Prenatal exomes and genomes – so much new and so much more to learn

The last 10–15 years have seen dramatic changes in the field of prenatal genetic diagnosis and screening thanks to the introduction of next generation sequencing (NGS). A few months ago, our Journal published a special topic issue (volume 41 issue 10) on ten years of cell-free DNA-based non-invasive prenatal screening, one of the prenatal uses of NGS. Here we are bringing two special topic issues (volume 42, issues six and seven) that focus on the introduction and rapidly expanding use of prenatal exome sequencing (pES) and genome sequencing (pGS), another important change in the practice of prenatal genetic diagnosis driven by access to NGS. The decision to devote two issues of Prenatal Diagnosis to prenatal exome sequencing was fueled by a substantial number of quality primary research papers from around the world, along with an overwhelming response to requests for contributions from experts in this field. This is strong testimony to the current impact and growing importance of genome-wide sequencing on prenatal diagnosis.

The first reported use of prenatal genome-wide sequencing was to resolve the breakpoints of a balanced translocation that disrupted CHD7 in a fetus with features suggestive of possible CHARGE syndrome¹. This was quickly followed by a report that showed an incremental diagnostic yield of 10% with trio exome sequencing on proband and parental samples for 30 euploid fetuses or neonates with structural anomalies that were identified on prenatal ultrasound². Until 2018, reported prenatal sequencing experience remained limited to a few smaller series, primarily for highly selected cases by a few research groups or diagnostic laboratories, showing a wide diagnostic range varying between 6 and 80%³. This left many questions about its clinical use unanswered and in 2018, the International Society for Prenatal Diagnosis (ISPD), along with the Society for Maternal-Fetal Medicine (SMFM) and the Perinatal Quality Foundation (PQF) published the first position statement on prenatal genome-wide sequencing⁴. This original statement urged caution for the diagnostic use of pES outside of a research setting and highlighted its benefits, pitfalls and complexities, along with a set of recommendations for diagnostic application and counseling. In the four years since, the experience with trio pES has exponentially increased and several larger series have now been reported. In early 2019, two important papers demonstrated an incremental diagnostic yield of 8.5–10% of trio pES for unselected fetuses with one or more prenatally identified structural anomalies, increased nuchal translucency and fetal growth restriction^5,6. Since then, many more papers have been published on pES. These were systematically reviewed by Mellis and colleagues, who show that the accumulated data is now large enough that we can begin to provide estimates of which prenatal phenotypes are more likely to be molecularly solved with pES, such as skeletal abnormalities, central nervous system defects, and multiple anomalies⁷. This data is paving the way for a more personalized approach to decisions regarding which tests to order first in the time-sensitive setting of prenatal diagnosis. These data, along with those of several of the primary papers in these two issues, also show that the overall incremental yield of pES for fetuses with structural birth defects (after non-diagnostic standard testing), exceeds that of chromosomal microarray (CMA). The overall diagnostic rate is 31%, ranging from 15% in unselected cases to 42% in more selected cases where the suspicion for single gene disorders is higher. This new aggregate data support the messages in the revised ISPD position statement on prenatal genome-wide sequencing⁸ included in this issue. The ISPD have updated their opinion to say that while trio pES should continue to be carefully considered and only done with expert guidance by genetic professionals, it can be offered clinically for fetuses with structural birth defects. The new position statement also lists many important technical laboratory and clinical standards, counseling requirements, pitfalls, and suggests future research directions to be considered.

The newly included primary research papers and reviews in our two special topic issues cover different aspects of fetal genome-wide sequencing. There are several new series on the diagnostic yield for various indications, adding to the collective knowledge. Baptiste and colleagues combine data on trio pES for fetal central nervous system (CNS) anomalies from two large cohorts and show an incremental diagnostic yield of 7.2% in isolated cases and 19% when there are multiple CNS anomalies. These authors also show that trio pES yielded pathogenic (P) and likely pathogenic (LP) variants in 13% of cases with isolated mild ventriculomegaly and in 30% of cases with isolated agenesis of the corpus callosum (ACC), higher than many might have expected⁹. The results for ACC align well with data from Lei and colleagues in a different population with callosal anomalies, where pES yielded diagnostic findings in 34%¹⁰. Such information changes counseling, since standard testing with CMA has lower yield for these diagnoses and also highlights the importance of collaboration and combining data from multiple cohorts. Other series from around the world are included in the second special issue to be published in June, and report sequencing yield for a variety indications of including hydrops, cardiac defects and renal anomalies^11–16, again showing significant incremental diagnostic yields. Another interesting report on NIHF by Reytan and colleagues describes a fetus with NIHF with compound heterozygous deletions of PIEZO1. This highlights the importance of comprehensive data analysis and the potential of copy number variant detection from exome data¹⁷. Although variant interpretation is becoming more automated, it still requires thoughtful consideration of the clinical scenarios and inheritance patterns within families.

Among the series we report, two are specifically exploring pGS as opposed to pES^12,13. pGS has the potential to more easily ascertain balanced and unbalanced chromosomal rearrangements and CNVs. pGS has the future potential of replacing the time-consuming sequential testing strategy of karyotype or CMA first, followed by sequencing, or the potential to reduce costs of those services currently running arrays and sequencing in parallel. Studies on pGS are currently ongoing and results are eagerly awaited, but this is an exciting prospect for prenatal diagnosis, where a fast turn-around time from sample collection to genetic diagnostic results is highly desirable.

These new reports, together with those published previously⁷, not only show the diagnostic capacity of pES, but also demonstrate how newly identified variants and genes associated with fetal phenotypes are being discovered at an increasingly rapid pace. While this new knowledge will help us provide answers for parents faced with complicated pregnancies, it is important that all perspectives of introducing genome-wide testing into prenatal care are considered. What do we do with secondary and incidental findings in a prenatal scenario? How does undergoing such testing with resulting diagnoses, that are sometimes unexpected and uncertain, and difficult decisions to be made affect parents, providers and other stakeholders? These are important questions that we need to continue to consider and research. Vears and Amor address the complex issue of what to do with secondary and incidental findings and propose an elegant practical framework for decision making on what to report and to whom¹⁸. Another study showed that when given the choice, the vast majority (84%) of parents accept receiving information on secondary findings, regardless of whether sequencing was done prenatally or after birth¹⁹. Several other studies also explore different perspectives of parents and other stakeholders on prenatal sequencing^20–22. The overall message these studies convey is that while pES is generally favorably perceived, there is a need to remain vigilant of the consequences of uncertainty and receiving complex results.

Finally, our tools and knowledge to gain all the potential benefits from sequencing are still incomplete. Strategies for analyzing data, such as inheritance filtering, can limit the results obtained⁸. Two studies compare results from sequencing on neonates to what could have been detected prenatally if we had more access to sequencing and better strategies to analyze the data and show that many conditions could be diagnosed earlier^23,24. Moreover, at least one study showed that even for fetuses with no observable structural anomalies, P and LP variants can be detected²⁵. While this can be concerning, it demonstrates how we can learn from other clinical settings where sequencing is done, and instills optimism about what future research will teach us about currently unknown prenatal presentations of genetic diseases. As highlighted by Liu and Vossaert, new sequencing and data analysis technologies are already used in other clinical scenarios and can be applied to the most elusive prenatal cases²⁶.

There are many more important papers in these two issues, not all of which can be highlighted here, but the editors of Prenatal Diagnosis hope that you will find this compilation of what is new in pES useful for your current practice, as well as inspiring and thought provoking about what the future holds. Although we have learned much in the last 5–10 years about prenatal diagnostic exome and genome sequencing, much is still unknown and more work is to be done. To achieve this, global collaboration, dialogue and input from different stakeholders is critical. Giordano and Wapner elegantly summarize some such ongoing efforts and research priorities that are needed to lead us there²⁷.