patients had a tripartite right ventricle, with three patients having right ventricular outflow tract reconstruction and atrial septal defect and palliative shunt closure. The fourth patient had the modified Fontan operation. A tricuspid valve with an annular circumference of less than 55% of normal probably is too hypoplastic to allow incorporation of the right ventricle into the repair. Nine of the 20 patients had anuli in this category, and all had the modified Fontan operation. Four of the nine patients had a tripartite right ventricle. The presence of three portions of the right ventricle therefore does not always signify adequacy of the tricuspid valve or right ventricle for standard right ventricular outflow tract reconstruction. Role of myocardial sinusoid-coronary artery communication. Little is known regarding the possible effect of definitive operation on patients with existing right ventricular myocardial sinusoid-coronary artery fistula. Although, by chance, none of the 20 patients in the present series had fistulae, among our other patients who are yet to have definitive operations, are two patients with large fistulous communication between the right ventricle and the left anterior descending artery. It

The Journal of Thoracic and Cardiovascular Surgery

4 6 2 Alboliras et af.

is conceivable that right ventricular decompression could adversely affect myocardial perfusion and produce myocardial infarction in patients with large myocardial sinusoid-to-coronary artery communication." 23 It is possible that such patients may benefit from ligation of the large fistulous communication. Right ventricular outflow tract reconstruction. Right ventricular outflow tract reconstruction with closure of atrial septal defect and existing left-to-right shunts can be done on patients who have sufficient tricuspid annular and right ventricular cavity size. Each of the 10 patients who had this operation had all three right ventricular portions present. This group had larger tricuspid annular circumferences (mean, 84% of normal) than the patients who had the Fontan operation (mean, 44.4% of normal). In this group, in one of the two patients who died (Case 10), the small tricuspid annular size (63.2% of normal) may have been one of the negative prognostic factors. De Leval and colleagues" reported an 80% mortality rate among their patients who had right ventricular outflow tract reconstruction and who had a tricuspid valve diameter below the authors' lower 99% confidence limit. However, their mortality rate was only 16% for patients with valve diameters above the lower 99% confidence limit. The authors also found a higher mortality (three of five patients) among patients without the trabecular portion of the right ventricle. The two methods available for outflow tract reconstruction (valveless outflow patch and valved conduit) were used, and the late results were excellent. Definitive repair may be performed without further infundibular resection before patch or conduit placement.' However, in our series, it was preferable in all cases to widen the right ventricular outflow tract through excision of portions of the infundibulum to provide optimal relief of the pressure gradient. If a valveless outflow patch is used, pulmonary insufficiency might be a concern with respect to late development of right ventricular failure.' However, there have been no overt signs of right ventricular failure in the four surviving patients in our series who had a valveless outflow patch and who had follow-up from 39 to 142 months. Furthermore, in patients with tetralogy of Fallot, in whom a transannular patch was used or extensive pulmonary valvotomy done, the resultant pulmonary insufficiency has been tolerated well in long-term survivors." If symptoms of right ventricular failure occur after a valveless outflow patch reconstruction, the option of reoperation and insertion of a valve or valved conduit exists. Pulmonary hypertension and pulmonary vascular

obstructive disease can occur in some patients who have had excessive pulmonary blood flow, as from a Waterston or a Potts shunt. One such patient has reportedly died of pulmonary vascular obstructive disease 8 months after right ventricular outflow tract repair.' In our two operative deaths in the group of 10 patients who had right ventricular outflow tract reconstruction (Cases 1 and 10), there was significant residual right ventricular hypertension; both patients had had Waterston shunts. Two additional patients (Cases 6 and 7) who had a prior Waterston shunt have persistent mild pulmonary hypertension. Fontan operation. The Fontan operation II is an alternative to right ventricular outflow tract reconstruction for patients with a small right ventricle and tricuspid valve anulus.v" 17 The most severe hypoplasia occurs in patients with a deficiency of one or two right ventricular portions." De Leval and colleagues" reported on 11 patients who had the Fontan operation for pulmonary atresia with intact ventricular septum. All patients had a tricuspid diameter below the authors' lower 99% confidence limit for normal, and all were without one or two right ventricular portions. Five of our 10 patients who required the Fontan operation had an incompletely formed right ventricle. However, we also had five patients with a tripartite right ventricle that was inadequate for right ventricular outflow tract reconstruction. Three had inadequate growth despite initial palliative procedures to open the right ventricular outflow tract (Tables V and VI). The techniques used at our institution to establish atriopulmonary artery continuity have been discussed in previous publications.P'P'" Because of our early experience with accelerated deterioration of the glutaraldehyde-preserved porcine valve after the Fontan operation," none of the 10 patients in this study had an inferior vena caval valve or a valved conduit placed in the atriopulmonary position. The operative mortality for the Fontan operation in this series (one of 10 patients) compares well with the mortality rate for right ventricular outflow tract reconstruction (two of 10 patients). Six surviving patients have had follow-up for 10 months or more and are active and improved from their preoperative condition. Conclusions Our total operative mortality in definitive operation for pulmonary atresia with intact ventricular septum was 15% (three of 20 patients), with one late death. Determining the presence or absence of an infundibulum is helpful in selecting the type of initial palliative procedure, especially in identifying those patients for

Volume 93 Number 3 March 1987

pulmonary valvotomy. The performance of a preliminary pulmonary valvotomy or patch reconstruction of the right ventricular outflow tract, without closure of the atrial septal defect, allows some ventricles to grow at a rate greater than somatic growth, thus increasing the patient's chances of being a candidate for subsequent complete repair. The approach to definitive repair is based primarily on the actual size of the tricuspid valve anulus, which is directly proportional to the size of the right ventricular cavity. Right ventricular outflow tract reconstruction can be done as a complete repair for patients who have adequate tricuspid annular and right ventricular cavity size. These patients generally have tricuspid anuli greater than 70% of normal and have the three right ventricular portions present. For patients with borderline tricuspid anuli (55% to 70% of normal), right ventricular outflow tract reconstruction may be performed in some. In other patients the Fontan operation will be required. The modified Fontan operation is the only option when the tricuspid valve anulus and the right ventricular cavity are severely hypoplastic (tricuspid anulus less than 55% of normal). This may occur not only in patients with a deficiency of one or two right ventricular portions but also in patients with a tripartite right ventricle that did not grow despite the initial palliative procedure to open the outflow tract. REFERENCES I. Bowman FO Jr, Maim JR, Hayes CJ, Gersony WM, Ellis K. Pulmonary atresia with intact ventricular septum. J THORAC CARDIOVASC SURG 1971;61:85-93. 2. Rigby ML, Silove ED, Astley R, Abrams LD. Pulmonary atresia with intact ventricular septum: open heart surgical correction at 32 hours. Br Heart J 1977;39:573-6. 3. Moulton AL, Bowman FO Jr, Edie RN, et al. Pulmonary atresia with intact ventricular septum: sixteen-year experience. J THORAC CARDIOVASC SURG 1979;78:527-35. 4. Weldon CS, Hartmann AF Jr, McKnight RC. Surgical management of hypoplastic right ventricle with pulmonary atresia or critical pulmonary stenosis and intact ventricular septum. Ann Thorac Surg 1984;37:12-24. 5. de Leval M, Bull C, Stark J, Anderson RH, Taylor JFN, Macartney FJ. Pulmonary atresia and intact ventricular septum: surgical management based on a revised classification. Circulation 1982;66:272-80. 6. de Leval M, Bull C, Hopkins R, et al. Decision making in the definitive repair of the heart with a small right ventricle. Circulation 1985;72(suppl 2):52-60. 7. 000011 ARC, Grignon A. Early and late results in pulmonary atresia. Ann Thorac Surg 1977;24:264-72. 8. Olley PM, Coceani F, Bodach E. E-type prostaglandins: a new emergency therapy for certain cyanotic congenital heart malformations. Circulation 1976;53:728-31. 9. Sade RM, Cosgrove DM, Castaneda AR. Infant and

Pulmonary atresia with intact ventricular septum 4 6 3

child care in heart surgery: clinical manual of the Department of Cardiovascular Surgery, Children's Hospital Medical Center, Boston. Chicago: Year Book Medical Publishers, 1977. 10. Lewis AB, Wells W, Lindesmith GG. Evaluation and surgical treatment of pulmonary atresia and intact ventricular septum in infancy. Circulation 1983;67:1318-23. II. Fontan F, Baudet E. Surgical repair of tricuspid atresia. Thorax 1971;26:240-8. 12. Gale AW, Danielson GK, McGoon DC, Mair DD. Modified Fontan operation for univentricular heart and complicated congenital lesions. J THORAC CARDIOVASC SURG 1979;78:831-8. 13. Choussat A, Fontan F, Besse P, Vallot F, Chauve A, Bricaud H. Selection criteria for Fontan's procedure. In: Anderson RH, Shinebourne EA, eds. Pediatric cardiology 1977. London: Churchill Livingstone, 1978:559-66. 14. Goor OA, Lillehei CWo Congenital malformations of the heart: embryology, anatomy, and operative considerations. New York: Grune & Stratton, 1975:11-4. 15. Bull C, de Leval MR, Mercanti C, Macartney FJ, Anderson RH. Pulmonary atresia and intact ventricular septum: a revised classification. Circulation 1982;66:26672. 16. Rowlatt UF, Rimoldi HJA, Lev M. The quantitative anatomy of the normal child's heart. Pediatr Clin North Am 1963;10:499-588. 17. Patel RG, Freedom RM, Moes CAF, et al. Right ventricular volume determinations in 18 patients with pulmonary atresia and intact ventricular septum: analysis of factors influencing right ventricular growth. Circulation 1980;61:428-40. 18. Graham TP Jr, Bender HW, Atwood GF, Page DL, Sell CGR. Increase in right ventricular volume following valvulotomy for pulmonary atresia or stenosis with intact ventricular septum. Circulation 1974;50(Pt 2):1169-78. 19. Gersony WM, Bernhard WF, Nadas AS, Gross RE. Diagnosis and surgical treatment of infants with critical pulmonary outflow obstruction: study of thirty-four infants with pulmonary stenosis or atresia, and intact ventricular septum. Circulation 1967;35:765-76. 20. Trusler GA, Freedom RM, Patel R, Williams WG. The surgical management of pulmonary atresia with intact ventricular septum. Pediatr Cardiol 1979;2:305-13. 21. Freedom RM, Wilson G, Trusler GA, Williams WG, Rowe RD. Pulmonary atresia and intact ventricular septum: a review of the anatomy, myocardium, and factors influencing right ventricular growth and guidelines for surgical intervention. Scand J Thorac Cardiovasc Surg 1983;17:1-28. 22. O'Connor WN, Cottrill CM, Johnson GL, Noonan JA, Todd EP. Pulmonary atresia with intact ventricular septum and ventriculocoronary communications: surgical significance. Circulation 1982;65:805-9. 23. Sauer U, Bindl L, PilossoffV, et al. Pulmonary atresia with intact ventricular septum and right ventricle-coronary artery fistulae: selection of patients for surgery. In: Doyle

464

Alboliras et al.

EF, Engle MA, Gersony WM, Rashkind WV, Talner NS, eds. Pediatric cardiology: Proceedings of the Second World Congress, New York: Springer-Verlag, 1986:56678. 24. Uretzky G, Puga FJ, Danielson GK, Hagler DJ, McGoon DC. Reoperation after correction of tetralogy of Fallot. Circulation 1982;66(Pt 2):1202-8. 25. Uretzky G, Puga FJ, Danielson GK. Modified Fontan procedure in patients with previous ascending aorta-

The Journal of Thoracic and Cardiovascular Surgery

pulmonary artery anastomosis. J THORAC CARDIOVASC SURG 1983;85:447-50. 26. Gale AW, Danielson GK, McGoon DC, Wallace RB, Mair DD. Fontan procedure for tricuspid atresia. Circulation 1980;62:91-6. 27. Danielson GK:· The modified Fontan procedure. In: Moulton AL, ed. Congenital heart surgery: current techniques and controversies. Pasadena, Calif.: Appleton Davies, 1984;317-29.