Current Controversies in Prenatal Diagnosis—Conference Debate 2024: All Fetuses Undergoing Fetal Therapy Should Have Exome Sequencing

ABSTRACT

This manuscript summarises the debate held at the 2024 annual meeting of The International Society for Prenatal Diagnosis (ISPD). Experts discussed whether all fetuses undergoing fetal therapy should undergo exome sequencing. Arguments in favor included that, with increasing experience and better clinical availability, exome sequencing can yield valuable diagnostic and prognostic information beyond what is available from karyotyping and microarray. This additional information is often helpful in counseling parents and provides a better understanding of fetal conditions, allowing for personalised medicine and supporting advancements in disease‐focused fetal therapies. On the contrary, however, significant concerns regarding availability and health equity were raised. Moreover, potential delays in care incurred by exome sequencing may negatively affect outcomes of fetal intervention. Finally, as the information gathered from genetic testing may or may not affect pregnancy management decisions beyond termination of pregnancy, many families may choose not to undertake testing. The arguments of both debaters document current controversies in exome sequencing and genetic testing in general. This was also reflected in a divided audience vote at the end of the debate.

Summary.

• What is already known about this topic?

  • ◦Many genetic diseases can underlie fetal anomalies, including single gene disorders.
  • ◦Ideally, fetal surgery is reserved for pregnancies in which the fetus would maximally benefit from intervention. Historically, such pregnancies were those with a single ‘isolated’ structural anomaly or those without a severe genetic disease in the fetus.

• What does this study add?

  • ◦This manuscript summarizes the arguments and highlights the controversies around exome sequencing for fetuses undergoing prenatal therapy.

1. Introduction (Tim van Mieghem, Moderator)

In this debate, two experts in the field of prenatal genetics and fetal therapy discussed whether fetuses undergoing in utero surgeries should undergo exome sequencing (ES). This discussion was very timely because over the past 3 decades, the number of fetuses undergoing fetal therapy has increased dramatically. Indeed, thanks to rapid advances in the field of ultrasound technology, and the implementation of routine ultrasound programs in most countries, more and more fetal structural anomalies are now diagnosed prenatally. A recent Cochrane review showed that mid‐trimester ultrasound screening in low‐risk populations identified approximately 50% of all anomalies. Addition of a first‐trimester anomaly scan can result in more than 80% of anomalies being detected [1]. Growing research in the field of embryology and fetal physiology has also given us a better understanding of the natural history of these anomalies, and we can now identify which fetuses are at higher risk of severe neonatal morbidity or mortality, or which conditions progress over the course of pregnancy. This high‐risk population is the ultimate target for fetal therapy as waiting to treat these fetuses postnatally may be suboptimal and the potential benefits of prenatal intervention can outweigh the risks of an intrauterine procedure.

Since the initial proof of concept in the 1960s that fetal therapy can rescue very sick fetuses affected by rhesus disease [2], dedicated research centers have expanded the scope of fetal therapy to now include a broad range of fetal conditions in virtually every single organ system. Some of these therapies are supported by level 1 evidence from randomized controlled trials (fetoscopic laser for twin‐twin transfusion [3], open fetal surgery for spina bifida [4], fetal tracheal occlusion for diaphragmatic hernia [5]), whereas others rely on a sound track record (e.g. intra‐uterine transfusion, thoraco‐amniotic shunting for pleural effusions [6]), or are still experimental (e.g. fetal cystoscopy [7], interventions for gastroschisis or vein of Galen malformations [8]).

The number of fetal therapy centers has increased in line with the number of fetuses eligible for prenatal therapy and almost every high‐income country now has one or more such centers. This has increased the access to care for patients but has also resulted in significant variability in institutional practices. Indeed, when exploring the potential benefits of fetal therapy, most research studies required fetuses to have an ‘isolated’ single fetal anomaly to be eligible for study participation. This means that fetuses with anomalies in multiple organ systems and those with associated genetic anomalies were typically excluded from trials. However, ‘real world’ implementation of fetal therapy, increased experience and a (sometimes irrational) patient or surgeon belief in the benefits of therapy has resulted in some centers having less stringent selection criteria for fetuses undergoing prenatal therapy, thereby also offering surgery to those with additional underlying structural or genetic anomalies [9].

In parallel with the advances in fetal therapy, genetic testing options have substantially evolved in recent decades. Genetic testing available in the prenatal setting has evolved from examination at the chromosomal level using karyotype or chromosomal microarray analysis (CMA) to next generation sequencing in the form of gene panels, ES, and even genome sequencing. ES has become increasingly available in many countries, and at some institutions, ES is routinely offered to all patients whose pregnancies are complicated by fetal anomalies not explained by karyotype or CMA.

Given the changing landscape in fetal therapy and prenatal genetic testing, we asked our experts whether all fetuses undergoing fetal therapy should undergo ES. Of note, for the sake of the debate, both speakers were asked to make their case as strongly as possible, irrespective of whether this was their clinical viewpoint.

2. The Case In Favor (Dr. Teresa Sparks)

Despite the evolution in both available genetic tests and fetal interventions, our approach to informing decision‐making about fetal interventions has remained stagnant. Chromosomal microarray (CMA) became the recommended genetic test for fetal anomalies over a decade ago [10], yet most institutions and clinical trials still require only karyotype before a fetal intervention. The goal of performing a karyotype in this situation has historically been to determine eligibility for fetal interventions and clinical trials. The types of genetic abnormalities that could preclude a fetal intervention are debated, with important considerations being the anticipated long‐term prognosis, expected impact of the intervention given the genetic diagnosis, and the unique goals and value set of each patient and family undergoing testing. Identifying a fetal genetic diagnosis in pregnancies already faced with a fetal anomaly has the potential to add valuable context for shared decision‐making and can inform key decisions such as whether to perform a fetal intervention, the cadence of fetal surveillance, and planned interventions after delivery.

2.1. Diagnostic Yield of Exome Sequencing in Conditions Eligible for Fetal Surgery

A large number of studies have been published in recent years, demonstrating the diagnostic yield of prenatal ES. It has become quite clear that ES is useful for identifying underlying single gene disorders when a fetal anomaly is present. Reported diagnostic yields range from the single digits to over 60% depending on the population studied and the types of fetal anomalies tested [11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22]. Further, the diagnostic yield of prenatal ES is approximately 11%–34% for anomalies that are often amenable to fetal interventions, such as non‐immune hydrops fetalis and other fetal effusions, congenital anomalies of the kidneys and urinary tract, left‐sided obstructive cardiac lesions, and congenital diaphragmatic hernia [18, 19, 20, 22, 23]. For example, fetal effusions and non‐immune hydrops fetalis may be amenable to intrauterine transfusions or placement of shunts, and the chance of finding an underlying single gene disorder with prenatal ES is approximately one third among those that are euploid [20, 23]. Among fetuses with left‐sided heart lesions for which aortic valvuloplasty might be considered, ES identifies a single gene disorder in about 18% [19].

The types of single gene disorders underlying fetal anomalies that are detected by ES vary widely, and many are well characterized and published [11, 12, 20]. As discussed further below, there is important clinical utility in identifying an underlying genetic disease, as this provides additional context for both patients and clinicians as they approach important decisions for these complicated pregnancies such as whether to perform fetal interventions. However, there can be uncertainty, such as when a rare genetic disorder is identified and prognostic expectations are less certain, or when a variant of uncertain significance (VUS) is detected which occurs in approximately 4%–9% of cases [12, 20, 23]. While the identification of uncertain information is certainly possible with prenatal ES, this is also the case with nearly every other test that we order. Cell free DNA can return atypical or ambiguous findings, prenatal ultrasound can uncover incidental or uncertain findings in the fetus or in the pregnant individual, and prenatal CMA can also return VUS or incidental findings. The course of pregnancies with fetal anomalies is often uncertain at baseline and perinatal outcomes cannot be predicted with certainty; therefore, the results of prenatal ES should be interpreted within this context. Finally, a negative result of prenatal ES is still the most likely outcome among pregnancies with fetal anomalies that undergo this testing. Negative findings do not preclude the presence of an underlying genetic disorder, but they do significantly lower the chance of a single gene disorder being present, which allows patients and providers to focus on the severity of the anomaly itself.

2.2. Clinical Utility and Impact of Prenatal Exome Sequencing

Several studies were published on the clinical utility of prenatal ES [24, 25, 26]. In one observational series, patients whose pregnancies were complicated by fetal anomalies were offered ES and clinicians were then surveyed to understand how the results impacted clinical decision‐making [24]. Clinicians reported that in 68% of cases, ES results impacted important prenatal decisions, whether they were positive or non‐diagnostic. Results were used to inform decisions about termination of pregnancy, fetal interventions, and peripartum interventions. Changes to prenatal and perinatal clinical management that may follow results of ES are quite broad, encompassing reproductive decision making and considerations about eligibility for in utero therapies, types and frequency of antenatal surveillance, and delivery decisions such as timing, location, and mode of delivery.

Prenatal ES has the potential to uncover not only a breadth of genetic diagnoses missed by conventional karyotype or CMA but also a breadth of diseases that were historically undiagnosed in utero. Prenatal diagnosis is inherently limited by ongoing fetal structural development, an incomplete understanding of fetal manifestations of genetic diseases, and the inability to detect certain features such as neurodevelopmental delay or hearing impairment. As a result, adjunctive testing with ES can be very beneficial for identifying an underlying genetic diagnosis, including in situations where there is uncertainty about the exact diagnosis, there are a combination of non‐specific features that are not pathognomonic for a particular disease, or when a specific diagnosis or set of diagnoses are suspected but molecular confirmation is needed. Establishing a genetic diagnosis in these already complicated pregnancies that may undergo a fetal intervention adds important context for not only patients but also providers as they consider the fetal and maternal risks of the procedure under consideration. A specific genetic diagnosis allows clinicians and patients to consider unique aspects of the particular disease and anticipated long‐term prognosis, and allows patients to more concretely consider their personal value sets and approach upcoming clinical decisions.

For example, consider these scenarios of three fetuses that presented quite similarly but had very different underlying genetic diagnoses uncovered by ES with widely variable prognoses. The first fetus presented in the second trimester with elevated middle cerebral artery (MCA) Doppler measurements, ascites, and organomegaly. Although serial intrauterine transfusions (IUTs) were performed to treat the fetal anemia, the ascites persisted. ES identified a heterozygous variant in PRF1 associated with hemophagocytic lymphohistiocytosis, which results from a severe immune response and carries a high risk of perinatal mortality [27]. A second fetus presented later in the third trimester with ascites, skin edema, echogenic bowel, and elevated MCA Doppler measurements. Serial IUTs were performed to treat mild fetal anemia, but the abnormalities described persisted. ES identified compound heterozygous NPC1 variants associated with Niemann Pick C disease, an inborn error of metabolism leading to neurologic and other manifestations along with early death [28]. A third fetus presented with elevated MCA Doppler measurements and ascites. Serial IUTs were performed and the ascites improved slightly. ES identified a heterozygous PIEZ01 variant associated with dehydrated hereditary stomatocytosis, which typically carries a much more favorable long‐term prognosis [29]. Thus, the anticipated outcomes for each of these three fetuses were quite different despite similar prenatal features. Without ES, it can be very challenging to accurately anticipate outcomes for pregnancies like these, and the specific genetic diagnoses provide additional context for families who are making decisions about fetal interventions and are faced with significant uncertainty.

It is important to highlight that there is also utility in a non‐diagnostic result of ES, as this significantly decreases the possibility of an underlying fetal genetic disease. For example, when clinicians and patients are considering the fetal and maternal risks versus benefits of placing a fetal thoraco‐amniotic shunt for pleural effusions, it can be helpful to know that there is not a clear ciliopathy, inborn error of metabolism, or other genetic disease that would further alter the fetal prognosis. A non‐diagnostic ES result can be informative for similar reasons when serial procedures such as IUTs are under consideration or when maternal risk is higher based on either the fetal intervention itself or a complex medical history.

2.3. Parental Perspectives

Our patients are often seeking answers and clarity amidst significant uncertainty when fetal anomalies are diagnosed during pregnancy. There are many reasons why patients wish to undergo prenatal ES, many of which overlap those discussed above under clinical utility from a provider perspective. Examples include closure or gaining more information about the reason for fetal anomalies, informing personal decisions around whether to proceed with termination of pregnancy, clarifying the postnatal prognosis that is likely to be expected, preparing to care for their child after birth, understanding recurrence risk as future pregnancies are considered, and importantly, guiding important decisions about prenatal care including whether to proceed with in utero interventions [30, 31, 32]. By providing prenatal ES to our patients undergoing fetal interventions, we offer them the opportunity to gain valuable information to support how they navigate these complicated pregnancies and reproductive decision‐making.

2.4. Improved Availability

Although in the early stages, ES was a difficult test to perform with limited availability, clinical implementation has shown that ES results can be returned as quickly as within 2 weeks. This, in association with the fact that fetal surgery is often not urgent—particularly now that anomalies can be diagnosed as soon as the first trimester of pregnancy—often allows time to wait for ES results prior to proceeding with surgery. ES can also be ordered with results pending at the time of an initial fetal intervention, especially in situations where repeated procedures such as IUTs are anticipated.

Although prenatal ES is not uniformly accessible, it is certainly becoming increasingly available in many countries across the world. There are many commercial companies that offer prenatal ES along with a large number of studies, academic medical centers, and national networks and government agencies. We must continue to look toward solutions to increase equity and accessibility of ES for fetal anomalies.

Baseline costs of sequencing have decreased over time, although additional costs such as those for bioinformatic analyses, labor and expertise, and reporting remain important for clinical sequencing [33]. While few studies have evaluated the cost‐effectiveness of prenatal ES in the setting of fetal anomalies, and none in the context of fetal interventions specifically, existing evidence suggests that ES helps to avoid extensive evaluations (the ‘diagnostic odyssey’) and is likely to be cost effective. One study, for example, modeled multiple strategies of gene panels, ES, and gene panels followed by ES for pregnancies complicated by fetal effusions with non‐diagnostic results of standard of care testing (chromosomal microarray and/or karyotype) [34]. This study found that proceeding directly to ES after non‐diagnostic results of standard of care testing was the most cost‐effective strategy.

2.5. Current Professional Society Recommendations

Prenatal ES is becoming increasingly recommended by professional societies, as evidenced by the following examples. An updated position statement from the International Society for Prenatal Diagnosis published in 2022 highlights the clinical utility of ES for prenatal diagnosis [35]. The authors emphasize that prenatal sequencing is beneficial for a major single anomaly or multiple anomalies when CMA does not identify a genetic etiology. The Canadian College of Medical Geneticists published a position statement in 2022 [36], which supports the consideration of genome wide sequencing for congenital anomalies that are isolated or that affect multiple systems. The authors also write that if understanding the cause and natural history of the condition could guide the degree of intervention, then prenatal genome wide sequencing may be appropriate. The American College of Medical Genetics and Genomics (ACMG) also published a statement in 2019 on the use of fetal ES for prenatal diagnosis [37]. The authors discuss that ES may be considered for a fetus with ultrasound anomalies when standard CMA and karyotype do not provide an answer, and that a genetic diagnosis can assist in determining the fetal prognosis and can inform decisions such as in utero therapy. Further, the ACMG published a practice guideline in 2021 on the use of exome and genome sequencing for pediatric patients with congenital anomalies [38]. While this document is focused on the pediatric population, there are important extensions to consider for the fetal population. The authors highlight that results of sequencing can obviate the need for more extensive and expensive work ups, clarify the expected prognosis, make a patient eligible for a targeted therapy, and allow clinicians to avoid recommending ineffective interventions that also increase risk.

2.6. Thinking Forward

Identifying a specific genetic disease in a fetus allows clinicians to pursue a precision‐based approach to current and future treatments, rather than continuing to offer fetal interventions in reaction to the ultrasound abnormality alone. A wide breadth of fetal therapies that are emerging and in clinical trials have the potential to change the face of fetal interventions [39, 40]. By adding ES to the diagnostic evaluations when a fetal abnormality is identified and a fetal intervention is under consideration, we will not only improve the precision of our care for individual patients but also move the fields of prenatal diagnosis and fetal precision medicine forward. For these reasons, I ask that you support the motion that all fetuses undergoing fetal therapy should have ES.

3. The Case Against (Dr. Rogelio Cruz‐Martinez)

3.1. Novel Tests Are Difficult to Access and Take Time to Complete

When the International Fetal Medicine and Surgery Society (IFMSS) established criteria for fetal therapy in the 1980's, one of the main selection criteria was the absence of chromosomal abnormalities or genetic disorders [41]. Prenatal genetic testing in that era, however, was very different from what is now available. Indeed, in the 1980s, standard prenatal genetic testing consisted of karyotyping. Fluorescence in situ hybridization (FISH) testing for aneuploidy became available on a larger scale in the 1990s, microarray in the early 2000s, and nowadays many centers are exploring exome or genome sequencing. More testing can result in a better understanding of disease but also has its limitations, particularly when looking at novel tests that are not yet well implemented. Indeed, exome sequencing, given its relative novelty, is not available to the same extent as microarray testing. The latter can be done at a reasonable cost in most countries, including many low‐ and middle‐income countries with a fairly short turnaround time (typically 7–10 days). Exome sequencing on the other hand is only available in high‐income countries, costs run in the thousands of dollars, pre‐ and post‐test counseling requires extensive expertise [35] and comes with a significant workload and results may take weeks to become available, particularly when a sequential testing strategy is followed (ES to be done only in those with a normal microarray).

The above raises significant concerns for 2 reasons.

First, from an equity of care perspective, when requiring all fetuses to undergo exome sequencing prior to fetal surgery, one automatically excludes a significant group of eligible patients from lifesaving interventions. Where one could suggest that in certain settings exome sequencing could be beneficial, a blanket statement to require exome sequencing for all undergoing fetal therapy, which is the topic at debate here, seems unreasonable.

Second, many fetal interventions cannot be delayed until exome testing results become available because the fetus will suffer irreparable damage or die while waiting for results. In keeping with this argument, some fetal pathologies such as massive pleural effusions with hydrops or hydropic fetuses with lung masses are at risk of imminent intrauterine fetal demise and therefore, such high‐risk cases are candidates for emergency fetal surgery (pleuroamniotic shunt [42] or laser ablation of the feeding artery). Even when non‐lethal, many fetal conditions in which prenatal therapy would be considered progress during pregnancy. Thus, any delay in achieving a genetic test result may be associated with poorer survival outcomes. For example, considering that the main objective of intrauterine spina bifida surgery is to decrease the risk of developing hydrocephalus after birth and to improve motor function, an intrauterine surgical repair in the presence of severe ventriculomegaly (ventricular width above 15 mm) or once significant nerve damage has occurred would not be of not benefit. Studies have clearly shown that earlier intervention for spina bifida results in improved outcomes [43, 44]. Similarly, lower urinary tract obstruction (LUTO) will result in ongoing renal damage and critical aortic stenosis can progress to hypoplastic left heart syndrome during fetal follow‐up [45]. Intervention prior to disease end‐stage is required to optimize infant outcomes. This is the main reason why some fetal surgery centers consider performing either vesico‐amniotic shunting or fetal cystoscopy before a genetic test in their clinical protocols [46, 47]. A last example here is fetal congenital diaphragmatic hernia (CDH). These fetuses have severe pulmonary hypoplasia and treatment with fetoscopic endoluminal tracheal occlusion (FETO) can improve survival. Intervention earlier in pregnancy (ideally prior to 30 weeks) results in more fetal lung growth than later in pregnancy, as evidenced by comparing the results of 2 randomized trials [48].

Particularly when considering the incidence of late diagnoses and inevitable delays in care caused by a need for referral to fetal therapy centers, adding additional delays in treatment while waiting for genetic testing results may have a profound impact on ultimate outcomes of fetal therapy.

Of note, in the different scenarios described above, genetic testing can be done on amniotic or pleural fluid during the fetal intervention, but obviously results become less relevant as only available after the fact. It is therefore questionable whether those tests should be undertaken at all.

3.2. Lower Yield of Exome Sequencing in Isolated Conditions

Where proponents of exome sequencing often quote its high diagnostic yield (up to 30%) and superiority over testing with genetic panels, the yield of exome sequencing in isolated structural anomalies is much more limited. Indeed, most professional societies support ES in complex fetal disorders, where anomalies are present in multiple organ systems, and in those at high risk of monogenic disease, but not in those with isolated lesions. However, these are not the fetuses undergoing in‐utero interventions. Indeed, fetal surgery is typically only offered to those where the lesion requiring therapy seems isolated on ultrasound. This, in combination with improvements in ultrasound technology and increased operator experience, has resulted in better fetal phenotyping with lower risks of ‘missing’ fetal anomalies. Therefore, a genetic test before fetal surgery may be unnecessary. When looking at a cohort of fetuses with diaphragmatic hernia for example, a condition well known to be associated with a high risk of genetic disease, exome sequencing only had a minimal incremental yield of 0%–10% in those with isolated lesions, compared to 10%–17% in those with non‐isolated disease [18, 49, 50]. Variability in rates reported may be caused by the level of ultrasound screening patients underwent, where less detailed phenotyping may result in lower rates of associated anomalies identified and a higher yield of genetic anomalies in ‘isolated’ anomalies.

Additionally, although these variants may explain the diaphragmatic hernia, the link with fetal and neonatal outcomes is not always clear and test findings may not affect fetal or neonatal care. Similarly, in isolated fetal heart disease, exome sequencing only identifies pathogenic and likely pathogenic variants in 9% of cases [51]. In fetuses with lung masses, genetic anomalies are only seen in 1% and in isolated spina bifida, the risk of genetic anomalies is less than 2%. Practically speaking, for the latter cohort, this would mean that we would delay 98 fetal procedures—and potentially worsen the outcomes of these fetuses—to detect only 2 with genetic anomalies. Genetic conditions are more common in those with primary hydrothorax (16%) and in fetuses with megacystis (30%), yet the majority of these are aneuploidies or anomalies that would be detected with microarray.

3.3. Parental Perspectives

Finally, let's not loose of the parental perspective. First, some parents may not be interested in ES as the results of the test would not affect their pregnancy decisions. Similar to what we see with first trimester aneuploidy screening, a significant number of parents will decline genetic testing and pregnancy termination, even in countries where this is legal. Cultural and religious factors may impact this. Second, even though ES may provide more information, the test can also result in unnecessary anxiety. The PAGE study has clearly shown that the risk of a variant of unknown significance (VUS) can be as high as 4% [12] and this can certainly bring confusion or cause parents to decline testing. Moreover, a non‐informative or even negative result can be overly reassuring [52].

Putting together the above arguments, parent preferences, low test yield in specific conditions as well as (unnecessary) delays in care and a lack of equity prevent us from stating that all fetuses undergoing fetal therapy should have exome sequencing. For these reasons, I ask that you reject the motion that all fetuses undergoing fetal therapy should have ES.

4. Conclusions

This excellent debate brought up the many pros and cons of exome sequencing. Exome sequencing may be useful to identify single gene disorders which would be ‘missed’ using microarrays and could provide additional information for decision‐making for both patients and clinical providers. Additionally, information from exome sequencing can provide deeper insight into fetal conditions and allow a personalized, precision‐based approach to management. On the other hand, current concerns with test availability and cost, delays in care when interventions are only provided after a normal result, as well as parental preferences may prevent general implementation. An audience vote at the end of the debate showed a divided response, highlighting the ongoing controversy.

Conflicts of Interest

Tim Van Mieghem and Teresa Sparks are associate editors of the journal Prenatal Diagnosis.

Data Availability Statement

The authors have nothing to report.