New Algorithms in Prenatal Diagnosis
The approach in prenatal diagnosis has been revolutionized by advances in prenatal molecular diagnostics. New algorithms in prenatal diagnosis are evolving and becoming increasingly complicated (Figure 1). The goal is to maximize the prenatal information for pregnant women and the families to make choices for the next generations.
Down syndrome screening has been the focus in prenatal diagnosis for a long time. However, many other chromosomal and structural foetal abnormalities were diagnosed incidentally during the screening programme for Down syndrome, coupled with the 18–22 weeks foetal anomaly ultrasound. Combined screening at 11–13 weeks by maternal age, foetal nuchal translucency (NT) thickness by ultrasound, maternal serum pregnancy-associated plasma protein A (PAPP-A) and free beta-human chorionic gonadotropin (β-hCG) can identify 90% of foetuses with trisomy 21 (Down syndrome) and other common aneuploidies (trisomies 18 and 13) at a false-positive rate (FPR) of 5%.1
The performance is highly reproducible worldwide and has also been demonstrated in the Hospital Authority universal Down syndrome screening programme in Hong Kong from 2010.2 Additional ultrasound and maternal serum markers with different contingent screening tests have been studied to further improve the detection rates and to reduce the FPR.
An important breakthrough in prenatal screening using maternal plasma cell-free foetal DNA as a noninvasive prenatal testing (NIPT) for foetal chromosomal abnormalities was discovered by Professor Dennis Lo from Hong Kong.3 The detection rate for Down syndrome using NIPT is >99% with an FPR of as low as 0.1%. NIPT can be performed using maternal blood sample from 10 weeks’ gestation onwards. NIPT is currently available as a secondary screening tool for pregnancies with positive conventional Down screening as well as for primary screening for Down syndrome.4-6
For prenatal diagnosis in the 21st century, traditional karyotyping is no longer adequate. Advances in prenatal molecular diagnostics including polymerase chain reaction (PCR) as rapid aneuploidy test, and chromosomal microarray (CMA) as molecular karyotyping including the detection of microdeletions and microduplications, are going to replace traditional karyotyping sooner or later.7 There are also new modalities being developed such as whole exome sequencing, WES (sequencing all the protein-coding genes in the genome) or whole genome sequencing, WGS (sequencing the entire genome). WES or WGS will not be discussed in this review article.
Despite the rapid ongoing development in prenatal molecular diagnostics, ultrasound maintains a pivotal role in the new algorithms (Figure 1), being the link between the various tests inside the algorithms. In modern prenatal diagnosis, continued utility of ultrasound (with corresponding counselling) could offer career sustainability for foetal medicine specialists.
This review article is focused on the new algorithms in prenatal diagnosis (Figure 1) from a clinical service/clinical pathway point of view. The current situation in Hong Kong will be referred to from time to time for illustration. It is beyond the author’s capacity to go in-depth into the laboratory aspects.
ADVANCED MATERNAL AGE IS NO LONGER AN ENTRY CRITERIA INTO THE ALGORITHMS OF PRENATAL DIAGNOSIS1
All pregnant women, irrespective of their age, should have equal opportunities for prenatal screening and diagnosis. The clinical service can be provided in public or private sectors, depending on resource allocation.
DATING ULTRASOUND IN THE FIRST TRIMESTER
An ultrasound measurement of the foetal crown-rump length (CRL) at 8–12 weeks would be most useful in dating the pregnancy. The first trimester ultrasound also confirms the foetal viability, intrauterine location, and number of foetuses. An accurate dating is essential to guide the conventional 11–13 weeks combined Down screening test as well as the NIPT for which the maternal blood sample should not be taken before 10 weeks’ gestation.8
CONVENTIONAL DOWN SYNDROME SCREENING PROGRAMME
In Hong Kong, a universal Down syndrome screening programme has been started in all the eight Hospital Authority Obstetric Units since 2010.2 All pregnant women were offered either first-trimester (11–13 weeks) combined test (NT, PAPP-A and free beta-hCG) or second-trimester (16–19 weeks) biochemical test (from double test including alpha-fetoprotein and total hCG, to quadruple test adding oestriol and inhibin-A).
The universal screening programme was effective, and achieved the expected detection rates (90% for first trimester test and 70–80% for second trimester test depending on the number of serum markers) and low FPR of 5%. Those with a trisomy 21 term risk of ≥1:250 were offered a diagnostic invasive test. Risks for trisomies 18 and 13 were also calcula-ted. This has formed a reasonable safety net for our local pregnant women. However, this government-funded conventional Down syndrome screening programme is considered inadequate with the availability of NIPT in the university and private sectors.
Why is the programme not adequate?
1. The 5% FPR – that means 5% of screened pregnant women will have a high-risk result. Traditionally, this high-risk group will be offered an invasive test, either chorionic villus sampling (CVS) or amniocentesis for PCR and/or karyotyping to confirm or exclude the prenatal diagnosis of Down syndrome. However, majority of these pregnant women do not carry a foetus with Down syndrome, with the odds of being affected given a positive result (OAPR) around 1 in 25 to 1 in 40. Thus, majority of them is subjected to unnecessary anxiety and the procedure-related risk of miscarriage due to the invasive test.
2. The procedure-related miscarriage risk following CVS (1% traditionally) and amniocentesis (0.5% traditionally) are substantial, considering that majority of the invasive tests are performed to confirm the absence of Down syndrome in the foetus. Recent studies showed that the procedure-related miscarriage risk of both CVS and amniocentesis are indeed much lower (0.1–0.2%),9 but many pregnant women would still prefer a noninvasive test unless they are at a very high risk of chromosomal abnormalities.
3. The 90% prenatal detection rate of Down syndrome suggests that 10% of foetuses with Down syndrome would be missed. Note that the NIPT can achieve a much higher detection rate of >99% with an FPR of as low as 0.1% in pregnant women at high or low risk of having a foetus with Down syndrome.
NIPT – MATERNAL PLASMA CELL-FREE FOETAL DNA1,8
It first started when Professor Lo reported the presence of Y-chromosome DNA in the blood of pregnant women carrying male foetuses, but not in pregnant women carrying female foetuses.3 At present, we understand that these cell-free foetal DNA molecules originate from the placenta. In maternal plasma, the foetal DNA circulates among a background of DNA mainly from the maternal blood cells. Foetal DNA accounts for about 10–15% of the total maternal plasma DNA. The cell-free foetal DNA in the maternal plasma forms the basis for the DNA-based NIPT. The foetal DNA is cleared from the maternal plasma in a matter of hours after delivery. A Down syndrome foetus would release an extra amount of chromosome 21 DNA into the maternal plasma when compared to non-affected foetuses. A sophisticated laboratory method called Massively Parallel Sequencing (MPS) has been used for DNA analysis in maternal plasma, rendering NIPT practically feasible and eventually becoming a real clinical service.4-6
There are now three laboratory approaches for NIPT:
1. “Shotgun” MPS followed by counting of DNA sequences;
2. “Targeted” MPS with counting of specific DNA sequences; and
3. A method based on the analysis of single nucleotide polymorphisms (SNPs).
NIPT end users (including foetal medicine specialists, obstetricians, midwives, genetic counsellors, and others) should understand which NIPT laboratory method to use given the difference in the range of findings.
In 2011, NIPT has been highly accepted by end users and pregnant women following its introduction for clinical use in Hong Kong.
Secondary screening for foetal Down syndrome refers to NIPT for the subgroup of pregnant women with high-risk conventional Down syndrome screening test (first or second trimester, FPR 5%). As the FPR of NIPT is only 0.1% with an OAPR of 1:1.2, a great majority of the (false-positive) high-risk cases classified by conventional Down syndrome screening test could become low risk again, thus avoiding an unnecessary invasive test. At present, NIPT is a self-financed item offered in private and university sectors in Hong Kong. Nonetheless, >50% of women with high-risk conventional Down syndrome screening will still choose NIPT for secondary screening.10
However, using NIPT as secondary screening alone would not improve the overall detection rate of foetal Down syndrome, unless the FPR of the conventional Down screening programme is further increased to 10 or 20% (ie, performing NIPT for the corresponding 10–20% of the screened positive pregnant women). For practical purposes, NIPT can be offered to the group with intermediate risk (eg, 1/250–1/1200).
Studies have shown that the performance of NIPT for primary screening is as good as the secondary screening. Today, more pregnant women are using NIPT as a primary screening tool for foetal Down syndrome. However, most of them are not at high risk of having foetus with Down syndrome.
Using the same test principles, NIPT could also be applied for trisomies 18 and 13, foetal gender, and sex chromosomal abnormalities (eg, monosomy X), and other rare trisomies (eg, 9, 16, and 22) as well as for subchromosomal microdeletions and microduplications such as DiGeorge syndrome (22q11.2 deletion) and Cri-du-chat syndrome (5p-).
CAN NIPT AS PRIMARY SCREENING REPLACE THE CONVENTIONAL 1ST TRIMESTER COMBINED DOWN SYNDROME SCREENING TEST?1,8,11
NIPT can be used as a primary screening in terms of detection rate, FPR and OAPR for prenatal screening and diagnosis of foetal Down syndrome. However, NIPT should not replace screening by ultrasound at 11–13 weeks. Foetuses with NT ≥3.5mm (>99th percentile) have a higher risk (5%) of other chromosomal abnormalities and copy number variants (CNV) apart from the common aneuploidies.12-13 This subgroup of pregnant women should be offered a direct invasive test to maximize the genetic information obtained. In lieu of NIPT, self-financed tests in Hong Kong such as the PCR and CMA are probable alternatives. Thick NT is also associated with major cardiac defects and a range of foetal structural and syndromal abnormalities.
Furthermore, early ultrasound could also pick up structural foetal abnormalities such as anencephaly, holoprosencephaly, cystic hygroma, foetal hydrops, gastroschisis, megacystis, body stalk anomaly, major spine abnormalities, and missing limbs. Therefore, the 11–13 weeks’ scan for NT and foetal abnormalities should not be omitted even if NIPT was performed.
NIPT AS AN ADDITIONAL REASSURANCE FOR ULTRASOUND SOFT MARKERS
Ultrasound soft markers eg, hypoplastic nasal bone, short long bones, choroid plexus cysts, intracardiac echogenic foci, tricuspid regurgitation, echogenic bowels, pyelectasis, single umbilical artery among others, are not uncommonly found in foetal anomaly ultrasound at 18–22 weeks. In the past, these ultrasound soft markers could result in an invasive test to exclude common aneuploidies. With the availability of universal Down syndrome screening, different odds ratios of individual soft markers are used to adjust the risk of aneuploidies. NIPT could be considered (if not performed yet) to relieve the anxiety (for both pregnant women and their family as well as for obstetricians) associated with the incidental findings of ultrasound soft markers.
SPECIAL CONSIDERATIONS FOR NIPT1,8,11
Regardless of their advantages, new molecular diagnostics have their own limitations warranting special considerations by end users. This knowledge is vital for pre-test and post-test counselling of pregnant women.
1. Foetal DNA fraction (>4%) as a quality control parameter. This is the reason why NIPT should not be performed <10 weeks’ gestation. After this gestation, the foetal DNA fraction is relatively constant throughout gestation.
2. No-result (failure to provide a result) rates can range from 0–12% for various reasons. There is controversial evidence whether there is a higher incidence of common aneuploidies in this no-result subgroup and whether an invasive test should be offered. However, if this approach is adopted in the algorithm, the advantage of NIPT in reducing the overall incidence of invasive tests with corresponding procedure-related miscarriage risks would be lost. The alternative is to switch back to the conventional Down syndrome screening – first trimester combined test or the second trimester screening test with lower detection rate.
3. Confined placental mosaicism (CPM) – mosaic chromosomal abnormalities found in placenta but not in the foetus. Note that NIPT assesses the circulating DNA of placental origin in maternal plasma, which could give rise to false-positive NIPT results.
4. False-positive NIPT results can also be related to abnormal maternal plasma DNA profiles such as mosaic sex chromosome aneuploidies, autoimmune diseases, and cancer.
5. Multiple pregnancies. NIPT is feasible in twin pregnancies. It is a good practice to specify and confirm with the laboratory first before maternal blood sample is sent. Note that circulating DNA for a vanishing twin can affect the interpretation of the NIPT result for the surviving twin.
6. Any abnormal NIPT result (even for trisomy 21 with 99% detection rate and 0.1% FPR, the positive predictive value (PPV) is only 50%) should be confirmed by an invasive test (CVS or amniocentesis) before any decision for termination of pregnancy is made.
MOLECULAR KARYOTYPING BY CMA7,14
Traditional karyotyping has been the gold standard for detecting chromosomal abnormalities in prenatal diagnosis since the 1960s. Genome-wide numerical and structural chromosomal abnormalities are detected at a resolution of 5–10 Mb. This low resolution has now become the major drawback for traditional karyotyping when the resolution of CMA in detecting genome-wide genomic imbalances or CNV can be as high as 50 kb.
There are different laboratory platforms for CMA such as oligonucleotide array comparative genomic hybridisation (aCGH) and SNP arrays. Similar to NIPT, CMA end users should understand which CMA laboratory platform will be used, particularly the corresponding resolution, given the difference in the range of findings.
At present, CMA is a self-financed item offered at the university and private sectors in Hong Kong.
INTERPRETATION OF PRENATAL CMA FINDINGS7,14
Unlike the application in postnatal (eg, paediatrics) setting, interpretation of CNV in prenatal diagnosis can be challenging because of limited phenotype information from ultrasound examination. CNV are usually categorized into three types in the prenatal setting according to various publicly available databases (ie, DECIPHER, ISCA, DGV, CHOP, etc) and published in-house datasets. The detailed information for CNV will be provided by the corresponding laboratory. Direct discussion on a case-by-case basis with the laboratory colleague together with a clinical geneticist would be most useful.
1. Normal molecular karyotype – no CNV identified;
2. Clinically significant CNV – a chromosome imbalance harbouring genes and/or overlapping with a known syndrome (eg, OMIM database). Parental study is necessary to further investigate whether the CNV is familial or de novo;
3. CNV of unknown clinical significance – a chromosome imbalance which has not been reported in public or in-house databases or literature. Parental study is also advised. This subgroup has created difficult scenarios in counselling and even medico-legal consequences. Note that the higher the resolution of the CMA performed, the higher incidence of CNV (both 2 and 3) will be identified.
The above information is important to be incorporated into the pre-test counselling and consent before the invasive test and CMA are performed as well as for post-test counselling.
COULD CMA REPLACE TRADITIONAL KARYOTYPING?7,14
There is a general tendency in principle to maximize the genetic information obtained once an invasive test (CVS or amniocentesis) is performed with an indication for prenatal diagnosis. The choice of pregnant women and their partners should be respected, although most of them would prefer access to as much information as possible.
CMA can replace traditional karyotyping given the characteristics of the various prenatal diagnostic tests available (Table 1). In particular, PCR (rapid aneuploidy test) could be performed as a first step to confirm or exclude common aneuploidies, which might account for 80% of chromosomal abnormalities in general, leaving behind the 20% cases for CMA. Compared to traditional karyotyping, CMA could identify an additional 2% clinically significant CNV when the indications for invasive prenatal test are positive screening or advanced maternal age, and as high as an additional 6% when ultrasound foetal abnormalities are present, which once again shows the importance of ultrasound.15
This review article gives an overview of the new algorithms in prenatal diagnosis that have surfaced due to the incorporation of the new prenatal molecular diagnostics – NIPT and CMA. Although expensive, the cost of these new molecular tests are likely to reduce with time.
To make the new algorithms cost-effective, it has to be a “give and take” exercise. The first to ‘give’ is the number of CVS and amniocentesis (significant reduction is already happening) as well as the corresponding experience in performing these invasive tests. Secondly, traditional karyotyping (expensive) is going to be ‘given up’ and replaced by PCR (rapid aneuploidy test) and CMA.
It would be more interesting if the new algorithms in prenatal diagnosis can be adopted as a territory-wide programme with extensive data collected prospectively for each pregnant woman. The prenatal diagnosis algorithms used in the programme will be assessed, monitored, and audited for continuous improvement and development of the new algorithms.16
About the author
Dr Wing Cheong Leung is a Consultant Obstetrician practising in the Department of Obstetrics and Gynaecology, Kwong Wah Hospital, Kowloon, Hong Kong SAR.