IN UTERO INTERVENTION FOR SEVERE CONGENITAL HEART DISEASE

HELENA M. GARDINER, MD, PHD

Introduction

Fetal therapy has been offered for a variety of disorders, including fetal anemia, lower bladder obstruction, twin-to-twin transfusion syndrome (TTTS), diaphragmatic hernia, and spina bifida. Following clinical trials, some therapies, such as bladder shunting, have not been recommended for inclusion into clinical practice¹ while others, such as laser photocoagulation for TTTS, form the bedrock of therapy for this disease.² Fetal therapies have been investigated using hypothesis driven, multicenter, and/or randomized trials. For example, the optimal management of TTTS was investigated in a trial comparing septostomy/amniodrainage and laser photocoagulation of placental anastomoses and its results provide an evidence base to inform clinical practice.² Unfortunately, evaluation of fetal cardiac therapies has resulted in only level 3 evidence to date and many questions remain to be answered.

The unifying intention of fetal therapy is to provide early correction of an underlying anatomical problem, or at least to reduce the amount of secondary damage likely to occur where there is abnormal anatomy and/or physiology. While the aim of fetal cardiac intervention is similar, finding a unifying solution to diverse cardiac morphology and pathophysiology is more challenging. Currently, the therapeutic approach is limited, based on our understanding of semilunar valve disease presenting at birth or in later childhood. The current literature lists selection criteria based on the clinical experience of 2 centers,^3–6 and the revision of the Boston criteria and current recommendations from Linz are shown in Tables 22.1 and 22.2, but to date no studies have been designed specifically to determine the potential for fetal cardiac therapy to alter outcomes for fetuses with severe semilunar valve stenosis. There is no good evidence that a child with a biventricular circulation has a better short- or medium-term outcome than a child with a good univentricular palliation.⁷ However, it is recognized that there are long-term issues with a univentricular circulation, including increased morbidity and the potential for heart failure to occur in a proportion, particularly where there is a systemic morphological right ventricle (RV). The thrust of fetal intervention is not only to increase the proportion of individuals with a biventricular circulation but also to improve the quality of the developing myocardium and pulmonary bed so that either postnatal surgical strategy will be more successful.

Table 22.1. Era change in retrospective guidelines for fetal aortic valvuloplasty

Table 22.2. Linz criteria for fetal aortic valvuloplasty*

^*Arzt et al. Ultrasound. Obstet Gynecol. 2011;37:689–695.

The original stimulus for fetal valvuloplasty was the recognition that both interventional and surgical outcomes for babies born with critical aortic stenosis (AS) were poor⁸ and at that time there was no successful palliative surgery for hypoplastic left heart syndrome (HLHS). These early attempts had mixed success, mostly because the catheter equipment available was not suitable for the small fetal heart. Technical improvements have led to renewed interest in fetal therapy and programes now report good technical success using coronary balloons.^3,5,9

Fetal valvuloplasty is also offered for critical pulmonary stenosis (PS) and pulmonary atresia with intact septum (PA-IVS), particularly when there is important tricuspid regurgitation and early hydrops, probably as a result of high systemic venous pressures.^10,11

Semilunar valvuloplasty and balloon atrial septostomy¹² have been introduced into clinical practice prenatally to ameliorate these lesions and in an attempt to prolong pregnancy complicated by hydrops.¹³ Fetal hydrops has a high risk of intrauterine demise and also occurs in fetuses with a thickened or intact atrial septum which often accompanies AS or HLHS.^14,15

Fetal pacing is still experimental. Isolated complete heart block occurs in hearts with concordant connections, thought due to maternal anti-Ro/La antibody transfer across the placenta.¹⁶ It is rare, with most pediatric cardiac centers only seeing one or two cases a year and it is unusual for fetuses presenting with complete heart block to benefit from any intrauterine therapy,^17,18 although the role of steroids has been proposed beneficial to prevent later cardiomyopathy.¹⁹ Sustained low fetal heart rates (below 55 beats per minute) may not be sufficient to support cardiac output and this often results in hydrops and fetal death. Cardiac pacing has been attempted in a few human fetuses with signs of hydrops, but so far success in the human has been limited to a temporary increase in heart rate.^20–23 Technical success in pacemaker implantation has been demonstrated in the sheep, but future utility in the human fetus remains to be evaluated.

Rationale for Intervention

Members of the cardiology community hold diverse opinions on the efficacy of fetal intervention for valve stenosis with few showing equipoise: some believe there are compelling theoretical reasons why fetal therapy may improve the outcome for an individual, while others feel it has no scientific merit and is meddlesome. Timing of surgery for a range of congenital heart defects (CHDs) has moved into the neonatal period in many centers; one of the initial driving forces behind this is evidence that earlier repair may improve the chances of ventricular and vascular remodeling and provide a better match between the enlarging myocardium and its blood supply.²⁴ Cardiac surgery is now performed routinely in the neonatal period with good medium- and long-term outcomes.^25,26 If improved adaptation alters outlook, then fetal therapy is a logical extension of current trends in postnatal surgical practice. The diagnosis of cardiac disease is now made earlier in the fetus, and progression of AS to HLHS has been observed to occur between 12 and 16 weeks’ gestation.²⁷ Recognition of in utero disease progression results in caution in counseling at the time of initial diagnosis as outcome may be worse than originally predicted. Secondary ventricular damage is often significant and may result in hypoplasia of the supporting ventricle and significant myocardial dysfunction due to altered coronary perfusion. Others believe the postnatal surgical route (a uni- or biventricular circulation) is likely to be determined already by the anatomy present in the second trimester and do not consider fetal intervention to be helpful in altering eventual outcome, unless there are signs of fetal hydrops. Thus, fewer cases are referred for consideration for fetal valvuloplasty than are diagnosed prenatally and this depends on the individual stance of congenital cardiology teams. All these factors affect eventual surgical choices and may worsen the outcome.²⁸

Natural History of Fetal Semilunar Valve Stenosis

The natural history of ductal-dependent critical aortic or PS or atresia is a neonatal death, unless immediate postnatal treatment is available. Valvar stenosis or atresia is usually managed by balloon valvuloplasty^29,30 if a biventricular circulation is thought possible or, if this does not seem possible, univentricular palliation using the Fontan or Norwood surgical pathway is planned.^31,32 These two surgical strategies are for right and left heart hypoplasia, respectively, and both aim to divert the returning systemic venous flow directly to the lungs and to use the heart as a systemic pump for the oxygenated blood. In contrast to single-center experiences, externally verified nationwide interventional and surgical data are now available in registries such as the National Institute for Cardiovascular Outcomes Research (NICOR) and Society of Thoracic Surgeons (STS) in the UK and USA, respectively. Current reports on surgical outcomes from these sources can inform the decision process for introducing any new technique.³³ Procedural mortality is now as low for neonatal balloon aortic valvuloplasty (1%–2%) as for most catheter-based procedures at any age (NICOR-congenital website: http://nicor4.nicor.org.uk/). Unfortunately, these databases do not reflect the experience of unselected series including fetal diagnoses³⁴; in one study, the overall survival was 46.9% (23/49) in the cohort of intention-to-treat cases diagnosed prenatally and followed through to 5 years.³⁵ One-year outcomes are difficult to determine from the NICOR registry, as about 25% of outcomes are recorded as unknown, but unselected national studies for PA-IVS report only about one-third have a 2-ventricle repair, and 5-year survival is only about 68%.³⁶

Natural History of Fetal Restrictive Interatrial Septum (IAS)

Patency of the foramen ovale (FO) is essential for flow returning from the placenta to fill the left ventricle (LV), and AS is associated with thickening and closure of this important interatrial communication. Overall, about 10%–12% of neonates with HLHS are born with a restrictive communication; an unknown proportion of these may have been cases of AS at the time of fetal anomaly screening in the second trimester. Some affected fetuses will not reach term and will develop hydrops with subsequent interuterine death.^14,15 Closure of the IAS is associated with pulmonary venous abnormalities which may lead to irreversible pulmonary venous hypertension. Neonates with pulmonary venous hypertension cannot be resuscitated easily at birth, nor readily weaned from a ventilator following surgery. A baby born with a closed IAS may experience postnatal pulmonary hemorrhage and neonatal death before they can reach a cardiac center for treatment. For survivors of a closed or very restrictive septum, damage to the pulmonary vasculature may be so severe in fetal life that conventional Norwood surgery may not be possible.¹⁴ The “hybrid” technique combining catheter and surgical techniques may allow high-risk neonates to undergo early palliative surgery and eventually complete a Norwood circulation.³⁷

Other Considerations Affecting Selection for Fetal Intervention

Critical PS and Pulmonary Atresia

In PA-IVS, fetal scans demonstrate the coexistence of RV to coronary artery fistulae in up to one-third of cases. The majority of fetuses with large fistulous communications have smaller unipartite RVs and so are unlikely to achieve a biventricular circulation and these cases would not be considered for an intervention. However, some fistulae are felt to be small and the RV a reasonable size, so their impact on postnatal surgical choices is uncertain. The importance of these communications is 2-fold. First, the coronary circulation depends in part on maintenance of the high pressure in the hypertrophic right ventricular cavity; after birth, if the pulmonary valve (PV) is opened and the right ventricular pressure falls, there may be coronary steal and myocardial perfusion is compromised. Second, high-pressure retrograde flow in the coronary system leads to stenosis and atresia of the coronary arteries and may result in fatal myocardial infarction in affected babies in the neonatal period.^36,38 In postnatal series, right ventricular-dependent coronary circulation is reported in 46%.³⁶ The gold standard of diagnosis is postnatal angiography, and so a degree of uncertainty may exist prenatally before this can be performed. As these individuals are at increased risk of neonatal death, an argument could be made for fetal intervention, the rationale being that early decompression of the high-pressure RV may allow regression of these communications and improve the postnatal chances of survival for these babies. If a coronary fistula is present and the right ventricular outflow tract is successfully opened in utero there will be no coronary steal at this time because of equivalent ventricular pressures. If a right ventricular-dependent coronary circulation is felt to be present and the PV had been successfully opened in utero, one option might be to plan for a surgical ligation or coil closure of the fistula in the perinatal period before steal occurs, perhaps at the same time as the placement of systemic to pulmonary artery (PA) shunt.

Critical AS

Critical AS progressively develops into HLHS in an indeterminate proportion of cases and, depending on this poorly defined time-line, fetuses will lie somewhere along this path at first detection. The triggers for growth failure have not been defined in the human, and retrospective definitions of the minimum size of aortic or mitral valves (MVs) and left ventricular length that might benefit from a fetal procedure have been proposed (Tables 22.1 and 22.2). The extent to which simple aortic valvuloplasty is able to influence developmental processes—for example, by improving growth velocities of the ventricle—seems limited; even if growth is improved, it is uncertain whether diastolic function will be normal in childhood. Persistent pulmonary hypertension is an important problem leading to significant long-term morbidity and childhood mortality.³⁹ Echocardiographic prediction of myocardial mineralization or endocardial fibroelastosis (EFE) is not thought to be accurate, but there are cases where the LV appears to be highly mineralized and stiff. This is an additional risk factor for tamponade, implying such cases should be avoided? The rationale behind treatment is to allow improved diastolic function, ventricular filling, and promote normal division of myocardial cells during the remainder of fetal life, thus increasing the proportion of healthy myocardial cells in the myocardium before delivery that may alter future outcome.⁴⁰ Such an improvement would theoretically be advantageous for a child, even if eventual left-sided morphology is too small for a biventricular management pathway after birth.

Interatrial Septostomy

A degree of interatrial restriction is helpful in the newborn period in cases of HLHS, but the fetal cardiologist tends to see more severely pathological cases in utero that may impact severely on perinatal morbidity and mortality because of secondary pulmonary damage.^14,15,39

Stay updated, free articles. Join our Telegram channel

Join