Noninvasive Prenatal Testing forFetal AneuploidyClinical Assessment and a Plea for Restraint

Mary E. Norton, MD, Nancy C. Rose, MD, and Peter Benn, PhD

The recent introduction of clinical tests to detect fetal

aneuploidy by analysis of cell-free DNA in maternal

plasma represents a tremendous advance in prenatal

diagnosis and the culmination of many years of effort by

researchers in the field. The development of noninvasive

prenatal testing for clinical application by commercial

industry has allowed much faster introduction into

clinical care, yet also presents some challenges regarding

education of patients and health care providers strug-

gling to keep up with developments in this rapidly

evolving area. It is important that health care providers

recognize that the test is not diagnostic; rather, it

represents a highly sensitive and specific screening test

that should be expected to result in some false-positive

and false-negative diagnoses. Although currently being

integrated in some settings as a primary screening test

for women at high risk of fetal aneuploidy, from

a population perspective, a better option for noninvasive

prenatal testing may be as a second-tier test for those

patients who screen positive by conventional aneuploidy

screening. How noninvasive prenatal testing will ulti-

mately fit with the current prenatal testing algorithms

remains to be determined. True cost–utility analyses will

be needed to determine the actual clinical efficacy of this

approach in the general prenatal population.

(Obstet Gynecol 2013;121:847–50)

DOI: http://10.1097/AOG.0b013e31828642c6

The potential to obtain and analyze fetal geneticmaterial from maternal plasma has stimulated vig-

orous research for more than three decades. InOctober2011, these efforts resulted in the first commerciallyavailable test to detect fetal aneuploidy by analysis ofcell-free DNA in maternal plasma. Over time, thistechnology is likely to be applied to the prenatal detec-tion of an ever-increasing range of genetic disorders.This commentary presents an overview of the devel-opment, clinical application, and limitations of nonin-vasive prenatal testing as currently available in clinicalpractice.

DEVELOPMENT OF NONINVASIVEPRENATAL TESTING

The development of cell-free DNA testing differs inimportant ways from that of other methods of prenataldiagnosis and screening. Amniocentesis, chorionicvillous sampling, and serum screening were largelydeveloped by academic investigators who were sup-ported by public funding. After the conduct ofindependent clinical trials, with results presented atscientific meetings and published in peer-reviewedjournals, clinical tests were introduced and supportedby national committee guidelines. This process wasslow but relied on the unbiased external review ofdata before widespread introduction of tests andchanges in the standards of clinical care. The techni-ques to analyze cell-free DNA in maternal plasmawere also originally developed in academic settingsbut were rapidly licensed to commercial companiestypically supported by venture capital or other privatefunding. The tests that were developed are nowpublicized through marketing directly to the clinician,

From the Department of Obstetrics and Gynecology, Stanford University Schoolof Medicine/Lucile Packard Children’s Hospital, Stanford, California; Inter-mountain Healthcare, University of Utah School of Medicine, IntermountainMedical Center, Maternal Fetal Medicine, Salt Lake City, Utah; and the Depart-ment of Genetics and Developmental Biology, University of Connecticut HealthCenter, Farmington, Connecticut.

Corresponding author: Mary E. Norton, MD, 300 Pasteur Drive, HH333,Stanford University/Lucile Packard Children’s Hospital, Stanford, CA 94305;e-mail: menorton@stanford.edu.

Financial DisclosureDr. Norton is a coprincipal investigator on clinical trial NCT0145167sponsored by Ariosa Diagnostics. The other authors did not report any potentialconflicts of interest.

© 2013 by The American College of Obstetricians and Gynecologists. Publishedby Lippincott Williams & Wilkins.ISSN: 0029-7844/13

VOL. 121, NO. 4, APRIL 2013 OBSTETRICS & GYNECOLOGY 847

and scientific presentations at times appear gearedtoward a specific product. Although such commercialdevelopment may allow for faster introduction intoclinical practice, to some degree, it bypasses somecomponents of a rigorous, unbiased vetting process,which helps to provide quality assurance.

The cell-free DNA field is still evolving, and theclinical tests have limitations that have not beenhighlighted in the rush to capture market share.Clearly, the sensitivity and specificity are very high,and this is likely to result in less patient anxiety asa result of false-positive results and a decrease ininvasive procedures. However, both false-positive andfalse-negative results do occur, and, therefore, theseshould be considered screening tests. The positivepredictive value varies substantially with other riskfactors and is relatively low in low-risk patients. Forexample, assuming a detection rate of 99% and a 0.5%false-positive rate, in a woman with a prior risk of onein 1,000 for Down syndrome, the likelihood thata positive result is a true-positive is only approxi-mately one in six. Even for a higher risk woman witha prior risk of one in 100, the chance of Downsyndrome with a positive result is two in three; onethird of positive results will be false positives. Onlineconversations among patients, health care providers,and even geneticists regarding false-positive resultsand other test characteristics demonstrate a lack ofunderstanding of some of these features of currentcell-free DNA tests. Further education is criticallyimportant to assure appropriate interpretation ofpositive test results.

The three clinically available tests offered bySequenom, Inc., Verinata Health, Inc., and AriosaDiagnostics, Inc. have quite different characteristics interms of how the sequencing data are analyzed andresults are reported. Sequenom presents results asdichotomous outcomes, in which they indicate thatincreased chromosome 21 material is either present ornot, implying to some health care providers (althoughnot explicitly stating) diagnostic accuracy. Ariosaprovides results as a risk, more typical of a screeningtest. Verinata reports three categories of results, includ-ing “aneuploidy detected,” “aneuploidy suspected,” or“no aneuploidy detected.” Another approach that pro-vides a case-specific estimate of risk is being developedby Natera Inc. This heterogeneity in approach contrastswith current serum screening tests, whose reporting andlaboratory standardization are much more consistentbetween laboratories.

The ease of obtaining blood samples for cell-freeDNA testing, together with the promise of early andaccurate clinical information and the intense marketing,

make it tempting to use these tests even in the absenceof comprehensive validation or understanding. Froma clinical provider’s standpoint, it is difficult and time-consuming enough to review standard information inestablished prenatal care without also having to explainto patients the limitations of these screening tests andthe differences between clinical validity and clinicaluse. However, given that pregnancy termination isa potential result of a positive test, and that cell-freeDNA tests will now be provided through general pre-natal practices rather than specialized prenatal diagno-sis centers, obstetric providers will absorb more of theburden of discussing these complex results.

QUALITY CONTROL AND REGULATION OFGENETIC TESTING

Unlike many European countries, the United States haslimited direct regulation of genetic testing. As a result,large professional societies, private medical insurers,and for-profit companies largely govern the uptake andintegration of new technologies into prenatal practice,and there is currently no legal obligation to demon-strate safety and effectiveness of cell-free DNA tests. Aslaboratory-developed tests, there is no requirement forpremarket approval by the U.S. Food and DrugAdministration.1 Although the laboratories providingnoninvasive prenatal testing in the United States arecompliant with the Clinical Laboratory ImprovementAmendments of 1988, Clinical Laboratory Improve-ment Amendments regulations are designed to overseecompliance at a laboratory operational level and do notextend to validation of specific tests.

The absence of guidelines for quality control andquality assurance for the laboratory testing is a signif-icant concern. The methodologies are highly complexand the testing is subject to intense commercialcompetitive pressures. Such competition has led toa series of patent infringement lawsuits between thedifferent companies; how these legal challenges will bedecided, and whether they will affect access to tests andviability of some of the companies, remains uncertain.

The Secretary’s Advisory Committee on Genet-ics, Health and Society has noted gaps in five mainareas of oversight that directly affect optimal use ofgenetic testing, including clinical laboratory qualityassurance, transparency of genetic testing, oversightof clinical validity of genetic tests, level of knowledgeregarding clinical use of genetic tests, and meeting theinformational needs of health professionals, the publichealth community, patients, and consumers. Thisgroup has noted that “.the growing use of genetictesting will require significant investment in evidence-based assessments to understand the validity and

848 Norton et al Noninvasive Prenatal Testing for Aneuploidy OBSTETRICS & GYNECOLOGY

utility of these tests in clinical and personal decision-making.”1

In the absence of direct regulation, other optionshave been suggested for oversight of these, and other,genetic tests. Professional societies could encouragevoluntary registration of new prenatal tests with theNational Institute of Health’s Genetic Test Registry,which would promote the documentation of clinicaland analytic validity of new tests before active use. Amodel that could be useful to evaluate new prenatalgenetic testing is one similar to the Secretary’s Advi-sory Committee for Heritable Diseases in Newbornsand Children. Chartered in 2003, this multidisciplinaryworking group is designed to report to the Secretary ofHealth and Human Services on “.the most appropri-ate application of universal newborn screening tests,technologies, policies, guidelines and standards” toeffectively reduce morbidity and mortality in newbornsand children having, or at risk for, heritable disorders.2

A similar multidisciplinary committee to oversee pre-natal genetic testing, which is likely to increase greatlyin scope and complexity in coming years, could be anappropriate venue for test evaluation and consider-ation. In the meantime, the Secretary’s Advisory Com-mittee for Heritable Diseases in Newborns andChildren suggests that professional societies shouldissue guidelines that encourage physicians to reviewvalidity data before ordering tests and likewise shouldrecommend minimum levels of clinical use.1

APPROPRIATE CLINICAL APPLICATION

Despite the complex background of cell-free DNAtesting, benefits for implementation into patient careclearly exist. However, the aggressive advertising, highvisibility in the press, and the initial introduction ofthese tests before publication of professional societyguidelines has left the general obstetrician with someconfusion on how to best incorporate this new tool intoclinical practice. To critically review the scientificpublications and noninvasive prenatal testing clinicaltrial reports requires an in-depth understanding ofgenetics, molecular biology, and statistics. Moreover,there are limited data on how well the tests will performin actual clinical practice, particularly for specificsubgroups of women in which noninvasive prenataltesting has not been adequately validated. The testshave primarily been validated on archived samples incarefully selected groups of high-risk women; suchstudies do not answer the question of clinical use inthe general population. Nevertheless, obstetricians maybe concerned about their liability if they do not offercell-free DNA tests as an available option. Conversely,given that pregnancy termination is a potential result of

cell-free DNA testing, the obstetrician bears theburden for accurate counseling and interpretationof test results.

Current American College of Obstetricians andGynecologists guidelines recognize that prenatal diag-nosis should be available to all women regardless ofmaternal age3 and that prenatal screening tests canhelp women decide whether to accept or reject inva-sive testing.4 First-trimester screening with nuchaltranslucency and maternal serum screening carriesadded benefits, including identification of aneuploidiesbeyond those currently detectable with noninvasiveprenatal testing.5 In a recent publication on noninva-sive prenatal testing in an average risk population ofwomen undergoing first-trimester screening, noninva-sive prenatal testing detected 55% of the total chromo-somal abnormalities (eight of eight cases of trisomy 21and two of three cases of trisomy 18), whereas first-trimester screening detected 100% (all T21 and T18 aswell as seven other deletions, duplications, and otherabnormalities).6 In addition, screening for neural tubeand ventral wall defects is an important component ofcurrent screening protocols. Optimal introduction ofnoninvasive prenatal testing would preserve the prin-ciple of providing prenatal screening and diagnosis ina way that maximizes women’s reproductive choiceand, in addition, ensure that noninvasive prenatal test-ing is not performed on women who are unaware ofthe purpose and scope of the test.7

Not surprisingly, the major companies providingthis testing in the United States are targeting large groupsof women for testing. Sequenom and Verinata advocatetesting on all high-risk women, which includes all womenof advanced maternal age. However, many older womenundergo other screening tests and most are determinedto be lower risk; currently available and recommendedserum and ultrasonographic protocols are better atidentifying high-risk women than just age alone.8 In gen-eral, the concept of “advanced maternal age” as a screen-ing test is now considered arbitrary and outdated.

Offering noninvasive prenatal testing to all preg-nant women eliminates the difficulties associated withoffering the test to only selected groups but at consider-able financial cost. Ariosa recommends screeningwomen of all ages and a complete replacement for allexisting aneuploidy screening but makes this recom-mendation in the absence of published in-depthresearch with regard to efficacy. They have carriedout limited studies on low-risk women and demon-strated that the test failure rate appears to be no higherand the false-positive rate comparable with that inhigh-risk women. Before concluding that conventionalscreening could be replaced by noninvasive prenatal

VOL. 121, NO. 4, APRIL 2013 Norton et al Noninvasive Prenatal Testing for Aneuploidy 849

testing, a detailed comparison of all costs and benefits,both direct and indirect, needs to be carried out.

RECOMMENDATIONSAND CONSIDERATIONS

So what is the general obstetric provider to do at thispoint in time? The American College of Obstetriciansand Gynecologists recent published guidelines indi-cating that cell-free DNA testing is of benefit in high-risk women.9 However, they caution that such testingshould not be part of routine care but should only beprovided after pretest counseling. Furthermore, theynote that the test has not yet been adequately evalu-ated in low-risk women, to whom it should not yet beoffered. In addition, we would suggest considering thefollowing: 1) Although noninvasive prenatal testinghas high detection rates and low false-positive rates,it is a screening test and a positive noninvasive pre-natal testing result must be confirmed by invasivetesting if pregnancy termination is being considered.2) In high-risk women, particularly those identified assuch through traditional screening, noninvasive pre-natal testing can be very useful but it should be madeclear that it screens for limited fetal aneuploidies atthis time (trisomies 21, 18, and 13). A diagnostic inva-sive test definitively identifies a much broader rangeof chromosome abnormalities (especially if microar-ray technology is used). 3) If patients are to undergononinvasive prenatal testing, they need to understandthe purpose and limitations of the test; this cannot beroutinely added to standard prenatal laboratory test-ing without adequate counseling. 4) Current serumscreening and ultrasonographic strategies that havebeen thoughtfully developed and evaluated and anyreplacement should be based on evidence of improvedclinical use, including considerations of the range ofabnormalities detected with each strategy. 5) A strategyof concurrently ordering both noninvasive prenataltesting and integrated screening cannot be advocated,because this seems likely to greatly increase costs with-out current evidence of incremental benefit.

At this time, the most appropriate use of non-invasive prenatal testing may be as a second-tier testfor those with screen-positive results from conventionalaneuploidy screening. As noninvasive prenatal testingexpands to screen for more disorders and there aremore data available on its use, it may replace currentscreening methods. However, at the present time, itdoes not replace all forms of prenatal screening.

Finally, the implications of failed tests need to beconsidered. Up to 5% of cases do not generate a result,often as a result of insufficient fetal cell-free DNA(especially for women with high body mass index).10 It

is unknown at what body mass index cutoff the test isno longer worth attempting and in what circumstancesa repeat attempt is worthwhile. Results of noninvasiveprenatal testing can take 1–3 weeks, and the implica-tions of the turnaround time, especially if a test result isultimately not provided, need to be considered, espe-cially for patients who are later in the second trimester,for whom it may be too late to pursue other screeningor diagnostic testing options.

In summary, noninvasive prenatal testing usingcell-free DNA is an exciting new technology withtremendous potential to benefit pregnant women.Undoubtedly, cell-free DNA technology will expandover the next few years and dramatically advance thefield of prenatal screening and diagnosis. Developingreasonable clinical management guidelines and edu-cation will be essential as the testing becomes moresophisticated. Like other new technologies, it shouldbe carefully assessed in an unbiased fashion before itcompletely replaces our current standard of care andbefore we change prenatal practice for the four millionpregnancies that occur in this country each year.

REFERENCES1. Ferreira-Gonzalez A, Teutsch S, Williams MS, Au SM,

Fitzgerald KT, Miller PS, et al. US system of oversight forgenetic testing: a report from the Secretary’s Advisory Commit-tee on Genetics, Health and Society. Per Med 2008;5:521–8.

2. Secretary’s Advisory Committee on Heritable Disorders inNewborns and Children. Available at: http://www.hrsa.gov/advisorycommittees/mchbadvisory/heritabledisorders/index.html.Retrieved October 23, 2012.

3. Screening for fetal chromosomal abnormalities. ACOG Prac-tice Bulletin No. 77. American College of Obstetricians andGynecologists. Obstet Gynecol 2007;109:217–27.

4. Invasive prenatal testing for aneuploidy. ACOG Practice Bulle-tin No. 88, December 2007. American College of Obstetriciansand Gynecologists. Obstet Gynecol 2007;110:1459–67.

5. Syngelaki A, Chelemen T, Dagklis T, Allan L, Nicolaides KH.Challenges in the diagnosis of fetal non-chromosomal abnor-malities at 11–13 weeks. Prenat Diagn 2011;31:90–102.

6. Nicolaides KH, Syngelaki A, Ashoor G, Birdir C, Touzet C.Noninvasive prenatal testing for fetal trisomies in a routinelyscreened first-trimester population. Am J Obstet Gynecol 2012;207:374.e1–6.

7. Benn PA, Chapman AR. Ethical challenges in providing non-invasive prenatal diagnosis. Curr Opin Obstet Gynecol 2010;22:128–34.

8. Malone FD, Canick JA, Ball RH, Nyberg DA, Comstock CH,Bukowski R, et al. First-trimester or second-trimester screening,or both, for Down’s syndrome. N Engl J Med 2005;353:2001–11.

9. Noninvasive prenatal testing for fetal aneuploidy. CommitteeOpinion No. 545. American College of Obstetricians andGynecologists. Obstet Gynecol 2012;120:1532–4.

10. Ashoor G, Poon L, Syngelaki A, Mosimann B, Nicolaides KH.Fetal fraction in maternal plasma cell-free DNA at 11–13 weeks’gestation: effect of maternal and fetal factors. Fetal Diagn Ther2012;31:237–43.

850 Norton et al Noninvasive Prenatal Testing for Aneuploidy OBSTETRICS & GYNECOLOGY