The promise and peril of new prenatal diagnostic technologies

Dr. Lockwood, Editor-in-Chief, is Dean of the Morsani College of Medicine and Senior Vice President of USF Health, University of South Florida, Tampa. He can be reached at DrLockwood@advanstar.com.

Our appreciation of the vast potential and major challenges posed by new genetic prenatal diagnostic testing seems to grow with each passing week. In particular, we now have a far better understanding of the utility and limitations of fetal cell-free DNA (cfDNA) testing and of expanded parental carrier testing. Beyond the ethical and economic issues raised by these new diagnostic modalities, their rapid introduction into clinical practice raises practical management questions that obstetricians must now address.

Is cfDNA the best screening test for common fetal aneuploidies?

A recent article in The New England Journal of Medicine appears to definitively answer this question.¹ Mary Norton and her colleagues reported the results of a large, industry-sponsored, prospective, blinded study comparing the effectiveness of cfDNA testing at 10 to 14 weeks with standard first-trimester screening using ultrasound nuchal translucency and maternal serum pregnancy-associated plasma protein A and human chorionic gonadotropin measurements for the detection of trisomies 21, 18, and 13. Of 18,955 women seeking routine prenatal screening, 15,841 were available for analysis with both screening modalities. Amongst these patients, cfDNA proved far more effective and efficient at detecting fetal trisomy. For example, fetal Down syndrome was detected using cfDNA testing in all 38 women bearing affected fetuses (100% sensitivity; 95%CI: 90.7–100) while standard testing identified only 30 of 38 women with affected fetuses (78.9% sensitivity; 95% CI: 62.7–90.4; P=0.008). Equally important, cfDNA had a 90-fold lower false-positive rate than standard testing (0.06%; 95% CI: 0.03–0.11 vs 5.4%; 95% CI: 5.1–5.8; P<0.001). Crucially, the positive predictive value of cfDNA for detecting fetal Down syndrome was an astonishingly high 80.9% (95% CI: 66.7–90.9) compared with only 3.4% (95% CI: 2.3–4.8) for standard first-trimester aneuploidy screening (P<0.001).

Also read: cfDNA testing predicts aneuploidy in low-risk pregnancies

Although the numbers of affected fetuses were too low to draw definitive conclusions, the efficiency of screening for trisomy 18 and 13 also appeared much higher for cfDNA. For trisomy 18, cfDNA yielded a higher sensitivity (90.0%; 95% CI: 55.5–99.7) and positive predictive value (90.0%; 95% CI: 55.5–99.7) than standard testing (80.0%; 95%CI: 44.4–97.5) and (14.0%; 95% CI: 6.2–25.8), respectively. For trisomy 13, cfDNA screening also yielded a higher sensitivity (100.0%; 95%CI: 15.8–100) and positive predictive value (50.0%; 95%CI: 6.8–93.2) than standard testing (50.0%; 95%CI: 1.2–98.7) and (3.4%; 95% CI: 0.1–17.8), respectively. Thus, cfDNA testing, when obtainable, is far more efficient than standard testing for the detection of trisomies 21, 18, and 13 in a routine obstetric population. Of course, standard testing provides other useful data such as the presence of potential skeletal dysplasias, cardiac abnormalities, and a number of other aneuploidies.

Reading between the lines of cfDNA testing 'failures'

The study by Norton and colleagues also provided other crucial insights that may alter our approach to aneuploidy screening. For example, they noted that 488 of 16,329 (3%) women otherwise eligible for cfDNA testing had to be excluded because of a lack of results due to low or undetectable fetal cfDNA fractions, high assay variance, or assay failure. Interestingly, 13 (1/38 or 2.7%) of these women were found to carry an aneuploid fetus: 4 with triploidy, 3 with trisomy 21, 2 with trisomy 13, and 1 each with trisomy 18, trisomy 16 mosaic, deletion 11p, and a structurally abnormal chromosome. For the subset of these pregnancies with a fetal cfDNA fraction less than the required cutoff of 4%, 9 in 192, or nearly 5%, had aneuploidy. Thus, inability to assay for cfDNA is a major risk factor by itself for fetal aneuploidy, the magnitude of which would seem to justify invasive fetal testing with chromosomal microarray studies. However, it should be noted that among the pregnancies in which cfDNA results were not available and in which the fetuses had 1 of the 6 most common aneuploides, all were detected by standard testing. Moreover, amongst the technologies used by the 5 companies currently providing cfDNA screening, there are differing threshold requirements for the fetal fraction.

Many false-positive cfDNA results are known to have clinically important causes, including a vanishing aneuploid twin, confined placental mosaicism, maternal mosaicism for aneuploidy, and even maternal cancers. Thus, in my opinion, “false-positive” cfDNA results require follow-up obstetric ultrasounds and a careful maternal history and physical examination to rule out maternal pathology.

Related: Discordant NIPT and ultrasound results from vanishing twin pregnancy 

The next new thing in cfDNA testing: A future fraught with promise and peril

As the genetic disorders potentially diagnosed by cfDNA testing continue to expand-and the potential here is almost unlimited-our ethical, economic, and practice management challenges also increase substantially. The ability of such testing to detect the 70% of aneuploidies beyond the most common (ie, trisomies 21, 18, and 13 and sex chromosome number abnormalities), such as microdeletions and microduplications, already exists.² Moreover, this technology can easily be leveraged to screen for relatively common single-gene defects such as achondroplasia. Thus, clinicians may soon be challenged by a myriad of screening offerings to present and explain to their patients. Does the couple want any prenatal diagnostic testing? If so, do they want just common trisomy and sex chromosome number screening? Do they also desire screening for microdeletions or microduplications – and should such testing include all such abnormalities or only those with clearly defined and severe phenotypes? Do they want all of these studies plus screening for known major single-gene defects, and if so, how many such disorders, and at what level of severity? How much will such testing cost? Who will pay-and will that cost be borne by health insurance and if not, what obligation do physicians, governments, and society have to ensure equitable access to such testing?

If all this is not worrisome enough, eventually whole genomic sequencing and assessment of the fetal methylome via cfDNA testing will be practical. But should it be made available? Such exhaustive diagnostic capabilities raise profound ethical challenges ranging from their impact on human evolution to its potential to exacerbate class and wealth distinctions (see Aldous Huxley’s Brave New World). The time to initiate this discussion is now, not after we obstetricians are overwhelmed by patient demands for services, counseling, detailed explanations, and “failure to diagnose” lawsuits.

Expanded carrier testing

Similar challenges exist as we begin to grapple with the complexities of expanded parental carrier testing. Traditionally, such testing has been very limited and focused on conditions with a specific ethnic and racial predilection (eg, Tay-Sachs and sickle cell screening). However, the United States is now a multiethnic, multiracial society and increasingly couples have many ethnicities and racial admixtures, often unbeknownst to them.³ For example, my grandson has Swedish, Norwegian, German, Irish, English, Welsh, Scottish, Scotch-Irish, Catalan, Spanish, French, Northern and Southern Italian, and Corsican ancestors!

Moreover, even among a given ethnic group, the number of diseases that can be tested for has expanded exponentially. Among the conditions that are often recommended to be screened for in couples of Ashkenazi Jewish ancestry are Tay-Sachs, Niemann-Pick A, Bloom syndrome, Canavan disease, familial dysautonomia, Fanconi anemia, Gaucher disease, mucolipidosis type IV, glycogen storage disease type 1a, maple syrup urine disease, nonsyndromic sensorineural hearing loss, and glycogen storage disease type III. However, even though Ashkenazi Jews are the most common carriers of Tay-Sachs disease alleles, because screening of such self-identified couples is so well utilized, most children affected by Tay-Sachs are not born to couples who identify themselves as Jewish.⁴ Thus, experts increasingly are urging pan-ethnic screening for many such disorders.

Expanded carrier testing using high-throughput sequencing, similar to that used for cfDNA testing, can simultaneously screen for a large number of parental carrier states at relatively low cost. While traditional carrier testing is focused on relatively common, clearly defined, severe and/or lethal autosomal-recessive and X-linked recessive conditions, expanded carrier testing panels often include conditions with far more variable phenotypes, such as Fragile X and hemochromatosis, or those that are far rarer. Paradoxically, amongst the latter conditions, calculation of residual risk is often impossible.⁵ The use of high throughput sequencing has the added complexity of identifying both pathological and potentially benign genetic variants, making post-test counseling more difficult. Conversely, such testing may uncover risks for severe or fatal conditions in the parent, including diseases that present later in life and for which there is no current prevention, such as Huntington’s, Parkinson’s, and Alzheimer’s diseases. Finally, while such expanded testing is now widely available, professional society recommendations have generally not kept pace with the proliferation of these tests. Moreover, some still propose limiting carrier testing to certain ethnic groups while others recommend pan-ethnic screening for all the reasons cited above.

Take-home message

We need to develop simple, clear strategies for offering prenatal and carrier testing that will benefit couples choosing to avail themselves of such studies. For the present we must better understand the advantages of both cfDNA and standard aneuploidy testing and develop algorithms that optimize each modality’s diagnostic strengths while minimizing costs. However, in the relatively near future, commercially available cfDNA diagnostic options are likely to vastly expand. When that occurs, we must develop a coherent set of cfDNA diagnostic options readily understandable by patients and readily implementable by practitioners.

One approach would be to offer a series of cfDNA testing sets of increasing complexity while ensuring that couples are fully informed of the scope and limitations of such screening. For example, patients might choose from a menu of cfDNA diagnostic offerings that include a basic trisomy screen, an expanded screen that includes a search for microdeletions and microduplications of known and clinically significant phenotype, or one that includes all these elements plus a screen for discrete single-gene conditions with severe phenotypes and high penetrance and expressivity. These tests could have variable pricing with payers determining the cost-effectiveness of coverage and market forces establishing residual pricing. Professional societies would need to collaborate to establish when a given condition should be added to the diagnostic paradigm. Couples would need to waive their right to access data of uncertain or trivial clinical significance. For patients in whom a cfDNA assay cannot be performed for technical reasons, consideration should be given to offering reflex standard testing or even invasive testing with chromosomal microarray analysis, depending on the parents’ wishes.

Expanded carrier screening also requires a standardized and parsimonious approach. A recent joint statement by American College of Obstetricians and Gynecologists, the American College of Medical Genetics and Genomics, the National Society of Genetic Counselors, the Perinatal Quality Foundation, and the Society for Maternal-Fetal Medicine outlines a cogent and practical set of principles for such testing.⁵ These include the necessary elements of pretest informed consent (eg, that risk assessment depends on an accurate knowledge of paternity, that a negative screen does not eliminate risk, and that screens for autosomal-dominant conditions may reveal prognostic information about them for which little can be done). Post-test counseling must also be available as well as a plan for sequential testing of the other parent for a given autosomal-recessive condition and information about prenatal diagnostic testing if so desired.

The Joint Statement authors also opine about which types of disorders should be tested for, such as those resulting in cognitive disability, those with substantial effects on quality of life, or those for which prenatal or obstetric interventions may improve outcomes. The authors stipulate that practitioners should have the freedom to not recommend screening for conditions with a highly variable adult phenotype (eg, alpha-1-antitrypisin deficiency) or conditions with low penetrance and expressivity (eg, Factor V Leiden or MTHFR mutations). The Statement also provides recommendations about which results should be reported (eg, reporting only those variants with clear phenotypes and the highest likelihood of being pathogenic).

We are at the dawn of a new age of prenatal testing that holds promise to provide expectant couples with important information about their fetus’ genetic heath. However, misapplication of these technologies will lead to stress, confusion, high costs and, potentially, social instability. Thus, we will need to find a practical equipoise amongst our obligations to: respect parental autonomy, personal choice, and individual freedom; ensure beneficence toward the fetus; and ensure a genetically diverse, just, and equitable society.

References

1. Norton ME, Jacobsson B, Swamy GK, et al. Cell-free DNA analysis for noninvasive examination of trisomy. N Engl J Med. 2015;372(17):1589-1597.

2. Zhao C, Tynan J, Ehrich M, et al. Detection of fetal subchromosomal abnormalities by sequencing circulating cell-free DNA from maternal plasma. Clin Chem. 2015;61(4):608-616.

3. Lee YL, Teitelbaum S, Wolff MS, Wetmur JG, Chen J. Comparing genetic ancestry and self-reported race/ethnicity in a multiethnic population in New York City. J Genet. 2010;89(4):417-423.

4. Intelihealth. Tay-Sachs Disease. http://www.intelihealth.com/article/tay-sachs-disease-0. Accessed May 16, 2015.

5. Edwards JG, Feldman G, Goldberg J, et al. Expanded carrier screening in reproductive medicine-points to consider: a joint statement of the American College of Medical Genetics and Genomics, American College of Obstetricians and Gynecologists, National Society of Genetic Counselors, Perinatal Quality Foundation, and Society for Maternal-Fetal Medicine. Obstet Gynecol. 2015;125(3):653-662.