In this article, I review some of the ethical issues that have arisen in the past when genetic testing has been done in newborns. I then suggest how whole genome sequencing may raise a new set of issues. Finally, I introduce a series of other articles in which the authors address different controversies that arise when whole genome sequencing is used in the newborn period.
Much has been written about ethical issues related to genetics. In fact, when the Human Genome Project was initiated in 1990, a component of the project was a focus on ethical, legal, and social implications of genomics. In the 1990s, whole genome sequencing (WGS) was not available. Thus, much of the research was focused on genetic testing by using other technologies. For newborns, much of the focus was on newborn metabolic screening, carrier screening, and genomewide association studies. WGS raises new and different ethical issues. To see why and how this is so, it is important to understand the ethical issues that arose from the use of these other forms of genetic testing.
Ethical Issues in Newborn Genomics Before WGS
Newborn Metabolic Screening
Newborn screening is an unusual form of clinical testing. In most states, it is mandated and done without parental consent. Some states allow parents to opt out, but most do not explicitly inform parents of this right.1
When newborn screening was first developed, the criteria were well defined and highly restrictive. Tests were designed “to identify infants with severe disorders that are relatively prevalent and treatable (or controllable).”2 The advent of new technology that made screening easier and less expensive led to the expansion of screening panels that can include diseases that are untreatable.3 Tests for such diseases were initiated before there was any empirical assessment of long-term outcomes, harms, and benefits.4
Newborn screening is associated with a number of well-recognized problems that are inherent to all screening tests.5,6 Compared with clinical testing, it leads to a high rate of false-positive results and the identification of children who may actually have a disease (true-positive results) but whose disease is milder than anticipated.7
Newborn metabolic screening can be used to test for a number of diseases at the same time, but each disease is tested for individually. Thus, there are a discrete number of tests that are done, and each test gives a result that is specific to that disease.
Thus, the ethical issues associated with newborn screening are focused primarily on determining which tests should be on the mandated panel of tests and how to ensure that parents understand the results and implications of a positive test result.
WGS, by contrast, is used to test for many diseases at the same time and is not used to specifically identify a single disease. Instead, it is used to identify variants in the genome that may or may not be pathologic. The complexity of interpretation of variants raises unique ethical issues.
Carrier Testing in Carefully Selected High-Risk Populations
Carrier testing has been conducted for a variety of conditions. Usually, people are selected for testing on the basis of their race, ethnicity, and family history. For example, screening for Tay-Sachs disease was originally conducted in Ashkenazi Jewish communities, screening for sickle cell disease was conducted in people of African descent, etc. There have been many studies of the psychosocial impact of such targeted testing for Tay-Sachs disease,8 muscular dystrophy,9 cystic fibrosis,10 and many other conditions. A meta-analysis of these studies revealed certain common themes.11 Such testing is predictably associated with anxiety and guilt in some people and euphoria or relief in others.
Currently, there is inconsistency in practices regarding communication about carrier status in different contexts. Many organizations oppose testing children for carrier status. Borry et al12 reviewed 14 policy statements from 24 different groups. They noted, “All the guidelines were in agreement that children preferably should not undergo carrier testing and that testing of children ideally should be deferred. All guidelines stated that it is in a child’s best interest for him [or her] to decide whether to be tested at some stage later in life.”
When carrier status is not the goal of screening but is instead an incidental finding, the consensus dissolves. The authors of 4 guidelines discussed situations in which carrier status was discovered incidentally (eg, during diagnostic testing, screening, or prenatal diagnosis or in a research context).13 Whereas guidelines from the British Medical Association14 and the American Academy of Pediatrics15 recommended that carrier status results obtained incidentally should be conveyed to parents, the American Medical Association16 and the German Society of Human Genetics17 recommended that this information should not be disclosed to parents or other third parties. Miller et al18 write, “The provision of carrier or predictive genetic testing is seen to infringe on the child’s autonomy and right to confidentiality because it forecloses on the child’s right to decide whether to seek this information and to whom it should be disclosed.”
WGS generates more incidental findings than most previous testing and thus can be used to screen for carrier status. If it is used in this way, it would raise all the same issues as single-gene testing for autosomal recessive conditions. If it is used for diagnosis, however, then researchers would not necessarily have to deal with the controversies about carrier testing and return of results.
Genetic Testing of Stored Samples in Biorepositories
There are a number of well-recognized ethical concerns about large-scale genomic biorepositories, including (1) the initial consent process, (2) concerns about privacy and confidentiality, and (3) reconsent for further studies or at the time when a child turns 18 years old.19
Consent issues are complicated because at the time when genetic material is first placed in a biobank, nobody knows what studies will be done on the material. In such situations, as Greely20 has noted, “It may be effectively impossible for them to get true informed consent; they can only get a redefined, watered-down version of informed consent.”
Concerns about privacy and confidentiality arise because to be useful, biobanks must have a way to link genetic material to the person from whom it came and provide detailed clinical information about that person that can be correlated with the genetic material. Most researchers have addressed this using computer technology that allows for identifiers to be “scrubbed” from medical records, creating essentially a deidentified clinical record that can still be linked to the matched but deidentified genomic material.21
Finally, the issue of when (if ever) to seek reconsent from people who were enrolled as minors but have now reached the age of majority is only beginning to be addressed. Some argue that parents should never be able to consent for the enrollment of minors.22 Others suggest that a robust process of recontact and reconsent at the age of majority will be sufficient.23 Given the rapid pace of change in the field, it is difficult to anticipate what we may be doing 5, 10, or 18 years from now both in terms of genomics and in terms of our ability to stay in touch with research subjects.
Ethics and WGS
When WGS became feasible in the clinical setting, it raised new and different questions. Should it be used as part of mandated newborn screening programs? If so, how should we deal with the massive amounts of information and uncertainty associated with the interpretation of WGS results? Botkin24 made a distinction between newborn screening and the screening of newborns: “The former is [used in the current public health system] to identify a range of actionable childhood diseases, whereas the latter is far more encompassing and could involve using sequencing (or other technologies) to identify adult-onset diseases and/or provide other information that would benefit patients throughout their lives.”
Since 2010, doctors have proposed that genomic testing of newborns be used not just for newborn screening, or the screening of newborns, but also for the clinical diagnosis of infants who are sick. That has been widely adopted. WGS is now used not only as a screening test in healthy children but also as a diagnostic test for children with illnesses that have been impossible to diagnose by using standard diagnostic testing. The nature of WGS is such that, however, such diagnostic testing will inevitably also give results about genomic variants that were not the target in testing. Thus, the diagnostic test is inevitably also a potential screening test.
The vast amount of information generated by using WGS leads to some well-recognized problems. Many researchers have developed algorithms to screen results and filter truly meaningless incidental findings from those that may have clinical significance. Solomon et al25 report an approach taken at the National Human Genome Research Institute for sorting through findings: “A working committee consisting of board-certified clinical geneticists, board-certified molecular geneticists, board-certified genetic counselors, bioethicists, and National Human Genome Research Institute [Institutional Review Board] members, as well as other genetic researchers, convened to discuss variants. … In several cases, experts in the study of individual genes and conditions were contacted when results remained equivocal.”25
These authors give an example to illustrate how difficult the process can be. They sought pathogenic genes in a pair of twins, 1 of whom had congenital anomalies and 1 of whom did not. Whole-exome sequencing was used to identify 79 525 genetic variants in the twins. After filtering artifacts and excluding known single-nucleotide polymorphisms and variants not predicted to be pathogenic, the twins had 32 novel variants in 32 genes that were thought to be likely to be associated with human diseases. Eighteen of these novel variants were associated with recessive diseases, and 18 were associated with dominantly manifesting conditions (variants in some genes were potentially associated with both recessive and dominant conditions). Only 1 variant seemed likely to be clinically relevant.
Clearly, WGS has a problematic signal/noise ratio. McKenna et al26 highlighted this, noting, “…a large development gap between sequencing output and analysis results.” Biesecker27 recognized the problems of information overload, saying, “A whole-genome or whole-exome result is overwhelming for both the clinician and the patient…(because) variants from [the] genome or exome range from those that are extremely likely to cause disease to those that are nearly certain to be benign and every gradation between these 2 extremes.”
At the same time, there are clearly some cases in which a signal can be discerned in the background noise. When that occurs, WGS may allow for difficult diagnoses to be made in a timely way that would be impossible to make by using any other diagnostic tools. Again, Biesecker27 said, “What is also clear is that some of these variants can be not only highly predictive of disease but their return can enable life-saving treatment.” And Green imagines “the routine use of genomics for disease prevention.”
The Specific Ethical Issues of WGS for Clinical Diagnosis
When WGS is used for clinical diagnosis, many of these problems that arose in other contexts are avoided. But some new issues are created. Ormond et al28 analyzed and summarized 3 sets of problems that are likely to arise in clinical WGS testing.
First, patients and parents need to understand the limitations of sequencing. For example, some methods do not reveal translocations, large duplications or deletions, copy number repeats, or expanding triplet repeats. When 2 variations are identified in the same gene, some WGS analyses cannot be used to establish whether those variations are in copies of the gene on different chromosomes or in the same copy of the gene, a distinction that is crucial for diagnosing recessive disorders.
Second, it is difficult to maintain up-to-date information on every known genetic disease. No centrally maintained repository of all rare and disease-associated variants currently exists.
Third, the availability of WGS might result in a large increase in testing by cautious physicians. Such an increase would not only raise health costs but would subject patients to the physical and psychological costs of increased testing.
It is hard to know how these theoretical issues will play out in the world of clinical medicine because WGS has not been used widely enough to yield the relevant outcome data. Furthermore, the technology for WGS is improving rapidly, as are the databases that allow for a more accurate interpretation of results. Thus, assessment is a moving target.
In this special supplement, I address some of these issues and report the results of early forays into the use of genomic sequencing for newborns.
Pereira and colleagues surveyed parents and clinicians about their knowledge and attitudes about WGS. They wanted to find out how these groups thought about WGS compared with traditional newborn screening.
Gyngell and colleagues analyze 3 issues that have arisen in their early forays into diagnostic WGS with a rapid return of results: (1) the value and meaning of gaining consent to a complex test in a stressful, emotionally charged environment; (2) the effect of rapid diagnosis on parent-child bonding and its implications for medical and family decisions, particularly in relation to treatment limitation; and (3) distributive justice and whether the substantial cost and diversion of resources to deliver rapid genomic testing in the NICU can be justified.
Shieh discusses some of the issues that arise as we discover more about incomplete penetrance and the complexity of communicating uncertainty to parents.
VanNoy and colleagues examine some of the controversies surrounding testing for carrier status.
Lantos discusses the conundrums that arise when a genetic variant is classified as clearly pathogenic but the patient remains asymptomatic.
Holm and colleagues discuss the issues that arise when a genetic variant is associated with an adult-onset condition but is discovered when testing a newborn.
Rubanovich and colleagues analyze the complexities of communication when a patient is on a “diagnostic odyssey” and the ways that WGS might be used to resolve some of those complexities.
Knapp and Lantos report the results of focus groups with neonatologists about their attitudes toward diagnostic WGS for infants who are critically ill in the NICU.
Taken together, these articles represent some of the first forays into understanding the ethical issues on the ground as WGS is rolled out in the clinical context of newborn care.
Dr Lantos wrote, reviewed, and revised the manuscript and approved the final manuscript as submitted.
FUNDING: Funded by the National Human Genome Research Institute (U19 grant U19HD077693; subaward agreement 3282-S1-A2).
POTENTIAL CONFLICT OF INTEREST: The author has indicated he has no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: The author has indicated he has no financial relationships relevant to this article to disclose.