Ethics of Cell-Free DNA-Based Prenatal Testing for Sex Chromosome Aneuploidies and Sex Determination

Introduction

Cell-free DNA-based prenatal testing (further: cfDNA testing) is a new technology that may both be used as a diagnostic test in the context of prenatal diagnosis of monogenetic disorders and as a second or first tier test in the context of prenatal screening for chromosome abnormalities. cfDNA testing is commonly also known as “noninvasive prenatal testing” (NIPT). When specifically used as a diagnostic test for monogenetic disorders, it is often referred to as “noninvasive prenatal diagnosis” (NIPD), whereas “noninvasive prenatal screening” (NIPS) has been proposed for its use as a screening test [ ]. However, in line with other chapters in this volume, we will speak of cfDNA testing.

In the context of screening for common autosomal aneuploidies, cfDNA testing is available in a growing number of countries [ ], either as a testing offer made available to patients through individual practitioners or practices in the private sector, or in the context of national or regional prenatal screening programs. While the emphasis in the literature is on how best to use this new technology for improving existing prenatal screening for Down syndrome (trisomy 21) and the other two common autosomal aneuploidies Patau and Edwards Syndrome (trisomy 13 and 18) [ ], part of the debate is also about how cfDNA testing may lead to a widening of the scope of prenatal screening beyond these conditions.

While with cfDNA testing technology it may eventually become possible to turn noninvasive prenatal screening into a comprehensive fetal genome scan, more immediate candidate conditions are sex chromosome aneuploidies (SCAs), microdeletions and -duplications and rare autosomal trisomies. In recent years, commercial providers have moved in this direction, optionally offering cfDNA testing for a wider range of chromosomal abnormalities, including for SCAs [ ].

With cfDNA analysis, the sex chromosomes of the fetus can be noninvasively identified much earlier in pregnancy than was previously the case with ultrasound. cfDNA testing allows for easy and safe sex determination already from 7 weeks of gestation. This has opened up possibilities for early fetal sex determination for medical reasons, for example, for women who are known carriers of a sex-linked disorder such as hemophilia or Duchenne muscular dystrophy [ ], but also as an add-on to prenatal screening for women wanting to know the sex of the fetus for curiosity or other nonmedical reasons [ ].

In this chapter we will first give background information on SCAs, briefly summarizing their etiology and diversity as well the benefit of early diagnosis. Next, we will provide information on how the sex chromosomes may be brought to light with cfDNA testing and what is known about the performance of this test for SCAs. In the two subsequent sections we will then discuss the ethical aspects of two main themes of this chapter: prenatal testing and screening for SCAs and the use of cfDNA testing for sex determination, followed by our conclusions.

Sex Chromosome Aneuploidies

Those born with one of the SCAs have an atypical number of the sex chromosomes, X and Y. A female has two X chromosomes and a male has a single X chromosome, the other sex chromosome in a male being the Y chromosome. The X chromosome is an average-sized chromosome carrying many genes that influence a wide range of developmental processes and physical and nervous system functioning. Some of the genes on this chromosome relate to sexual development and reproductive function but most do not. The Y chromosome is a very small chromosome whose function relates especially to sexual development: the presence of the Y chromosome in the embryo leads to development as a male and is required for male fertility. The genes involved in these functions are passed from father to son and to grandson; they never pass through a female. One can appreciate immediately that a few genes involved in being male are never part of the genetic content of a female whereas many genes, important for a wide range of biological functions, are present in two copies in the female and one copy in the male. Mammals cope with this difference between the sexes in the dosage of many genes by inactivating large parts of one X chromosome in any female cell that has two: X chromosome inactivation (XCI) is the mammalian approach to X chromosome dosage compensation [ ]. Different solutions to this problem are found in other classes of the animal kingdom.

A small number of other genes are present on the X and Y chromosomes that are not inherited in a sex-linked fashion. These “pseudo-autosomal” genes are inherited as if they were on the usual type of chromosome (an autosome) because both copies are active in a female and the Y chromosome has an active equivalent, so there is no need for a mechanism of dosage compensation. Changes in the number of the sex chromosomes therefore lead to changes in the number of copies of active genes not involved solely in sex and reproduction. If the process of XCI applied to the whole of the X chromosome, and if the Y chromosome dealt only with male-specific traits, such SCAs would have little if any phenotypic effect. As it is, however, there is a range of consequences associated with this group of conditions although, because XCI affects much of the X chromosome, these effects are much less marked than might be expected for an equivalent block of autosomal chromatin. The sex chromosome trisomies are rather well tolerated and their effects are often rather mild or subtle; Turner syndrome (often 45,X) is mostly well tolerated in the small proportion of affected conceptions that survive the pregnancy. There are other types of SCA with more marked effects but these are much less common and will not be considered further here.

Common Types of SCA

There are four relatively common types of SCA, all of which can have consequences although these will often be mild. Combined together, these occur at a birth incidence of approximately 1 in 1000, being somewhat more common in phenotypic males than females. This compares to a birth incidence for trisomy 21 of approximately 1 in 700 live births. One of the genes in the pseudo-autosomal region of the X chromosome is SHOX , which promotes skeletal growth. People with only one sex chromosome—who have Turner syndrome (45,X)—are on average shorter than women who have two X chromosomes, while people with any of the three conditions with an additional chromosome (XXX, XXY, and XYY) tend to be taller. Dosage differences at other loci will contribute to the other effects of the SCAs, only some of which can be fully accounted for in a simple “gene dosage–phenotype” relationship.

People with Turner syndrome are usually infertile and may have one or more congenital anomalies (such as coarctation of the aorta or other cardiovascular defects, renal anomalies or cystic hygroma, that develops in utero but usually resolves by the time of birth to leave webbing of the neck). Girls with Turner syndrome may also be fully affected by sex-linked disorders, as if they were male, as they are hemizygous for all X chromosome genes. It should be remembered that Turner syndrome can also be caused by deletions and other rearrangements affecting the X chromosome. People with Klinefelter syndrome (47, XXY) are also usually infertile; people with XYY and XXX may have subfertility. All four of these conditions are associated with a modest drop in mean IQ but most of the individuals involved have an IQ in the normal range. Subtle neurocognitive problems may also be found in Turner syndrome. There is also a modest association with some behavioral problems in the SCAs, with XYY syndrome associated with autistic spectrum disorder. People with four or more X chromosomes have more serious cognitive and behavioral difficulties.

Benefit of Early Diagnosis

Early diagnosis is very helpful in Turner and Klinefelter syndromes because it enables the prompt recognition and appropriate management of endocrine problems. In Turner syndrome, expert management is important if growth is to be optimized. As well as growth hormone in childhood, the prescription of estradiol and then also progesterone triggers puberty with its growth spurt and enables pseudo-menstrual cycles. In Klinefelter syndrome, prompt treatment with supplementary testosterone permits a normal male puberty with normal male musculature; it may also have helpful behavioral effects.

It is difficult to make objective quality of life assessments of these conditions as there has been a long-standing problem with ascertainment bias. Although some unbiased, population-based cytogenetic studies have been carried out, much clinical experience is skewed to those with more marked difficulties and many affected individuals are probably never diagnosed [ ]. However, experience indicates that an early diagnosis is on balance helpful; it enables the optimization of medical and educational support [ ] and avoids the sudden discovery of a diagnosis as a cause of infertility in adult life [ ]. Putting to one side the question of infertility in most of the people with Turner and Klinefelter syndromes, these conditions are often compatible with “normal,” happy, and fulfilled lives.

Identification of SCAs in Prenatal Testing

When not deliberately sought for in an antenatal screening program, SCAs may still be identified antenatally in three circumstances: (i) triggered by ultrasound findings: Turner syndrome may be suspected—and then tested for—in the presence of fetal hydrops, cystic hygroma, or coarctation of the aorta found at fetal ultrasound scan; (ii) as an additional finding of karyotyping or molecular analysis after amniocentesis or chorion villus sampling performed for a different indication; (iii) inadvertently when cfDNA testing is performed, either as part of population screening or for a different specific purpose, when it has been decided to conduct the analysis in such a way that SCAs are identified. SCAs may be sought as part of a package of cfDNA testing conducted primarily to screen for the autosomal trisomies but included among the additional options. One should note that the performance of cfDNA-based tests is often only moderate and sometimes rather poor [ ]. Also, a discordance between sex predicted by cfDNA based testing and phenotype of the external genitalia is likely to occur in 1 in 1500–2000 pregnancies [ ]. In addition to technical reasons, there can be several biological reasons for such discordances, for instance, a Y-signal picked up by cfDNA testing in a pregnancy of a female fetus may be due to a vanishing twin. Many of those discordant results may also be caused by complex disorders of sexual differentiation [ ]. This is further discussed in Chapter 5 .

Strategies for Detecting the Sex Chromosomes in CELL-FREE DNA Testing

There are three principal molecular strategies to detect the sex chromosomes in cfDNA testing: (i) PCR amplification of sequences from the Y chromosome, as used in the offer of fetal sexing by cfDNA testing. It is essentially a Yes/No test that aims to detect Y chromosome sequences, often the sex-determining SRY gene. It is good at determining whether or not there is a Y chromosome present but is not so effective at counting how many Y chromosomes there are, and it cannot determine how many X chromosomes are present. The other two methods are based on sequencing either (ii) unselected cfDNA in maternal plasma, as performed in whole genome sequencing, or (iii) cfDNA that has been enriched for sequences (i.e., chromosomes) of interest, often by means of a microarray that contains target sequences from the X and Y chromosomes.

The sequencing methods rely on counting the copies of the sequences of interest, to determine their ratio against sequences of reference chromosomes. Sequencing has to be performed to a sufficient depth: enough copies have to be sequenced that one can be confident of the ratio of the placenta-derived sequences in the maternal plasma.

The key is the calibration of the test on pregnancies of known chromosomal constitution, using the sequences that are to be detected and counted when performing the test in a clinical situation. This is technically simpler but demands more sequencing if one sequences unselected cfDNA—method (ii)—than if one enriches the cfDNA for specific chromosomal regions-method (iii). With the latter method one can achieve statistical significance more readily, and much more cheaply [ ]. The disadvantage of the enrichment step is that it could introduce a bias in sequence representation unless great care is taken at the design stage and in each analytical step, hence the need for careful calibration.

In the absence of a pregnancy, the woman's plasma will contain roughly equal proportions of DNA from her two X chromosomes but in a pregnancy there will usually be a slight excess of DNA from one of the mother's X chromosomes, along with DNA from either a Y chromosome or from the father's X chromosome. To detect deviations from these two normal scenarios, a 46,XX or a 46,XY fetus, will require quantification of a large number of DNA fragments. Simply examining the ratio of autosomes:X chromosome:Y chromosome will answer some questions. More precise information about the origin of the chromosomal nondisjunction can be obtained if sequences are generated from polymorphic regions where the mother's two X chromosomes differ from one another and from the father's X chromosome. Given a fetal fraction of 5%–10%, the ratio expected between sequences derived from the two maternal and one paternal X chromosome in different scenarios will depend upon both the chromosome constitution of the fetus and the precise meiotic error underlying the SCA. In 47,XXY, for example, the additional X chromosome sequences may derive from meiosis I or II of the mother or come from the father.

Let us take Klinefelter syndrome as an example and consider one X-specific sequence and one Y-specific sequence. If we assume that the fetal fraction of the cfDNA in maternal plasma is 12%, so that the maternal fraction is 88%, then one has to distinguish between a ratio of 94:6 (expected X:Y copy ratio if the fetus is 46,XY) and 96:4 (expected X:Y ratio if the fetus is 47,XXY). Many copies of these sequences have to be counted for us to distinguish these ratios with confidence, more sequence information being required if the fetal fraction is less.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here