Advances in Fetal Cardiac Imaging and Intervention

Abstract

The field of fetal cardiology has evolved significantly in recent years. This review focuses on specific advances in fetal cardiac imaging and intervention that are increasingly used in clinical practice. On the imaging frontier, updated screening guidelines and artificial intelligence hold promise for improving prenatal detection of congenital heart disease. Advances in ultrasound technology and magnetic resonance imaging techniques have enabled greater diagnostic and prognostic accuracy of fetal heart disease from the first to third trimesters, and maternal hyperoxygenation can offer additional physiological insights. Fetal cardiac therapy has also seen great progress, with advances in transplacental pharmacologic treatments, infusions of enzyme replacement therapy, and fetal surgery for select rare and severe conditions.

Fetal cardiology is a relatively new subspeciality with the first reports of widespread clinical use of fetal echocardiography starting in the late 1980s.¹ The field has remained highly innovative with advances in imaging technology alongside an expanding range of interventional approaches to the fetus with heart disease.

Advances in Fetal Cardiac Imaging

Prenatal screening for congenital heart disease

Prenatal diagnosis of congenital heart disease (CHD) has been demonstrated to reduce morbidity and mortality.2, 3, 4, 5 Other established benefits of a prenatal diagnosis include time for counselling, informed decision-making, and preparation for postnatal management or termination of pregnancy.⁶ Because fetal CHD may occur in low-risk pregnancies with no known risk factors, the identification of fetal cardiac abnormalities depends on a high-quality ultrasound screening programme. Of all areas of fetal cardiology, advances in screening have the potential to have the greatest benefit.

Although first described in 1997, the 3-vessel view has only been introduced into mandatory screening programmes in countries around the world in recent years. The 3-vessel view demonstrates the superior vena cava, ascending aorta, and main pulmonary artery in a cross-sectional plane⁷ (Fig. 1). The view allows identification of conditions including tetralogy of Fallot, transposition of the great arteries, and a range of arch abnormalities.⁸ In Canada, the view was included in screening guidelines by the Society of Obstetricians and Gynaecologists of Canada in 2009⁹ and by the Canadian Association of Radiologists 2010.¹⁰ In 2019, the American Institute of Ultrasound in Medicine added the 3-vessel view “if technically feasible” to the 4-chamber and outflow tract views,¹¹ and finally in 2022, the International Society of Ultrasound in Obstetrics and Gynecology mandated the 3-vessel view.¹² Where mandatory inclusion has been introduced, detection rates of abnormalities affecting the great vessels have significantly increased, including transposition of the great arteries in Canada.¹³ In the Netherlands, there was increased detection of coarctation of the aorta from 25.7% to 42.2%¹⁴ and tetralogy of Fallot from 44.2% to 82.4%.¹⁵

Figure 1.

Normal 3-vessel view with the superior vena cava (SVC), aorta (Ao), and pulmonary artery (PA) arranged in a linear fashion from right (R) to left (L) with each vessel sequentially larger in size. Although this view is not new, it has only recently been incorporated into national and international obstetrical screening guidelines.

Despite an increasing number of ultrasound views, prenatal diagnosis of CHD remains suboptimal.¹⁶ This is partly due to limited access to high-quality ultrasound imaging and interpretation. Recent advances in artificial intelligence may help to identify fetuses with major forms of CHD.^17,18 However, the technology is currently in its infancy, and human-artificial intelligence interactions will remain essential.

Early fetal echocardiography

The benefits of an accurate fetal cardiac diagnosis may have even a greater impact when diagnosed earlier in gestation, including more time for consultation, decision-making, and/or safer, less-traumatic termination procedures.¹⁹ First trimester fetal echocardiography by transvaginal ultrasound was first reported in the early 1990s.²⁰ By 9 weeks of gestation, all of the major heart structures are formed,²¹ and by 12 weeks, the fetal heart is situated with a normal leftward cardiac axis.²² The optimal timing for first trimester echocardiography is 12- to 13-week gestation.²³ Figure 2 depicts images of normal and abnormal cardiac anatomy at this gestational age.

Figure 2.

Early fetal echocardiographic images demonstrating normal 4-chamber (A) and 3-vessel views (B) at 12 + 1 weeks of gestation by transabdominal imaging. Images (C) and (D) depict a hydropic fetus with a borderline left ventricle in the 4-chamber view and a hypoplastic arch in the 3-vessel view at 12 + 4 weeks, respectively. This fetus was found to have 45X0 (Monosomy X or Turner syndrome).

With improvements in ultrasound technology, such as transducers with high frequency range, harmonic imaging, and advancements in color Doppler,²⁴ early fetal echocardiograms can be highly accurate.²⁵ A meta-analysis of studies published between 2004 and 2021 found that early fetal echocardiography detected approximately 80% of major cardiac abnormalities identified prenatally in high-risk patients.²⁶ International Society of Ultrasound in Obstetrics and Gynecology updated their guidelines for the performance of fetal ultrasound at 11- to 14-week gestation in 2023.²⁷ Although the guidelines have minimum requirements for the evaluation of the fetal heart, the potential for more extensive screening views is acknowledged.

There are no indications that early fetal echocardiography is unsafe. Concerns over exposure to thermal and mechanical stress are particularly important in first trimester ultrasound. Both pulse wave and color Doppler, which are important components of fetal echocardiography, require greater power output and average intensity and, therefore, increase the potential risk of heating compared with B-mode. Bidirectional Doppler (known as Power Doppler or HD Doppler depending on the vendor) has also been introduced to aid in visualization of vascular structures. The risks associated with these techniques can be mitigated by the operator applying the “as low as reasonably achievable” principle.²⁷

Despite the advances in technology that have made early fetal echocardiography both feasible and safe, uptake into routine practice is not universal. Currently, specialized centres offer the service to high-risk patients, such as those with nuchal translucency >3.5 mm, known or suspected genetic diagnosis, major extracardiac anomaly, exposure to teratogens, maternal haemoglobin A1C >7.5%, or family history of CHD in a first-degree relative. Important limitations include inability to visualize certain structures (ie, pulmonary veins) and the potential for evolution of fetal cardiac disease. Among continued pregnancies, a standard second trimester fetal echocardiogram is recommended. As evidence continues to emerge and indications crystallize, early fetal echocardiography is likely to expand.

Maternal hyperoxygenation

Transitional physiology at birth is characterized by a fall in pulmonary vascular resistance with exposure to oxygen, along with a rise in systemic vascular resistance due to separation from the placenta. Acute maternal hyperoxygenation (MH) attempts to mimic these changes by lowering fetal pulmonary vascular resistance and increasing pulmonary blood flow (it is not possible to mimic the increase in systemic vascular resistance). In so doing, acute MH may help to better stratify risk at birth and plan for neonatal intervention. MH protocols involve baseline fetal echocardiography, followed by administration of 60%-100% oxygen via a non-rebreather mask over 10-15 minutes at which time specific images and measurements are repeated.

Fetuses with hypoplastic left heart syndrome with intact or severely restrictive atrial septum are a subgroup of patients known to be at risk for urgent intervention at birth,²⁸ where MH testing could be particularly useful. Szwast et al.²⁹ found that a significant fall in pulmonary artery pulsatility index with MH was reassuring, whereas a less than 10% reduction predicted the need for immediate atrial septum intervention. More recently, velocity time integral of the pulmonary veins with MH has been used to predict restrictive interatrial septum. Mardy et al.³⁰ found that a pulmonary vein forward to reverse velocity time integral ratio of <6.5 with MH had 100% sensitivity and specificity for the need for emergent atrial septostomy at birth.

Restriction of the interatrial septum is also a key question in planning for delivery of neonates with transposition of the great arteries. Conventional assessment of the atrial septum evaluates the size and location of the foramen ovale, mobility of the atrial septum primum, and direction of shunting. MH has been proposed to augment diagnosis. Lack of a significant change in the pulmonary artery pulsatility index and direction of shunting at the atrial level have been suggested as predictors of the need for emergency septostomy.³¹ However, predictive parameters are not yet well established, and MH in transposition of the great arteries is an area of active research.

A further lesion where MH may be a useful adjunct is total anomalous pulmonary venous connection. Without MH, a vertical vein Doppler peak velocity >0.74 m/s³² and lower variability in the pulmonary veins (maximum-minimum/mean velocity) have been reported as predictors of obstruction.³³ However, prediction of obstruction can be difficult due to low pulmonary blood flow in the fetus. MH can be used to augment pulmonary blood flow and, therefore, “unmask” obstruction to aid in delivery planning.³¹

Acute MH is also being used to evaluate changes in cerebral blood flow.^34,35 It is currently unknown if, and how, such changes may be able to predict neurodevelopmental outcome. Chronic MH (continuous use of oxygen at home) has also been studied for both specific congenital heart lesions (small left heart structures)³⁶ and the effects on the brain. A report of smaller biparietal diameter in fetuses undergoing chronic MH has led to caution;³⁷ however, regulated studies remain ongoing.

There have not been reports of adverse events with acute MH. There is a return to baseline of fetal haemodynamics within 10-15 minutes of discontinuation of oxygen administration. Other than longer overall scan duration, there are relatively few barriers to implementing acute MH testing in a clinical environment. The introduction of testing inevitably involves a learning curve and a need to standardize indications for testing and imaging protocols.

Fetal cardiac magnetic resonance imaging

Fetal cardiac magnetic resonance (CMR) imaging arose from sophisticated advancements in fetal imaging techniques. CMR is challenging in the fetus for a variety of reasons. The fast fetal heart rate and small cardiovascular structures present challenges for temporal and spatial resolution. A proportionally larger field of view is needed to avoid wrap artefact from the mother. In addition to involuntary maternal movements, fetal CMR must contend with fetal cardiac and respiratory motion as well as unpredictable larger fetal movements.³⁸ A further limiting factor is that, for some pregnant patients, long scan times in a confined space may not be tolerable.³⁹ Despite these challenges, there have been remarkable advancements in fetal CMR, and it is now a useful adjunct to imaging in a range of circumstances.

Traditional CMR imaging typically uses electrocardiogram gating via an MR compatible electrocardiogram or a pulse oximeter to synchronize acquisition to the cardiac cycle. Unfortunately, these techniques cannot effectively trigger from the fetal cardiac cycle in the magnetic resonance imaging (MRI) environment, and initially, fetal CMR was limited to nongated techniques better suited to extracardiac anatomy. A breakthrough in the advancement of fetal CMR was the development of metric optimized gating (MOG).⁴⁰ MOG takes an oversampled data set and retrospectively applies a hypothesized pseudo-waveform. Image artefact from misgating is then measured against a set image quality metric and the hypothesized “trigger” iteratively adjusted until optimized.⁴¹ MOG was initially developed for the assessment of flow but has subsequently applied to a range of applications, including cine imaging (Fig. 3).

Figure 3.

Cardiac magnetic resonance images of a fetal heart with tricuspid regurgitation, right atrial dilation, and cardiomegaly in the third trimester obtained by metric optimized gating.

More recently, a reliable MR compatible Doppler-ultrasound device⁴² has become commercially available, potentially widening access to fetal CMR to a far larger number of centres. Practical advantages of the Doppler-ultrasound device include that it avoids the need for computationally intensive reconstruction, which allows images to be reviewed during a scan and repeated, if necessary.³⁸ It can also be used with existing vendor-supplied sequences.⁴³

Alongside these advances, multiple approaches have been developed to accelerate image acquisition and integrate motion correction. Fetal CMR is now able to provide a full assessment of 2-dimensional (2D) and 3D anatomy with volumetric reconstruction. Adapting techniques initially developed for fetal brain MRI, reconstruction of multiple scattered 2D acquisitions has been used to create 3D data sets encompassing the whole fetal heart. One advantage of this approach as compared with earlier single slice imaging is that it is relatively robust to fetal movements. A further development has been the integration of cine imaging to create 4D whole heart images.⁴⁴

Phase-contrast MRI (PCMR) is considered the reference standard technique for noninvasively measuring flow.⁴⁵ Fetal PCMR using the gating techniques outlined above can be used for assessments of flow in the fetal vasculature and calculation of left and right ventricular cardiac output.^46,47 An exciting new development in fetal CMR is the development of sequences for 4D flow assessment (flow in 3 dimensions with the fourth dimension of time).⁴⁸ Such sequences have the potential advantage of a more comprehensive assessment in a shorter duration.

CMR also has the unique capability to noninvasively assess oxygenation. Using the T1 and T2 relaxation time, it is possible to calculate both oxygen saturations and the hematocrit of blood within a vessel. Combining this information with PCMR flow data and using the Fick equation, assessment of oxygen delivery is both feasible and accurate.⁴⁹ Fetal CMR has revealed important findings in the circulations of fetuses with CHD with implications for brain growth and neurodevelopment.⁵⁰ The addition of acute and chronic MH to these investigations will reveal important insights in the near future.

Fetal CMR may be of benefit where conventional fetal echocardiographic approaches have proven inconclusive. In a series of 85 patients studied between 2015 and 2017 using 2D and motion corrected 3D volume acquisitions, Lloyd et al.⁵¹ identified 10 patients with anatomic abnormalities that had not been detected by fetal echocardiography. Similarly, in a mixed cohort of 31 patients with a wide range of CHD studied between 2017 and 2020 in Lund, the authors found that fetal CMR added clinically relevant information in 84% of cases.⁴⁷

One specific area where fetal CMR has been of demonstrable benefit is coarctation of the aorta. Coarctation represents around 7% of all live births with CHD⁵² and can be difficult to diagnose prenatally, with detection rates as low as 40%.⁵³ The need for postnatal intervention can also be difficult to determine. Using a combination of high-resolution 3D vascular imaging to assess the angle of displacement of the aortic isthmus and aortic flow, fetal CMR was able to correctly predict the need for intervention in 93% of neonates.⁵⁴ If reproducible and cost-effective, risk stratification of possible coarctation by fetal CMR could have a significant impact on clinical practice.

Individual case reports and series point to other potentially valuable clinical applications for fetal CMR, including the assessment of vascular rings,⁵⁵ aortopulmonary collaterals, and fetal cardiac tumours.^56,57 Fetal MRI may also be used for the assessment of extracardiac anatomy relevant to the cardiac diagnosis, for example, in cases of cardiac malposition,⁵⁸ or for the evaluation of the presence of “nutmeg lung” signifying pulmonary lymphangiectasia.⁵⁹

Fetal MRI has no known harm to the developing fetus. Official guidelines do not place a lower limit on gestation for scanning or specify scan duration; however, for practical reasons, fetal CMR is often performed in the third trimester when the fetus is larger, and commercial MRI systems limit the specific absorption rate to 4 W/kg.⁶⁰ Until recently, factors including scanning time, expense, and reliance on bespoke sequences have limited fetal CMR to few academic centres. However, some of the advances outlined above will likely translate to more widespread uptake. Ultimately, fetal CMR will not replace fetal echocardiography but will be increasingly useful as a complementary imaging modality as in postnatal life.

Advances in Fetal Cardiac Intervention

Fetal cardiac interventions may be pharmacologic, whereby administration of a drug to mother passes to the fetus via the placenta; percutaneous, involving fetal infusions or transcatheter intervention; or open fetal surgery. Pharmacologic intervention has existed for decades and is often directed at treatment of fetal arrhythmias. Newer approaches have been proposed in recent years, including the use of nonsteroidal anti-inflammatory drug (NSAID) therapy in patients with Ebstein anomaly or tricuspid valve dysplasia (EA/TVD) with circular shunting and mammalian target of rapamycin (mTOR) inhibitors for cardiac rhabdomyomas. Enzyme replacement therapy via fetal infusion for Pompe disease will also be addressed. Finally, the management of pericardial teratomas, including open fetal surgery, will be discussed.

Percutaneous fetal cardiac intervention is performed for select lesions^61,62 and will not be discussed in detail in this review. Briefly, fetal aortic valvuloplasty may be performed for severe midgestation aortic stenosis with evolving hypoplastic left heart syndrome in order to “salvage” the left ventricle and achieve a biventricular circulation by the time of birth. Large centres have reported that approximately half of patients achieve a biventricular outcome after technically successful intervention.63, 64, 65 Fetal pulmonary valvuloplasty may be considered for pulmonary atresia with intact ventricular septum with evolving hypoplastic right heart syndrome; however, the heterogeneous natural history of this lesion with 1-, 1.5-, and 2-ventricle outcomes renders criteria for candidacy difficult.⁶⁶ Established hypoplastic left heart syndrome with a highly restrictive or intact atrial septum is the most high-risk single ventricle lesion. Opening the atrial septum in utero provides better postnatal oxygenation,⁶⁷ but survival benefit has not been clearly demonstrated across centres.⁶⁸ This may be explained, in part, by chronic lung changes and pulmonary lymphangiectasia that is not reversed by relief of left atrial hypertension.

As in all areas of fetal medicine, maternal health and prevention of harm to the mother and fetus are paramount.⁶⁹ Percutaneous fetal cardiac intervention has a risk of fetal demise of approximately 10%. More innovative therapies may bring unknown challenges. Therefore, it is essential to have rigorous criteria for all types of fetal therapy and intervention by a multidisciplinary team of maternal and fetal specialists as well as ongoing surveillance of outcomes.

NSAID therapy for fetuses with Ebstein anomaly or tricuspid valve dysplasia with circular shunting

EA/TVD produces varying degrees of tricuspid regurgitation in the fetus. In cases at the most severe end of the spectrum, the development of pulmonary regurgitation results in a circular shunt. With this physiology, a significant proportion of blood flow is recirculated through the duct into the heart instead of to the rest of the fetus and the placenta. This often leads to insufficient systemic perfusion, organ failure, hydrops, and demise. Conventionally, management has been expectant with preterm delivery and early intervention; however, outcomes are poor with significant morbidity and mortality.^70,71

The duct is a critical part of the circular shunt and responds to maternal intake of NSAIDs with constriction. Therefore, it represents a target to mitigate circular shunting in the most severe cases (Fig. 4). By reducing circular shunting, cardiac output and systemic perfusion improve, prolonging pregnancy and thereby increasing the chance of neonatal survival.

Figure 4.

Diagram depicting circular shunting physiology in a fetus with severe tricuspid regurgitation from Ebstein anomaly or tricuspid valve dysplasia. The “X” in the PDA indicates where NSAIDs target. By constricting the PDA, fetal haemodynamics will improve. Ao, aorta; LA, left atrium; LV, left ventricle; NSAID, nonsteroidal anti-inflammatory drug; PA, pulmonary artery; PDA, patent ductus arteriosus; PFO, patent foramen ovale; RA, right atrium; RV, right ventricle.

Several cases and case series have reported the successful use of NSAIDs (indomethacin and/or ibuprofen) to reduce circular shunting. The largest series reports on 21 patients, of whom 15 accepted NSAID therapy between the 20th and 34th week of pregnancy. Most, 80%, of treated fetuses responded with ductal constriction and improved pulmonary regurgitation and haemodynamics. Of those fetuses where there was ductal constriction, 92% survived as compared with only 44% survival among fetuses who did not receive NSAIDs, or constriction was not achieved. Moreover, treated fetuses had pregnancy outcome at 36.1 weeks as opposed to 33.0 weeks in the untreated or unresponsive cohort.⁷² Torigoe et al.⁷³ reported a case series of 4 fetuses, supplemented by fetal CMR findings, to demonstrate the reduction in circular shunting volume with indomethacin and/or ibuprofen. Other authors have reported success with metamizole.⁷⁴

Fetuses with EA/TVD and circular shunting also have cerebral hypoperfusion and may be at risk for neurodevelopmental impairment. This presents a further argument for the use of fetal therapy in this population. There are potential risks, however, which include effects on fetal renal function and the development of oligohydramnios. As an emerging therapy for a relatively rare condition with low case numbers at individual centres, an international multicentre registry is underway to guide optimal treatment.

mTOR inhibitors for fetal cardiac rhabdomyomas

Cardiac rhabdomyoma is the most frequently occurring fetal cardiac tumour. Most cardiac rhabdomyomas are identified in the late second trimester of pregnancy and grow throughout the remainder of gestation before stabilizing and spontaneously regressing in postnatal life.⁷⁵ In a minority of cases, cardiac rhabdomyomas cause fetal compromise, including ventricular dysfunction, hydrops, and fetal demise. Cardiac rhabdomyomas >20 mm in diameter have been associated with poor outcomes.⁷⁶

In up to 80% of cases, cardiac rhabdomyomas are associated with tuberous sclerosis complex (TSC). TSC is typically caused by variants in the TSC1 or TSC2 genes.⁷⁷ TSC1 and TSC2 encode the protein products hamartin and tuberin, respectively, which act as a tumour suppressor complex to downregulate mTOR-mediated signalling that leads to cell growth and proliferation. Mutations in these genes lead to loss of tumour suppressor function. An understanding of this pathway led to the use of mTOR inhibitors (ie, sirolimus and everolimus) to address tumour burden in patients with TSC.⁷⁸

After several case reports of large, haemodynamically significant cardiac rhabdomyomas treated successfully with mTOR inhibitors postnatally, its first use as successful fetal therapy via transplacental passage was published in 2018.⁷⁹ Several groups have since published cases of successful treatment of rhabdomyomas causing fetal cardiac compromise (Fig. 5), including inflow/outflow obstruction, impaired function, effusions, and/or arrhythmias. Treatment has been started at gestations ranging from 23 to 35 weeks.⁷⁹ No significant maternal adverse events have been reported.

Figure 5.

Large rhabdomyoma with outflow tract obstruction at 26 + 2 weeks of gestation. The patient was treated with sirolimus both pre- and postnatally.

Sirolimus pharmacokinetics have significant inter- and intrapatient variability. Where cord blood has been sampled at delivery, higher fetal than maternal serum levels have been found, suggesting reduced clearance in the fetus and emphasizing the need to find the minimum effective dose.⁸⁰ Serial maternal serum monitoring is required after initiation; trough levels are typically targeted at 10-15 ng/mL.⁸¹ Monitoring of cell counts, liver function tests, and cholesterol is important because mTOR inhibitors may cause immune compromise, transaminitis, and hyperlipidaemia. Because of concerns regarding impairing wound healing, therapy is discontinued a couple of weeks before anticipated delivery. Optimal dosing and treatment protocols are not yet established; however, with interest in this form of fetal therapy both from cardiology and neurology perspectives in TSC,⁸² standardized guidelines may emerge.

Fetal enzyme replacement therapy for Pompe disease

Pompe disease, also known as glycogen storage disease type II, is an autosomal recessive condition caused by acid alpha-glucosidase deficiency, which results in progressive glycogen accumulation.⁸³ Unique among lysosomal storage disorders, Pompe disease may present with cardiac manifestations early in life.⁸⁴ In the most severe infantile onset form, hypertrophic cardiomyopathy presents within the first 12 months of life. Without treatment, children often die by 2 years of age due to respiratory insufficiency and/or left ventricular outflow tract obstruction.⁸⁵

Fetal echocardiography may identify a subset of patients with ventricular hypertrophy in the second trimester.⁸⁶ Those who present prenatally have a significant burden of morbidity and mortality.⁸⁷

Enzyme replacement therapy is the intermittent intravenous administration of specific recombinant enzymes. It has been a successful treatment strategy for lysosomal storage disorders starting with Gaucher disease in the 1990s.⁸⁸ In 2022, the first case of fetal enzyme replacement therapy was reported in a patient confirmed to have a known familial pathogenic variant by chorionic villus sampling. Two previous siblings in the family had died from the condition, 1 diagnosed postnatally and 1 prenatally with hypertrophic cardiomyopathy. Alglucosidase alfa was infused via the umbilical artery at a 2-weekly interval between 24+5 and 34 + 5 weeks of gestation. Throughout pregnancy and up to 13 months of age, echocardiography was reported as normal.⁸⁹

A phase I clinical trial to determine feasibility and safety of fetal enzyme replacement therapy in storage disorders is now underway. While globally a rare condition, this novel approach to treatment of a metabolic condition with serial infusions illustrates the breadth of possibility in the current era of fetal medical therapy.

Management of fetal pericardial teratomas, including open resection

Teratomas account for up to 15% of fetal cardiac tumours. Most commonly, they arise from the intrapericardial space at the junction of the superior vena cava, right atrium, and aorta.⁹⁰ Far more rarely, they may occur in the atrial or ventricular myocardium. Teratomas can grow rapidly and are frequently associated with pericardial effusions (Fig. 6). In up to 80% of cases, there is associated fetal hydrops. Before the current era of more aggressive management, mortality was high with approximately 50% of cases surviving the neonatal period.⁹⁰ After resection, this group of patients has excellent long-term outcomes.⁹¹

Figure 6.

Large pericardial teratoma with associated pericardial effusion at 32 + 4 weeks of gestation. This patient underwent successful postnatal resection.

For the majority of fetuses, management involves close monitoring of tumour progression and the accompanying pericardial effusion. If the pericardial effusion becomes significant, then ultrasound-guided aspiration or serial aspiration may be performed.⁹² The goal of this strategy is to prolong gestation until the fetus can be safely delivered and postnatal surgery performed.⁹³ Successful resection has also been reported in premature neonates.^93,94

In rare instances, there can be rapid tumour progression with a haemodynamically significant pericardial effusion in the second trimester. In such cases, open fetal cardiac surgery has been performed to resect the tumour in utero.⁹⁵ The first case of open resection of a pericardial teratoma was reported over 20 years ago in a hydropic fetus at 24 weeks.⁹⁶ The resection was successful, but the hydrops did not resolve. Mirror syndrome ensued necessitating caesarean section with neonatal demise shortly thereafter. In a recent case series of 8 fetuses with a prenatal diagnosis of suspected cardiac teratoma, Rychik et al.⁹⁵ reported successful in utero resection of a 4.9 cm teratoma causing haemodynamic compromise in a 24-week fetus. The rest of the pregnancy was uneventful with delivery at 37 weeks of gestation. In another case reported by Nassr et al.,⁹⁷ resection was successful, but the fetus still demised. Importantly, the subsequent pregnancy was complicated by uterine rupture at 33 weeks with neonatal death, highlighting the morbidity incurred from open fetal surgery. At the Hospital for Sick Children, we performed a successful resection in 2021; however, the fetus demised weeks later of unknown cause (unpublished).

The field of fetal surgery is moving from open to minimally invasive, particularly for repair of fetal myelomeningocele where similar safety profiles have been demonstrated.⁹⁸ As instrumentation and techniques advance, minimally invasive pericardial teratoma resection may become an option. Regardless of the approach, careful consideration of timing, fetal indication, and goals of the mother for future pregnancies is essential before proceeding with intervention.

Conclusions

Recent years have seen an acceleration in both fetal imaging techniques and approaches to fetal intervention. Challenging questions remain that continue to drive innovation, such as the adaptation of the fetus to postnatal life and the complex interplay among the heart, placenta, and brain. Enhanced understanding of the underlying mechanisms and physiology, along with advances in next-generation sequencing, will open doors for therapies to improve cardiac and noncardiac outcomes. Collaborative approaches have been and will continue to be essential to address these questions.