Noninvasive fetal diagnosis

Dr. Simpson is Executive Associate Dean for Academic Affairs and Professor of Obstetrics and Gynecology and Human and Molecular Genetics at Florida International University College of Medicine, Miami. He is also a member of the Contemporary OB/GYN editorial board.

The ability to characterize cell-free fetal DNA in maternal blood and the recognition of its potential was first appreciated 15 years ago.¹ However, in June 2012 Kitzman and colleagues, in their report of noninvasive sequencing of the entire genome of an 18.5-week fetus using samples of maternal blood and paternal saliva, took the technology beyond detection of chromosomal or single-gene disorders to contemplation of genome-wide prenatal diagnosis of both recessive and dominant Mendelian disorders.²

Screening for and (essentially) detecting fetal trisomies is being accomplished from cell-free fetal DNA in maternal blood, based initially on studies of achieved samples derived from trisomic and nontrisomic pregnancies.^3-5

In 2011, and later in 2012, came reports of maternal blood samples that were analyzed “blindly” while traditional chorionic villus sampling (CVS) or amniotic fluid analysis was performed.^6-11 The results were highly salutary. The initial reports involved women already slated for CVS or amniocentesis (ie, high risk), blood samples being obtained before the invasive procedures. More recently, a similar study design reported favorable results from younger (mean age,
31.8 years) obstetric patients more analogous to the routine (low-risk) obstetric population.¹⁰

It is not an understatement to say that the science of prenatal genetic diagnosis is moving at a pace that would have seemed unimaginable just a few years ago. For ob/gyns, the challenge is not so much keeping up with the technical literature but assessing its applicability to patient care and determining when current evidence is sufficient to warrant use in a clinical setting.

Trisomy 18 and 21 detection: beginnings

More than 20 years ago, our own research group was among those pioneering detection of fetal trisomy via analysis of nucleated fetal red blood cells from maternal blood using fluorescence in situ hybridization (FISH) with chromosome-specific probes.¹² Our results were confirmed by Bianchi and colleagues and by Gänshirt-Ahlert and associates in 1992, and 1993, respectively.^13,14

With publication of a more detailed report on the technology in 1993, it seemed as though clinical application would be imminent.¹⁵ Unfortunately, fetal cell enrichment for diagnosis required and was made feasible only with fluorescence-activated cell sorting or magnetic-activated cell sorting. This proved laborious and inconsistent. Still, we obtained 74% sensitivity for detection of analyzable trisomy 21 fetuses in a multicenter collaborative study funded by the National Institute of Child Health and Human Development (NICHD).¹⁶

Fetal cell-free DNA

Given the inability to translate fetal red blood cell recovery readily into clinical use, the focus began to change to cell-free fetal DNA in the plasma of pregnant women, albeit complicated by admixture with cell-free maternal DNA.¹ The detection of Y chromosome DNA sequences in maternal plasma confirmed derivation from a male fetus and validated proof of principle.

During the aforementioned NICHD study,¹⁶ concomitant detection of Y sequences was in fact the sine qua non for verifying presence of intact fetal (male) cells. Detecting a qualitative difference in DNA sequence between mother and fetus is thus not difficult, because a novel (eg, Y) sequence DNA in the mother’s blood means that it had to have been of fetal origin, transmitted from the father. This general approach-detecting a qualitative difference between cell-free maternal and cell-free fetal DNA-was soon applied to detecting transmission to the fetus of any paternally transmitted mutation.¹⁷

Of special obstetric relevance was the noninvasive ability to detect when a paternally derived Rh (D) allele was transmitted to the fetus. This is of particular importance in families in which the father was heterozygous for the Rh gene and the mother was Rh negative. Use of cell-free fetal DNA analysis is now well accepted in Europe for determining whether Rh immune globulin should be given prenatally to Rh-negative women.¹⁸

Detecting fetal trisomy from cell-free DNA

Detecting fetal trisomy from the 5% or more of cell-free fetal DNA in maternal blood that is fetal in origin is more difficult than detecting paternally derived single-gene disorders. Unlike detecting the qualitative differences discussed earlier, trisomy detection requires distinguishing quantitative DNA differences between affected (trisomies) and unaffected (euploid) pregnancies because, of course, both mother and fetus have chromosome 21 DNA sequences.

Differentiating trisomic from normal pregnancies is still possible, however, because a pregnancy with a trisomy 21 fetus has more chromosome 21 DNA sequences than one with a normal fetus. The trisomic fetus has 3 21 chromosomes, rather than the usual 2. The small but finite quantitative difference in the total number of chromosome 21 sequences in maternal blood should be detectable. Given that the fetus is the source of perhaps 5% of cell-free DNA in maternal plasma, blood from a mother carrying a trisomy 21 pregnancy should have 2.5% more chromosome 21 sequences than if her fetus were not trisomic. This small but finite difference serves as the basis for detecting fetal trisomy (aneuploidy) by analysis of maternal blood. Two general approaches are possible.

Massively parallel genomic sequencing

One technology used in the United States by Sequenom (San Diego, California) and Verinata Health (Redwood City, California) is massively parallel genomic sequencing (MPGS). In MPGS, all cell-free fetal DNA (maternal plus fetal) is sequenced. From among some 25 million sequences (reads), those relevant to the chromosome being assessed (21, 13, 18) are quantitatively compared to number of “reads” expected.

Several reports validated a threshold using archived or accumulated samples of maternal plasma. In 7% of 576 plasma samples studied in 2011 by Chiu and colleagues,⁶ informative results were not possible, yet all 86 trisomy 21 cases in the remaining samples were successfully detected. In 2011 Ehrich and associates reported detection of 39 of 39 trisomy samples derived from 449 analyzable maternal plasma samples.⁷ There were no false positives.

In 2012 Bianchi and colleagues at Verinata detected 89 of 89 trisomy 21 pregnancies in analyzable samples.⁸ Of these samples, 3% had insufficient cell-free fetal DNA. Their method detected 35 of 36 trisomy 18 cases and 11 of 14 trisomy 13 cases. There were no false positive cases for these autosomal trisomies. Also in 2012, researchers in Germany reported on successful use of MPGS of DNA using the technique previously described by Chiu⁶ to identify correctly fetal trisomy 21 in all 8 samples from women carrying affected fetuses out of a total of 42 samples.¹⁹ In the same publication, investigators reported efforts to optimize quantification of trisomy 21 sequences with new algorithms for calculation of the z score.

These studies provide confidence that cell-free fetal DNA can be exploited to detect trisomy 21 pregnancies using a quantitative threshold
model. Using MPGS, detection rates are less than 100% for trisomies 13 and 18 but are still
very high.

Targeted DNA sequencing

Others are using a different approach for detecting fetal trisomies: targeted DNA sequencing. Rather than accumulating and sequencing maternal and fetal cell-free DNA from all chromosomes, only sequences from those chromosomes necessary for diagnosis are recovered and quantified-chromosomes 21, 13, and 18, for example. One can retrieve selected sequences using probes.

A progression of publications from 1 group (Ariosa; San Jose, California) illustrates salutary results based on either a 99% likelihood (threshold) for trisomy 21 or, in contrast, less than 1% (normal).⁵ Ariosa derives a specific risk figure based not only on quantitative values for the number of sequences for a given chromosome but also on maternal age and percentage of fetal DNA. Advancing maternal age is relatively more likely to be associated with a true positive than a false positive value (ie, sensitivity increases with incidence). Likewise, quantitative accuracy should directly correlate with amount of fetal DNA.

A set of trisomy 21 cases and controls first yielded 100% detection of trisomic samples.⁴ Norton and colleagues next reported a cohort study involving a sample of women (mean age, 34.3 years) undergoing mostly a second trimester procedure (mean gestational age, 17 weeks).⁹ The noninformative rate was 4.6% (1.8% inadequate fetal DNA; 2.8% assay failure). The detection rate was 81 of 81 trisomy 21 cases and 37 of 38 trisomy 18 cases. One false-positive case was observed (0.3%).

In the latest report using this approach, “routinely screened” first-trimester samples were assessed in women of maternal age of 31.8 years.¹⁰ The noninformative rate was again near 5%. Eight of 8 trisomy 21 cases and 2 of 3 trisomy 18 cases were detected. (The relatively low number of cases was expected, given maternal age.) The false-positive rate was 0.1%.

The other group in the United States using a targeted approach (Natera; San Carlos, California) bases its diagnosis on parental single nucleotide polymorphisms (SNPs),¹¹ an approach they previously showed to be of value in preimplantation genetic diagnosis.²⁰ Knowing parental SNPs allows one to enumerate all possible trisomic, disomic, and monosomic fetal genotypes. Using Bayesian calculations, predictive values for individual cases can be determined.

Zimmerman and associates studied 166 samples using this approach, targeting 11,000 SNPs on chromosomes 13, 18, 21, X, and Y.¹¹ The achieved samples studied included 2 trisomy 13 cases; 3 trisomy 18 cases; 11 trisomy 21 cases; 2 45,X cases; and 2 47,XXY cases. The correct chromosome number was reported in all cases.

A potential advantage of this approach is applicability using smaller percentages of cell-free fetal DNA. In the initial report, the noninformative rate was 12.6%, but the decision of whether to call a sample informative was based on the ability to obtain information on all 5 chromosomes, not the fewer number in other studies. The noninformative rate for the 3 autosomal trisomies alone (13, 18, 21) is comparable to other reports.

What can a reasonable obstetrician conclude?

At present, one could conclude that cell-free fetal DNA has a very high detection rate for trisomy 21: 99% or 100%. Detection is slightly less but still impressive for trisomies 13 and 18. Compared with the 85% to 90% modeled detection rate using maternal serum analytes and nuchal translucency (NT), cell-free DNA for aneuploidy screening would seem better. The false-positive rate is reported to be less than 1%, much lower than the 5% false-positive rate predetermined in maternal serum analyte/NT screening. Overall, cell-free fetal DNA for aneuploidy detection would be considered safer because fewer invasive procedures would be necessary.

Before universal adoption, it would be desirable to have a “real-time” series in which a lab processes samples on an ongoing basis and provides results that are acted on in that pregnancy. We have yet to know whether the more than 99% detection rate for trisomy 21 will remain as high as observed in specimens studied retrospectively in batch. And will the 1% or lower false-positive rate persist?

 Cell-free fetal DNA for detection of aneuploidy would, however, still be an advance with less than the near perfection reported to date for trisomy 21 detection in validation studies and studies using archived samples. Nuchal translucency would probably remain in use to detect a significant NT diameter (>3 mm or
>4 mm) that would indicate cardiac anomalies or warrant CVS for chromosomal analysis without further screening. However, maternal serum analytes might no longer be used as a primary screen for fetal aneuploidy.

False positives probably have a biologic basis and thus are irreducible. Explanations include a “vanishing” aneuploid co-twin, confined placental mosaicism, or somatic mosaicism in the mother. Confirmatory CVS or amniocentesis will be necessary. That is, cell-free fetal DNA analysis will remain a screen, not a test requiring no additional assays before a management decision. Confirmatory CVS or amniocentesis will, however, show a normal karyotype (false-positive screen) far less often than in maternal serum analyte/NT screening.

Noninformative results

At least 5% of samples are yielding noninformative results. One explanation is insufficient cell-free fetal DNA, based on a defined percentage of cell-free fetal DNA. An obvious solution might be repeating the assay 1 week later, but this will not necessarily yield better results because cell-free fetal DNA levels do not increase with advancing gestational age between 10 and 21 weeks. The other general explanation is assay failure. Quality-control problems may decrease with experience but also could be counterbalanced by problems arising as labs process more samples.

Currently, diagnosis can be offered only from 10 weeks’ gestation, and there may be insufficient DNA before this. With respect to early detection, it is interesting that correct diagnosis of cystic fibrosis and spinal muscular atrophy was recently reported in 63 consecutive cases by analysis of fetal trophoblasts recovered from maternal blood and “blindly” analyzed in the first trimester concurrent to carrying out CVS results that were acted on clinically. This process (Rarecells Diagnostics; Paris) involves molecular analysis of individual trophoblasts, which can be isolated on a filter beginning at 5 weeks gestation.²¹

Cost

Cost is a challenge to widespread use of cell-free DNA analysis for prenatal diagnosis. One vendor posts a list price approximating $2,700 using MPGS technology but takes into account what an individual insurance carrier allocates. Costs should ideally be comparable to those incurred with maternal serum analytes; however, costs need not be precisely identical given that if the false-positive rate is truly 1% or less, savings would occur as a result of the need to perform fewer invasive procedures.

Ready or not?

Suppose none of these described issues alone proves a serious impediment to implementation. Should the ob/gyn community alter its approach to prenatal aneuploidy screening and universally adopt cell-free fetal DNA for fetal aneuploidy screening? After all, detection rates using cell-free fetal DNA seem 10% to 15% higher than the 85% to 90% detection rate predicted from screening with maternal serum analyte and NT. Incidentally, maternal serum analyte screening was applied clinically after analysis of only archived samples; large-scale cohort studies came later. The former corresponds to the current state of the art using cell-free fetal DNA.

Should cell-free fetal DNA analysis be offered as a stand-alone screening assay? It should, but doubtless some will propose combination with selected components of extant maternal serum analysis screening programs. This is likely to be unnecessary and surely unwieldy.

Cell-free fetal DNA screening for aneuploidy should be robust enough to be offered without previous screens. A positive screen should proceed directly to a confirmatory invasive test, without intervening protocols. However, ultrasound (U/S) NT measurements in the first trimester to detect increased NT might remain in use for detection of nontrisomic fetal cardiac defects. Neural tube detection will also be necessary, either by second-trimester U/S or maternal serum alpha-fetoprotein.

Will patients accept cell-free fetal DNA as a fetal aneuploidy screen? Of course. The low false-positive rate alone would be a winner. Already we are seeing patient interest in minimizing false-positive scenarios. These are unexpected and anxiety provoking and require invasive procedures. Patients will welcome the higher detection rate.

A recent report from the United Kingdom, where noninvasive prenatal diagnosis based on cell-free fetal DNA is a well-established service for patients at high risk for sex-linked genetic disorders, indicates that patients are receptive to the technology but that expert counseling may be important before and after testing.²²

There is 1 final consideration: the diagnostic limitations at present using cell-free fetal DNA analysis. Cell-free fetal DNA does not yet permit application of the exciting new approach of array comparative genome hybridization (CGH).²³ Array CGH identifies not only all aneuploidies but also microdeletions and microduplications too small to be visualized by traditional karyotype. Thus, prenatal array CGH has far greater diagnostic breadth than simply detecting trisomies 13, 18, and 21 with cell-free fetal DNA.

In pregnancies with fetal anomalies on U/S, array CGH detects 6% additional cytogenomic abnormalities than karyotype alone and 1.7% more even in women of advanced age without fetal anomalies.²³ Performing array CGH in cell-free fetal blood is theoretically possible, but at present prenatal array CGH will probably require intact fetal cell(s) derived from CVS or amniocentesis or intact fetal trophoblasts from maternal blood.²¹

The future of fetal chromosomal testing

Initial reports in 2011 and 2012 are allowing us to conclude confidently that cell-free fetal DNA analysis is valid for detecting trisomy 21. Detection to date is near 100%. The false-positive rate is less than 1%, far lower than the 5% expected with maternal serum analyte/NT screening. Fewer invasive procedures will thus be needed.

If results continue to be so favorable, there will be much to recommend cell-free fetal DNA as the primary screening approach for fetal aneuploidy. Initially, guidelines unavoidably will recommend this approach for “high risk” populations, given reports primarily having involved such samples.

The new committee opinion from the American College of Obstetricians and Gynecologists (ACOG) reflects the literature to date.²⁴ That the field is accelerating is evident even in this recent publication, because not cited is the 1 study involving a low-risk population.¹⁰ Revision will be appropriate as similar reports emerge.

Indeed, in December 2012 Beijing Genetics Institute reported a favorable experience in a study of 11,105 pregnancies, both low and high risk, over 2 years.²⁵ Of 49 centers, 7 enrolled patients irrespective of prior risk assessments. All 143 trisomy 21 cases were detected and all 47 trisomy 18 cases. A single false positive trisomy 21 case and a single false positive trisomy 18 were observed. Given this report, published after the ACOG committee opinion,²⁴ a strong case can now be made for ACOG guidelines to extend to the entire obstetrical population.

Lurking like a figure in the background of a Magritte painting is prenatal array CGH. Array CGH can increase greatly the number of detectable disorders beyond fetal trisomy. Is it then still appropriate to detect only trisomies when array CGH can detect much more? At what cost should an invasive procedure be eschewed? Will the detection of greater numbers of disorders by array CGH (or sequencing the entire fetal genome) trump the noninvasive approach of cell-free fetal DNA, thus actually generating more invasive tests? Or can cell-free fetal DNA evolve to achieve most of the diagnoses now possible only by array CGH?

On the other hand, if intact fetal trophoblasts can be recovered readily, would this be a more attractive option than studying cell-free fetal DNA unavoidably admixed with maternal DNA? The prenatal landscape is indeed likely to continue to change.

References

1. Lo YM, Corbetta N, Chamberlain PF, et al. Presence of fetal DNA in maternal plasma and serum. Lancet. 1997;350(9076):485-487.

2. Kitzman JO, Snyder MW, Ventura M, et al. Noninvasive whole-genome sequencing of a human fetus. Sci Transl Med. 2012;4(137):137ra76.

3. Fan HC, Blumenfeld YJ, Chitkara U, Hudgins L, Quake SR. Noninvasive diagnosis of fetal aneuploidy by shotgun sequencing DNA from maternal blood. Proc Natl Acad Sci U S A. 2008;105(42):16266-16271.

4. Ashoor G, Syngelaki A, Wagner M, Birdir C, Nicolaides KH. Chromosome-selective sequencing of maternal plasma cell-free DNA for first-trimester detection of trisomy 21 and trisomy 18. Am J Obstet Gynecol. 2012; 206(4):322.e1-322.e5.

5. Sparks AB, Struble CA, Wang ET, Song K, Oliphant A. Noninvasive prenatal detection and selective analysis of cell-free DNA obtained from maternal blood: evaluation for trisomy 21 and trisomy 18. Am J Obstet Gynecol. 2012;206(4):319.e1-319.e9.

6. Chiu RW, Akolekar R, Zheng YW, et al. Non-invasive prenatal assessment of trisomy 21 by multiplexed maternal plasma DNA sequencing: large scale validity study. BMJ. 2011;342:c7401.

7. Ehrich M, Deciu C, Zwiefelhofer T, et al. Noninvasive detection of fetal trisomy 21 by sequencing of DNA in maternal blood: a study in a clinical setting. Am J Obstet Gynecol. 2011;204(3):205.e1-205.e11.

8. Bianchi DW, Platt LD, Goldberg JD, et al; Maternal Blood is Source to Accurately Diagnose Fetal Aneuploidy (MELISSA) Study Group. Genome-wide fetal aneuploidy detection by maternal plasma DNA sequencing. Obstet Gynecol. 2012;119(5):890-901.

9. Norton ME, Brar H, Weiss J, et al. Non-Invasive Chromosomal Evaluation (NICE) Study: results of a multicenter prospective cohort study for detection of fetal trisomy 21 and trisomy 18. Am J Obstet Gynecol. 2012;207(2):137.e1-137.e8.

10. Nicolaides KH, Syngelaki A, Ashoor G, Birdir C, Touzet G. Noninvasive prenatal testing for fetal trisomies in a routinely screened first-trimester population. Am J Obstet Gynecol. 2012;207(5):374.e1-374.e6.

11. Zimmermann B, Hill M, Gemelos G, et al. Noninvasive prenatal aneuploidy testing of chromosomes 13, 18, 21, X, and Y, using targeted sequencing of polymorphic loci. Prenat Diagn. 2012;32:1-9.

12. Price JO, Elias S, Wachtel SS, et al. Prenatal diagnosis with fetal cells isolated from maternal blood by multiparameter flow cytometry. Am J Obstet Gynecol. 1991;165(6 part 1):1731-1737.

13. Bianchi DW, Mahr A, Zickwolf GK, Houseal TW, Flint AF, Klinger KW. Detection of fetal cells with 47,XY,+21 karyotype in maternal peripheral blood. Hum Genet. 1992;90(4):368-370.

14. Gänshirt-Ahlert D, Börjesson-Stoll R, Burschyk M, et al. Detection of fetal trisomies 21 and 18 from maternal blood using triple gradient and magnetic cell sorting. Am J Reprod Immunol. 1993;30(2-3):194-201.

15. Simpson JL, Elias S. Isolating fetal cells from maternal blood. Advances in prenatal diagnosis through molecular technology. JAMA. 1993;270(19):2357-2361.

16. Bianchi DW, Simpson JL, Jackson LG, et al. Fetal gender and aneuploidy detection using fetal cells in maternal blood: analysis of NIFTY I data. National Institute of Child Health and Development Fetal Cell Isolation Study. Prenat Diagn. 2002;22(7):609-615.

17. Saito H, Sekizawa A, Morimoto T, Suzuki M, Yanaihara T. Prenatal DNA diagnosis of a single-gene disorder from maternal plasma. Lancet. 2000;356(9236):1170.

18. Clausen FB, Christiansen M, Steffensen R, et al. Report of the first nationally implemented clinical routine screening for fetal RHD in D- pregnant women to ascertain the requirement for antenatal RhD prophylaxis. Transfusion. 2012;52(4):752-758.

19. Stumm M, Entezami M, Trunk N, et al. Noninvasive prenatal detection of chromosomal aneuploidies using different next generation sequencing strategies and algorithms. Prenat Diagn. 2012;32(6):569-577.

20. Johnson DS, Gemelos G, Baner J, et al. Preclinical validation of a microarray method for full molecular karyotyping of blastomeres in a 24-h protocol. Hum Reprod. 2010;25(4):1066-1075.

21. Mouawia H, Saker A, Jais JP, et al. Circulating trophoblastic cells provide genetic diagnosis in 63 fetuses at risk for cystic fibrosis or spinal muscular atrophy. Reprod Biomed Online. 2012;25(5):508-520.

22. Lewis C, Hill M, Skirton H, Chitty LS. Fetal sex determination using cell-free fetal DNA: service users’ experiences of and preferences for service delivery. Prenat Diagn. 2012;32(8):735-741.

23. Wapner RJ, Martin CL, Levy B, et al. Chromosomal microarray versus karyotyping for prenatal diagnosis. N Engl J Med. 2012;367(23):2175-2184.

24. Committee opinion no. 545: noninvasive prenatal testing for fetal aneuploidy. Obstet Gynecol. 2012;120(6):1532-1534.

25. Dan S, Wang W, Ren J, et al. Clinical application of massively parallel sequencing-based prenatal noninvasive fetal trisomy test for trisomies 21 and 18 in 11 105 pregnancies with mixed risk factors. Prenat Diagn. 2012;32(13):1225-1232.

Take-home messages

The science of prenatal genetic diagnosis is moving at a pace that would have seemed unimaginable just a few years ago.

Cell-free fetal DNA has a very high detection rate for trisomy 21: 99% or 100%.