Impact of Emerging Technologies in Prenatal Genetic Counseling

Abstract

For decades, prenatal testing has been offered to evaluate pregnancies for genetic conditions. In recent years, the number of testing options and range of testing capabilities has dramatically increased. Because of the risks associated with invasive diagnostic testing, research has focused on the detection of genetic conditions through screening technologies such as cell-free DNA. Screening for aneuploidy, copy number variants, and monogenic disorders is clinically available using a sample of maternal blood, but limited data exist on the accuracy of some of these testing options. Additional research is needed to examine the accuracy and utility of screening for increasingly rare conditions. As the breadth of prenatal genetic testing options continues to expand, patients, clinical providers, laboratories, and researchers need to find collaborative means to validate and introduce new testing technologies responsibly. Adequate validation of prenatal tests and effective integration of emerging technologies into prenatal care will become even more important once prenatal treatments for genetic conditions become available.

In prenatal genetic counseling, counselors offer pregnant women and their partners information about genetic risks to their developing fetus and provide psychological counseling about test results and their implications. Prenatal genetic counseling involves understanding a patient's needs and values, assessing risk factors, discussing testing options, interpreting test results, and providing counseling to address the psychological responses of parents. Although genetic counseling in the prenatal setting may not result in testing, advances in prenatal testing technology have shaped the practice of prenatal counseling. As early as the 1950s, the ability to use amniocentesis with cytogenetic and biochemical analysis allowed couples to learn their fetus was affected by a genetic condition. By the 1970s, amniocentesis was routinely offered to women over 35 yr of age or with a family history of a genetic condition (Resta 2002). The availability of amniocentesis and its risks and limitations were the impetus for the development of the earliest iteration of prenatal genetic counseling, which is now evolving to face new challenges related to the development and expansion of cell-free DNA screening technologies.

Although technological advancements in genetics have revolutionized many facets of health care, the impact in obstetrics has novel attributes that make genetic counseling critical to prenatal care. First, there are few treatments for genetic conditions prior to birth and a prenatal diagnosis may not impact the management of a pregnancy or delivery. Therefore, the decision to undergo prenatal genetic testing depends on a patient's values, needs, and desires. Further, the decision to undergo diagnostic genetic testing requires invasive testing, with a procedure-related risk of miscarriage. Finally, prenatal genetic testing for some families involves facing a decision about whether or not to terminate a wanted pregnancy if a genetic condition is identified. As a consequence of these issues, prenatal testing is notably different from other types of tests performed during pregnancy. Genetic counseling by an expert provider is warranted to properly address the potential benefits, risks, limitations, and consequences of testing.

This review aims to provide a snapshot of prenatal genetic counseling in 2019, an overview of the use of prenatal diagnostic and screening tests, the challenges of prenatal testing, and a glimpse into future directions of reproductive genetic testing and counseling.

GENETIC COUNSELING

Prenatal genetic counseling is relevant to all pregnancies as there is a 3%–5% baseline risk for a birth defect or genetic condition (Centers for Disease Control and Prevention 2008). Genetic counseling can be provided by a variety of health-care providers including genetic counselors, obstetricians, maternal–fetal medicine specialists, geneticists, and nurses. Historically, referrals for genetic counseling were limited to patients with advanced maternal age, abnormal screening results, a significant family history, or suggestive ultrasound findings. Although this referral practice continues to predominate, the new, more accurate, and lower-risk genetic tests have increased the number of pregnant women with fewer or none of these risk factors attending genetic counseling for risk assessment, discussion of genetic testing options, decision-making, and counseling centered on values and beliefs.

A typical prenatal genetic counseling session includes the following (the information is tailored to each patient based on the indication and delivered with sensitivity to her educational, emotional and cultural needs):

1. Contracting: Exploring the patient's needs and goals through a dialog called “contracting” is useful for understanding the purpose of the visit and the patient's values and needs. This understanding can be used by the counselor to tailor information and help facilitate patient decision-making. Contracting includes an assessment of the patient's concerns, psychological well-being, and understanding, all of which leads to shared goal setting.

2. Risk Assessment: To facilitate informed decision-making, a genetic counselor evaluates and explains risks to the fetus. Personal and family medical history is assessed by collecting health information for a three-generation pedigree. Medical records are also reviewed for maternal age, screening results, and a history of any teratogen exposures. Additionally, if an ultrasound has been done, certain findings may indicate a higher risk for the fetus to have a condition. Collectively, this information is used to determine a baseline fetal risk assessment.

3. Testing Review: General population testing is available for all pregnant patients. Additional genetic testing may be indicated based on the individualized risk assessment. There is often a range of testing options available and pretest counseling by a knowledgeable provider is important. Key components of pretest counseling include a discussion of the purpose of testing, how the test is performed, risks and benefits of testing, possible results, turnaround time, and communication of results (Janssens et al. 2017). Counseling and decision-making are discussed in Goldman (2019).

4. Psychological Counseling: In addition to roles as educator and testing facilitator, genetic counselors also explore the psychological needs of patients as they often experience feelings of shock, anxiety, disbelief, guilt, grief, and mourning when a risk factor is identified or when results are abnormal and unexpected, particularly when significant uncertainty about the future health and well-being of the fetus exists (Raymer 2004; Bernhardt et al. 2013; Werner-Lin et al. 2016)

5. Plan for Follow-Up: Expectations for follow-up are communicated to the patient, if needed. This may include plans for results disclosure (in-person vs. over the phone), referrals to a specialist, and communication back to the referring provider. Written information and/or contact information for patient support groups and advocacy networks are also helpful resources, depending on the nature and outcome of the session.

OVERVIEW OF GENETIC TESTING OPTIONS

The discussion of prenatal genetic testing options with patients is essential as the number and types of testing options change often. A knowledgeable provider with expertise in the risks, benefits, and limitations of prenatal genetic testing should facilitate this discussion.

Prenatal Screening Tests

Screening tests evaluate the risk for a specific condition or group of conditions but cannot definitively diagnose or rule them out. In the prenatal context, screening tests provide information with minimal risk to the pregnancy as they are typically performed on maternal blood or via ultrasound. Prenatal screening tests are available for a variety of genetic conditions, but the most common tests screen for chromosome conditions (such as Down syndrome) or autosomal recessive conditions (such as sickle cell anemia and cystic fibrosis).

Screening for Fetal Aneuploidy

Chromosomal aneuploidy is extra or missing chromosomal material. Full aneuploidy refers to an extra or missing chromosome that results in a condition. Risk factors for full aneuploidy include increasing maternal age, previous child with aneuploidy, and ultrasound abnormalities (Gardner et al. 2012). Because of their sporadic nature, screening for chromosome abnormalities should be available to all pregnant patients, and their decision to test or not to test should be supported. Various screening methods can be used depending on the patient preference, insurance coverage, and indication.

Multiple Marker Screening: Traditional screening tests utilize biochemical markers in maternal blood that are produced by the fetus and placenta. Patterns of abnormal analyte levels are associated with conditions such as Down syndrome, trisomy 18, and open neural tube defects. Additionally, less common conditions, such as Smith–Lemli–Opitz syndrome and triploidy, may cause abnormal maternal serum markers. Multiple marker screening is designed to provide a risk assessment with detection rates ranging from 69% to 96%, depending on the number of analytes assessed, the use of first- and/or second-trimester maternal blood samples, and the use of ultrasound markers evaluated in conjunction with maternal analytes. The false-positive rate for these various screening methods is typically ∼5% (Practice Bulletin No. 163 2016).

Cell-Free DNA (cfDNA)/Noninvasive Prenatal Testing (NIPT)/Noninvasive Prenatal Screening (NIPS): First introduced in 2011, this novel screening test evaluates the likelihood of fetal chromosome abnormalities from cfDNA in maternal blood. After ∼10 wk gestation, enough cfDNA from the trophoblast cells of the placenta is present in maternal blood and can be sequenced to determine the chromosome of origin (Committee Opinion No. 640 2015). Using a counting method, the number of fragments from the chromosome of interest can be evaluated and samples with a significantly greater or lesser number of these counts are reported out as increased risk. Alternatively, single-nucleotide polymorphisms (SNPs) can be analyzed on the chromosomes of interest to determine whether an aneuploidy is suspected. Both methods have shown very high detection rates and low false-positive rates (Table 1).

Table 1.

Detection rate and false-positive rates of cell-free DNA screening

Condition	Detection rate (95% CI)	False-positive rate (95% CI)
Trisomy 21	>99% (99.1%–99.9%)	0.04% (0.02%–0.07%)
Trisomy 18	97.9% (94.9%–99.1%)	0.04% (0.03%–0.07%)
Trisomy 13	99% (65.8%–100%)	0.04% (0.02%–0.07%)
Monosomy X	95.8% (70.3%–99.5%)	0.14% (0.05%–0.38%)
Sex chromosomes abnormalities (other than monosomy X)	>99% (83.6%–100%)	0.004% (0%–0.08%)

Data in table adapted from Gil et al. 2017.

Initially, cfDNA screening was offered exclusively to women at high risk. However, it is increasingly utilized in average-risk populations as it has been shown to have similar sensitivity and specificity in low-risk pregnancies (Pergament et al. 2014). Although the test performance is similar in low-risk populations, the meaning of a screen positive result can differ greatly because of differences in positive predictive value (PPV). Positive predictive value is the likelihood that any given positive screen is a true positive; PPV is impacted not only by sensitivity and specificity of the test but also by the prevalence of the condition at issue. For example, if NIPT has a 99.9% sensitivity and specificity, and a 45-yr-old woman gets a positive screen for Down syndrome, she has a 98% PPV. A 25-yr-old woman, on the other hand, has a 49% PPV even though the NIPT test she took has the same sensitivity and specificity. All other things being equal, women with a higher a priori risk for Down syndrome will have a higher PPV than women at low risk. To aid providers in calculating the positive and negative predictive values of screening tests, a calculator was created by the National Society of Genetic Counselors and the Perinatal Quality Foundation (PQF) (https://www.perinatalquality.org/Vendors/NSGC/NIPT/). Using the prevalence, sensitivity, and specificity of the condition, positive and negative predictive values can be generated to aid in posttest counseling.

ACOG and the Society of Maternal–Fetal Medicine (SMFM) recommend that laboratories report PPVs to assist with posttest counseling (Committee Opinion No. 640 2015). The PPV can also be calculated manually if the test sensitivity, specificity, and prevalence of the condition are known. Sensitivity and specificity data are readily available for conditions that have been well-studied, such as trisomy 21 and trisomy 18, but limited data are available on the sensitivity and specificity for other conditions, such as sex chromosome abnormalities and microdeletion syndromes. Despite the existence of a PPV and NPV calculator, accurate postscreen risk assessment cannot be performed without reliable sensitivity and specificity data.

False-positive and false-negative cfDNA results do occur, albeit less commonly than with multiple marker screening (Gil et al. 2017). Several maternal and fetal factors can lead to false-positive or -negative screening results, including mosaicism (Grati et al. 2014; Hartwig et al. 2017). Mosaicism is defined as two or more cell lines with different genetic compositions (Gardner et al. 2012). NIPT results reflect the makeup of the placenta (Lo et al. 1997), and if the placenta contains abnormal cells, then it can lead to a positive NIPT result. If the mosaicism is only in the placenta, which is known as confined placental mosaicism (CPM), the NIPT results may not reflect the fetal karyotype (Van Opstal and Srebniak 2016). On the other hand, if the placenta contains cells with normal chromosomes, but the fetus is chromosomally abnormal or mosaic, NIPT results will yield a false-negative result.

Other causes of false-positive NIPT results include the demise of a twin with a chromosome abnormality, a maternal chromosome abnormality, maternal mosaicism for a chromosome abnormality, and a maternal condition such as cancer (Bianchi et al. 2015; Hartwig et al. 2017; Leonard et al. 2018; Kim et al. 2019; Yu et al. 2019).

A false-negative result, the failure to detect a fetal chromosome abnormality, is commonly due to low fetal fraction (FF). The FF is the percentage of cfDNA in maternal blood derived from the placenta. Pregnant women have on average 10% FF, meaning that 90% of the cfDNA is maternal in origin (Ashoor et al. 2013). An insufficient level of cfDNA from the pregnancy make is less likely to detect a fetal abnormality. Low FF has been associated with many factors such as increased maternal weight and early gestational age, but also with maternal conditions such as lupus and maternal therapy with blood thinners (Ashoor et al. 2013; Burns et al. 2017; Dabi et al. 2018). Low FF has also been found to be associated with genetic conditions, such as trisomy 18, trisomy 13, and triploidy (McKanna et al. 2019). Therefore, when a low FF is found and results cannot be reported, diagnostic testing is recommended over repeat testing via NIPT, because a repeat test is not guaranteed to yield a result and may further delay obtaining diagnostic information (Committee Opinion No. 640 2015).

Screening for Microdeletions

NIPT has the capacity to screen for countless numbers of genetic conditions as the entire nuclear genome from trophoblastic cells exists in maternal blood. As previously described, aneuploidy is evaluated by detecting an abundance or deficiency of cfDNA from a particular chromosome or by observing a trisomic or monosomic SNP pattern. Using this same available data, pregnancies can also be evaluated for microdeletions and microduplications, such as 22q11 deletion syndrome. Some argue that it is appropriate to screen the general population for microdeletion syndromes as they are more prevalent than Down syndrome for women in their early and mid-20s.

Most clinical laboratories offer the option of microdeletion screening for a select number of microdeletion syndromes, such as 22q11, 1p36, 4p, 5p, 8q, 11q, and 15q deletions. Sensitivity and specificity data are limited because of the rarity of these conditions. For example, Helgeson et al. (2015) reported 100% PPV for detection of 4p and 11q deletions (95% CI 5.5%–100%), but only detected one case of each in their study population, whereas two of six cases (95% CI 24.1%–94%) of 5p deletion cases were found to be false positives. Another study reports PPVs ranging from 0% to 66.7%, but 40% of cases had an unknown outcome (Martin et al. 2018). To address the limited evidence on detection rates for rare conditions, laboratories have created artificial plasma mixtures or simulated samples to evaluate screening accuracy (Wapner et al. 2015). It is unknown whether these spiked samples can accurately assess how the test will function with actual patients.

It is important to note that even if high sensitivity and specificity are confirmed, the PPV will be low in the absence of other risk factors (family history, ultrasound findings, etc.) because of the rarity of these conditions. For example, Wolf–Hirschhorn syndrome is a condition caused by a deletion of chromosome 4p. The prevalence of Wolf–Hirschhorn is ∼1 in 20,000. Even with a 99.9% sensitivity and specificity, the PPV of a positive cfDNA screen would be only 5%.

Screening for Monogenic Disorders via Cell-Free DNA

Evidence regarding the accuracy of clinically available monogenic disorder screening is even harder to find (Zhang et al. 2019). Screening for autosomal dominant conditions via maternal blood is clinically available for select de novo or paternally inherited conditions, but not maternally inherited dominant conditions. Screening for de novo or paternally inherited conditions is hypothetically feasible because the testing evaluates the presence of pathogenic variants that are not present in the maternal genome. Thus, the presence of the variant in maternal blood indicates its presence in the fetus (Hudecova and Chiu 2017).

Data appear promising for the detection of de novo or paternal variants; however, outcome data were available in <50% of reported cases in the single peer-reviewed article available to date (Zhang et al. 2019). Furthermore, positive cfDNA results were detected only in pregnancies with abnormal ultrasound findings consistent with the screening result or those with a positive paternal family history. None of the pregnancies tested for other indications (abnormal screening results, advanced maternal age, advanced paternal age, other positive family history) was found to have abnormal results. In other words, all cases with positive results had a high a priori risk based on ultrasound findings or based on a 50% risk to inherit the paternal mutation. Proof of clinical utility in the general population requires further study and additional research is needed to confirm the accuracy of screening for monogenic conditions via cfDNA.

Screening for known autosomal recessive, autosomal dominant, and X-linked conditions via cfDNA is also clinically available, but there are no peer-reviewed publications that demonstrate the accuracy of this method for that purpose (https://www.progenity.com/tests/resura).

Reproductive Carrier Screening

Screening parents for heritable autosomal recessive and X-linked conditions they may carry is far simpler than screening the fetus for de novo conditions, as parental DNA can be sequenced or genotyped for inherited pathogenic variants. This type of reproductive carrier screening has been available for decades; however, the number of conditions for which carrier screening is available continues to expand. All individuals are thought to carry on average four to five autosomal recessive conditions (Morton et al. 1956), and a lack of family history is not an effective screening tool, as both parents must carry the same genetic condition to be at increased risk and therefore children with recessive conditions are typically born to families without a family history.

Traditionally, carrier screening has been offered according to ethnic group (Gregg and Edwards 2018). For example, African–American women were offered sickle cell anemia screening and Caucasian women were offered cystic fibrosis screening. With the rising affordability of genetic testing and the high rates of mixed or unknown ancestry, expanded carrier screening (ECS) is now offered on a routine basis. Expanded carrier screening entails screening for more conditions than are recommended by professional guidelines and enables the same screening test to be offered all patients, regardless of ethnicity or ancestry (Gregg and Edwards 2018).

If an individual is found to be a carrier of an autosomal recessive condition, protocol requires offering the partner carrier screening. The fetus is at risk of being affected only when both parents carry a variant in the same gene. If the partner also tests positive and both parents carry a disease-causing variant, there is a 25% risk to each fetus of being affected. With a negative test, the risk of the partner being a carrier is reduced but not eliminated because carrier screening does not detect 100% of carriers. The detection rate for carrier screening varies depending on the condition of the patient, ethnicity of the patient, and testing methodology. Carrier screening can utilize targeted genotyping panels or genome sequencing to identify pathogenic variants among many genes. The advantage of sequencing is a higher detection rate of pathogenic variants. Yet sequencing also identifies likely pathogenic variants and variants of uncertain significance, which are changes that are uninterpretable based on available evidence and thus have unclear or unknown pathogenicity. The reporting of likely pathogenic variants or variants of uncertain significance is inconsistent and at the discretion of the laboratory that conducts the screening.

Carrier screening panels have grown to include hundreds of autosomal recessive and X-linked recessive conditions. As of 2017, screening for cystic fibrosis and spinal muscular atrophy was recommended for all patients (Committee on Genetics 2017). Additional screening for hemoglobinopathies, thalassemia or conditions that occur more often among individuals of Ashkenazi Jewish background were recommended based on ethnicity. Fragile X syndrome screening was recommended for those with a positive family history (Committee on Genetics 2017). Expanded panels have increased the number of conditions available for screening but panels vary by laboratory. Multiple publications have put forth guidelines from which the selection of conditions on expanded panels should be based, but interpretation of these guidelines can be challenging and adherence to these guidelines is not universal (Grody et al. 2013; Edwards et al. 2015; Stevens et al. 2017).

As panels expand, the chances of detecting carrier status in a patient increase. Although the ability to diagnose at-risk couples may be valued as a benefit, risks and limitations of expanded screening also exist. Possible drawbacks include the psychological impact of receiving positive carrier screen results, cost and limited availability of follow-up testing for the father, and the cost of time spent on follow-up. More evidence is needed to evaluate expanded panels for their added benefit to current standards for carrier screening.

Prenatal Diagnostic Tests

Diagnostic tests require an invasive procedure to obtain a direct fetal sample for testing. This direct testing allows for more definitive results and the ability to test for more conditions than most screening tests. The major disadvantage is that the procedures required to obtain a diagnostic sample come with risks to the pregnancy, including miscarriage.

There are two primary diagnostic procedure currently in use for diagnostic genetic testing: chorionic villus sampling (CVS) and amniocentesis (amnio). CVS entails a biopsy of the chorionic villi from the placenta in the first trimester of pregnancy (10–14 wk). The sample can be obtained by passing a catheter transcervically or inserting a needle transabdominally into the placenta. Amniocentesis is done by sampling amniotic fluid transabdominally and can be performed later in pregnancy, from 16-wk gestation on. Both procedures are associated with high diagnostic accuracy and low procedure-related risks.

The types of tests that can be performed using CVS or amniocentesis samples are summarized in Table 2. In addition to cytogenetic and molecular genetic testing, amniotic fluid can also be evaluated for viral infections, elevated alpha-fetoprotein and acetylcholinesterase levels indicative of open neural tube defects and biochemical abnormalities.

Table 2.

Common prenatal diagnostic tests

Test	Evaluated conditions	Examples
Karyotype	Aneuploidy, large deletions and duplications (>5 Mb), chromosome rearrangements	Down syndrome, Wolff–Hirschhorn, translocations
Fluorescence in situ hybridization (FISH) for aneuploidy	Common aneuploidies (21, 18, 13, X, Y)	Down syndrome, Turner syndrome
Chromosome microarray (CMA)	Microdeletion and microduplication syndromes (<5 Mb)	22q11 deletion syndrome, Williams syndrome
Next-generation sequencing (single-gene testing, NGS panels, exome sequencing)	Monogenic disorders	Noonan syndrome, cystic fibrosis

(Mb) Megabase.

Prenatal Diagnosis of Chromosome Aneuploidy

Karyotyping and fluorescence in situ hybridization (FISH) testing are routinely ordered to detect aneuploidy. Indications for this testing may include advanced maternal age, a previous child with aneuploidy, or a positive screening test. Regardless of the indication, testing for additional chromosome conditions such as microdeletion and microduplication syndromes may also be considered when invasive testing is pursued.

Prenatal Diagnosis of Microdeletions and Microduplications

Chromosome microarray is used to detect smaller chromosome deletions and duplications and is increasingly standard, particularly following the detection of fetal abnormalities on ultrasound (Durham et al. 2019). Studies have shown that women 35 yr of age and older and those who have a positive screening test result, have a 1%–2% chance that they will learn their fetus is affected by a condition that would not be detected by routine karyotyping. If ultrasound abnormalities are detected, the chance of a significant finding in the fetus increases to 6%. It is important to note that a chromosome microarray also has a 1%–2% chance to detect a change that cannot be characterized as benign or pathogenic, also known as a variant of uncertain significance (Wapner et al. 2012). Patients who receive uncertain results report intense negative reactions, demonstrating the need for adequate pretest counseling regarding the possibility of abnormal or uncertain results (Bernhardt et al. 2013).

Prenatal Diagnosis of Monogenic Disorders

Although karyotyping and chromosome microarray results can effectively identify chromosome abnormalities, sequencing is needed to identify variants within genes. The majority of sequencing tests use next-generation sequencing (NGS), a quicker, cost-effective sequencing approach. A variety of sequencing tests are available prenatally depending on the indication. For example, a couple known to carry variants in the CFTR gene can undergo testing for the known variants. On the other hand, testing a fetus found to have congenital anomalies may require a multigene panel depending on the differential diagnosis. These panels could include a number of genes associated with a single condition, such as Noonan syndrome, or a number of genes related to a phenotype, such as a skeletal dysplasia panel.

When sequencing panels have not identified a cause or there is not an appropriate panel for the indication, exome sequencing may be considered. Exons are the translated parts of the genetic code and they make up the exome, which comprises ∼1%–2% of the genome. The majority of disease-causing variants are found in the exome, so targeting the exome for sequencing is a more feasible method to detect sequence variants (Ng et al. 2009; Stenson et al. 2009; Bainbridge et al. 2010).

Contemporary Use of Prenatal Screening and Diagnostic Testing

Practice guidelines recommend that all pregnant women be offered prenatal screening and diagnostic testing, regardless of maternal age or risk factors. Previously, guidelines recommended invasive diagnostic testing only for women at increased risk. The change in recommendations is largely based on a belief that patients and their partners should weigh the risks and benefits of testing in light of their preferences and values to determine the use of prenatal genetic testing.

1. Screening: ACOG recommends that women be offered some form of screening for aneuploidy. There is no one superior prenatal screen for all circumstances, so the provider and patient together must make an individualized determination of the best screening option, if any (Cai et al. 2016). ACOG specifically states in a 2015 Committee Opinion (Committee Opinion No. 640 2015) and a 2019 Practice Advisory (https://www.acog.org/Clinical-Guidance-and-Publications/Practice-Advisories/Cell-free-DNA-to-Screen-for-Single-Gene-Disorders) that screening for microdeletion syndromes and monogenic disorders is not recommended because of lack of clinical validation data.

   Carrier screening for recessive conditions should also be offered all women who present for care, ideally prior to conceiving. There is no one carrier screening panel (ethnicity based vs. expanded) that is ideal for all patient populations, so the risks, benefits and limitations of each panel option should be reviewed with patients to facilitate informed decision-making.

2. Diagnostic Testing: Uptake depends on many factors, but the rate of diagnostic testing has dropped significantly since the introduction of cfDNA screening despite its diagnostic advantages over screening tests (Warsof et al. 2015; Stevens et al. 2019). When CVS or amniocentesis is desired, patients should routinely be offered both karyotyping and chromosome microarray along with a discussion of the risks, benefits, and limitations of each (Committee on Genetics and the Society for Maternal-Fetal Medicine 2016). Additional testing options may be offered depending on the testing indication, such as NGS panels for monogenic disorders, biochemical testing for open neural tube defects or biochemical testing for metabolic conditions.

   Exome sequencing (ES) is not recommended by ACMG for routine testing of the fetus. However, when other recommended prenatal tests have not arrived at a diagnosis, ES may be considered. Limitations of ES should be emphasized, including longer turnaround times, false positives, false negatives, and variants of unknown significance (ACMG Board of Directors 2012).

   Similarly, the International Society of Prenatal Diagnosis (ISPD), PQF, and SMFM do not recommend prenatal ES be used routinely, and they note that when it is performed, it should be done under a research protocol. However, use of prenatal ES can be considered when a genetic condition is suspected and the case is managed with expert guidance of a genetic professional (International Society for Prenatal Diagnosis 2018). Prenatal ES is used almost exclusively during pregnancy when ultrasound abnormalities are detected that increase suspicion for a genetic condition. Although this will likely change in the future, laboratories are not offering prenatal ES for apparently healthy pregnancies. The diagnostic yield of prenatal ES varies depending on the indication for testing, but ranges from 10% to 30%.

3. Introduction of New or Expanded Clinical Testing: It is not uncommon for national guidelines on genetic testing to lag behind clinical care. For example, cfDNA screening was introduced in clinical practice in 2011 after publication of a multicenter research study performed on more than 4000 pregnancies by 27 centers, including more than 200 pregnancies diagnosed with Down syndrome (Palomaki et al. 2011). ACOG did not endorse it as a screening option for high-risk pregnancies until December of 2012 (American College of Obstetricians and Gynecologists Committee on Genetics 2012). It is still not specifically endorsed for low-risk populations (Committee Opinion No. 640 2015).

Subsequent expansions of cfDNA, including screening for microdeletions and monogenic disorders, have been introduced clinically without sufficient validity testing. For example, screening for microdeletions, including 22q11, 5p, 4p, and 1p36 deletions, became clinically available in late 2013. Peer-reviewed validation data were published two years later in 2015 (Helgeson et al. 2015; Wapner et al. 2015; Zhao et al. 2015), and these studies used a combination of simulated or artificial samples as well as maternal plasma DNA with microdeletions. Similarly, genome-wide NIPT was introduced in 2015 by one company after showcasing data in a poster presentation at the ISPD conference. Clinical validation was subsequently reported in a peer-reviewed journal in August of 2016 (Lefkowitz et al. 2016), followed by data on the first 10,000 cases published December of 2017 (Ehrich et al. 2017). Screening for monogenic disorders became clinically available in 2017 and only a white paper published by the performing laboratory was made available (https://bcm.box.com/shared/static/eehibpd8d7bpobn7wypiwoef3iv9z8hn.pdf) until 2019 when the first peer-reviewed research was published on the performance of cfDNA screening for these monogenic conditions (Zhang et al. 2019).

Up until the introduction of cfDNA for common aneuploidies, research was conducted and validation data was published prior to the clinical introduction of a screening test. The expansion of cfDNA screening to include microdeletions, genome-wide screening, and monogenic disorders have upended this standard of care. Clinical samples and patients undergoing clinical testing are now used to gather validation data. This industry-driven approach to prenatal screening presents major challenges for patients and providers, leaving the latter to offer screening tests whose risks and detection rates are poorly understood. If a provider offers nonvalidated tests, patients do not have sufficient information to make informed decisions and may not be aware that their results will be the source of data to validate screening.

Major obstacles to conducting prospective validation studies for very rare conditions are costs and time. A validation study for detection of trisomy 21, the most common chromosome abnormality, took 27 centers enrolling more than 4000 pregnant patients (Palomaki et al. 2011). Significantly more participants are required for a condition that occurs in 1 in 5000 pregnancies. One may argue that unlike Down syndrome and trisomy 18, there are no alternative screening methods for conditions like 22q11 deletion syndrome, and a somewhat accurate test may be better than no screening test. Therefore, as long as adequate pretest counseling is performed to ensure the limitations of expanded testing are understood, arguably patients seeking information should be able to obtain as much as possible while avoiding procedure-related risks. However, there are disadvantages and risks associated with offering inadequately validated screening tests to patients that informed consent may not be adequate to address.

As prenatal screening expands to rarer conditions using more advanced technology, patients, clinicians, laboratories, researchers, and policy makers must work together to integrate testing responsibly into clinical care. Missing outcome data and unknown sensitivity and specificity for clinically available tests is not appropriate and a solution is imperative as the pace of new technology and patient demand continue to accelerate. Infrastructure to facilitate data collection and pregnancy outcomes is needed as are patient-reported preferences for delivery of information. Ideally, nonvalidated tests should only be offered to patients under a research protocol in a prospective manner. If outcome data can be collected on participating patients, then sensitivity and specificity data can be determined. If the option of expanded testing is limited to research protocols through testing laboratories and testing is provided without additional cost to the patient, interested patients and professionals may be likely to participate.

Past validation studies of prenatal chromosome microarray demonstrate the success of such efforts (Wapner et al. 2012). Professional organizations will be unable to recommend expanded screening tests without evidence and insurance providers are unlikely to cover services that are not recommended by professional organizations. In addition to test accuracy, data on patient experiences with newly introduced tests that assess understanding and psychological well-being can be used to guide the offer and disclosure of prenatal testing results in a sensitive and responsible way. Public policy on use of prenatal testing may consider allocation of funds for data collection to offset costs fronted by testing laboratories.

FUTURE DIRECTIONS IN PRENATAL GENETIC TESTING

Despite technological advances, it is not possible to diagnose all suspected genetic conditions using current genetic testing techniques. As tests improve, the ability to interpret genetic and genomic data must keep pace, because once these tests are established in the pediatric setting they will be rapidly introduced in the prenatal setting. In the prenatal setting, uncertainty generated by testing is further compounded by limited designation of a fetal phenotype, which is typically restricted to ultrasound imaging. Thus, identifying exome or genome sequencing data relevant to the fetal phenotype can be challenging. Advancements in fetal imaging are needed along with advancements in genetic testing and interpretation of results.

Studies have shown that women are interested in learning about the health of their fetus, but do not want to put their pregnancy at risk from the procedure (Kalynchuk et al. 2015; Sullivan et al. 2019). For this reason, much research has focused on the ability to test for genetic conditions using maternal blood. Some of the emerging areas of research are outlined below.

1. Improved Cell-Free DNA Screening for Monogenic Diseases: Screening tests for select de novo dominant conditions and known inherited variants are currently available. Although screening for the presence of a de novo or paternally inherited variant is relatively straightforward, cfDNA screening for the maternal variant requires more sophisticated technology as the fetal variant must be detected in the presence of maternal cfDNA containing the same variant. Screening for the paternal variant is more challenging if the parental variants are identical, as may be the case in consanguineous couples or among those carrying founder mutations.

   Two testing approaches, relative mutation dosage by droplet digital PCR and relative haplotype dosage by massively parallel sequencing (MPS), are techniques used to screen for monogenic disorders using cfDNA. Relative mutation dosage analyzes the mutant allele proportion to the wild-type allele proportion in maternal serum, expecting greater mutant allele proportion in an affected pregnancy, greater wild-type proportion for an unaffected pregnancy and a balanced ratio when the fetus is a carrier. Relative haplotype dosage involves determining maternal and paternal haplotypes from genomic regions linked to the mutation. Then SNPs heterozygous in the mother and homozygous in the father are selected to determine whether the mutation or wild-type allele was passed on by the mother. Both methods are highly dependent on a sufficient FF (Hudecova and Chiu 2017).

   Several cases of prenatal detection of monogenic disorders using these two methods have been reported, including identification of β-thalassemia, sickle cell anemia, methylmalonic acidemia, hemophilia, cystic fibrosis, hearing loss, congenital adrenal hyperplasia, Duchenne muscular dystrophy, Gaucher, and α-thalassemia (Chiu et al. 2002; Ding et al. 2004; Li et al. 2005; Tungwiwat et al. 2006, 2007; Papasavva et al. 2008, 2013; Ho et al. 2010; Galbiati et al. 2011, 2016; Tsui et al. 2011; Yan et al. 2011; Barrett et al. 2012; Phylipsen et al. 2012; Sirichotiyakul et al. 2012; Ge et al. 2013; Gu et al. 2014; Ma et al. 2014; New et al. 2014; Xiong et al. 2015, 2018; Xu et al. 2015; Yoo et al. 2015; Zeevi et al. 2015; Parks et al. 2016; Vermeulen et al. 2017; Chang et al. 2018; Guissart et al. 2018).

   A 2019 study by Luo et al. evaluated the feasibility of performing cfDNA screening for aneuploidy, copy number variants, and monogenic disorders to demonstrate that it is possible to have a 3 in 1 cfDNA screen that is cheaper and faster than doing separate analyses. One of the challenges of combining chromosome screening with screening for monogenic disorders is the greater depth of sequencing necessary for identification of sequencing variants. Luo and colleagues assessed chromosome aneuploidy, copy number variants and mutation hotspots in six genes of the three of the most common single-gene disorders in China: β-thalassemia, hearing-impairment, and phenylketonuria. Every aneuploidy (7/7) and deletion/duplication >20 Mb (3/3) was detected and 86.4% of the monogenic disorders were accurately detected (19/22) (Luo et al. 2019). Malcher et al. (2018) also developed a comprehensive cfDNA product that demonstrated 100% sensitivity and 98.53% specificity for trisomy 21, demonstrated 100% accuracy for fetal sex, and detected five of seven skeletal dysplasia cases.

2. Exome/Genome Sequencing by Cell-Free DNA Screening: Testing for targeted variants associated with a known family history is achievable by the methods previously described, but testing for multiple targets is less feasible using these methods. Genome sequencing of a fetus using maternal blood has been accomplished but with several limitations, primarily involving the amount of fetal DNA available for analysis. For this reason, some studies have been performed using genome sequencing in the second or third trimester when FF levels are higher (Chan et al. 2016).

   Several approaches have also been used to enrich the fetal DNA analyzed for genome sequencing (Rabinowitz et al. 2019). Fetal fragments of cfDNA are typically shorter than maternal fragments because of differences in nucleosome position (Chan et al. 2004; Fan et al. 2010, Lo et al. 2010; Yu et al. 2014). This approach has been used to screen for chromosome abnormalities (Cirigliano et al. 2017; Sun et al. 2017) and genome sequencing (Rabinowitz et al. 2019). However, enrichment based on DNA strand size can lead to PCR amplification concerns that may skew allele ratios. A new method, circulating single-molecule amplification and resequencing technology (cSMART), has been introduced to remove amplification bias. The cSMART method was implemented in a 2015 study that correctly classified four pregnancies at risk for Wilson disease (Lv et al. 2015).

   Other applications of exome or genome sequencing may be the identification of carriers of genetic conditions, particularly for couples at higher risk such as consanguineous couples or those who have had multiple children/pregnancies affected with a suspected genetic condition. Additionally, couples undergoing in vitro fertilization and testing of embryos via preimplantation genetic testing (PGT) may also consider GS/ES.

3. Whole-Cell Identification: Because of many situational challenges presented by cfDNA screening, such as assessment of twin pregnancies, inability to detect Robertsonian translocation versus trisomy, abnormal cfDNA from maternal malignancy, placental mosaicism, and low FF, efforts to identify intact fetal cells in maternal blood for analysis have continued. The biggest challenge to this approach is the low quantity of fetal cells in maternal circulation, estimated at 1–45 cells per 30 mL maternal blood (Beaudet 2016). Fetal cells are identified and isolated using unique fetal receptors and antibodies. Although studies have shown successful isolation and diagnosis of chromosome abnormalities and microdeletions, significant challenges remain (Feng et al. 2018).

Patient Perspectives

Multiple studies have demonstrated patient interest in learning more about the genetic health of their unborn baby. A population of more than 200 pregnant patients presenting for genetic testing were studied in 2018 and the majority of respondents (83%) felt ES should be offered prenatally and about half of participants were interested in pursuing it. However, only 17% reported they would undergo amniocentesis to obtain the information. This study demonstrates positive patient response to noninvasive screening for a wide variety of genetic conditions (Kalynchuk et al. 2015).

Another study of more than 500 pregnant women demonstrated the desire for information about their pregnancy primarily “to prepare financially, medically, or psychologically for a child with special needs.” The majority of patients wanted information regarding serious treatable childhood-onset conditions (89.7%) and were least likely to want information about nonmedical traits (40%) (Sullivan et al. 2019).

It is important to note that these surveys were hypothetical offers for genetic testing and did not entail detailed counseling about the potential for variants of uncertain significance. Therefore, it is unclear if patients will demonstrate as much interest in genetic testing as these studies indicate.

If cfDNA and fetal cell isolation technologies continue to improve, the risk for miscarriage associated with invasive testing will no longer be a decision-making factor. However, patients must still contemplate the risks, benefits and limitations of information obtained from genetic testing. The risk to obtain variants of uncertain significance or discovery of incidental findings will continue to be decision-making factors. Additional barriers may also include cost and insurance coverage. In addition to research on the accuracy of noninvasive testing, research must focus on the psychological impact of the decision-making process, the experience of receiving positive results and the impact of genetic testing on outcomes.

Changing Paradigm: Emergence of Prenatal Treatments

Prenatal genetic testing is optional and requires pretest counseling because the risks and benefits of the information obtained from genetic testing are weighed differently in light of patients’ values, needs, and desires. The goal of testing for most patients is to get reassurance from normal results, but learning of abnormal results can also be beneficial for some. Prenatal identification of a genetic condition can help decrease uncertainty during the pregnancy, and can allow for termination of pregnancy or mental preparation for a genetic condition, adjustments to the birth plan (i.e., monitoring during labor or plans for vaginal delivery vs. Caesarian section), changes in location and timing of delivery, and immediate neonatal management. For example, a 2017 study of ES performed on neonates <100-d-old revealed that medical management was affected in >50% of cases, which demonstrates the value of early diagnosis and highlights the benefit of prenatal diagnosis (Meng et al. 2017). Despite these potential benefits, the outcome for many individuals with genetic conditions will not change because of a prenatal diagnosis, and testing for the sake of knowing ahead of time may not be perceived as a benefit for all families. Some expectant parents find information obtained through prenatal testing more harmful than beneficial. The decision-making process with regard to prenatal testing is complex and is affected by many personal factors including a patient's “need to know,” feelings about pregnancy termination, and availability of interventions.

As new treatments and preventive measures become available for genetic conditions, prenatal testing will become increasingly relevant during pregnancy as a means to improve health outcomes. Presumably, use of prenatal testing will increase, given that pregnant patients have been shown to be more motivated to pursue testing if a condition is treatable prenatally or in the immediate postnatal period (Kalynchuk et al. 2015).

Conditions that are ideal candidates for prenatal intervention include conditions that result in irreparable damage prior to delivery, that do not have effective postnatal treatment, when small fetal size is an advantage for treatment efficacy (i.e., higher treatment to target cell ratio), or when early delivery of a vector or transgene could produce a decreased immune response and/or a future immune tolerance which could improve outcomes (David and Waddington 2012; AJMG 2018).

Prenatal gene therapy in mouse models have shown promise in several conditions: hemophilia A and B, congenital blindness, Crigler–Najjar type 1 syndrome, and Pompe disease (glycogen storage disease type II). Other conditions that are potential gene therapy candidates include cystic fibrosis, Duchenne muscular dystrophy, lysosomal storage disorders, spinal muscular atrophy, urea cycle defects, α-thalassemia, severe combined immunodeficiency, epidermolysis bullosa, and even intrauterine growth restriction and congenital diaphragmatic hernia (David and Waddington 2012).

Gene editing is an emerging gene therapy that can be utilized in the fetal period; however, much research is needed to determine the safety of such technology. Research in animal models is already underway and the β-thalassemia phenotype has apparently been cured in a mouse model (McClain and Flake 2016; AJMG 2018).

In the case of conditions with an effective postnatal treatment, prenatal diagnosis can allow for early diagnosis so that postnatal treatment can be initiated without the need to wait for testing results following delivery, such as either of the two FDA approved treatments for spinal muscular atrophy.

When prenatal genetic therapies are ready for clinical use, the impact on prenatal genetic testing will be profound. Multifaceted research is imperative, not only on efficacy and safety of treatments, but also on long term outcomes, provider education, determining effective patient counseling models, service delivery models for increased genetic testing needs and patient experiences and attitudes toward testing in light of treatment options.

CONCLUSION

We have witnessed prenatal testing evolve drastically as we have progressed from the discovery of trisomy 21 to ES on prenatal samples. Screening has also advanced significantly from the observation of abnormal maternal analyte levels in pregnancies with Down syndrome. Today, we have the ability to screen for a multitude of chromosome and monogenic disorders with cfDNA. Despite these advances, many genetic conditions are not yet diagnosed in the prenatal setting and many challenges exist when expanding screening to more rare conditions. Furthermore, for those that are diagnosed, a prenatal diagnosis does not often lead to major improvements in postnatal outcomes. Benefits of prenatal diagnosis are often limited, with only some perceiving as beneficial advance knowledge of the diagnosis or the option to terminate the pregnancy. As therapies and treatments for genetic testing emerge, the purpose and goals of prenatal testing will shift, and the need for the accurate prenatal diagnosis will be essential.

As genetic testing continues to expand and treatments are developed, risks and benefits to prenatal testing will still coexist. Ensuring patient understanding of these benefits, risks, and also limitations will be of utmost importance. The roles for genetic counselors are numerous as these advancements are made, including nonclinical and clinical research involvement, the introduction of new technologies into clinical care, provider education, patient advocacy, obtaining informed consent, facilitating of testing and treatment decisions, delivery of results, and psychological support of patients and their patients throughout the process.