Whole genome sequencing for newborns—The devil is in the details

Almost as if there is a ‘you are here’ pin on a moving timeline, 2022 appears to be an important year for genomics, both historically and looking forward. This year marks the 200th birthday of Gregor Mendel, the father of genetics, who discovered principles of heredity long before the concept of genes was known. His work remains influential to such an extent that the word ‘Mendelian’ was coined to describe single-gene traits that are transmitted from parents to children. Earlier this year, the first, gapless, telomere-to-telomere (T2T-CHM13) assembly of the human genome was published—a landmark achievement that finally completed end-to-end sequencing of whole chromosomes, filling in the many gaps remaining in the current human genome reference assembly (GRCh38). A complete human genome assembly will undoubtedly render a better understanding not only of the natural diversity of healthy human genomes, but also of the variation that is associated with the roughly 7000–8000 rare diseases that are estimated to exist.

Through the Precision Medicine Initiative in the United States and other similar national programmes, the potential of genomics as one of the keystones of healthcare has been brought to centre stage and is supported by plummeting sequencing costs, driving this technology within arm's reach. With more questions than answers remaining concerning technical, ethical, clinical and societal impacts, one of the most obvious, yet controversial, efforts to maximise the benefit derived from genomics would be the use of genome sequencing at birth. For example, in newborns manifesting symptoms of a suspected inherited disorder or in those who may develop a disorder later in childhood, genomic sequencing is a powerful diagnostic modality. DNA sequencing of newborns was already foreshadowed in 1990 by Walter Gilbert, a 1980 Nobel Prize winner in Chemistry, who extrapolated, from exponential growth of DNA sequencing, that all newborns would have their genomes sequenced by 2030 or 2040. These ideas have been echoed in the years since, asserting that knowledge of genetic risk empowers critical early interventions in life. Studies have since looked at the possibilities of using genomic sequencing to replace, enhance or supplement the battery of newborn screening tests that are done shortly following birth that screen for serious but rare, mostly treatable conditions. Many others have plunged into the numerous complex issues expected to emerge from using DNA sequencing technology in the newborn setting. This commentary aims to dissect some of these points by referencing a recently published pilot study that genome sequenced the DNA of 321 unselected newborns and analysed over 240 genes to compare the added benefit of genome sequencing versus routine newborn screening.

Before proceeding, it is essential to draw a clear distinction between sequencing healthy versus ill babies. There are sensational examples of lives being saved in the neonatal period due to deployment of rapid whole genome sequencing happening at exceptional, sometimes record-breaking speed, where a genetic diagnosis immediately leads to life-saving medical interventions. In apparently healthy babies with borderline newborn screening results, genomic sequencing has uncovered genetic variants that expand the detectable phenotypic spectrum with added genomic context and triggers reconsideration of preventative treatments. One example came from the BabySeq Project that applied exome sequencing in a cohort of healthy and ill newborns, where a newborn with borderline results for biotinidase deficiency (OMIM: 253260) was identified as harbouring biallelic variants in BTD, one of which was classified as a variant of uncertain significance that was immediately reclassified as likely pathogenic based on diagnostic enzyme levels. In the Jian et al. study, one baby was identified with phenylketonuria (OMIM: 261600) at newborn screening that was confirmed with biallelic variants in PAH (OMIM: 612349). Further examples show a clear benefit of diagnosing disorders that will likely manifest not directly at birth but later in childhood, such as a newborn who was identified with a KCNQ4 (OMIM: 603537) variant, causing autosomal dominant, postlingual sensorineural hearing impairment. The Jian et al. study uncovered many carriers for hearing loss and even two newborns with biallelic likely pathogenic/pathogenic variants in GJB2 (OMIM: 121011) and four with pathogenic MT-RNR1 (OMIM: 561000) variants, all in children who reportedly passed universal newborn hearing screening evaluation. Emerging gene therapies for hearing impairment require a morphologically intact auditory system, as suggested by normal newborn hearing screening results, revealing a potential critical therapeutic window for two newborns with one of the most common forms of pediatric onset deafness (GJB2) or preventative knowledge about aversion to ototoxic drugs (MT-RNR1) in an additional four newborns.

While these findings highlight tangible benefits of applying genome sequencing in healthy babies, these appear to be the exception rather than the rule, and it can be extremely complex without a family history of a disorder. Deep phenotyping yields an improved success rate if genomic medicine is initiated with a specific symptom or illness and can be very helpful to end a so-called ‘diagnostic odyssey’, allowing ‘children to be children’ and avoid spending their childhood at numerous specialist appointments. When timing is critical, in instances when a diagnostic answer comes later in childhood, there is a risk that the delay may have drastic consequences. A diagnosis at birth allows opportunities to break cycles of poor health and inequality. If this is essentially flipped, as is the case in a healthy newborn, the frame of reference of what to look for in a genome is lost. Without phenotypic context, it will be much more challenging to glean something relevant—after all, there is no diagnostic odyssey in a healthy newborn.

Genomics is not exceptional to other areas of medicine but requires several unique considerations, such as a balance of expected outcomes, recognition of uncertainties and complexities of genomic information, appreciation of broader implications for the family and other numerous ethical issues. Considerations with respect to information that has an immediate diagnostic value at the time of birth versus the possibility to support research and therapeutic development following the newborn period should be distinctly separated. Genome sequencing provides a resource that can be re-analysed for evolving questions over time without requiring new samples or new laboratory tests and, therefore, can be helpful to address a variety of questions later in life. For example, pharmacogenetic risk, of which nearly all (313/321, 97.5%) newborns in the Jian et al. study were carriers, may eventually be useful knowledge. Discussions around genomic testing usually happen in a genetic counselling setting where a patient or family is referred due to a history suggesting a rare condition. The chance of finding something meaningful is increased due to the presence of specific clinical features, while the predictive value of genome sequencing in healthy newborns is still under investigation. If a genetic cause of an adult onset disorder is identified in the genome of a child, it is possible that other family members either currently have or are at risk for the disorder, making the family the possible immediate beneficiary instead of the child, complicating the individual's right ‘(not) to know’. It is, therefore, important to study the ethical implications of deciding which results are fed back to a family for a child who may not want to know these results before becoming of consenting age. In instances when a finding does not emerge in the diagnostic screening and the genome is allowed to be fully analysed in a research context, a finding may fall out of the research side. Identifying what to report and when to report it is crucial and an area that merits further research.

While the authors of the Jian et al. study did not discuss consent and impact on healthcare setting, I would like to add a few brief words about these essential topics. A genome contains extremely private and sensitive information, making an understanding of the risks and harms brought on by genome sequencing fundamental. Complete anonymisation of a genome is impossible due to its uniqueness, and consent of this private information is an evolving task. Informed consent that was initially granted by the parents or legal guardians will need to be revisited at the age of 18 years, if not sooner. Additionally, when the child grows into a consenting individual, there is a question of where the consent responsibility should move, for example, from the maternity ward at initial parental consent to another department that would remain to be defined. Gene selection methodology would need justification; if mainly centered around ‘actionability’ what exactly does that mean (i.e., diet vs. cheaper interventions vs. costly interventions vs. interventions in clinical trials)? However, expansion of genomic analysis from a restricted newborn panel to a broader set of genes, even for which we may not have current therapies, may provide copious benefits to children and their families, as demonstrated by rare disease communities. Health systems are chronically stretched thin, and it should be expected that there is a different timescale for genomics versus routine newborn screening methods that are currently in place. While additional personnel would certainly be needed to run a newborn genome sequencing programme at scale, it is anticipated that entire pathways of care would be impacted and could end up saving healthcare costs in the long run, although the extent of which requires further investigation.

There are still many open questions around this engaging topic that are actively being researched—what is the benefit of knowing a given diagnosis at birth? Are there any effects on parent-baby bonding or, if parents are in denial following a diagnosis, does this risk delaying treatment? What are the benefits to participants and society for implementing pilot programmes and how to ensure that they remain equitable? How will it be possible to work with industry for acceleration of treatment and therapeutic development while controlling privacy and confidentiality through data access? The dialogue around genomics is a reflective, on-going and quickly evolving process that needs to consider the stakeholders at every part of the pathway. At its core is an alignment of expectations of participants and the public. One of the places that may uncover answers to these and many other questions likely will be from a pilot National Newborn Genomes Program that Genomics England, in partnership with the National Health Service, is launching in 2023. This study is uniquely positioned to address the effects of newborn genome sequencing on a national scale and includes research into aspects such as feasibility, consent procedures, use of data for research in future diagnostics and treatments and the long-term implications of storing and using a genome over an individual's lifetime. As genome sequencing is an extremely private dataset, there is a great need to handle these data correctly with room for stakeholders at the table during the design and pilot phase to ensure that such a programme is rolled out with the full trust and support of the public and with infrastructure and know-how in place.

Newborn genome sequencing has the power to potentially transform diagnosis and treatment of the sickest children when they are born. As not every newborn screening programme tests the same disorders in a uniform fashion in every country, there would be reason to believe that through genome sequencing, there would be potentially quick gains to be achieved via genomics. One could argue, as with other technologies and with significant improvements to sequencing technologies, that it is likely that a laboratory would, at some point before the newborn reaches 18 years of age, prefer newer data with newly developed methodologies that overcome the numerous current technological bottlenecks. The reality of genome sequencing is that it is not possible to openly discover the future. Specific questions enhance the potential output of a genome. The Jian et al. paper provides a basis on which serious investment on expanded newborn genome sequencing pilot programmes should build and was one of the first to pave the path for future work.

AUTHOR CONTRIBUTIONS

Dr. Barbara Vona prepared and collected the original literature, as well as wrote and edited the manuscript. Dr. Barbara Vona was responsible for the structural design, scientific quality and writing.

ACKNOWLEDGEMENTS

Not applicable.

FUNDING INFORMATION

Dr. Barbara Vona is supported through the German Research Foundation DFG VO 2138/7-1 grant 469177153 and through the Multiscale Bioimaging Cluster of Excellence (MBExC).

CONFLICT OF INTEREST

The author declares no conflict of interest. The paper was handled by editors and has undergone a rigorous peer-review process. Dr. Barbara Vona was not involved in the journal's review of/or decisions related to this manuscript.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

ETHICAL APPROVAL

Not applicable.

References

Nurk S, Koren S, Rhie A, et al. The complete sequence of a human genome. Science. 2022;376(6588):44‐53. [DOI: https://dx.doi.org/10.1126/science.abj6987]