Bidirectional Cavopulmonary Shunt with Additional Pulmonary Blood Flow

Introduction

A bidirectional cavopulmonary shunt (BCPS) is frequently used as an intermediate step before the Fontan operation in patients with univentricular heart syndrome. ^1,2 Advantages of a BCPS include effective pulmonary blood flow without increased volume overload, low pulmonary arterial (PA) pressure, and less PA distortion. ^1,3 However, the BCPS provides less pulmonary blood flow than completed Fontan or normal circulation, which may lead to limited PA growth before Fontan completion. There are some controversies regarding the usefulness of leaving additional pulmonary blood flow (APBF) at the time of BCPS. ^{3 –7} APBF has some benefits including improved postoperative arterial oxygen saturation, enhanced PA growth, and reduced collateral vessel formation. ^8,9 However, APBF may also increase pressure in the superior vena cava (SVC) with volume overload on the single ventricle, and it may not improve the outcome after BCPS or eventual Fontan completion. We have a policy of leaving APBF at the time of BCPS, with adjustment in each patient to reduce the disadvantages of APBF. This study evaluated our results retrospectively.

Patients and Methods

From April 2002 to September 2008, 13 patients underwent BCPS with fine adjustment of APBF in our hospital. Ten boys and 3 girls, aged 9 months to 6 years (mean age, 24 ± 16 months), weighing 6 to 17.7 kg (mean, 10 ± 3 kg) underwent BCPS as an intermediate step before a Fontan-type operation. All patients had a functional single ventricular heart: 4 had double-outlet right ventricle, including one Taussig-Bing anomaly; 2 had univentricular heart; 3 had tricuspid atresia; 2 had transposition the great arteries (types II and III); one had pulmonary atresia with intact ventricle; and one had corrected transposition of the great arteries. A total of 18 previous palliative procedures had been performed in these 13 patients (Table 1).

OPEN IN VIEWER

Table 1.

Characteristics of patients undergoing bidirectional cavopulmonary shunt procedure

Flow	Patient No.	Age (months)	Weight (kg)	Diagnosis	Previous Surgery
Low	1	24	11.2	Double-outlet right ventricle, severe pulmonary stenosis Dextrocardia, superior/inferior ventricle, heterotaxy	Left BT 4-mm at 2 months
	2	18	10.5	Single atrium, univentricular heart, transposition of the great arteries Common atrioventricular valve, PAPVR	Right BT 4-mm at 2 months
	3	16	9.0	Double-outlet right ventricle, pulmonary stenosis	TAPVR repair + RVOTR at 1 month
				Total anomalous pulmonary venous return, single atrium, heterotaxy	Left BT 4-mm at 3 months
	4	15	8.0	Pulmonary atresia with intact ventricular septal defect Sinusoidal connection	Right BT at 1 month Left BT at 8 months
	5	22	9.9	Transposition of the great arteries (III), pulmonary stenosis Large ventricular septal defect	Right BT occlusion Left BT 4-mm at 3 months
	6	25	11.5	Corrected transposition of the great arteries Pulmonary stenosis, ventricular septal defect	Right BT 4-mm at 1 month
	7	72	17.7	Univentricular heart, pulmonary atresia Transposition of the great arteries, dextrocardia	Right BT at 2 months Left BT at 8 months
	8	9	6.0	Tricuspid atresia (IIa), pulmonary atresia Pulmonary coarctation of aorta	Left BT 4-mm at 5 months
High	9	21	9.5	Double-outlet right ventricle, pulmonary stenosis, dextrocardia Common atrioventricular valve, transposition of the great arteries	Right BT 4-mm at 2 months
	10	21	9.5	Taussig-Bing malformation	PA band 22-mm
	11	36	8.1	Tricuspid atresia (IIc), cerebral damage (hypoxemia)	PA band (unknown size) at 3 months
	12	16	7.5	Tricuspid atresia (Ic)	PA bands: 23-mm at 3 months 2 × 24-mm at 5 months
	13	17	7.4	Transposition of the great arteries (II) Large ventricular septal defect	PA bands: 21-mm at 1 month 2 × 23-mm at 3 months

BT = modified Blalock-Taussig shunt, PA = pulmonary artery, PAPVR = partial anomalous pulmonary venous return, TAPVR = total anomalous pulmonary venous return, RVOTR, right ventricle outflow tract reconstruction.

The BCPS procedure was undertaken through a median sternotomy, using cardiopulmonary (CPB) bypass in all patients. When atrial septectomy was required during the procedure in one patient, ventricular fibrillation was induced for a short period. Central venous pressure (CVP) was monitored by a catheter placed in the SVC. APBF was controlled using the following methods to keep CVP <18mm Hg immediately after the BCPS procedure. If the patient had PA banding with a 3-mm Teflon tape before BCPS, the band was tightened by shortening the tape. If the patient had a Blalock-Taussig (BT) shunt, it was left open at BCPS. After weaning from CPB, if CVP was >18mm Hg and there was hemodynamic instability, the BT shunt flow was reduced using a clip. After BCPS, dipyridamole was routinely given until Fontan completion. Cardiac catheterization was performed before and at 13 ± 2.9 months after BCPS to estimate suitability for Fontan completion. In suitable patients, total cavopulmonary connection was carried out using a graft of 16–18 mm in diameter or atriopulmonary connection. In 9 patients who completed the Fontan procedure, CVP, duration of chest tube drainage, intubation time, and use of nitric oxide were recorded.

Data are expressed as mean ± standard deviation. The PA index, PA pressure, cardiothoracic ratio, left ventricular end-diastolic pressure, and oxygen saturation before and after BCPS were compared using the paired t test. Statistical significance was defined as a p value less than 0.05. StatView version 5.0 for Macintosh (SAS Institute, Cary, NC, USA) was used for analysis.

Results

There was no operative or hospital death. Mean CPB time was 42.2 ± 11.7 min and operative time was 199.2 ± 29.5 min. The median duration of mechanical ventilation was 30.4 h (range, 1–120 h). CVP just after BCPS completion was 16.6 ± 3.8 mm Hg. In 5 patients, CVP exceeded 18 mm Hg after CPB, without hemodynamic instability; it gradually decreased to <18 mm Hg several hours later in the intensive care unit. There were no pleural or pericardial effusions requiring drainage procedures, no SVC syndrome, and no mediastinitis. In one case of common atrioventricular valve insufficiency, there was no change in the degree of regurgitation after BCPS. Major pulmonary arteriovenous malformations were not seen in any patient. APBF could be left in all patients (Table 2). Previously created BT shunts in 7 patients were left open, the shunt was ligated and divided and the stenotic pulmonary trunk was left open in one, and another had a BT shunt constructed after 2 months because of low pulmonary blood flow due to pulmonary stenosis. During follow-up, the PA grew well and APBF became excessive in one patient; the previously created BT shunt was separated, and the PA was banded with Teflon tape (20 mm in length). The PA band was tightened in 2 patients and left alone in one. In 3 patients who required repeat PA banding, the length of tape (in mm) was almost double the patient's weight (in kg). One patient had unbalanced pulmonary blood flow because the PA band had migrated. The proximal right PA was stenotic from the migrated PA band; in contrast, the left PA had excessive flow. This patient had a BCPS constructed on the right PA, and left PA banding with a 15-mm length tape. Systemic arterial oxygen saturation after BCPS was significantly higher than before BCPS (74.5% ± 7.4% vs. 84.6% ± 1.9%; Figure 1A). There was a significant difference in cardiothoracic ratio before and after BCPS (55.9% ± 6.1% vs. 53.2% ± 3.4%; Figure 1B). There was significant difference in left ventricular end-diastolic pressure before and after BCPS (11 ± 2.6 vs. 7.8 ± 3.0 mm Hg; Figure 1C). PA index was similar (503 ± 226 vs. 558 ± 237), but mean PA pressure decreased significantly after BCPS (14.7 ± 7.5 vs. 10.2 ± 3.1 mm Hg;Figure 2). Nine patients have undergone Fontan completion (interval from BCPS to Fontan, 15.6 ± 4.4 months) and 3 are awaiting the procedure. Fontan completion was refused in one patient because he had cerebral damage due to hypoxemia before PA banding. CVP in the operating room just after Fontan completion was 14.6 ± 2.9 mm Hg; none of these patients required fenestration. Complications after the Fontan operation are listed in Table 2. In 2 patients (No. 3 and 7), nitric oxide was used to decrease PA pressure and stabilize hemodynamics. Intubation times ranged from 1.5 h to 16 days. All patients who have completed the Fontan procedure are doing well in New York Heart Association functional class I. The 3 patients awaiting Fontan completion are in a stable condition.

OPEN IN VIEWER

Table 2.

Operative data of BCPS and Fontan operations

Patient No.	BCPS (min)	CPB (min)	Additional Flow	Additional Procedures	CVP * (mm Hg)	Intubated	Fontan	CVP * (mm Hg)	Intubated	Postoperative Complications
1	165	41	Left		21	2 h	Extracardiac	15	5 h	Thoracentesis
			Left BT (4-mm)
2	260	70	Left		18	4 days	Extracardiac	13	6 days	Thoracentesis
			Left BT (4-mm)	PAPVR repair
3	216	37	Left		15	1 day	Extracardiac	15	7 days	Prolonged drainage
			Left BT (4-mm)	Right PV ligation
4	156	47	Left	Right BT takedown	25	1 day	Scheduled
			Left BT (4-mm)
5	175	36	Left Left BT (4-mm)		15	1 day	Extracardiac	10	16 h	Prolonged drainage
6	200	40	Left		19	1 day	Scheduled
			Right BT (4-mm)
7	180	32	Left	Right BT takedown	13	1 h	RA to PA	13	1.5 h	Thoracentesis
			Left BT (4-mm)	Right PA plasty
8	207	27	Left Left BT (4-mm)		18	5 days	RA to PA	20	16 days	Prolonged drainage
9	226	53	Redo PA band (20-mm)	Right BT takedown	17	2 h	Extracardiac	15	3 days	Prolonged drainage
10	206	49	Redo PA band (18-mm)		15	6 h	Extracardiac	16	1 day	Prolonged drainage
11	166	48 †	Left PA band	ASD creation	15	1 day	Refused
12	216	32	Redo PA band (17-mm)		15	1 day	Scheduled
13	216	33	Left PA band (15-mm)		10	1 day	Extracardiac	14		Bronchomalacia

*

After CPB.

†

Ventricular fibrillation for 6 min. ASD = atrial septal defect, BCPS = bidirectional cavopulmonary shunt, CPB = cardiopulmonary bypass, CVP = central venous pressure, PA = pulmonary artery, PV = pulmonary vein, RA = right atrium.

OPEN IN VIEWER

Figure 1.

(A) Changes in systemic arterial oxygen saturation, (B) cardiothoracic ratio, and (C) left ventricular end-diastolic pressure before and after bidirectional cavopulmonary shunt (BCPS) procedures. *p < 0.05 vs. preoperative value.

OPEN IN VIEWER

Figure 2.

(A) Changes in pulmonary artery index and (B) mean pulmonary arterial pressure before and after bidirectional cavopulmonary shunt (BCPS) procedures. *p < 0.05.

Discussion

The first successful SVC-PA anastomosis was reported by Glenn ¹⁰ in 1958. This procedure has now been changed to a bidirectional anastomosis, and widely accepted as an intermediate stage before a Fontan-type operation. ^1,2 APBF was devised to increase pulmonary blood flow after BCPS, leading to higher oxygen saturation, lower mortality, improved cardiac function, prevention of arteriovenous fistulas, and better PA growth. ^11,–13 However, excessive APBF causes volume overload, increases atrioventricular valve regurgitation, and may raise PA pressure. Yoshida and colleagues ¹⁴ reported that appropriate APBF, based on CVP during the operation, provides better results than uncontrolled APBF. CVP of 16 mm Hg or less intraoperatively was considered ideal, because it decreased by 2–3 mm Hg on extubation after BCPS; as a result, CVP decreased to 12 ± 2.3 mm Hg after extubation with controlled APBF. ¹⁴ There were similar findings in our series. We controlled CVP at <18 mm Hg for adjustment of pulmonary blood flow, and there was no ventricular overload or elevated PA pressure after BCPS. Also, the PA index did not decrease, and arterial oxygen saturation increased after BCPS. There are few reports of regulation of ABPF. ¹⁵ If pulmonary blood flow can be measured perioperatively in real time, ABPF can be regulated to preserve systemic blood flow equal to pulmonary blood flow, and avoid excessive APBF. Appropriate ABPF is achieved when CVP after BCPS is around 16 mm Hg. ¹⁴

The optimal timing of the BCPS remains unclear. In this series, the mean age at BCPS was higher than the generally accepted ideal age. Our policy on age at BCPS is approximately 1.5 years if the patient does not have atrioventricular valve regurgitation and single right ventricle, as in hypoplastic left heart syndrome. Early BCPS will result in early volume unloading of the single ventricle, reducing ventricular end-diastolic volume, and thus preserving ventricular function. ¹⁶ Cleuziou and colleagues ¹⁷ reported excellent results of BCPS at the age of 6 months or less; however, PA growth after the BCPC remained too small for the relevant body surface area. They concluded that PA size does not influence the clinical outcome. In these patients, early completion of a total cavopulmonary connection leads to early volume unloading of the systemic ventricle and minimizes the risk of severe intrapulmonary right-to-left shunting.

The criteria of PA size have been modified during the past 3 decades. More patients with small central PA sections are accepted for the Fontan operation, and the risk level of the PA index has been lowered. ¹⁸ Especially in patients after the Norwood I operation, oxygen saturation decreases with time, and thus they may require BCPS earlier. Hosein and colleagues ¹⁹ demonstrated that that high PA pressure, small PA, right atrial isomerism, and poor ventricular function were preoperative risk factors for a post-Fontan surgical or catheter re-intervention. Recently, Ovroutski and colleagues ²⁰ noted that PA growth after the Fontan operation is clearly reduced despite somatic growth. The absence of PA growth may contribute to an increase in vascular resistance. They found a correlation between low central PA indices and unfavorable Fontan outcomes. Although there are several criteria for a Fontan-type operation, PA size is still a very important. Studies have shown that APBF has no adverse effect on outcome after cavopulmonary anastomosis and Fontan-type procedures. ^{2 –5} These reports mention that augmented PA growth before the Fontan operation should improve the long-term outcome. Although fine adjustment of APBF during the BCPS procedure may not contribute to the short-term Fontan outcome, it has been shown to maintain or improve PA size and long-term results of the Fontan operation.

This study is limited by the small number of patients, the homogeneity of the study group, and its retrospective nature. In our series, there were no patients with small PA and hypoplastic left heart syndrome. In those with a small PA at the time of BCPS, the effect of APBF is unclear. This study also did not compare BCPS with and without APBF. Further large cohorts and long-term studies are certainly needed to clarify whether BCPS might need APBF to improve the long-term Fontan prognosis. However, the adjustment of APBF during BCPS is useful for avoiding a decrease in PA size without volume overload, and this procedure is expected to improve the long-term results of Fontan-type operations.