Professional Documents
Culture Documents
CURRENT
OPINION Noninvasive prenatal diagnosis for single
gene disorders
Stephanie Allen, Elizabeth Young, and Benjamin Bowns
Purpose of review
Noninvasive prenatal diagnosis for single gene disorders is coming to fruition in its clinical utility. The
presence of cell-free DNA in maternal plasma has been recognized for many years, and a number of
applications have developed from this. Noninvasive prenatal diagnosis for single gene disorders has
lagged behind due to complexities of technology development, lack of investment and the need for
validation samples for rare disorders.
Recent findings
Publications are emerging demonstrating a variety of technical approaches and feasibility of clinical
application. Techniques for analysis of cell-free DNA including digital PCR, next-generation sequencing and
relative haplotype dosage have been used most often for assay development. Analysis of circulating fetal
cells in the maternal blood is still being investigated as a viable alternative and more recently transcervical
trophoblast cells. Studies exploring ethical and social issues are generally positive but raise concerns
around the routinization of prenatal testing.
Summary
Further work is necessary to make testing available to all patients with a pregnancy at risk of a single gene
disorder, and it remains to be seen if the development of more powerful technologies such as isolation and
analysis of single cells will shift the emphasis of noninvasive prenatal diagnosis. As testing becomes
possible for a wider range of conditions, more ethical questions will become relevant.
Keywords
cell-free fetal DNA, fetal cells, next-generation sequencing, prenatal diagnosis, single gene disorder
1040-872X Copyright ß 2017 Wolters Kluwer Health, Inc. All rights reserved. www.co-obgyn.com
Table 1. Techniques for analysis of cell-free fetal DNA for noninvasive prenatal diagnosis of single gene disorders
dysplasia
COLD-PCR and microarray CF, beta thalassemia [19] 1
Fragment analysis Huntington disease [20] 1
Allele-specific real-time PCR b-thalassemia [21,22] 2
cSMART Wilson disease [23] 1
Articles within the review period reporting NIPD of SGDs by analysis of cell-free fetal DNA. These are categorized into methods used for testing and disorders
tested. Source: original. CF, cystic fibrosis; COLD, coamplification at lower denaturation temperature; cSMART, ciruclating single-molecule amplification and
resequencing technology; NGS, next-generation sequencing; NIPD, noninvasive prenatal diagnosis; SGD, single gene disorder.
FIGURE 1. Principles of relative haplotype dosage. (a) Relative haplotype dosage analysis for X-linked disorders showing how
over-representation of informative single nucleotide polymorphisms on the X chromosome is used to show whether the fetus has
inherited the mutant or normal copy of the gene. Adapted from [8 ] (open access). (b) Relative haplotype dosage analysis for
&
tests was considered prohibitively high in terms of single gene disorders study that was aimed at devel-
clinical application. oping tests with a level of affordability to enable
The haplotype-based approach was taken fur- clinical implementation. Two publications from the
ther during the noninvasive prenatal diagnosis for study detail the development of assays for NIPD of
1040-872X Copyright ß 2017 Wolters Kluwer Health, Inc. All rights reserved. www.co-obgyn.com 75
Duchenne/Becker muscular dystrophy (DMD/BMD) of monogenic disorders. They performed NIPD SNP
&
and SMA [8 ,9]. Methods were similar to those genotyping in 55 maternal plasma samples, achiev-
above, except that analysis of cfDNA to determine ing 100% accuracy for detection of paternal alleles
the fetal haplotype was done by relative haplotype and 96% accuracy for diagnosis of alleles with
dosage (RHDO). The predicted fetal haplotypes were maternal origin by RMD. They conclude that
in agreement with known outcomes in each of the their study demonstrates the potential of ddPCR
nine and 16 pregnancies tested. Importantly, by for NIPD fetal genotyping independently of inher-
combining these two assays into a panel [and with itance pattern.
assays for cystic fibrosis (CF) and congenital adrenal
hyperplasia (CAH) in development], the authors
state that it is possible to test up to three patients BESPOKE/TARGETED NEXT-GENERATION
simultaneously for any of three single gene dis- SEQUENCING
orders. This significantly reduces the cost to the Two recent publications describe the use of targeted
point that the DMD/BMD and SMA tests are now next-generation sequencing (NGS) for NIPD of
&
available as a clinical service. skeletal dysplasias [17,18 ]. Skeletal dysplasias may
The RHDO method has also been adopted in a be diagnosed by ultrasonography, and therefore a
recent study for NIPD of CAH [10], with the fetal noninvasive test is beneficial for diagnosis without
genotype correctly predicted in all five cases putting the pregnancy at risk. NIPD can be a useful
reported so far. Haplotype-based NIPD of CAH has aid to clinical management as it allows a definitive
been explored previously with similarly promising diagnosis, differentiation between lethal and non-
results, as summarized in a recent review article by lethal forms, and the option of a surgical termin-
Kazmi et al. [11]. ation as a postmortem is not required. It can also be
Haplotype-based methods have been used in a helpful in twin pregnancies to avoid putting the
slightly different way by Zeevi et al. [12]. They unaffected twin at risk. Alternatively, there may
demonstrated a universal strategy for NIPD in the be a family history of the disorder, and where this
Ashkenazi-Jewish population using SNPs to con- is apparently de novo (i.e. not detected in either
struct a haplotype around a prevalent mutation parent), or paternally inherited, then NIPD for the
on a common haplotype background in that mutation can be offered very early in pregnancy.
particular population. Chitty et al. report on the development of an NGS
panel covering 29 known disease-causing mutations
in the FGFR3 gene, associated with achondroplasia
DROPLET DIGITAL PCR and thanataphoric dysplasia. Dan et al. developed a
There have been several studies that have utilized targeted capture sequencing method for de-novo
droplet digital PCR (ddPCR) for NIPD. ddPCR is mutation detection in 16 lethal skeletal dysplasia
based on partitioning of the sample into tens to genes. They analyzed maternal and paternal
hundreds of thousands of droplets, which each con- genomic DNA alongside maternal plasma cfDNA
tains one or no copies of the target gene. When for each pregnancy, using a bioinformatics algor-
compared with real-time PCR, the partitioning ithm developed for low-frequency mutation detec-
increases sensitivity and specificity because of the tion and a strict variant interpretation strategy to
reduced effects of competition of target molecules. A identify de-novo mutations in the plasma data.
positive reaction in a droplet indicates the presence Their assay was shown to be successful in three
of a single molecule, and the proportion of positive pregnancies in which variants identified were
droplets is used to estimate the concentration shown to be identical to those identified in amniotic
of target. fluid from the pregnancy.
The majority of studies using ddPCR have dem-
onstrated detection of paternal or de-novo disease-
causing mutations including CF (CFTR gene) [13], OTHER METHODS
neonatal diabetes (KCNJ11) [14] and achondroplasia Galbiati et al. [19] looked at the application of
&
(FGFR3) [15]. Perlado et al. [16 ] have taken this one COLD-PCR and microarray analysis for NIPD for
step further and have also looked at detection of both b-thalassemia and CF. They tested 87 and 40
maternal alleles by relative mutation dosage prenatal diagnoses, respectively, studying detection
(RMD). One difficulty with validating NIPD assays of the paternal allele only. Their results showed
is the availability of clinical samples for rare disorders complete concordance in all cases with invasive
and individual mutations, which they have prenatal diagnosis. They comment that they would
addressed by carrying out a proof of principle study recommend that NIPD be performed by two inde-
using a SNP strategy to mimic the inheritance pattern pendent methods and this is why they investigated
this approach, and that both methods are inexpen- through amplification of the SRY gene and fluor-
sive and easy to develop and perform in clinical escence in-situ hybridization to determine fetal sex
laboratories. and to demonstrate trisomy 13, 18 or 21. The
To date, very few studies have reported on the authors estimated that the cost of the clinical pro-
development of NIPD for trinucleotide repeat vision of this analysis would be equivalent to cur-
expansion disorders. This is largely due to the tech- rent NIPT techniques; however, they acknowledge
nical issues associated with detecting repeat expan- that it is currently impractical for use as a screening
sions within highly fragmented cfDNA. Direct technique.
detection of paternally inherited normal and inter-
mediate-sized CAG repeats in maternal plasma
has been demonstrated ([20] and references TRANSCERVICAL TROPHOBLASTIC CELLS
therein); however, there is difficulty in detecting Bolnick et al. [26] first published their trophoblast
large Huntington disease repeats in the expansion retrieval and isolation from the cervix (TRIC) in
size range. 2014, in which they demonstrated that they could
Several groups have recently published details of use trophoblast cells accumulating in the endocer-
assays developed for NIPD of b-thalassemia. These vical canal at the beginning of pregnancy to accu-
reports concerned the use of allele-specific real-time rately determine fetal sex from 5-week gestation. In
PCR with and without high-resolution melt analysis contrast with cffDNA, it has been shown that the
to detect the presence or absence of the paternal yield of fetal cells from TRIC is not affected by
mutation in the maternal plasma [21,22]. These gestational age (5–20 weeks), maternal BMI, parity
assays showed a high level of accuracy but again or maternal age [27]. The same group has since used
are only applicable to paternal inheritance and this technique to identify fetal sex in a pregnancy at
require different mutation alleles to be present in risk of CAH [28], determine fetal genotypes using
the mother and father. Despite the limitations, there NGS [29] and to evaluate altered biomarkers to
may be benefit in considering these types of assays predict manifestation of perinatal disease [30].
in suitable scenarios, due to their relative simplicity Although this technique does not appear to have
and affordability. the associated risk of adverse clinical outcomes [28]
that are seen with traditional invasive testing
methods, it could still be considered to be a some-
CIRCULATING FETAL CELLS &
what ‘invasive’ procedure. Pfeifer et al. [31 ] have
Although much of the recent work has focused on developed a technique that extracts trophoblastic
cffDNA as a source of material for NIPT, the use of cells from the external parts of the cervix using the
fetal cells is still an attractive proposition as it offers Papanicolaou test technique and were able to dem-
pure fetal material (not mixed with maternal) and onstrate that these cells could be used as a source of
nonfragmented DNA, which is more amenable to pure fetal DNA, which could be used for NIPD.
whole genome analysis. Previously published
methods for isolating circulating fetal nucleated
red blood cells have included density gradients, ETHICAL AND SOCIAL CONSIDERATIONS
magnetic-activated cell sorting and fluorescence- NIPD using cffDNA in maternal plasma allows safer
activated sorting; however, these techniques have and earlier prenatal testing, which is likely to trans-
drawbacks including purity and recovery rate that form prenatal care for couples with pregnancies at
mean that they are not suitable for routine prenatal risk of SGDs. Therefore, it is important that the
testing. More recently, there have been several pub- ethical and social issues of the implementation of
lications that describe techniques capable of isolat- NIPD into clinical practice are fully considered.
ing fetal cells at yields and purities sufficient for Several studies have been conducted using focus
NIPT. These use a two-step enrichment with a groups and interviews that have included adult
positive enrichment process based on red blood cell patients [32], support group representatives [33]
hyperaggregation, followed by negative enrichment and carriers of autosomal recessive diseases, includ-
using microfluidic technology [24]. The potential ing those who have an affected child (living or
for these cells to be used for subsequent fetal genetic deceased) and those who have used NIPD [34–36].
analysis was demonstrated by determining fetal sex Although opinion is generally positive, with
through expression of SRY genes. the main benefits being seen as the low risk and
&
Kanda et al. [25 ] have used a combination of a early provision of testing, several themes of com-
modified lectin-based erythroblast isolation tech- mon concerns were identified. These included
nique and automated image analysis to isolate accuracy of results, routinization of testing,
pure fetal erythroblasts. Fetal sex was demonstrated informed consent, increased pressure to have
1040-872X Copyright ß 2017 Wolters Kluwer Health, Inc. All rights reserved. www.co-obgyn.com 77
prenatal testing, increase in terminations and clinical practice. Studies are ongoing to investigate
moral issues related to prenatal testing in general the cost and ethical considerations of testing.
(need for guidelines for when testing should be It remains to be seen if the development of more
offered; paternal rights) [32– 36]. powerful technologies such as single cell sequenc-
Much of the NIPT currently available is for ing, a reduction in sequencing costs and evolution
conditions that affect the baby at birth or early in of more robust methods for isolation of single cells
life. However, given the pace at which the technol- will shift the emphasis and scope of NIPD [40]. In
ogy is moving, it seems likely that testing will be addition, as testing becomes possible for a wider
available in the near future for a much wider range range of conditions, many more ethical questions
of conditions, including those with varying pene- are likely to be raised.
trance and/or adult onset. This raises many more
ethical questions, including which conditions it is Acknowledgements
ethically justifiable to test for prenatally and how is None.
this decided and if we are willing to offer an invasive
test for a condition, should we also offer NIPT? Financial support and sponsorship
Deans et al. [37] make the point that if parents are None.
using noninvasive testing for ‘information only’,
are we invading the privacy of that future child’s Conflicts of interest
right to make their own decisions about genetic
There are no conflicts of interest.
testing? Many of these questions are addressed
in a study by Bennett et al. [38] looking at the
views of health professionals regarding the use REFERENCES AND RECOMMENDED
of NIPD for BRCA1/2 mutations in breast and READING
ovarian cancer. Papers of particular interest, published within the annual period of review, have
been highlighted as:
& of special interest
&& of outstanding interest
14. De Franco E, Caswell R, Houghton JA, et al. Analysis of cell-free fetal DNA for 26. Bolnick JM, Kilburn BA, Bajpayee S, et al. Trophoblast retrieval and isolation
noninvasive prenatal diagnosis in a family with neonatal diabetes. Diabet Med from the cervix (TRIC) for noninvasive prenatal screening at 5 to 20 weeks of
J Br Diabet Assoc 2016. [Epub ahead of print] gestation. Fertil Steril 2014; 102:135–142.
15. Orhant L, Anselem O, Fradin M, et al. Droplet digital PCR combined with 27. Fritz R, Kohan-Ghadr HR, Sacher A, et al. Trophoblast retrieval and isolation
minisequencing, a new approach to analyze fetal DNA from maternal blood: from the cervix (TRIC) is unaffected by early gestational age or maternal
application to the noninvasive prenatal diagnosis of achondroplasia. Prenat obesity. Prenat Diagn 2015; 35:1218–1222.
Diagn 2016; 36:397–406. 28. Bolnick AD, Fritz R, Jain C, et al. Trophoblast retrieval and isolation from the
16. Perlado S, Bustamante-Aragonés A, Donas M, et al. Fetal genotyping in cervix for noninvasive, first trimester, fetal gender determination in a carrier of
& maternal blood by digital PCR: towards NIPD of monogenic disorders congenital adrenal hyperplasia. Reprod Sci Thousand Oaks Calif 2016;
independently of parental origin. PLoS One 2016; 11:e0153258. 23:717–722.
A study investigating droplet digital PCR for single gene disorders as a proof of 29. Jain CV, Kadam L, van Dijk M, et al. Fetal genome profiling at 5 weeks of
principle, using single nucleotide polymorphism genotyping. gestation after noninvasive isolation of trophoblast cells from the endocervical
17. Chitty LS, Mason S, Barrett AN, et al. Noninvasive prenatal diagnosis of canal. Sci Transl Med 2016; 8:363re4.
achondroplasia and thanatophoric dysplasia: next-generation sequencing 30. Bolnick JM, Kohan-Ghadr H-R, Fritz R, et al. Altered biomarkers in trophoblast
allows for a safer, more accurate, and comprehensive approach: Noninvasive cells obtained noninvasively prior to clinical manifestation of perinatal disease.
prenatal diagnosis using next-generation sequencing. Prenat Diagn 2015; Sci Rep 2016; 6:32382.
35:656–662. 31. Pfeifer I, Benachi A, Saker A, et al. Cervical trophoblasts for
18. Dan S, Yuan Y, Wang Y, et al. Non-invasive prenatal diagnosis of lethal & noninvasive single-cell genotyping and prenatal diagnosis. Placenta 2016;
& skeletal dysplasia by targeted capture sequencing of maternal plasma. PLoS 37: 56–60.
One 2016; 11:e0159355. This study describes the extraction of trophoblastic cells from the external parts of
A study that demonstrates the feasibility of using a sequencing panel for NIPD of fetal the cervix in women between 8 and 12 weeks gestation and demonstrated that
skeletal dysplasia in which there is no known diagnosis – sequencing 16 genes. they could be used as a source of material for NIPD of cystic fibrosis and spinal
19. Galbiati S, Monguzzi A, Damin F, et al. COLD-PCR and microarray: two muscular atrophy.
independent highly sensitive approaches allowing the identification of fetal 32. Hill M, Twiss P, Verhoef TI, et al. Noninvasive prenatal diagnosis for cystic
paternally inherited mutations in maternal plasma. J Med Genet 2016; fibrosis: detection of paternal mutations, exploration of patient preferences
53:481–487. and cost analysis. Prenat Diagn 2015; 35:950–958.
20. van den Oever JME, Bijlsma EK, Feenstra I, et al. Noninvasive prenatal 33. Hill M, Compton C, Karunaratna M, et al. Client views and attitudes to
diagnosis of Huntington disease: detection of the paternally inherited ex- noninvasive prenatal diagnosis for sickle cell disease, thalassaemia and cystic
panded CAG repeat in maternal plasma. Prenat Diagn 2015; 35:945–949. fibrosis. J Genet Couns 2014; 23:1012–1021.
21. Zafari M, Gill P, Kowsaryan M, et al. High-resolution melting analysis for 34. Lewis C, Hill M, Chitty LS. Noninvasive prenatal diagnosis for single gene
noninvasive prenatal diagnosis of IVS-II-I (G-A) fetal DNA in minor beta- disorders: experience of patients. Clin Genet 2014; 85:336–342.
thalassemia mothers. J Matern-Fetal Neonatal Med 2016; 29:3323–3328. 35. Skirton H, Goldsmith L, Chitty LS. An easy test but a hard decision: ethical
22. Ramezanzadeh M, Salehi M, Farajzadegan Z, et al. Detection of paternally issues concerning noninvasive prenatal testing for autosomal recessive
inherited fetal point mutations for b-thalassemia in maternal plasma using disorders. Eur J Hum Genet 2015; 23:1004–1009.
simple fetal DNA enrichment protocol with or without whole genome ampli- 36. Pisnoli L, O’Connor A, Goldsmith L, et al. Impact of fetal or child loss on
fication: an accuracy assessment. J Matern-Fetal Neonatal Med 2016; parents’ perceptions of noninvasive prenatal diagnosis for autosomal reces-
29:2645–2649. sive conditions. Midwifery 2016; 34:105–110.
23. Lv W, Wei X, Guo R, et al. Noninvasive prenatal testing for Wilson disease by 37. Deans Z, Clarke AJ, Newson AJ. For your interest? The ethical acceptability of
use of circulating single-molecule amplification and resequencing technology using noninvasive prenatal testing to test ‘purely for information’. Bioethics
(cSMART). Clin Chem 2015; 61:172–181. 2015; 29:19–25.
24. Byeon Y, Ki C-S, Han K-H. Isolation of nucleated red blood cells in maternal 38. Bennett J, Chitty L, Lewis C. Noninvasive prenatal diagnosis for BRCA
blood for noninvasive prenatal diagnosis. Biomed Microdevices 2015; 17:118. mutations – a qualitative pilot study of health professionals’ views. J Genet
25. Kanda E, Yura H, Kitagawa M. Practicability of prenatal testing using lectin- Couns 2016; 25:198–207.
& based enrichment of fetal erythroblasts. J Obstet Gynaecol Res 2016; 39. Verhoef TI, Hill M, Drury S, et al. Noninvasive prenatal diagnosis (NIPD) for
42:918–926. single gene disorders: cost analysis of NIPD and invasive testing pathways.
Through a combination of lectin-based erythroblast isolation and automated image Prenat Diagn 2016; 36:636–642.
analysis, this group were able to isolate sufficient fetal erythroblasts for clinical 40. Zhu W, Zhang X-Y, Marjani SL, et al. Next-generation molecular diagnosis:
diagnosis and demonstrated that they could be used to detect fetal aneuplodies by single-cell sequencing from bench to bedside. Cell Mol Life Sci 2016. [Epub
FISH. ahead of print]
1040-872X Copyright ß 2017 Wolters Kluwer Health, Inc. All rights reserved. www.co-obgyn.com 79