Definition and Morphology
Pulmonary atresia with intact ventricular septum (PAIVS) was first described by John Hunter in 1783. It is a rare congenital cardiac malformation with considerable morphologic heterogeneity, and until recently, relatively poor outcome. There is complete atresia of the pulmonary valve in conjunction with a variable degree of hypoplasia of the tricuspid valve (TV) and right ventricular (RV) cavity. Invariably there is a usual atrial arrangement with concordant atrioventricular and ventriculoarterial connections. The morphologic diversity documented at birth has profound consequences on long-term outcome.
The right ventricular cavity is usually small with thick myocardium and suprasystemic RV pressures. The TV is highly variable and often small and dysplastic with stenosis and/or regurgitation. Its size is often denoted by a negative “Z-score” (number of standard deviations of a measurement from the population mean for a given body surface area). The median Z-score for the TV in this condition is about −5 at birth but with considerable variation (range: −18 to +9). The Ebstein anomaly coexists in 10% of cases (see Chapter 43 ).
The RV also shows considerable variation, ranging from a tiny hypertensive cavity with severe hypertrophy to a hugely dilated, thin-walled cavity ( Fig. 50.1A ). While appreciating that all three ventricular components are always present in this condition (inlet, trabecular, and outlet), there can be variable intracavity muscular overgrowth. In most cases, all three components can be identified and the ventricle is termed tripartite (see Fig. 50.1B ). In a bipartite ventricle the inlet and outlet can be identified with overgrowth of the apical trabecular portion (see Fig. 50.1C ). In a unipartite ventricle, only the inlet can be identified with overgrowth of the infundibular and apical trabecular portions (see Fig. 50.1D ). The occurrence of each type at birth is 59%, 33.5%, and 7.5%, respectively. In addition, when there is complete muscular infundibular obliteration, it is termed muscular atresia (25%) (see Fig. 50.1D ); when there is a patent infundibulum with complete fusion of the valve leaflets, it is termed membranous atresia (75%) (see Fig. 50.1B and 50.1C ).
Those individuals with the worst prognosis are the minority with severe tricuspid regurgitation, producing the so-called wall-to-wall heart (see Fig. 50.1A ). The right ventricle is thin and dilated and at very low pressure, resembling a Uhl anomaly. These cases account for one-sixth of the overall group. The tricuspid valve has an Ebstein anomaly or is severely dysplastic. The pulmonary valve is usually imperforate. The grossly enlarged heart occupies much of the thorax, thus preventing lung development during the latter part of fetal development.
The persistently hypertensive RV may be associated with communications between the RV and the coronary arteries. These are termed fistulae or RV-to-coronary connections (see Figs. 50.1D, 50.2, and 50.3 ) and are present in nearly 50% of patients at birth. In 17% of patients, overt abnormalities are found in the coronary arteries including stenoses, gross ectasia (see Figs. 50.1D to 50.3 ), and interruptions. These are presumed to be caused by the long-standing effects of very high RV pressure on the coronary arteries. This situation is termed an RV-dependent coronary circulation and implies that RV decompression would lead to coronary artery steal, ischemia, and/or sudden death.
Systemic-to-pulmonary collateral vessels are very unusual. Rarely, the right and left pulmonary arteries are nonconfluent, both being supplied by individual ducts.
Such tremendous diversity prevents the recommendation of a standard preferred surgical or catheter intervention, and this has led to the concept of management strategies tailored to individual morphologic subtype ( Table 50.1 ).
| Tricuspid valve diameter (Z score) |
| Tricuspid valve function: stenosis and/or regurgitation |
| Tripartite, bipartite, or unipartite right ventricle |
| RV cavity size |
| RV outflow obstruction: membranous or muscular |
| Presence of RV to coronary artery fistulae |
| RV-dependent coronary circulation: coronary ectasia, stenoses, and interruptions |
Genetics and Epidemiology
PAIVS is a relatively uncommon disease accounting for about 3% of congenital heart disease with an incidence of 7 to 8 cases per 100,000 live births. In the United Kingdom, this has fallen to 4.5 cases as a consequence of prenatal diagnosis and consequent termination of affected individuals with PAIVS; Sweden has reported similar findings. Sex incidence is equal. Progression from pulmonary stenosis to PAIVS in utero has been documented, and there may be a similar etiology. Both conditions may be found in recipients of twin-to-twin transfusion. Rare cases have been found in siblings. Controversy surrounds the etiology with proponents of both genetic and acquired causes. It may be that the disease is the endpoint of differing causes. Even the morphogenesis is controversial, with some suggesting that the primary event is pulmonary atresia with fistulae due to persistence of primitive RV-to-coronary connections and others suggesting that large RV-to-coronary connections are primary and lead to a progressive atresia of the pulmonary valve.
Fetal Presentation
PAIVS may be diagnosed in the second trimester of pregnancy owing to an abnormal four-chamber view. The presence of a hypoplastic or dilated RV may be evident at routine screening (18 to 22 weeks), although features of critical pulmonary obstruction can be seen as early as 12 weeks’ gestation. Fetal diagnosis can provide parents with important information; clinicians may also use data to alter timing and method of delivery to optimize neonatal treatment. Fetal intervention with in utero perforation of the atretic pulmonary valve aims to be a life-saving or disease-modifying intervention, with the rationale that growth of the right-sided structures may occur in the remainder of pregnancy if right ventricular outlet obstruction is relieved, enabling an eventual biventricular repair. The Boston group previously reported their experience with 10 fetuses. There was an initial learning curve followed by technical success in the most recent 6 cases, with improved right-sided heart growth and postnatal outcome. Guidelines on percutaneous fetal balloon valvuloplasty were produced in 2006 by the UK National Institute for Health and Clinical Excellence (NICE) ( www.nice.org.uk/guidance ), with acknowledgment that it was difficult to ascertain the efficacy owing to the small numbers with varying anatomy and limited long-term follow-up data.
Early Presentation and Management
PAIVS is a duct-dependent lesion, and presentation with cyanosis occurs when the duct closes shortly after birth. Findings include a single second heart sound, oligemic lung fields on chest radiography, and a normal QRS axis with precordial R wave progression consistent with a dominant LV on the electrocardiogram. After birth, a prostaglandin infusion is begun and the morphologic variables obtained by echocardiography are carefully evaluated. Angiography may be necessary to document coronary artery abnormalities. Management depends on the unique constellation of morphologic features present and can have long-lasting implications. The main thrust of the investigations is to predict suitability for long-term biventricular or univentricular repair. When this is not clear-cut, the strategy is usually to aim toward a biventricular repair, establish RV-to–pulmonary artery continuity, and maximize RV growth, because an arterial shunt alone may preclude this in the future. Management options are listed in Table 50.2 .
| Initial Strategy | Procedure Sequence | |||
|---|---|---|---|---|
| Biventricular repair | Catheter procedure Wire/laser/radiofrequency perforation of pulmonary valve + Balloon valvuloplasty | |||
| +/– Surgery Surgical systemic to pulmonary shunt (modified Blalock-Taussig shunt) | Catheter procedure Device occlusion of systemic to pulmonary shunt Device occlusion of residual ASD | |||
| Surgery Pulmonary valvotomy/valvectomy +/–Transannular patch +/– Monocusp homograft +/– Surgical systemic to pulmonary shunt (modified Blalock-Taussig shunt) +/– Hybrid procedure Stenting of patent ductus arteriosus | ||||
| Univentricular repair | Balloon atrial septostomy | Surgery Systemic to pulmonary shunt (modified Blalock-Taussig shunt) | Surgery Superior cavopulmonary anastomosis (bidirectional Glenn) | Surgery Total cavopulmonary connection |
Biventricular Strategy
The long-term aim of biventricular strategy is to achieve separated pulmonary and systemic circulations with two pumping chambers. This requires a reasonably sized RV, without significant coronary artery abnormalities (ie, RV-to–coronary artery fistulae are permissible; coronary stenosis, interruptions, and ectasia are not permissible) and with adequate TV size and function (see Fig. 50.1 ). Traditionally, this involved surgical intervention to reconstruct the RV outflow tract (see Table 50.2 ). When there is concern as to whether the RV can generate sufficient pulmonary blood flow, then a systemic-to-pulmonary shunt is created in addition (modified Blalock-Taussig shunt). Since the 1990s it has become feasible to achieve transcatheter radiofrequency perforation of the atretic pulmonary membrane when there is membranous rather than muscular atresia ( Fig. 50.4 ). A total of 40% to 60% of patients subsequently require a surgical arterial shunt because of continuing cyanosis, leading some groups to stent the arterial duct routinely. Some also ultimately need a surgical RV outflow tract procedure. Although a biventricular circulation is usually achieved, multiple procedures may be required to achieve it.
1.5-Ventricle Repair
In some cases in which the RV is of borderline size, it becomes apparent that the RV will not be capable of totally supporting the pulmonary circulation alone. In such cases, after an initial RV outflow tract procedure, a superior bidirectional cavopulmonary anastomosis can be fashioned to provide an additional source of pulmonary blood flow. Once any arterial shunts and/or atrial septal defects (ASDs) are closed, this is known as a 1.5-ventricle repair.
Univentricular Strategy
The long-term aim of a univentricular strategy is to achieve separated pulmonary and systemic circulations with only one contributory pumping chamber (the left ventricle). This strategy tends to be performed in those with small RVs (see Fig. 50.1D ) or an RV-dependent coronary circulation (see Figs. 50.1D, 50.2 and 50.3 ). A balloon atrial septostomy is initially performed to enable the obligatory right-to-left shunt and prevent obstruction, followed by an arterial shunt, a superior cavopulmonary anastomosis (bidirectional Glenn procedure), and finally a total cavopulmonary connection (TCPC) ( Fig. 50.5 ).
Severely Dilated Right Ventricles
When there is gross dilatation of the RV at presentation, often due to severe tricuspid regurgitation, the mortality is high with little improvement despite advances in fetal diagnosis and surgical management. The operative strategy must include TV repair or, in many cases, occlusion of the TV with construction of a systemic-to-pulmonary arterial shunt (Starnes procedure) followed by a univentricular route.
There is substantial debate as to which initial strategy should be adopted. Most groups use single or multiple morphologic features on presentation as a guide to the initial management. These include (selected from many others):
- •
Tricuspid size (often expressed as a Z score)
- •
Partite classification of the RV (tripartite, bipartite, or unipartite)
- •
Infundibular size
- •
Indices of right ventricular size
- •
Presence of coronary artery stenoses
Late Outcome
Survival and Functional Status
Survivors are now reaching adulthood. Their numbers are few (but increasing), and this relates to the rarity of the disease and high early mortality. Consequently, there are scant data about survival and functional status in adulthood. The Toronto Hospital for Sick Children reported 10-year survival of 43%, the Congenital Heart Surgeons study reported 15-year survival at 58%, the Swedish Collaborative study noted 10-year survival of 68%, but more reassuringly, a series from the University of California, Los Angeles (UCLA), reported a 10-year survival of 86%. In the future, an increasing number of patients will be expected to reach adulthood.
Biventricular Repair
Mortality with biventricular repair tends to occur in the first 6 months of life, and the survival curves then flatten. A Japanese study documented 14-year survival after biventricular repair at 86%. Twenty percent of patients had late arrhythmias, and right atrial dilatation was found in all patients. In a UK cohort of patients with initial radiofrequency- or laser-assisted pulmonary valvuloplasty (the majority of whom ultimately had a biventricular circulation), no arrhythmias were reported at median follow-up of 9 years (range 2 to 21 years).
Although intuitive, there is limited actual evidence that biventricular repair is better than univentricular repair. Sanghavi et al. found no statistical difference in exercise capacity between those with biventricular versus univentricular repair using programmed bicycle ergometry. Most patients in both groups had subnormal peak oxygen consumption and a trend toward impaired performance with increasing age. Similarly, Ekman-Joelsson et al., in the Swedish Collaborative Study, found no difference in exercise capacity in patients after biventricular versus univentricular repair, although the group with RV-to-coronary artery fistulas and a biventricular repair did worse. Decreased lung function was noted in all groups. Karamlou et al. found a trend toward higher VO 2 in patients with biventricular and 1.5-ventricle repairs compared to univentricular patients. However, increased performance was strongly associated with an initial tricuspid valve Z-score, rather than conferred by repair type. Peak VO 2 and maximum heart rate were lower in survivors of PAIVS than controls regardless of their type of repair. The study also demonstrated an interesting dichotomy whereby patients with PAIVS believe they are doing well despite important physical limitations.
There are several possible explanations for these findings. It is known that after biventricular repair, patients have evidence of RV diastolic dysfunction with restrictive RV physiology. With atrial contraction there is retrograde flow in the superior vena cava and antegrade flow in the pulmonary artery. In addition, widespread perfusion defects have been found using myocardial perfusion scintigraphy, particularly in the ventricular septum. Mi and Cheung have documented abnormalities of both RV and left ventricular (LV) long-axis function in patients late after biventricular repair.
Until more specific data are available, prediction of longer-term outcome can be made only by drawing parallels with other diseases. After a biventricular repair for PAIVS, patients show many similarities with patients who have undergone definitive repair of pulmonary stenosis and tetralogy of Fallot (see Chapter 45 , Chapter 47 ), particularly those with restrictive physiology. When there is minimal residual hemodynamic disturbance, the long-term outlook is probably excellent with good quality of life. Mild residual pulmonary stenosis should be well tolerated; if it is more severe it may lead to arrhythmia from atrial dilatation. Long-standing pulmonary regurgitation has been shown to be detrimental to RV function in tetralogy of Fallot (see Chapter 47 ) and may be expected to cause similar problems in PAIVS. Little is known about the risks of sudden death and ventricular tachycardia in patients with PAIVS in adult life, but they may be expected in patients with residual pulmonary regurgitation. Long-term complications are shown in Table. 50.3 .
