Right ventricle-pulmonary artery shunt in first-stage palliation of hypoplastic left heart syndrome

Right Ventricle-Pulmonary Artery Shunt in First-Stage Palliation of Hypoplastic Left Heart Syndrome Shunji Sano, Kozo Ishino, Masaaki Kawada, and Osami Honjo This article reviews our experience using a prosthetic conduit between the right ventricle (RV) and the pulmonary artery (PA), in lieu of the more traditional aortopulmonary shunt, for infants undergoing surgical palliation of hypoplastic left heart syndrome. Thirty-three consecutive infants underwent Norwood procedure between February 1998 and November 2003, using an RV-PA conduit. There were 31 hospital survivors (94%) and 25 patients have undergone a stage II bidirectional Glenn anastomosis with an additional two deaths. Nine patients have undergone completion Fontan. This technique provides reproducible results, simplifies postoperative management, and improves outcome, especially for “low volume” programs. © 2004 Elsevier Inc. All rights reserved. Key words: Hypoplastic left heart syndrome (HLHS), Norwood procedure, Sano procedure.

H

ypoplastic left heart syndrome (HLHS) is a spectrum of underdevelopment of the left-sided heart structure, characterized by aortic valve atresia or severe stenosis with a small ascending aorta and hypoplasia of the left ventricle.1 Without surgical treatment, practically all infants with this complex cardiac anomaly die within the first month of life.2,3 Since Norwood et al4 reported successful surgical palliation in 1981, multistaged reconstructive surgery involving the Norwood procedure (stage I) followed by bidirectional Glenn (BDG; stage II) and Fontan operations (stage III) has been the established treatment for infants with HLHS. Despite successful aortic reconstruction, most deaths occur in the first 24 to 48 hours after surgery because of hemodynamic instability secondary to the unpredictable rapid fall in pulmonary vascular resistance.2,3 Maldistribution of cardiac output associated with the systemic-topulmonary shunt has been implicated as a major From the Department of Cardiovascular Surgery, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan. Address reprint requests to Shunji Sano, MD, Department of Cardiovascular Surgery, Okayama University Graduate School of Medicine and Dentistry, 2-5-1 Shikata-cho, Okayama-City 700-8558, Japan. © 2004 Elsevier Inc. All rights reserved. 1092-9126/04/0701-0006$30.00/0 doi:10.1053/j.pcsu.2004.02.023

22

cause of early death after stage I palliation.5,6 A review of 122 postmortem cases after the Norwood procedure at the Children’s Hospital Boston7 indicated that the most important causes of death were impairment of coronary perfusion, excessive pulmonary blood flow, and obstruction to pulmonary blood flow. This maldistribution of blood flow was implicated in more than 60% of postoperative mortality. Over the last decade, therefore, many efforts to achieve a balanced circulation have focused on limiting pulmonary blood flow and increasing systemic oxygen delivery. These measures have included reduction in the size of the shunt,6,8 use of systemic vasodilators,9,10 and induction of hypoxemia and hypercarbia by ventilator manipulations6,10-12 and/or hypoxic admixtures.9,11 However, such delicate postoperative management requires a great deal of manpower on a 24-hour basis. The recent analysis of a database of 1,986 neonates with HLHS treated between 1988 and 1997 in the United States13 showed that the proportion of patients treated with the Norwood procedure increased over time while the use of comfort care, still the most common treatment (⬎50%), decreased. Although several institutions with a large surgical volume have achieved early survival following the Norwood procedure (stage I palliation) between 68% and 94%,6,8,10,11 this procedure has remained a challenging step with unacceptably high mortality

Pediatric Cardiac Surgery Annual of the Seminars in Thoracic and Cardiovascular Surgery, Vol 7, 2004: pp 22-31

RV-PA Shunt in First-Stage Palliation of HLHS

for many less-experienced centers.14,15 Particularly in Japan, only 200 neonates underwent stage I palliation for HLHS between 1998 and 2000, with hospital mortality of constantly higher than 60% (annual survey by The Japanese Association for Thoracic Surgery).

History of Right VentriclePulmonary Artery Shunt in FirstStage Palliation of Hypoplastic Left Heart Syndrome The purpose of the stage I palliation for HLHS is to provide unobstructed systemic blood flow from the right ventricle to the aorta and coronary arteries, to re-establish a pulmonary blood supply, and to relieve obstruction to pulmonary venous return through the atrial septum. The right ventricle-pulmonary artery (RV-PA) shunt to reestablish pulmonary blood supply in stage I palliation for HLHS was first introduced by Norwood et al in 1981.4 The shunt materials they used were relatively large for neonates; 8 mm nonvalved polytetrafluroethylene (PTFE) tubes in two patients and 12 mm valved conduits in another two. These patients all died within 11 hours after surgery either from excessive pulmonary blood flow or from right ventricular failure. In Japan, Kishimoto et al16 revived the RV-PA shunt using a xenopericardial valved conduit in seven patients and reported stable postoperative hemodynamics with a high diastolic blood pressure. However, their long-term results were unsatisfactory and no other surgeons had used the procedure. The findings of Kishimoto et al16 stimulated us to attempt a RV-PA shunt, but with a small nonvalved PTFE conduit in first-stage palliation. We constructed a nonvalved PTFE shunt between the right ventricle and the pulmonary artery (RV-PA shunt) as an alternative pulmonary blood source. This first attempt was applied to as an 8-day-old girl with aortic atresia, mitral atresia, and hypoplasia of ascending aorta using 4 mm PTFE graft on February 9, 1998. This patient survived the procedure, however, she developed sudden severe cyanosis 4 months after surgery and died. A second attempt was performed on February 15,1998 on a 17-day-old boy with aortic atresia and mitral atresia. He underwent second-

23

stage BDG anastomosis at 4 months of age and successfully completed third-stage Fontan operation at 2.8 years of age. Subsequently, we have used the RV-PA shunt on all the patients with HLHS, although using a small nonvalved PTFE conduit. Our results were first reported at the annual meeting of the Japanese Association of Cardiovascular Surgery in 1999.17 The first description of our technique at a major international meeting was at the 3rd World Congress of Pediatric Cardiology and Cardiac Surgery in Toronto in 2001,18 followed by the 43rd Annual Meeting of American Association for Thoracic Surgery in 2002.19,20

Okayama Experience With Right Ventricle-Pulmonary Artery Shunt in a First-Stage Palliation of Hypoplastic Left Heart Syndrome The results of the present series have clearly shown differences in the postoperative hemodynamics between the RV-PA shunt and the systemic-to-pulmonary shunt. Without requiring delicate manipulations of the systemic and pulmonary vascular resistance, the RV-PA shunt (using a nonvalved PTFE conduit) provides not only adequate pulmonary blood flow but also stable systemic circulation after stage I reconstruction. In this article we describe our developing surgical techniques and 5 years’ institutional experience with a modified Norwood procedure using an RV-PA shunt as stage I palliation for HLHS. We also analyzed intermediate outcome for the 33 infants with this modification and the results were also compared with those for the preceding patients who had undergone the classic Norwood operation using an systemic-pulmonary (SP) shunt.

Patients and Methods Patient Population Between February 1998 and November 2003, 33 consecutive infants (20 boys, 13 girls) with HLHS or a variant underwent a modified Norwood procedure using a RV-PA shunt at Okayama University Hospital (Okayama, Japan). The diagnosis of HLHS was based on 2-dimensional echocardiographic or angiographic findings. Twenty-five pa-

Sano et al

24

tients had classic HLHS, including aortic atresia or stenosis, mitral atresia, or stenosis with a poorly developed left ventricle, normally related great arteries, and an intact ventricular septum. The remaining eight patients had variants of HLHS with left ventricular and aortic arch hypoplasia; four had a double-outlet right ventricle with aortic and subaortic stenoses, two had heterotaxy syndrome (asplenia) with aortic atresia, one had aortic and mitral stenoses with a ventricular septal defect, and one had aortic atresia with type-B interrupted aortic arch and a ventricular septal defect. Associated anomalies included a persistent left superior vena cava in six patients, vascular sling in one, total anomalous pulmonary venous connection in one, and an aberrant right subclavian artery in one. The diameters of the ascending aorta ranged from 1.6 to 7.6 mm (mean, 3.8 mm; median, 3.0 mm) and were 2 mm or less in five patients (15%). Right ventricular fractional shortening ranged from 19% to 43% (mean, 33%; median, 32%). Tricuspid regurgitation documented by color flow Doppler echocardiography was absent or trivial in six patients, mild in 16, moderate in nine, and severe in two. Age at operation ranged from 3 to 57 days (mean, 14 days; median, 10 days). Weight at operation ranged from 1.6 to 3.9 kg (mean, 2.8 kg; median, 2.8 kg) and five patients (15%) weighed less than 2.0 kg. Twenty-three patients underwent mechanical ventilation before the operation, and all but one patient received an infusion of prostaglandin E1. One patient required peritoneal dialysis because of acute renal failure secondary to ductal shock.

Operative Procedure A detailed procedure for the surgery has been reported elsewhere.20 Briefly, we describe key points at each step of the modified Norwood procedure with the RV-PA shunt. Cardiopulmonary bypass. To avoid the use of total circulatory arrest as much as possible, cardiopulmonary bypass was established by dual arterial perfusion and single atrial or bicaval drainage cannulations. One perfusion cannula was inserted into a 3.0 mm PTFE tube that was anastomosed to the innominate artery to maintain continuous cerebral perfusion in all patients. The other perfusion cannula was inserted into the descending aorta from the ductus arteriosus.

Figure 1. The entire aortic arch and ascending aorta were reconstructed by direct anastomosis of the proximal main pulmonary artery. A small right ventriculotomy was made in the outflow tract for proximal anastomosis of the RV-PA shunt.

During aortic reconstruction, this cannula was removed and circulation in the lower body was arrested at 20 to 22°C. In one patient who had suffered from renal failure before operation, an additional perfusion cannula was directly inserted into the descending aorta in the lower posterior mediastinum to maintain whole systemic circulation.13 Cardioplegic solution was administered from a side port of the perfusion cannula during temporary total circulatory arrest. Modified ultrafiltration was routinely used following weaning from cardiopulmonary bypass. Aortic reconstruction and atrial septectomy. Because homograft is not available in Japan, aortic reconstruction without patch supplementation6 was the procedure of choice. In all patients, duct tissue was completely excised and a neoaorta was constructed by direct anastomosis of the descending aorta and a proximal main pulmonary artery to an opened tiny native aorta (Fig 1). A small piece of pericardial patch was necessary in some patients. An atrial septectomy was performed during the period of total circulatory

RV-PA Shunt in First-Stage Palliation of HLHS

25

tients weighing less than 2 kg received the 4-mm shunt.

Other Techniques Deep hypothermic total circulatory arrest has been commonly used during aortic reconstruction in stage I palliation. Even surgeons at institutions with high surgical volume required the mean durations of circulatory arrest of 44 to 57 minutes.6,10,21 It is obvious that less experienced surgeons need longer circulatory arrest, predisposing critically ill infants to neurologic deficits and multiorgan failure. Continuous cerebral perfusion with or without descending aortic perfusion used in many centers20,22 minimized or avoided the use of total circulatory arrest and, at least in part, contributed to the improved outcome of stage I palliation for HLHS.

Postoperative Management Figure 2. After completion of aortic reconstruction and atrial septectomy, proximal anastomosis of the RV-PA shunt was performed with the heart beating. arrest for infusion of cardioplegic solution. One patient underwent repair of the common atrioventricular valve during the Norwood procedure. Right-ventricle-pulmonary artery shunt. Before establishment of cardiopulmonary bypass, the shunt material was prepared. A 5-mm PTFE tube was cut in an appropriate length as a shunt tube. A cuff for an anastomosis to the distal end of the main pulmonary artery was tailored by vertically opening a piece of the same PTFE tube. The center of the cuff was punched out in 5-mm diameter, and the tube was anastomosed to this opening. During the cooling phase, the main pulmonary artery was transected just proximal to the bifurcation, and the PTFE cuff was anastomosed to the distal stump of the main pulmonary artery. For the proximal anastomosis of the shunt, a small right ventriculotomy was made in the right ventricular outflow tract 1 to 2 cm below the pulmonary valve. To prevent late obstruction at the shunt anastomosis, it is of great importance to slice off a piece of ventricular muscle underlying the ventriculotomy. The size of the shunt used was 4 mm in six patients, 5 mm in 26, and 6 mm in one (Fig 2). Three of five pa-

The sternum was routinely left open and patients underwent delayed sternal closure on postoperative day 1 to 11 (median, 3 days). Recently, the sternum was primarily closed in selected patients. Postoperative intensive care of patients with the RV-PA shunt was basically the same as that for neonates undergoing other types of surgery. Thus, delicate manipulations to control pulmonary and systemic vascular resistance were not necessary. Ventilator settings were adjusted to keep arterial oxygen saturations higher than 75% and carbon dioxide levels lower than 45 mm Hg. It shoud be kept in mind that hypercarbia, which is often used to increase pulmonary vascular resistance in patients with the systemic-pulmonary shunt, easily causes hypoxemia and subsequent hemodynamic deterioration in those with the RV-PA shunt. The inotropic drugs most often used were dopamine or dobutamine 5 ␮g/kg per minute and epinephrine 0.05 to 0.1 ␮g/kg per minute; if necessary, calcium chloride 0.25 mmol/h was also used.

Stage II Palliation (Bidirectional Glenn) Stage II BDG anastomosis was performed with a partial cardiopulmonary bypass or with a temporary bypass between the superior vena cava and the right atrium using a centrifugal pump. The RV-PA shunt at this stage was left open in all but one patient who had an interrupted inferior vena cava with hemiazygos continuation.

Sano et al

26

Stage III Palliation (Fontan) Stage III Fontan was performed by a lateral tunnel technique with a right atrial roll14,15 or by an extracardiac technique, and the RV-PA shunt was transected. Modified ultrafiltration was routinely performed for 10 to 15 minutes after termination of cardiopulmonary bypass at each operation.

Echocardiographic Assessment and Late Management To evaluate diastolic reverse flow and the pressure gradient across the nonvalved RV-PA shunt, sequential Doppler echocardiography was performed in all patients postoperatively at 1 and 4 months before stage II palliation. A spectral time-velocity display was obtained at the middle point of the RV-PA shunt. The velocity-time integrals of reverse and forward flow were measured by tracing the curves above and below the spectral baseline, respectively. The velocity-time integrals of the diastolic reverse flow were divided by the velocity-time integrals of systolic forward flow for calculation of the Doppler flow reversal ratio. Elective cardiac catheterization was carried out 3 to 4 months after stage I palliation.

Evaluation of Ventricular Function and Pulmonary Artery Growth Ventricular function was evaluated by 2-dimensional echocardiography before each operation. Right ventricular fractional shortening, degree of tricuspid regurgitation, right ventricular end-diastolic diameter, and tricuspid valve annulus diameter were measured. Pulmonary artery sizes were measured at cardiac catheterization before BDG and Fontan procedures. Measurements of the right and left pulmonary artery diameters were made just proximal to the takeoff of the branching vessels in the anteroposterior projection, and each pulmonary artery diameter was indexed to the patient’s body surface area (PA diameter/m2).23 Right-to-left pulmonary artery diameter ratio was calculated by dividing indexed right pulmonary artery diameter by indexed left pulmonary artery diameter. The pulmonary artery index (Nakata) was calculated as the summation of the right and left pulmonary artery cross-sectional areas indexed to the patient’s body surface area.24 All parameters mea-

sured in 25 patients who underwent stage II palliation and in nine patients who subsequently underwent Fontan operation were compared with those in the preceding five patients who undergone Fontan operation after the classic Norwood procedure with an SP shunt.

Results Stage I Norwood Procedure The mean cardiopulmonary bypass time was 158 minutes (range, 119 to 254 minutes), the mean myocardial ischemic time was 45 minutes (range, 27 to 94 minutes), and the mean renal ischemic time was 48 minutes (range, 0 to 99 minutes). The mean total circulatory arrest time in 21 patients was 9 minutes (range, 3 to 28 minutes). All patients were weaned from cardiopulmonary bypass. Without any particular ventilatory manipulation, hemodynamic instability never occurred in any of the 26 patients. There were 31 hospital survivors (94%), including five patients weighing less than 2 kg. One patient died from sudden cardiac arrest on the next day, and the other from septicemia after 2 weeks. There were two late deaths among the first four patients. Both died from severe hypoxemia caused by obstruction of the RV-PA shunt (3 and 4 months after surgery). Elective cardiac catheterization was performed in all but one survivor (N ⫽ 25) at a median age of 4 months before stage II. Balloon angioplasty was necessary for stenosis at proximal and/or distal anastomotic sites of the RV-PA shunt in six patients, for recoarctation of the aorta in three, and for a restrictive atrial septal defect in one.

Stage II BDG Anastomosis Twenty-five patients underwent BDG at a mean age of 5.7 months (median, 5.6 months; range, 1.1 to 11.3 months). Nine patients needed instrumental dilation of stenosis for branch pulmonary arteries. Concomitant procedures included repair of total anomalous pulmonary venous connection associated with heterotaxy syndrome in one patient, tricuspid valve repair in one, and enlargement of an intraatrial communication in one. Two of the six patients with persistent left superior vena cava underwent a staged bilateral BDG: a right BDG through a median sternotomy and a left BDG through a lateral thoracotomy after 1

RV-PA Shunt in First-Stage Palliation of HLHS

27

Figure 3. Current status of all hospital survivors after stage I reconstruction of hypoplastic left heart syndrome.

and 3 months, respectively. One patient received a systemic-to-pulmonary shunt 10 months after surgery for possible future biventricular repair. There were two hospital deaths: one from viral pneumonia and the other from progressive hypoxia, and no late deaths.

Stage III Fontan Operation In nine patients, elective cardiac catheterization was performed at a median age of 1.8 years. At echocardiographic study before stage III, the RV-PA shunt flow was detected in five patients and was not detected in four. The Fontan operation has been performed at a median age of 2.5 years (range, 1.8 to 3.2 years). Concomitant procedures included tricuspid valve replacement because of severe tricuspid regurgitation in one patient, and patch augmentation of the pulmonary artery in two. One patient who had aortic atresia with interrupted aortic arch and a ventricular septal defect underwent successful biventricular repair. There were no hospital or late mortalities.

Follow-Up The current status of all hospital survivors after stage I palliation is shown in Fig 3. Median follow-up period for 27 survivors was 20 months (mean, 23 months; range, 5 to 66 months). One patient underwent reoperation for complete occlusion of the right pulmonary artery after Fontan operation. At the latest echocardiographic evaluation, the mean right ventricular fractional

shortening of the 27 survivors was 33% (range, 28% to 43%), and tricuspid or common atrioventricular valve regurgitation was absent or trivial in nine patients, mild in 13, and moderate in five. Actuarial survival for the entire cohort of patients was 82% (95% confidence interval: 65% to 97%) at 1 year and 80% (95% confidence interval: 65% to 95%) at 2 years (Fig 2). Echocardiographic assessment and late management. Mean Doppler flow reversal ratio in the RV-PA shunt decreased from 0.28 ⫾ 0.03 at postoperative 1 month to 0.16⫾0.02 at 4 months (n ⫽ 7; P ⫽ .027), while the mean pressure gradient across the shunt increased from 34 ⫾ 10 mm Hg to 58 ⫾ 14 mm Hg (P ⫽ .018). At the same points of time, the mean transcutaneous oxygen saturation decreased from 82% ⫾ 7% to 70% ⫾ 8% (P ⫽ .018). Cardiac catheterization and cineangiography before stage II palliation showed that the mean pulmonary artery pressure was 11 ⫾ 2 mm Hg (n ⫽ 13; range, 9 to 15). One of the concerns of this technique was related to the degree of flow reversal in the nonvalved RV-PA shunt. Doppler echocardiography demonstrated the greatest reversal flow ratio within 1 month after surgery, but all patients tolerated this volume overload well, maintaining reasonable oxygen saturation levels. Evaluation of ventricular function and pulmonary artery growth: Comparison with the classic Norwood. Parameters evaluated by echocardiography and cardiac catheterization were listed

Sano et al

28

Table 1. Comparison of Hemodynamics Before Stage II and Stage III Before Stage II

Variable

Before Stage III

RV-PA RV-PA BT (N ⫽ 5) (N ⫽ 25*) P value BT (N ⫽ 5) (N ⫽ 9*) P value (mean value (mean value (Mann-Whitney’s (mean value (mean value (Mann-Whitney’s ⫾ SD) ⫾ SD) Test) ⫾ SD) ⫾ SD) Test)

RVEDP (mm Hg) 10 ⫾ 5 Ao-Sat (%) 77 ⫾ 4 Qp:Qs 0.88 ⫾ 0.1 RPA/m2 27 ⫾ 7 LPA/m2 21 ⫾ 8 Rt/Lt ratio 1.3 ⫾ 0.2 Nakata index 335 ⫾ 200 RVFS (%) 28 ⫾ 2 RVEDd (mm) 32 ⫾ 5

6⫾3 72 ⫾ 5 0.66 ⫾ 0.2 21 ⫾ 4 20 ⫾ 6 1.1 ⫾ 0.3 201 ⫾ 86 35 ⫾ 7 26 ⫾ 5

.12 .126 .008* .196 .785 .083 .103 .018* 0.063

7⫾2 82 ⫾ 5 0.56 ⫾ 0.2 22 ⫾ 1 14 ⫾ 3 1.7 ⫾ 0.4 232 ⫾ 28 32.9 ⫾ 2.7 34 ⫾ 7

8⫾3 82 ⫾ 3 0.76 ⫾ 0.2 18 ⫾ 3 15 ⫾ 4 1.3 ⫾ 0.4 209 ⫾ 44 32.5 ⫾ 5.1 34 ⫾ 3.3

.610 .806 .086 .011* .954 .395 .395 .806 .713

Abbreviations: RVEDP, right ventricular end-diastolic pressure; Ao-Sat, aorta oxygen saturation; Qp:Qs, pulmonary-tosystemic blood flow ratio; Rt/Lt ratio, indexed right-to-left pulmonary artery diameter ratio; RPA/m2, indexed right pulmo nary artery diameter; LPA/m2, indexed left pulmonary artery diameter; RVFS, right ventricular fractional shortening; RVEDd, right ventricular end-diastolic diameter. *One patient who underwent systemic-to-pulmonary shunt and subsequent biventricular repair was excluded.

in Table 1. In patients with RV-PA shunt, pulmonary-to-systemic blood flow ratio (Qp/Os) was lower before stage II, and was slightly higher before stage III, probably because of an additional blood flow through the RV-PA shunt. Right ventricular end-diastolic pressure tended to be lower in patients with the RV-PA shunt before stage II. There was a trend toward a smaller right ventricular end-diastolic diameter and tricuspid valve annulus in patients with the RV-PA shunt before stage II. Right ventricular fractional shortening was higher in patients with the RV-PA shunt before stage II, but did not differ between the groups before stage III. Although the Nakata index and the indexed diameters of branch pulmonary arteries were relatively smaller in patients with the RV-PA shunt before stage II, there was no difference in pulmonary artery growth between the groups before stage III. The right-to-left pulmonary artery diameter ratio of the branch pulmonary arteries was slightly higher in the Blalock-Taussig group before stage II, while there was no difference in the groups before stage III (Table1).

Discussion Coronary hypoperfusion caused by a diastolic runoff through the SP shunt has been proposed to be a cause of late mortality after the Norwood operation.24-27 The fundamental advantages of

RV-PA shunt were that elimination of diastolic run-off into the pulmonary circulation resulted in high diastolic pressure which was constantly over 40 mm Hg throughout the entire postoperative period, and coronary perfusion was independent of the pulmonary-to-systemic flow ratio. Independent coronary perfusion resulted in a very stable postoperative course.26 The volume load is directly associated with right ventricular performance in patients with HLHS following stage I palliation. A single right ventricle with a volume load imposed by an SP shunt showed significant increase in ventricular end-diastolic volume and ventricular mass, but decrease in ejection fraction.28 Furthermore, persistent volume loading to the right ventricle might also be associated with interim mortality29 in patients with HLHS following stage I with the SP shunt. In addition, systemic tricuspid valve is exquisitely sensitive to volume loading of the right ventricle, especially when required to eject at systemic arterial pressure.30 Our results showed that the patients with the RV-PA shunt had significantly lower Qp/Qs associated with smaller right ventricular end-diastolic diameter and tricuspid valve annulus than those in patients with SP shunt. The decrease in volume load on the right ventricle associated with smaller ventricular dimension and tricuspid annulus may be advantageous to prevent progres-

RV-PA Shunt in First-Stage Palliation of HLHS

sion of congestive heart failure and tricuspid regurgitation, and consequently may contribute to reduce interim mortality. Therefore, we believe that the RV-PA shunt provides physiologic coronary circulation and proper volume load on the right ventricle, thus leading to improved survival of children with HLHS. The effects of a right ventriculotomy for the placement of the RV-PA shunt on the systemic right ventricular function are of great concern. The necessity for a right ventriculotomy, which may affect contractile function of the systemic ventricle, is a potential disadvantage of the RV-PA shunt. Norwood et al4 attempted to use the RV-PA shunt in stage I palliation with relatively large materials (such as an 8-mm nonvalved PTFE tube or 12-mm valved conduit) but all died because of excessive pulmonary blood flow or right ventricular failure. The same findings were observed in Kishimoto’s report.16 Thus, a large ventriculotomy in neonates certainly impairs right ventricular function.31 Another concern related to a ventriculotomy is the complication of ventricular arrhythmia. Our results indicate that a small right ventriculotomy made in the upper outflow tract for an anastomosis with a 4 to 6 mm conduit exerts little influence on ventricular function and does not become arrhythmogenic at least until Fontan completion32; however, long-term follow-up should be necessary. Previous studies have shown that extensive manipulation of the pulmonary arteries and shunt placement at stage I palliation33 and compression of left pulmonary artery by a neoaortic arch8,23 are able to alter the symmetric growth of the pulmonary arteries and result in pulmonary arteries distortion. Alboliras et al33 suggested that patients with a central shunt have more symmetric growth of pulmonary artery rather than that of patients with a SP shunt following stage I palliation, even in the late postoperative period. Because the RV-PA shunt was inserted in the central portion of the pulmonary arteries, it is hypothesized that the RV-PA shunt may provide a symmetric growth of pulmonary artery. Despite the possibility of compression of left pulmonary artery by the neoaorta, particularly those constructed by a direct anastomosis without patch supplementation, our results showed that RV-PA group had well-balanced pulmonary ar-

29

tery configuration. Because of frequent occurrence of stenosis in distal anastomotic site,20,31 it is important to relieve the stenosis at stage II for future pulmonary artery growth. Although a pulsatile blood flow from an RV-PA shunt to pulmonary arteries seems to improve the pulmonary artery growth before stage II, our results showed that pulmonary artery growth was limited and the Nakata index before stage II was lower in patients with the RV-PA shunt. Limited growth of the pulmonary vasculature is one of important disadvantages of the RV-PA shunt, which results from several reasons: first, the Qp/Qs provided by the RV-PA shunt was certainly lower than that in patients with SP shunt.9-12 Second, a nonvalved RV-PA shunt has some flow reversal to the right ventricle,9 which causes decrease in effective pulmonary blood flow. Finally, the RV-PA shunt becomes obstructive over time.20,31 Because of these hemodynamic characteristics of the RV-PA shunt, pulmonary artery growth was not sufficient before stage II. However, our results showed that there was slight increase in the Nakata index before stage III in patients with the RV-PA shunt, while there was significant reduction in the Nakata index in patients with SP shunt. Our results showed no evidence of adverse effects of the additional pulmonary blood flow after BDG, and indicated that the additional pulmonary blood flow through the stenotic RV-PA shunt, in part, plays a role to pulmonary artery growth during interstage before Fontan operation. The results of analysis assessing the effects of pulmonary artery size on clinical outcome after Fontan operation have been conflicting.34,35 However, lower pulmonary resistance may be more important factor than pulmonary artery size to establish Fontan circulation. In patients with the RV-PA shunt, the pulmonary vascular bed may be protected from excessive pulmonary blood flow, which results in lower pulmonary resistance than that in patients with SP shunt, and Fontan circulation could be established in patients with HLHS following stage I palliation with an RV-PA shunt, despite the smaller pulmonary artery size. The lessons learned from our 5 years of experience with the modified Norwood procedure using the RV-PA shunt are two-fold. First, the nonvalved PTFE RV-PA shunt becomes obstructive

30

Sano et al

over time, particularly 3 months or more after surgery. Because two patients in the early series died from progressive shunt obstruction after hospital discharge, we check the patients’ saturations every week at an outpatient clinic. When patients present with progressive desaturation, cardiac catheterization should be performed as early as possible. Our current strategy first involves attempting to perform balloon angioplasty for significant stenoses in the RV-PA shunt or branch pulmonary arteries. If angioplasty fails or the patients’ saturations again drop, we then perform the BDG anastomosis. Our results indicate that lower pulmonary vascular resistance makes subsequent Fontan procedures feasible even in patients with small pulmonary arteries. In conclusion, the RV-PA shunt in stage I reconstruction provides stable hemodynamics without requiring extensive postoperative medical intervention. Therefore, we believe that improvement in the survival of infants with HLHS can be accomplished by many less experienced surgeons by applying the modified Norwood procedure with the RV-PA shunt.

References 1. Noonan JA, Nadas AS: The hypoplastic left heart syndrome: An analysis of 101 cases. Pediatr Clin North Am 5:1026-1056, 1958 2. Sinha SN, Rusnak SL, Sommers HM, et al: Hypoplastic left ventricle syndrome: Analysis of thirty autopsy cases in infants with surgical consideration. Am J Cardiol 21:166173, 1968 3. Hawkins JA, Doty DB: Aortic atresia: morphologic characteristics affecting survival and operative palliation. J Thorac Cariovasc Surg 88:620-626, 1984 4. Norwood WI, Lang P, Castaneda AR, et al: Experience with operation for hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 82:511-519, 1981 5. Murdison KA, Baffa JM, Farrell PE Jr, et al: Hypoplastic left heart syndrome: outcome after initial reconstruction and before modified Fontan procedure. Circulation 82:IV199-207, 1990 (suppl 4) 6. Ishino K, Stumper O, De Giovanni JJV, et al: The modified Norwood procedure for hypoplastic left heart syndrome: early to intermediate results of 120 patients with particular reference to aortic arch repair. J Thorac Cardiovasc Surg 117:920-930, 1999 7. Bartram U, Grunenfelder J, Van Praagh R: Causes of death after the modified Norwood procedure: a study of 122 postmortem cases. Ann Thorac Surg 64:1795-1802, 1997 8. Bove EL, Lloyd TR: Staged reconstruction for hypoplastic left heart syndrome. Ann Surg 224:387-395, 1996 9. Jobes DR, Nicolson SC, Steven JM, et al: Carbon dioxide

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

prevents pulmonary overcirculation in hypoplastic left heart syndrome. Ann Thorac Surg 54:150-151, 1991 Tweddell JS, Hoffman GM, Mussatto KA, et al: Improved survival of patients undergoing palliation of hypoplastic left heart syndrome: Lessons learned from 115 consecutive patients. Circulation 106:I-81-I-89, 2002 (suppl 1) Mahle WT, Spray TL, Wernovsky G, et al: Survival after reconstructive surgery for hypoplastic left heart syndrome: A 15-year experience from a single institution. Circulation 102:III-136-III-141, 2000 (suppl 3) Hoffman GM, Ghanayem NS, Kampine JM, et al: Venous saturation and the anaerobic threshold in neonates after the Norwood procedure for hypoplastic left heart syndrome. Ann Thorac Surg 70:1515-1521, 2000 Chang RR, Chen AY, Klitzner TS: Clinical management of infants with hypoplastic left heart syndrome in the United States, 1988-1997. Pediatrics 110:292-298, 2002 Kern JH, Hayes CJ, Michler RE, et al: Survival and risk factor analysis for the Norwood procedure for hypoplastic left heart syndrome. Am J Cardiol 80:170-174, 1997 Jacobs ML, Blackstone EH, Bailey LL: The Congenital Heart Surgeons Society. Intermediate survival in neonates with aortic atresia: A multi-institutional study. J Thorac Cardiovasc Surg 116:417-431, 1998 Kishimoto H, Kawahira Y, Kawata H, et al: The modified Norwood palliation on a beating heart. J Thorac Cardiovasc Surg 118:1130-1132, 1999 Sano S, Kawada M, Kino K, et al: Modified Norwood procedure with the use of RV-PA shunt in first stage palliation of hypoplastic left heart syndrome. Presented at the Twenty-ninth Annual Meeting of the Japanese Association for Cardiovascular Surgery Sano S, Ishino K, Kawada M, et al: The modified Norwood operation for hypoplastic left heart syndrome: Using right ventricle-to-pulmonary artery shunt. Cardiol Young 11: 21, 2001 (suppl 1) Sano S, Ishino K, Kawada M, et al: Right ventriclepulmonary artery shunt in first-stage palliation of hypoplastic left heart syndrome. Presented at the Eightysecond Annul Meeting of the American Association for Thoracic Surgery, Boston, MA, May 2003 Sano S, Ishino K, Kawada M, et al: Right ventricle-topulmonary artery shunt in first-stage palliation of hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 126:504-510, 2003 Gaynor JW, Mahle WT, Cohen MI, et al: Risk factors for mortality after the Norwood procedure. Eur J Cardiothorac Surg 22:82-89, 2002 Imoto Y, Kado H, Shiokawa Y, et al: Experience with the Norwood procedure without circulatory arrest. J Thorac Cardiovasc Surg 122:879-882, 2001 Weber HS, Myers JL: Association of asymmetric pulmonary artery growth following palliative surgery for hypoplastic left heart syndrome with ductal coarctation, neoaortic arch compression, and shunt-induced pulmonary artery stenosis. Am J Cardiol 91:1503-1506, 2003 Nakata S, Imai Y, Takanashi Y, et al: A new method for quantitative standardization of cross-sectional areas of the pulmonary arteries in congenital heart diseases with decreased pulmonary blood flow. J Thorac Cardiovasc Surg 88:610-619, 1984

RV-PA Shunt in First-Stage Palliation of HLHS

25. Pizarro C, Malec E, Maher KO, et al: Right ventricle to pulmonary artery conduit improves outcome after stage I Norwood for hypoplastic left heart syndrome. Circulation 108:II-155-II-160, 2003 (suppl 2) 26. Malec E, Januszewska K, Kolcz J, et al: Right ventricleto-pulmonary artery shunt versus modified Blalock-Taussig shunt in the Norwood procedure for hypoplastic left heart syndrome – Influence on early and late hemodynamic status. Eur J Cardiothorac Surg 23:728-734, 2003 27. Mair R, Tulzer G, Sames E, et al: Right ventricular to pulmonary artery conduit instead of modified BlalockTaussig shunt improves postoperative hemodynamics in newborns after the Norwood operation. J Thorac Cardiovasc Surg 126:1378-1384, 2003 28. Kuroda O, Sano T, Matsuda H, et al: Analysis of the effects of the Blalock-Taussig shunt on ventricular function and the prognosis in patients with single ventricle. Circulation 76:III-24-28, 1987 29. Altmann K, Printz BF, Solowiejczyk DE, et al: Two-dimensional echocardiographic assessment of right ventricular function as a predictor of outcome in hypoplastic left heart syndrome. Am J Cardiol 86:964-968, 2000

31

30. Karl T: Staged reconstruction for HLHS: the bidirectional cavopulmonary shunt, in Rychik J, (ed): Hypoplastic Left Heart Syndrome. Boston, MA, Kluwer, 2003, pp 129-148 31. Sano S, Ishino K, Kado H, et al: Outcome of right ventricle-to-pulmonary artery shunt in first stage palliation of hypoplastic left heart syndrome: A multi-institutional study. Ann Thorac Surg (in press) 32. Sano S, Ishino K, Kawada M, et al: Five-year experience with RV-PA shunt in first-stage reconstruction of hypoplastic left heart syndrome. Cardiol Young (in print) 33. Alboliras ET, Chin AJ, Barber G, et al: Pulmonary artery configuration after palliative operation for hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 97:878885, 1989 34. Reddy VM, McElhinney DB, Moore P, et al: Pulmonary artery growth after bidirectional cavopulmonary shunt: is there a cause for concern? J Thorac Cardiovasc Surg 112:1180-1192, 1996 35. Knott-Craig CJ, Julsrud PR, Schaff HV, et al: Pulmonary artery size and clinical outcome after the modified Fontan operation. Ann Thorac Surg 55:646-651, 1993