Debates on fetal fraction measurement and DNA ‐based noninvasive prenatal screening: time for standardisation?
2016; Wiley; Volume: 123; Issue: S3 Linguagem: Inglês
10.1111/1471-0528.14197
ISSN1471-0528
AutoresTuangsit Wataganara, T-H. Bui, Kwong Wai Choy, TY Leung,
Tópico(s)Parvovirus B19 Infection Studies
ResumoBJOG: An International Journal of Obstetrics & GynaecologyVolume 123, Issue S3 p. 31-35 CommentaryFree Access Debates on fetal fraction measurement and DNA-based noninvasive prenatal screening: time for standardisation? T Wataganara, T Wataganara Division of Maternal Fetal Medicine, Department of Obstetrics and Gynaecology, Faculty of Medicine Siriraj Hospital, Bangkok, ThailandSearch for more papers by this authorT-H Bui, T-H Bui The Karolinska Institute, Centre for Molecular Medicine, Clinical Genetics Unit and Centre for Fetal Medicine, Karolinska University Hospital, Stockholm, SwedenSearch for more papers by this authorKW Choy, KW Choy Department of Obstetrics and Gynaecology, The Chinese University of Hong Kong, Hong Kong, ChinaSearch for more papers by this authorTY Leung, Corresponding Author TY Leung [email protected] Department of Obstetrics and Gynaecology, The Chinese University of Hong Kong, Hong Kong, ChinaCorrespondence: Dr TY Leung, Department of Obstetrics and Gynaecology, The Chinese University of Hong Kong, Prince of Wales Hospital, First Floor of Block E (Special Block), Shatin, NT, Hong Kong, China. Email [email protected]Search for more papers by this author T Wataganara, T Wataganara Division of Maternal Fetal Medicine, Department of Obstetrics and Gynaecology, Faculty of Medicine Siriraj Hospital, Bangkok, ThailandSearch for more papers by this authorT-H Bui, T-H Bui The Karolinska Institute, Centre for Molecular Medicine, Clinical Genetics Unit and Centre for Fetal Medicine, Karolinska University Hospital, Stockholm, SwedenSearch for more papers by this authorKW Choy, KW Choy Department of Obstetrics and Gynaecology, The Chinese University of Hong Kong, Hong Kong, ChinaSearch for more papers by this authorTY Leung, Corresponding Author TY Leung [email protected] Department of Obstetrics and Gynaecology, The Chinese University of Hong Kong, Hong Kong, ChinaCorrespondence: Dr TY Leung, Department of Obstetrics and Gynaecology, The Chinese University of Hong Kong, Prince of Wales Hospital, First Floor of Block E (Special Block), Shatin, NT, Hong Kong, China. Email [email protected]Search for more papers by this author First published: 14 September 2016 https://doi.org/10.1111/1471-0528.14197Citations: 14AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Introduction Next-generation sequencing (NGS) allows for a superior DNA-based prenatal screening for common aneuploidies. The percentage of cell-free fetal DNA (cff-DNA) in maternal plasma, or the fetal fraction (FF), is crucial for the detection of a small increment of fragmented chromosome sequences resulting from fetal aneuploidies. The rate of detection of fetal trisomy 21 (T21) drops from over 99% to <75% at FFs below 8%.1 The American College of Obstetricians and Gynecologists and Society for Maternal–Fetal Medicine recently issued a joint committee opinion in September 2015 not only acknowledging the importance of FF for an accurate result but also recognising the wide variation in FF measurements among DNA laboratories: 'Fetal fraction, the amount of the cell-free DNA in the maternal blood that is of fetal origin, is essential for accurate test results. Some laboratories require a fetal fraction of at least 4% for a reportable result. Other laboratories, however, do not measure or report the fetal fraction'.2 The method for quantifying FF has not been standardised. Some established cff-DNA service laboratories do not include the measurement of FF in their quality control metrics. Their reported accuracy and test failure drawn from a large cohort of commercial samples are apparently comparable with those of laboratories which routinely measure FF.3 This commentary aims to revisit the basic principles of how FF contributes to test reliability and the need to standardise the measurement of FF. Methods for DNA-based prenatal aneuploidy screening and the relationship between FF and the accuracy of the test There are numerous cell-free DNA fragments of both maternal and fetal (placental) origin circulating in the plasma of a pregnant woman. As chromosome 21 (chr21) is the shortest chromosome, its derived sequence accounts for only 1.3% of total plasma DNA fragments. This ratio will be slightly increased if a fetus carries an extra copy of chr21. Assuming that the FF in the specimen is 15%, the ratio should be 1.3% and 1.4% of total DNA fragments for unaffected and affected fetuses, respectively. To reliably distinguish these small percentage differences, 10 million DNA molecules must be counted. A 'shotgun' approach has been used to 'randomly select' DNA molecules from the population of all cell-free DNA polymerase chain reaction (PCR) products for sequencing.4 Then, counting enough DNA molecules in the pool requires massively parallel sequencing (MPS). If the FF is below a critical level or limit of detection (LOD), the difference in the ratio of chr21 sequences can be too small to distinguish.5 MPS is a method for nonselective sequencing and analysis of all the circulating DNA of both maternal and fetal origin in maternal plasma, and it is technically possible to derive the FF in the same MPS assay with no additional cost.6 There are also targeted approaches [chromosome-selective sequence analysis and single nucleotide polymorphism (SNP)-based analysis] which are directed against specific regions on the major chromosome of interest before sequence analysis. With technical modification, some commercial DNA laboratories that use the MPS approach can avoid reporting the incidental detection of other chromosomal abnormalities of unknown clinical relevance. (However, some commercial DNA laboratories intentionally look for and report other chromosomal abnormalities for an extra charge.7) The true benefits of this approach will be hard to prove due to the extremely low prevalence of other chromosome abnormalities. Methods of measuring FF and their accuracy The level of cff-DNA in a maternal plasma specimen can be expressed either as genome equivalent (GE) per millilitre, which reflects the concentration of the total amount of fetal-derived DNA sequences in the plasma volume, or as fetal fraction percentage, which reflects the proportion of fetal-derived DNA sequences in the pool of maternal cell-free DNA. These two surrogates for the level of cff-DNA are increased with gestational age at measurement. Increased maternal weight and haemolysis (which usually occurs when there is a delay in freezing a specimen) decrease the levels. FF in T21 and trisomy 13 (T13) pregnancies may be higher than those in trisomy 18 (T18).8 Currently, there are five different approaches being adopted for the routine measurement and reporting of FF by large commercial DNA laboratories. There is a significant discrepancy in principle: the target domain of the fetal-derived sequence being measured by the assay, specificity, reproducibility and the cost of the FF quantification assay. Table 1 summarises each approach to quantifying FF by comparing their principles, the domain on the fetal-derived sequence that the assay is estimating, and the technical and clinical issues for each assay as well as the supporting literature. A summary of the different approaches to the measurement of FF adopted by major service laboratories is shown in Table 2. At present there are no data on the coefficient of variation or any large-scale comparisons between different measurement methods—therefore none of them can be considered a 'standard' FF measurement. Table 1. Summary of characteristics of five major molecular quantification assays that have been used to estimate fetal fraction (FF) in the test specimen. The assays from these tests recognise different fetal-specific domains. Therefore, the FF estimates obtained from different quantitative approaches cannot be directly compared, and individual lower limits of detection (LODs) can be significantly different from one DNA laboratory to another. Even if all the FF results share the same unit (%), direct comparison cannot be made because they do not reflect the same domain Quantification assay Target of measurement Details References Number of Y-chromosome DNA fragments Y-chromosome-specific DNA markers, such as DYS-1, DYS-14 Only 50% of plasma samples from pregnant women with male fetus will be eligible for analysis Wataganara et al.19 Bisulphite sequencing Differential methylation patterns between the trophoblasts (heavily methylated) and maternal DNA (less methylated) Due to the complexity of the assay, there is a chance of technical failure There is a chance of a false negative from the expected hypomethylated domain on the placenta There is a chance of a false positive from the presence of an additional 152 differentially methylated regions in maternal kidney, liver and endothelium Jensen et al.11 SNP ratio Allelic ratios of SNPs on chromosomes that are never trisomic or monosomic in a viable pregnancy Theoretically, SNP sequencing is the most accurate because foreign SNPs (paternal/fetal derived) are readily apparent in the woman's plasma Some companies may not measure foreign SNPs separately and up front, but instead embed them in the SNP sequencing process to keep the cost down Sparks et al.12, Brar et al.13 Size fractionation Circulating fetal DNA molecules that are shorter than the corresponding maternal DNA Plasma DNA size analysis using paired-end massively parallel sequencing and microchip-based capillary electrophoresis Yu et al.14 Sequence read approach (SeqFF) Circulating fetal DNA molecules that are shorter than the corresponding maternal DNA. This is a spin-off from the size fractionation approach SeqFF is based on the principle of size fractionation, but it obtains the volume of shorter DNA fragments (of presumed fetal origin) within the routine assay of unselected counting (massive parallel sequencing). The number of reads is then aligned within specific autosomal regions before being analysed with a weighting scheme derived from a multivariate model. This novel approach yields a comparable result to the SNP-based approach with Pearson correlation as high as 0.921 Kim et al.6 Table 2. Summary of the fetal fraction (FF) quantification assays routinely used by major commercial DNA service laboratories. Note that one company neither performs a FF assay nor puts FF on their reports. The other four major commercial DNA laboratories which routinely measure and report FF have adopted four different assays, and established their own 'lower FF cut-off' between 2.8 and 4%. (The information in this table was updated on 28 April 2016) Laboratory Report of FF Assay Lower FF cut-off for detection of aneuploidy References Sequenom Yes Bisulphite conversion N/A Ehrich et al.9, Nygren et al.10 Ariosa Yes 576 SNP DANSR assays 4% Spark et al.12, Brar et al.13 Natera Yes Embedded in SNP sequencing 2.8% Personal communication Illumina/Verinata No N/A N/A Personal communication BGI Yes (since April 2016) Size fractionation 3.5% Yu et al.14 DANSR, digital analysis of selected regions; N/A, not available. Inconsistency of results for FF using different quantitative approaches is an important consequence when there are many quantitative assays using different principles and no standardisation. Different FFs have been obtained from aliquots of the same plasma specimen tested at two separate commercial DNA services using different FF quantification assays. There are many reports of markedly different results from FF quantification assays. A recent example was from a small experiment conducted by a group in Belgium to assess the quality of two DNA laboratories in Belgium and another two outside the country.15 Each laboratory received two blood specimens taken from two women with twins at 12 and 17 weeks of gestation, respectively. One fetus from the 17-week twin pair had T21 (proven by genetic amniocentesis). None of the service laboratories that tested the specimen picked up the T21, and their reported FFs varied. It is also noteworthy from this study that even the two service laboratories that reported FFs as high as 10% and 16.4%, respectively, still could not detect fetal T21. Counter-intuitively, it is a false reassurance that a high FF guarantees an accurate result. Pros and cons of checking FF routinely Measurement of FF before testing can ensure that a specimen contains adequate fetal DNA for meaningful analysis so as to minimise false negatives The laboratories analysing cff-DNA have been individually regulating their own threshold, and it is impossible to standardise them, as they use different methods. It has been estimated that 2% of the specimens obtained from pregnant women between 10 and 21 weeks of gestation may contain a FF below the critical level of 4%.5 That figure was deduced from 22 384 commercial specimens from a cff-DNA laboratory which routinely measured FF with a SNP approach.5 False negatives cannot be completely avoided, even with routine measurement of FF A report from another cff-DNA service laboratory that routinely measures FF using the SNP approach showed two false-negative (FN) results from 17 885 screen-negative tests with complete follow-up data.16 Of note, the complete follow-up rate in that paper was only 58.24% (17 885 of 30 705 eligible samples for analysis), and the FF of these two FN cases was not specified in the report. This test performance was not remarkably different from the results of a recent large cohort reported from another service laboratory that does not routinely measure FF.3 The report from this latter laboratory showed nine FN results (six cases of T21 and three cases of T18) from 112 669 tests. The sensitivity, positive predictive value and negative predictive value were as high as 99.17, 92.19 and 99.99%, respectively. Commercial DNA laboratories claim that it is more important for the test performance to carry out deeper DNA sequencing and a higher number of reads to compensate for specimens that have a low fetal contribution. However, the company's claim and the published numbers have to be interpreted with caution. The reported FN figures might be underestimated as they relied on voluntary self-report. Spontaneous miscarriage is more common in cases of fetal aneuploidy, and a karyotype study may not always be available. The service laboratory in the report that did not routinely measure FF was located in mainland China, where the incidence of maternal obesity is much lower than in the United States. Hence the probability of missing T21 due to obesity-related low FF is lower. Subsequently measurement of FF on the leftover samples of eight of the nine FN cases in the latter cohort revealed levels of fetal-derived sequences from 5.18 to 13.39%; this is above quality control cut-off in most cff-DNA laboratories.3 The FF measurement assay was not specified. Evidently, FN can occur even in specimens with high FF. Possible explanations for this include placental mosaicism and vanishing twins, which are inherent and cannot be corrected regardless of assay modification. How can FN and litigation be minimised if FF is not routinely measured? In theory, minimisation of low-FF-related FN results can be achieved by deeper sequencing. This strategy is used by some cff-DNA service laboratories that do not routinely measure FF. Clinicians should counsel women about the limitations of noninvasive prenatal screening (NIPS) with regard to low FF. They should check with their patients whether they have any risk factors for low FF such as early gestation or maternal obesity, and avoid offering them NIPS that does not include checking of the FF. Newly developed measurement assays for FF are increasingly accurate and not costly.6 From the point of view of public health, a public screening test must be cost-effective in that it must take account of the balance between a higher detection rate and additional laboratory costs including administrative costs for recall. If offered as a private service, clients should be fully informed about different testing methods and their cost before making their decision. To avoid mistakenly sequencing plasma obtained from a nonpregnant woman, which is a waste of resources as well as a potential matter for litigation, ultrasound examination is bundled with the test in many places. Ultrasound examination not only can confirm the state of pregnancy but also can differentiate between a singleton and twins. This information is crucial because a different algorithm will be applied for twin pregnancies. The reliability of DNA-based prenatal aneuploidy screening in twins has not been adequately established, but a combination of Y-chromosome sequencing and polymorphisms may be useful.17 What if FF is measured but it is repeatedly low? A low FF causes clinical concern. Approximately 75% (15/20) of aneuploid samples with 'no call' results from a SNP-based approach had either low FF or insufficiently clear data. Of these 20 aneuploid samples with a 'no call' result, half had a FF of <3.4% (below the 1.5th percentile).18 A critically low FF level can increase the chance of fetal aneuploidy up to six-fold.18 A 'no call' result should be counselled appropriately, and invasive tests should be offered according to prior risks and ultrasound findings. Conclusions FF should represent a unique marker of fetal presence. Test reliabilities between those laboratories that do and do not routinely measure FF are not significantly different, possibly due to lack of assay standardisation. Every laboratory should have a scientific approach to set its LOD, which is subject to constant monitoring and adjustment every time the sequencing technique, the number of reads and the algorithms are updated. Deeper sequencing and an increased number of reads should lower FF LOD, improve test performance, and reduce specimen rejection and the failure rate. Determination of the best techniques for measuring FF and the best sequencing protocols by a head-to-head comparison may not be possible due to a lack of corporate funding and legal protection of proprietary assays. If a 'gold standard' FF measurement cannot be agreed upon in the near future, could the FF LOD perhaps be presented in the form of a percentile? Regardless of FF level, fetal, placental and maternal mosaicism remains an Achilles heel for test performance across all platforms. Disclosure of interests Full disclosure of interests available to view online as supporting information. Contribution to authorship TW, THB, KWC and TYL conceived the idea, wrote and approved the manuscript. Details of ethics approval No ethical approval was necessary for this article as there were no human subjects. Funding None. Acknowledgements This commentary is based on the debate session hosted at the 11th Asia Pacific Conference in Maternal Fetal Medicine on 28 November 2015 in Taipei, Taiwan. References 1Canick JA, Palomaki GE, Kloza EM, Lambert-Messerlian GM, Haddow JE. The impact of maternal plasma DNA fetal fraction on next generation sequencing tests for common fetal aneuploidies. Prenat Diagn 2013; 33: 667– 74. 2 Medicine ACoOaGCoGSfM-F. Committee opinion no. 640: cell-free DNA screening for fetal aneuploidy. Obstet Gynecol 2015; 126: e31– 7. 3Zhang H, Gao Y, Jiang F, Fu M, Yuan Y, Guo Y, et al. Non-invasive prenatal testing for trisomies 21, 18 and 13: clinical experience from 146,958 pregnancies. Ultrasound Obstet Gynecol 2015; 45: 530– 8. 4Fan 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 USA 2008; 105: 16266– 71. 5Wang E, Batey A, Struble C, Musci T, Song K, Oliphant A. Gestational age and maternal weight effects on fetal cell-free DNA in maternal plasma. Prenat Diagn 2013; 33: 662– 6. 6Kim SK, Hannum G, Geis J, Tynan J, Hogg G, Zhao C, et al. Determination of fetal DNA fraction from the plasma of pregnant women using sequence read counts. Prenat Diagn 2015; 35: 810– 15. 7Lefkowitz RB, Tynan JA, Liu T, Wu Y, Mazloom AR, Almasri E, et al. Clinical validation of a noninvasive prenatal test for genomewide detection of fetal copy number variants. Am J Obstet Gynecol 2016; 215: 227.e1–16. 8Wataganara T, Bianchi DW. Fetal cell-free nucleic acids in the maternal circulation: new clinical applications. Ann N Y Acad Sci 2004; 1022: 90– 9. 9Ehrich M, Deciu C, Zwiefelhofer T, Tynan JA, Cagasan L, Tim R, 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: 205.e1–11. 10Nygren AO, Dean J, Jensen TJ, Kruse S, Kwong W, van den Boom D, et al. Quantification of fetal DNA by use of methylation-based DNA discrimination. Clin Chem 2010; 56: 1627– 35. 11Jensen TJ, Kim SK, Zhu Z, Chin C, Gebhard C, Lu T, et al. Whole genome bisulfite sequencing of cell-free DNA and its cellular contributors uncovers placenta hypomethylated domains. Genome Biol 2015; 16: 78. 12Sparks 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: 319.e1–9. 13Brar H, Wang E, Struble C, Musci TJ, Norton ME. The fetal fraction of cell-free DNA in maternal plasma is not affected by a priori risk of fetal trisomy. J Matern Fetal Neonatal Med 2013; 26: 143– 5. 14Yu SC, Chan KC, Zheng YW, Jiang P, Liao GJ, Sun H, et al. Size-based molecular diagnostics using plasma DNA for noninvasive prenatal testing. Proc Natl Acad Sci USA 2014; 111: 8583– 8. 15Bevilacqua E, Guizani M, Cos Sanchez T, Jani JC. Concerns with performance of screening for aneuploidy by cell-free DNA analysis of maternal blood in twin pregnancy. Ultrasound Obstet Gynecol 2016; 47: 124– 5. 16Dar P, Curnow KJ, Gross SJ, Hall MP, Stosic M, Demko Z, et al. Clinical experience and follow-up with large scale single-nucleotide polymorphism-based noninvasive prenatal aneuploidy testing. Am J Obstet Gynecol 2014; 211: 527.e1–17. 17Struble CA, Syngelaki A, Oliphant A, Song K, Nicolaides KH. Fetal fraction estimate in twin pregnancies using directed cell-free DNA analysis. Fetal Diagn Ther 2014; 35: 199– 203. 18Pergament E, Cuckle H, Zimmermann B, Banjevic M, Sigurjonsson S, Ryan A, et al. Single-nucleotide polymorphism-based noninvasive prenatal screening in a high-risk and low-risk cohort. Obstet Gynecol 2014; 124: 210– 18. 19Wataganara T, Chen AY, LeShane ES, Sullivan LM, Borgatta L, Bianchi DW, et al. Cell-free fetal DNA levels in maternal plasma after elective first-trimester termination of pregnancy. Fertil Steril 2004; 81: 638– 44. Citing Literature Volume123, IssueS3Special Issue: Women's Health in ChinaSeptember 2016Pages 31-35 This article also appears in:2016 Supplements ReferencesRelatedInformation
Referência(s)