plasma nucleic acid analysis by massively parallel...

INVITED REVIEWJournal of PathologyJ Pathol (2011)Published online in Wiley Online Library(wileyonlinelibrary.com) DOI: 10.1002/path.2960

Plasma nucleic acid analysis by massively parallel sequencing:pathological insights and diagnostic implicationsYM Dennis Lo1,2* and Rossa WK Chiu1,2

1 Li Ka Shing Institute of Health Sciences and2 Department of Chemical Pathology, Chinese University of Hong Kong, Hong Kong SAR, China

*Correspondence to: YM Dennis Lo, Department of Chemical Pathology, Prince of Wales Hospital, 30-32 Ngan Shing Street, Shatin, NewTerritories, Hong Kong SAR, China. e-mail: [email protected]

AbstractOver the past 15 years there has been increasing interest in the biology and diagnostic applications of circulatingDNA in the plasma of human subjects. In particular, DNA from a fetus, a tumour, a transplanted organ andinjured tissues has been found in the plasma of pregnant women, cancer patients, transplant recipients andpatients suffering from multiple pathologies, respectively. The advent of massively parallel sequencing has givenus a quantitative and powerful tool for studying circulating DNA on a genome-wide level. Using this approach,fetal chromosomal aneuploidies can be robustly detected using maternal plasma. Furthermore, a genome-widegenetic map of a fetus can also be constructed using this approach. This method has also allowed one to identifytumour-associated chromosomal translocations, which can then be detected in plasma. The direct application ofmassively parallel sequencing to the serum of cancer patients has also allowed quantitative aberrations that areassociated with malignancy to be detected in serum. The use of massively parallel sequencing on the plasma oftransplantation recipients has opened up an approach for detecting rejection. The application of circulating DNAsequencing has also opened up a new method for elucidating the quantitative aberration of circulating DNA inmany pathological conditions. Such developments would provide new modalities for molecular diagnostics andwould improve our understanding of the biology of circulating nucleic acids.Copyright ! 2011 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Keywords: circulating DNA; next-generation sequencing; diagnosis

Received 13 May 2011; Revised 25 June 2011; Accepted 27 June 2011

Conflict of interest statement. The authors hold patents and have filed patent applications on the diagnostic applications of plasma nucleic acids.Part of this portfolio has been licensed to Sequenom. YMDL is a consultant to, hold equities in and has research support from Sequenom.

Introduction

The presence of cell-free nucleic acids in the plasmaof human subjects was reported as early as 1948 [1].There had only been occasional interest in this areaover the first few decades of its history. However, withthe emergence of powerful molecular biological tech-niques, the diagnostic applications of such circulatingnucleic acids have blossomed over the past 15 years[2]. Thus far, four groups of diagnostic applicationshave received the most interest in the field: first, thedetection of fetal nucleic acids in maternal plasma fornon-invasive prenatal diagnosis [3]; second, the anal-ysis of tumour-derived nucleic acids in plasma for thedetection and monitoring of cancer [4,5]; third, thedetection of donor-derived nucleic acids in the plasmaof transplantation recipients for rejection monitoring[6]; and fourth, the measurement of circulating nucleicacid concentrations for prognostication of tissue dam-age [7]. Most of the studies in this field have been per-formed using polymerase chain reaction (PCR)-basedtechniques, which typically analyse one or a small

number of genomic loci. However, the recent adventof massively parallel sequencing [8] has opened upthe possibility of analysing plasma nucleic acids at agenome-wide level.

Massively parallel DNA sequencing is a recentlyemerged technology in which single DNA moleculesare spatially segregated and then either clonallyamplified and sequenced (for the so-called ‘second-generation technology’) or directly sequenced withoutamplification (for the so-called ‘third-generation tech-nology’). Depending on the platform, typically hun-dreds of thousands to millions or even billions ofsequence reads can be obtained in a single run of thesequencer. Here we review the implications of mas-sively parallel sequencing for developing novel diag-nostic applications using circulating nucleic acids andfor understanding its biology.

Fetal nucleic acids in maternal plasma

The presence of cell-free fetal DNA in maternal plasmawas first reported in 1997 [3]. This discovery has

Copyright ! 2011 Pathological Society of Great Britain and Ireland. J Pathol (2011)Published by John Wiley & Sons, Ltd. www.pathsoc.org.uk www.thejournalofpathology.com

YMD Lo and RWK Chiu

Figure 1. Schematic diagram showing the use of massively parallel sequencing of maternal plasma DNA for non-invasive prenataldiagnosis. Arrows marked 1 and 2 indicate the applications for the detection of fetal chromosomal aneuploidies and single gene diseases,respectively.

opened up new possibilities for non-invasive prenataldiagnosis. Early clinical applications have focused onthe detection of fetal-specific targets that are paternallyinherited and that are not present in the pregnantmother’s genome. Examples of such targets includemarkers present on the Y chromosome for detectinga male fetus [9,10] and the RHD gene for detecting aRhD-positive fetus carried by a RhD-negative pregnantmother [11,12]. However, the use of maternal plasmaDNA for detecting the presence of fetal chromosomalaneuploidies, eg trisomy 21, is much more technicallychallenging. For illustration, if fetal DNA represents10% of the DNA that is present in a particular maternalplasma sample, then for one to detect the presence of atrisomy 21 fetus, the analytical method has to be ableto discriminate a 5% difference [13]. In 2007, Lo et al[13] outlined the principle whereby such a precisioncould be reached by approaches that allow single DNAmolecules to be counted. This group demonstratedthe feasibility of the concept using digital PCR [14].However, the performance of tens of thousands or>100 000 digital PCRs for each case is extremelylabour-intensive.

With the development of massively parallel sequenc-ing, a much more efficient way of analysing millions ofDNA molecules has become available [8] (Figure 1). In2008, two groups independently showed that the mas-sively parallel sequencing of maternal plasma DNAwould allow a fetus with trisomy 21 to be detectedwith high accuracy [15,16]. These results have recentlybeen validated in two large-scale studies involving 753samples [17] and 449 samples [18]. Both Chiu et aland Ehrich et al applied massively parallel sequencingto the plasma of pregnant women who had previouslybeen classified as high risk by conventional screeningmodalities. Chiu et al [17] reported that fetal trisomy21 could be detected with a sensitivity of 100% anda specificity of 98%. The data by Ehrich et al [18]were consistent with those of Chiu et al and showed

a sensitivity of 100% and a specificity of 99.7%. Theprecision of massively parallel sequencing for the mea-surement of the dosage of chromosomes 13 and 18,on the other hand, is lower [16,19]. Special bioinfor-matics procedures have been demonstrated to improvethis precision [20,21].With further validation in largecohorts, it is likely that massively parallel sequencingof maternal plasma DNA would be used clinically inthe near future for the screening of the common fetaltrisomies.

Recent data have indicated that the massively paral-lel sequencing of maternal plasma DNA has diagnosticapplications outside of trisomy detection. Lo et al [22]have shown that, by combining the maternal plasmaDNA sequencing data with genotype information ofthe father and haplotype information of the mother, onecould deduce a genome-wide genetic map of an unbornfetus (Figure 1). These investigators have demonstratedthe diagnostic utility of this approach for the pre-natal diagnosis of !-thalassaemia. Indeed, in theory,virtually all monogenic diseases could be investigatedprenatally using this approach. However, before thetechnology can be widely used, a number of issueswould need to be explored. First, as reported by Loet al [22], the technology for such whole-genome fetalgenetic scanning is expensive. Second, the requirementof maternal haplotype information would potentiallyreduce the ease with which this technology could beused, as such information is typically generated throughpedigree analysis. Third, the social and ethical implica-tions of prenatal fetal whole-genome scanning wouldneed to be discussed by all stakeholders.

A number of the above concerns are already beingaddressed. With regard to the cost issue, it is unlikelythat clinical needs would routinely require the entirefetal genome to be scanned prenatally. A probablescenario would be for the technology to be used forselected genetic diseases prevalent in a particular pop-ulation. In this regard, Liao et al [23] have recently


Plasma nucleic acid analysis by massively parallel sequencing

shown that, through the prior capturing of plasma DNAmolecules from specific genomic regions, one couldfocus the sequencing power to these regions. Sequencecoverage of the targeted regions has been found tobe increased by over 200 times. This strategy has thepotential to greatly reduce the cost of implementingthis technology clinically. Furthermore, with the rapidreduction in sequencing costs, it is likely that costswould no longer be of major concern in the years tocome. With regard to the requirement for maternal hap-lotype information, there are recent reports describingnew methods for genome-wide haplotyping [24,25].Such approaches would allow haplotype informationto be more readily generated. Concerning the socialand ethical implications, discussion on this topic hasalready started and is expected to intensify in the future[26,27].

Apart from the above-mentioned diagnostic impli-cations, the advent of massively parallel sequencinghas given us an unprecedentedly high-resolution pic-ture of plasma DNA. With the use of paired-endsequencing, in which sequences from both ends of aDNA molecule are elucidated, one could deduce thelength of the molecule by comparison with the refer-ence human genome sequence. Through the paired-endsequencing of billions of molecules, Lo et al [22] haveshown that circulating DNA molecules in the plasmaof a pregnant woman have a size distribution that isclosely related to the structure of a nucleosome. Suchanalysis has also provided a molecular explanationfor the longstanding observation that circulating fetalDNA molecules are generally shorter than the mater-nal DNA molecules [28]. Thus, through paired-endsequencing, it has been found that circulating mater-nal DNA molecules have a predominant size peak at166 bp [22], corresponding to the size of the DNAwrapping round the core histones (!143 bp) plus alinker fragment (!20 bp) [29]. This peak is followedby a series of peaks from 143 bp downwards, occurringwith an approximately 10 bp periodicity. Circulatingfetal DNA, on the other hand, is relatively depletedfor the 166 bp peak, but has retained the peaks from143 bp and below. These data suggest that circulatingfetal DNA molecules might have their linker fragment‘trimmed’. These observations might shed light on themechanisms involved in the clearance of circulatingnucleic acids and might have implications on the devel-opment of methods for selectively enriching circulatingfetal DNA.

Circulating DNA in cancer patients

Historically, the interest in circulating DNA was ini-tially focused on its quantitative and qualitative aberra-tions in cancer [4,5,30]. Beck et al [31] used massivelyparallel sequencing to profile DNA isolated from theserum of a number of apparently healthy individuals.They found that the proportions of different classes of

sequences (eg genes, protein-coding sequences, etc.)were generally reminiscent of those in the genome,except that certain classes of repeat sequences (eg theAlu repeats) were over-represented in serum DNA.The observation concerning Alu repeats is reminiscentto that made using real-time PCR [32]. This groupthen applied this approach to study the serum DNAisolated from patients with breast cancer and con-trol subjects [33]. They found that the serum DNAof patients with breast cancer contained significantlymore repetitive elements than that from the controlsubjects. The molecular basis of these observationsis unclear at the present time. One possible expla-nation might be epigenetic aberrations in the cancergenome, which would then lead to an alteration innucleosome positioning. Such changes might lead toa differential degradation of DNA derived from can-cer and normal cells in serum. In this regard, it isimportant to note that the studies by Beck et al werebased on serum, rather than plasma DNA [31,33].The choice of serum, rather than plasma, might intro-duce an additional complexity, as it has previouslybeen shown that blood clotting during serum forma-tion would lead to DNA release into the serum [9]. Itis currently unknown whether the clotting process incancer and healthy subjects might have subtle differ-ences that might lead to aberrations in the compositionof serum DNA.

The advantage of the method by Beck et al is thatit seems to allow the detection of cancer withoutneeding access to tumour tissues. Other investigators,however, have taken a different route. Two groups haveused massively parallel sequencing of DNA extractedfrom tumour tissues to identify translocations in thetumour genome [34,35]. They then developed highlyspecific PCR-based assays to detect and measure suchtumour-specific translocations in the plasma of patientsfrom whom the tumours were originally obtained. Thisstrategy has been found to allow personalized, highlysensitive and specific monitoring of tumour-derivedDNA in the circulation. With the further reduction inthe cost of DNA sequencing, this approach is expectedto be increasingly accessible for clinical care. Thedisadvantage of this approach is the fact that tumourtissues are necessary for the initial characterization ofthe translocations and thus this approach, as currentlydescribed, is not usable for cancer screening.

Korshunova et al [36] have explored the use ofmassively parallel sequencing for detecting tumour-associated DNA methylation changes in the serum ofbreast cancer patients. Following bisulphite conver-sion, in which methylated and non-methylated cyto-sine residues would be unchanged and changed touracil, respectively, Korshunova et al generated multi-ple amplicons corresponding to four genomic regionsin which their previous work had demonstrated dif-ferential methylation in breast cancer tissues. Theythen sequenced these amplicons using massively par-allel sequencing. Using this approach, these authorswere able to observe a difference in methylation


Howard

Highlight

Howard

Highlight

YMD Lo and RWK Chiu

levels in breast cancer and normal tissues. How-ever, they were not able to observe a convincingdifference between the methylation levels in serumsamples from breast cancer patients and control sub-jects. The most likely explanation for these relativelydisappointing results was that the fractional concen-trations of tumour DNA in the serum samples werebelow the sensitivity of the analytical approach. Theuse of serum, instead of plasma samples, was likelyto be a contributory factor, because DNA releasefrom nucleated blood cells into the serum would beexpected to further reduce the fractional tumour DNAin the sample [9]. The future analysis of plasma DNAfor methylation changes using much deeper sequenc-ing might yield more useful data for the ‘plasmaepigenome’.

Plasma DNA for transplantation monitoring

In 1998, it was shown that, following solid organtransplantation, the DNA from the donor could befound in the plasma of the recipient [6]. This phe-nomenon was first demonstrated in female recipientsof organs from male donors [6]. Sequences from the Ychromosome could be observed in 100% of such sex-mismatched liver transplantation recipients and 80%of the kidney transplantation recipients. The ease ofdetecting such donor-derived sequences appears to berelated to the size of the donated organ, with the trans-planted liver releasing readily detectable amounts ofdonor-derived DNA when compared to the situationfor kidney and heart transplantation [37]. Furthermore,in sex-mismatched bone marrow transplantation cases,it was found that the sex of the plasma of the recipientwas converted to the sex of the donor [38]. This latterobservation has suggested that haematopoietic cells arethe predominant source of plasma DNA.

As it is widely accepted that DNA is released intoplasma when cells die, it has been proposed that themeasurement of donor-derived DNA in the plasma oftransplantation recipients could be used for monitoringgraft rejection [6]. This hypothesis has subsequentlybeen shown to be correct by a number of groups in bothhumans and animal models [39–41]. Thus, there isan elevation in the concentration of circulating donor-derived DNA during rejection episodes. The use ofY-chromosomal sequences as a marker of the donorhas limited such an approach for female recipientswith male donors. Several workers have attempted toovercome this disadvantage by using markers from thehuman leukocyte antigen (HLA) region [39]. While thislatter method has expanded the population coverage ofthis approach for monitoring rejection, it requires thedevelopment of multiple quantitative assays.

Recently, it was shown that massively parallelsequencing of the plasma DNA obtained from trans-plantation recipients might provide a universal solutionfor this problem [42]. Snyder et al used heart trans-plantation as their model system. They employed a

bead-based system for carrying out genome-wide geno-typing of single nucleotide polymorphisms (SNPs) ofconstitutional DNA of the donor and recipient. Theythen analysed the sequencing data from the recipients’plasma for reads containing such SNP signatures. Theywere then able to calculate the fractional concentrationof donor-derived DNA in the plasma. They observedthat rejection episodes were associated with an eleva-tion in the fractional concentration of donor-derivedDNA in plasma.

Future work would be needed to test the stabilityof the concentration threshold for diagnosing rejectionepisodes. It is likely that such a threshold would needto be adjusted for different types of transplantation andfor different immunosuppressive regimens. As prompttreatment of rejection episodes is needed, future workwould also be needed to reduce the turnaround time ofsuch an assay, which currently takes days to complete.

Plasma DNA for tissue damage

The relationship between plasma DNA and tissue dam-age, explored above in the context of graft rejection,can be generalized to other types of tissue damage.Hence, it has been shown that the concentration ofplasma DNA would increase in a number of acutemedical conditions, eg trauma [7], cardiac ischaemia[43], stroke [44] and sepsis [45]. The measurements ofplasma DNA in such applications are typically per-formed by relatively simple quantitative PCR-basedtechniques. With the development of massively paral-lel sequencing, researchers have attempted to explorewhether abnormalities in plasma DNA can be seen ina number of chronic disorders. In this regard, it isintriguing that one group of workers have reportedthe aberration representation of selected coding andrepeat sequences in the serum of patients with multiplesclerosis [46]. The biological basis of this observa-tion remains unclear at the moment. One possibilitymay be related to changes in the methylation statusof DNA [47]. In cattle, massively parallel sequenc-ing has identified disease-specific patterns in circu-lating DNA in animals infected with transmissiblespongiform encephalopathies [48]. Thus, it is possiblethat massively parallel sequencing of circulating DNAwould provide us with new tools for the monitoring ofchronic neurological disorders and shed light on theirpathogenesis.

Conclusion

The recent availability of massively parallel sequencinghas given us unprecedented precision and ability toanalyse DNA in plasma. This development is expectedto impact on the future development of moleculardiagnostics and to provide us a window on molecularprocesses that are occurring in tissues that have hitherto


Plasma nucleic acid analysis by massively parallel sequencing

been difficult to sample, eg those from the fetus,tumour, transplanted organ or the nervous system.

Author contributions

Both authors contributed to the planning, writing andrevision of this review.

References

1. Mandel P, Metais P. Les acides nucleiques du plasma sanguin chezl’homme. C R Acad Sci Paris 1948; 142: 241–243.

2. Tsang JC, Lo YMD. Circulating nucleic acids in plasma/serum.Pathology 2007; 39: 197–207.

3. Lo YMD, Corbetta N, Chamberlain PF, et al . Presence of fetalDNA in maternal plasma and serum. Lancet 1997; 350: 485–487.

4. Nawroz H, Koch W, Anker P, et al . Microsatellite alterations inserum DNA of head and neck cancer patients. Nat Med 1996; 2:1035–1037.

5. Chen XQ, Stroun M, Magnenat JL, et al . Microsatellite alterationsin plasma DNA of small cell lung cancer patients. Nat Med 1996;2: 1033–1035.

6. Lo YMD, Tein MSC, Pang CCP, et al . Presence of donor-specificDNA in plasma of kidney and liver-transplant recipients. Lancet1998; 351: 1329–1330.

7. Lo YMD, Rainer TH, Chan LY, et al . Plasma DNA as a prognosticmarker in trauma patients. Clin Chem 2000; 46: 319–323.

8. Schuster SC. Next-generation sequencing transforms today’s biol-ogy. Nat Methods 2008; 5: 16–18.

9. Lo YMD, Tein MS, Lau TK, et al . Quantitative analysis of fetalDNA in maternal plasma and serum: implications for noninvasiveprenatal diagnosis. Am J Hum Genet 1998; 62: 768–775.

10. Costa JM, Benachi A, Gautier E. New strategy for prenatal diag-nosis of X-linked disorders. N Engl J Med 2002; 346: 1502.

11. Lo YMD, Hjelm NM, Fidler C, et al . Prenatal diagnosis of fetalrhd status by molecular analysis of maternal plasma. N Engl J Med1998; 339: 1734–1738.

12. Finning K, Martin P, Summers J, et al . Effect of high throughputRHD typing of fetal DNA in maternal plasma on use of anti-RhDimmunoglobulin in RhD-negative pregnant women: prospectivefeasibility study. Br Med J 2008; 336: 816–818.

13. Lo YMD, Lun FM, Chan KC, et al . Digital PCR for the moleculardetection of fetal chromosomal aneuploidy. Proc Natl Acad SciUSA 2007; 104: 13116–13121.

14. Vogelstein B, Kinzler KW. Digital PCR. Proc Natl Acad Sci USA1999; 96: 9236–9241.

15. Chiu RWK, Chan KCA, Gao Y, et al . Noninvasive prenatal diag-nosis of fetal chromosomal aneuploidy by massively parallelgenomic sequencing of DNA in maternal plasma. Proc Natl AcadSci USA 2008; 105: 20458–20463.

16. Fan HC, Blumenfeld YJ, Chitkara U, et al . Noninvasive diagnosisof fetal aneuploidy by shotgun sequencing DNA from maternalblood. Proc Natl Acad Sci USA 2008; 105: 16266–16271.

17. Chiu RWK, Akolekar R, Zheng YW, et al . Non-invasive prenatalassessment of trisomy 21 by multiplexed maternal plasma DNAsequencing: large-scale validity study. Br Med J 2011; 342: c7401.

18. Ehrich M, Deciu C, Zwiefelhofer T, et al . Noninvasive detectionof 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.

19. Chiu RWK, Sun H, Akolekar R, et al . Maternal plasma DNA anal-ysis with massively parallel sequencing by ligation for noninvasiveprenatal diagnosis of trisomy 21. Clin Chem 2010; 56: 459–463.

20. Fan HC, Quake SR. Sensitivity of noninvasive prenatal detectionof fetal aneuploidy from maternal plasma using shotgun sequencingis limited only by counting statistics. PLoS ONE 2010; 5: e10439.

21. Sehnert AJ, Rhees B, Comstock D, et al . Optimal detection offetal chromosomal abnormalities by massively parallel DNAsequencing of cell-free fetal DNA from maternal blood. Clin Chem2011; (in press).

22. Lo YMD, Chan KCA, Sun H, et al . Maternal plasma DNAsequencing reveals the genome-wide genetic and mutational profileof the fetus. Sci Transl Med 2010; 2: 61ra91.

23. Liao GJ, Lun FM, Zheng YW, et al . Targeted massively parallelsequencing of maternal plasma DNA permits efficient and unbiaseddetection of fetal alleles. Clin Chem 2011; 57: 92–101.

24. Fan HC, Wang J, Potanina A, et al . Whole-genome molecularhaplotyping of single cells. Nat Biotechnol 2011; 29: 51–57.

25. Kitzman JO, Mackenzie AP, Adey A, et al . Haplotype-resolvedgenome sequencing of a Gujarati Indian individual. Nat Biotechnol2011; 29: 59–63.

26. Greely HT. Get ready for the flood of fetal gene screening. Nature2011; 469: 289–291.

27. de Jong A, Dondorp WJ, de Die-Smulders CE, et al . Non-invasiveprenatal testing: ethical issues explored. Eur J Hum Genet 2010;18: 272–277.

28. Chan KCA, Zhang J, Hui AB, et al . Size distributions of maternaland fetal DNA in maternal plasma. Clin Chem 2004; 50: 88–92.

29. Lewin B. Nucleosomes: Gene IX. Sudbury: Jones and Bartlett,2008; 757–95.

30. Leon SA, Shapiro B, Sklaroff DM, et al . Free DNA in the serumof cancer patients and the effect of therapy. Cancer Res 1977; 37:646–650.

31. Beck J, Umovitz HB, Riggert J, et al . Profile of the circulatingDNA in apparently healthy individuals. Clin Chem 2009; 55:730–738.

32. Stroun M, Lyautey J, Lederrey C, et al . Alu repeat sequences arepresent in increased proportions compared to a unique gene inplasma/serum DNA: evidence for a preferential release from viablecells? Ann NY Acad Sci 2001; 945: 258–264.

33. Beck J, Urnovitz HB, Mitchell WM, et al . Next generationsequencing of serum circulating nucleic acids from patients withinvasive ductal breast cancer reveals differences to healthy andnonmalignant controls. Mol Cancer Res 2010; 8: 335–342.

34. Leary RJ, Kinde I, Diehl F, et al . Development of personalizedtumor biomarker using massively parallel sequencing. Sci TranslMed 2010; 2: 20ra14.

35. McBride DJ, Orpana AK, Sotiriou C, et al . Use of cancer-specificgenomic rearrangements to quantify disease burden in plasma frompatients with solid tumors. Genes Chromosomes Cancer 2010; 49:1062–1069.

36. Korshunova Y, Maloney RK, Lakey N, et al . Massively parallelbisulphite pyrosequencing reveals the molecular complexity ofbreast cancer-associated cytosine-methylation patterns obtainedfrom tissue and serum DNA. Genome Res 2008; 18: 19–29.

37. Lui YYN, Woo KS, Wang AYM, et al . Origin of plasma cell-free DNA after solid organ transplantation. Clin Chem 2003; 49:495–496.

38. Lui YY, Chik KW, Chiu RWK, et al . Predominant hematopoieticorigin of cell-free DNA in plasma and serum after sex-mismatchedbone marrow transplantation. Clin Chem 2002; 48: 421–427.

39. Gadi VK, Nelson JL, Boespflug ND, et al . Soluble donor DNAconcentrations in recipient serum correlate with pancreas–kidneyrejection. Clin Chem 2006; 52: 379–382.

40. Moreira VG, Garcia BP, Martin JMD, et al . Cell-free DNA as anoninvasive acute rejection marker in renal transplantation. ClinChem 2009; 55: 1958–1966.


YMD Lo and RWK Chiu

41. Martins PN, Mashreghi MF, Reutzel-Selke A, et al . Quantificationof donor-derived DNA in serum: a new approach of acute rejectiondiagnosis in a rat kidney transplantation model. Transpl Proc 2005;37: 87–88.

42. Snyder TM, Khush KK, Valantine HA, et al . Universal noninva-sive detection of solid organ transplant rejection. Proc Natl AcadSci USA 2011; 108: 6229–6234.

43. Rainer TH, Lam NY, Man CY, et al . Plasma !-globin DNA as aprognostic marker in chest pain patients. Clin Chim Acta 2006;368: 110–113.

44. Rainer TH, Wong LK, Lam W, et al . Prognostic use of circulatingplasma nucleic acid concentrations in patients with acute stroke.Clin Chem 2003; 49: 562–569.

45. Saukkonen K, Lakkisto P, Pettila V, et al . Cell-free plasma DNAas a predictor of outcome in severe sepsis and septic shock. ClinChem 2008; 54: 1000–1007.

46. Beck J, Urnovitz HB, Saresella M, et al . Serum DNA motifspredict disease and clinical status in multiple sclerosis. J Mol Diagn2010; 12: 312–319.

47. Liggett T, Melnikov A, Tilwalli S, et al . Methylation patterns ofcell-free plasma DNA in relapsing-remitting multiple sclerosis.J Neurol Sci 2010; 290: 16–21.

48. Gordon PM, Schutz E, Beck J, et al . Disease-specific motifs canbe identified in circulating nucleic acids from live elk and cattleinfected with transmissible spongiform encephalopathies. NucleicAcids Res 2009; 37: 550–556.


plasma nucleic acid analysis by massively parallel...

Documents