Summary
An early, primary, in utero cardiac abnormality may prevent normal heart development and cause irreversible secondary structural changes. The idea of foetal cardiac intervention stems from this understanding and focuses on antenatal intervention targeting the primary abnormality to allow normal flow and haemodynamics and thus normal heart development. Crucial aspects of foetal vascular access, varying foetal lie and structural complexity make it very hard to set procedural standards. The procedures are complex and are associated with significant maternal and foetal morbidity and mortality. The high risk-benefit ratio clearly explains the investigational nature of such therapies. With the development of minimally invasive techniques and continued animal experiments, foetal interventional therapy may see a low rate of morbidity and mortality, improving the prognosis of newborns with congenital heart disease previously considered incurable.
Résumé
Une anomalie cardiaque précoce intra-utérine peut empêcher le développement normal du cœur et causer des changements structurels secondaires irréversibles. L’idée d’intervention cardiaque fœtale prend sa source dans la théorie du flux qui indique qu’une cavité se développe au prorata du sang qui la traverse. L’anomalie précoce est ici la réduction du flux, c’est-à-dire l’obstacle à l’éjection. Les interventions prénatales ciblent cette anomalie primaire pour permettre le rétablissement d’un écoulement et d’une hémodynamique normale et in fine d’un développement normal de cœur. Les procédures sont complexes, associées à une morbidité significative et à une mortalité maternelles et fœtales. Le rapport bénéfice/risque actuel en faveur du risque explique clairement la nature expérimentale de telles thérapies. Les résultats décevants sont expliqués par la difficulté d’accéder au mobile fœtal avec du matériel non dédié sur des malformations d’anatomie très variables dont l’histoire naturelle est mal connue et controversée. Avec le développement des techniques mini-invasives, de matériel approprié et des expérimentations animales, la thérapie interventionnelle fœtale pourrait voir ce rapport augmenter au bénéfice de la sécurité avec une réduction de la morbidité et de la mortalité, améliorant le pronostic des nouveau-nés avec la maladie cardiaque congénitale précédemment considérée incurable.
Introduction
Percutaneous, invasive, per utero, foetal therapies have been developed to correct and prevent progression of pathologies diagnosed in utero (such as spina bifida, diaphragmatic hernia, etc.). Such techniques are, however, quite challenging to apply in clinical practice. The same applies to the art and science of foetal cardiac interventions . Congenital heart disease (CHD) is the most common inborn defect, occurring in about 19/1000 live births . Advancements in imaging technology over the past couple of decades, with the development of high-frequency, high-resolution probes, have made accurate diagnosis of most CHDs possible during foetal life. Colour flow and echocardiography-based haemodynamic foetal cardiac assessment in pregnancy allows a better understanding of the pathophysiology of even complex CHDs . Currently, the diagnosis of CHD is possible only after the end of the first trimester of pregnancy . Such scans are crucial for foetuses at risk of CHD due to a strong family history, maternal exposure to toxic drugs or chemicals, specific maternal conditions and nuchal translucency above the 99th percentile . The diagnosis of CHD at an early stage of pregnancy allows enough time for a team approach, with prenatal and genetic counselling by the expert paediatric cardiologist, discussion of the optimal management plan, delivery and postnatal care, and planning for future pregnancies. For example, once diagnosed, duct-dependent CHD should be referred to a maternity hospital with the appropriate neonatal intensive cardiac support to reduce perinatal morbidity and mortality . For conditions with no possibility of biventricular repair, the expectant parents may choose between postnatal palliative management (such as the Norwood procedure for hypoplastic left heart syndrome [HLHS]) or for interruption of pregnancy. The latter option is chosen by most young couples in our country nowadays. Factors like these and our insufficient knowledge of the underlying mechanisms have resulted in slow progress in developing effective intrauterine therapies for the management of CHD.
In utero natural history and therapeutic perspectives
The impact and outcomes of invasive, per utero, foetal therapies are still debated and there are differing opinions among the various experts . It is hard to predict the natural history and intrauterine course of events. There is a school of thought that the natural history of CHD worsens over time and that it is best to perform corrective surgery as early as possible (previously performed in childhood and now carried out in the neonatal period or infancy) . Others believe that the outcome of CHD is definitively defined from an anatomical point of view at the second trimester in most cases and that foetal intervention would not alter the natural history, except in case of foetal hydrops. The first school of thought is based on the flow-driven mechanistic approach that might help the understanding of the pathophysiology of some severe heart diseases. With this understanding, it is believed that myocardial chamber growth is related to the increasing blood flow into the cavities. Thus, with in utero alteration of haemodynamics and blood flow – and the resultant morphological and functional changes – CHD may progress and may be worse at birth than originally predicted.
Valvular stenosis may result in poor chamber development by reducing the blood flow. Thus, aortic valve stenosis may be the root cause and instigating defect leading to HLHS, small or virtual left ventricle, mitral atresia and aortic arch abnormalities. Once the crucial phase of chamber development is reached, it becomes irreversible, despite correction of stenosis. These findings have inspired the idea of in utero catheter intervention, restoring the flow through the stenotic valve in order to allow the development of the respective chamber and thus leave the option for a biventricular repair after birth. Similar interventions could improve the global prognosis of such heart malformations. The feasibility of an interventional approach is heavily dependent on the access to the foetal vascular compartment and on acceptable maternal and foetal morbidity and mortality.
Access to foetal vascular compartment
Cordocentesis
Ultrasound-guided cordocentesis was first described in 1983 by Daffos et al. . The procedure involves percutaneous access to the foetal vascular compartment to draw foetal blood for cytological, biochemical and cytogenetic studies; it is also used for foetal transfusion in cases of foetal anaemia due to maternofoetal alloimmunization. Furthermore, cordonal access allows perfusion of therapeutic agents to the foetus for sedation analgesia during technical procedures on the foetus, such as setting up a pleuroamniotic drain in the management of hydrothorax or for intracardiac interventions. The risk of foetal loss related to the cordocentesis is estimated at 1–2%. The length of the umbilical cord and its highly tortuous course make progression with a guide both dangerous (due to risk of perforation) and technically impracticable.
Cardiocentesis
Percutaneous, per utero, foetal, transthoracic, direct ventricular access was proposed as an alternative to cordonal access for foetal transfusion in cases of maternofoetal rhesus alloimmunization when cordonal access was technically impossible or after failure . Cardiocentesis is performed with local anaesthesia to the maternal skin and direct needle puncture of the foetal chest wall to enter the foetal left ventricle.
Besides technical problems such as difficult access with varying foetal lie and catheter/balloon fragmentation, the rate of foetal loss is estimated at around 5.5–6.5%. Currently, cardiocentesis is seldom or not used in the indication of foetal transfusion. Its current indications are mainly for foetal reduction in dichorionic twin pregnancies if one of the foetuses satisfies the indication for selective termination of pregnancy, such as aneuploidy. This technique is also used for therapeutic acts in utero on foetuses with CHD. It requires experience in percutaneous puncture under ultrasound guidance as well as in foetal echocardiography. The access has to be focused on the area to be treated (e.g. transventricular access in case of valvular lesion). The right ventricle is usually more accessible given its subparietal location. It is essential to precisely locate atrioventricular valves, valve chordae and interventricular septum (containing the bundles of conduction) in order to avoid injury to these structures if the pregnancy is expected to continue.
Foetoscopic access
Kohl et al. first described foetoscopic access in a foetal sheep: with three to four trocars percutaneously placed in the uterus and videofoetoscopic equipment, they achieved a limited thoracotomy to obtain minimally invasive, direct, foetal cardiac access . This approach was tested on 15 foetal sheep for foetal cardiac pacing or antegrade foetal cardiac catheterization. The technique was achieved in 10 foetal sheep (five failures because of bleeding or technical complications); eight were alive at the end of the procedure. Six ewes continued gestation; three non premature lambs (20% of cases) were born. Foetal death occurred in seven cases, maternal death by sepsis in two cases and technical complications, such as bleeding of the puncture site, and technical difficulty in identifying the subxyphoid region were also noted. Foetoscopic access has some advantages compared with maternal hysterectomy but this technique is not done in humans because of the high rate of complications.
To provide alternative approaches for human foetal cardiac interventions, the same team described transumbilical foetal cardiac catheterization in sheep by minimally invasive foetoscopy, guided by foetal transoesophageal echocardiography . The umbilical cord was punctured in 6/6 cases, with the possibility of anterograde catheterization into the foetal heart. However, all foetuses died secondary to dissection, total thrombosis of umbilical vein or blood loss after sheath dislodgment or removal. Interestingly, a suture at the sheath insertion site ensured foetal survival for 1–2 weeks. The translation of cordonal cardiac catheterization to a human foetus seems inconceivable at present; its interest lies in the simplicity of ultrasound guidance but its morbidity and mortality rates in animals is close to 100%.
Transhepatic access
The technical deadlock in performing transumbilical foetal cardiac catheterization led to the suggestion of transhepatic access . The aim was to reproduce the conditions of catheterization using Seldinger’s technique as it is done in the postnatal period and to improve foetal tolerance.
The feasibility of anterograde, echocardiography-guided, cardiac catheterization through a transhepatic approach of the intra-abdominal foetal vessels was tested on ten foetal lambs. Access to the subdiaphragmatic portion of the inferior vena cava was performed by ultrasound guidance via a transhepatic approach. A guide was placed into the subhepatic vein and cardiac catheterization was performed according to Seldinger’s technique, allowing the heart chambers to be reached in all cases, with atrial or ventricular pacing in six foetuses and ballooning of the pulmonary valve in nine cases ( Fig. 1 ). Three foetuses died after the procedure and five foetuses were born at term, with an autopsy showing no significant cardiac or peritoneal injury. The simplicity of the procedure, with its shallow learning curve, has successfully lowered complication rates in subsequent studies. In the future, this method could become an alternative to percutaneous transventricular catheterization. The perioperative bleeding risk could be reduced by peritoneal reabsorption of red blood cells. However, variations in foetal positions and hepatic vein diameter add to the challenges; the latter limits this method to the second part of pregnancy when a needle can be safely inserted.
