search for




 

Validation of fetus aneuploidy in 221 Korean clinical samples using noninvasive chromosome examination: Clinical laboratory improvement amendments-certified noninvasive prenatal test
J Genet Med 2015;12:79-84
Published online December 31, 2015;  https://doi.org/10.5734/JGM.2015.12.2.79
© 2015 Korean Society of Medical Genetics.

Min-Jeong Kim, Chang Hyuk Kwon, Dong-In Kim, Hee Su Im, Sungil Park, Ji Ho Kim, Jin-Sik Bae, Myunghee Lee, and Min Seob Lee*

Eone-Diagnomics Genome Center (EDGC), Inc., Incheon, Korea
Min Seob Lee, Ph.D. Eone-Diagnomics Genome Center (EDGC), Inc., 291 Harmony-ro, Yeonsu-gu, Incheon 22014, Korea. Tel: +82-32-210-2341, Fax: +82-32-210-2349, E-mail: mlee@edgc.com
Received November 11, 2015; Revised December 14, 2015; Accepted December 24, 2015.
cc This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract

Purpose:

We developed and validated a fetal trisomy detection method for use as a noninvasive prenatal test (NIPT) including a Clinical Laboratory Improvement Amendments (CLIA)-certified bioinformatics pipeline on a cloud-based computing system using both Illumina and Life Technology sequencing platforms for 221 Korean clinical samples. We determined the necessary proportions of the fetal fraction in the cell-free DNA (cfDNA) sample for NIPT of trisomies 13, 18, and 21 through a limit of quantification (LOQ) test.

Materials and Methods:

Next-generation sequencing libraries from 221 clinical samples and three positive controls were generated using Illumina and Life Technology chemistries. Sequencing results were uploaded to a cloud and mapped on the human reference genome (GRCh37/hg19) using bioinformatics tools. Based on Z-scores calculated by normalization of the mapped read counts, final aneuploidy reports were automatically generated for fetal aneuploidy determination.

Results:

We identified in total 29 aneuploid samples, and additional analytical methods performed to confirm the results showed that one of these was a false-positive. The LOQ test showed that the proportion of fetal fraction in the cfDNA sample would affect the interpretation of the aneuploidy results.

Conclusion:

Noninvasive chromosome examination (NICE), a CLIA-certified NIPT with a cloud-based bioinformatics platform, showed unambiguous success in fetus aneuploidy detection.

Keywords : Prenatal diagnosis, High-throughput nucleotide sequencing, Computational biology, Cell-free DNA, Trisomy, Aneuploidy
Introduction

Recent technological advancements in next generation sequencing (NGS) and bioinformatics have led to a novel noninvasive prenatal test (NIPT) for analyzing fetus aneuploidy using cell-free DNA (cfDNA) in the plasma of pregnant women. Whereas other NIPT technologies perform an analysis of specific regions of the chromosome by a targeted sequencing method, such as single nucleotide polymorphism genotyping or haplotype-based analysis [1-3], this novel chromosome analysis uses massive parallel sequencing data obtained by an NGS technology and applies statistical normalization to each chromosome read count [4,5]. Sequencing reads are mapped to the human reference genome (GRCh37/hg19) and quantified according to their genomic locus. After normalizing the read count in a specific bin of a genomic region, Z-scores are calculated for each chromosome to determine fetal aneuploidy status.

We developed the noninvasive chromosome examination (NICE) test to improve the processing and success rate of the NIPT, and we implemented the data processing, report generation, and file storage on a secure, cloud-based platform. The NICE process, including the cloud-based bioinformatics and reporting system, was developed to meet the quality control requirements of clinical laboratory standards and has been fully validated and accredited under the Clinical Laboratory Improvement Amendments (CLIA) in the United States.

In the present study, we performed a validation of the new test by comparing the outcome of a NICE analysis to the known results of 221 clinical samples and three proprietary control samples we developed. We generated 227 NGS sequencing libraries using Illumina and Life Technology sequencing chemistries, including two libraries for each three positive aneuploid controls, and we then tested and validated NICE using the automated cloud-based bioinformatics platform.

Materials and Methods

1. Cell-free DNA extraction from maternal blood

Blood samples (10 mL) were collected from 221 participants into Cell-Free DNA BCT tubes (Streck, Omaha, NE, USA) and centrifuged at 1,800 ×g for 15 min at 4°C. Plasma samples were transferred into 1.7-mL tubes and then centrifuged at 18,000 ×g for 15 min at 4°C. The separated plasma samples were transferred to new tubes and stored at ?80°C. cfDNA was isolated from 2 mL plasma by using the QIAamp Circulating Nucleic Acid Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions.

2. Next-generation sequencing and data analysis

Isolated cfDNA samples (<100 ng) were used as input DNA for either Illumina or Life Technology sequencing reactions. HiSeq 100-bp paired-end (PE), 50-bp single-read (SR), or Proton SR sequencing libraries were prepared according to the manufacturer’s instructions (Illumina, San Diego, CA, USA; Life Technology, Carlsbad, CA, USA). All sequencing data were then uploaded to the cloud-based bioinformatics platform to perform the aneuploidy test and to generate results based on Z-score calculations for review.

First, we mapped the sequencing reads to the human reference genome (GRCh37/hg19) using Burrows-Wheeler Aligner (BWA) software; we then sorted the mapped reads and normalized the read count using Samtools. This was followed by detection and removal of duplicated reads. Finally, we calculated the Z-score with a GC correction and mapping correction using the ReadDepth R module [6-8]. Upon approval of the CLIA director, final reports were automatically generated from the NICE portal (Fig. 1).

Results

1. Limit of quantification (LOQ) test

Although NGS-based NIPT aneuploidy tests have high sensitivities and specificities, the accuracy of each test is heavily influenced by the amount of fetal DNA present in maternal plasma. A LOQ test is very important to assess the reliable lowest quantity of fetal DNA in a sample to be differentially detected because the concentration of fetal DNA in maternal plasma can vary depending on maternal health factors, gestational age, and health of the placenta [9,10]. In order to evaluate the sensitivity of the NIPT, an LOQ test is essential for the analytical procedure.

For the test, we used non-fetal mother (Cat. No. AG09387), T13 (GM02948), T18 (GM01359), or T21 (AG09394) cell lines purchased from the Coriell Institute for Medical Research (Camden, NJ, USA). The LOQ test was performed with different concentrations of trisomy genomic DNA samples extracted from the cells, and each DNA sample was mixed with the genomic DNA solution of the non-fetal mother in concentrations of : 0.5%, 2%, 4%, 6%, 8%, and 10% fetal DNA. We also prepared a DNA sample extracted from the normal brother (AG09392) of T21 as a negative control and used the same serial concentrations. Three replicated NGS libraries for each sample were prepared, and sequencing data for all 72 libraries were generated using a Proton sequencer (Life Technology).

From the results of mapping the sequencing data to the human genome (see Materials and Methods), LOQ was calculated with a 95% confidence level based on the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines. We found that the proportion of trisomy 13 fetal DNA should be more than 2.5% of the sample to detect T13 aneuploidy. More than 4.1% and 5.2% of the chromosome 18 and 21 trisomies in the samples could be detected, respectively, meaning that lower concentrations of trisomy DNA in the samples may lead to false negatives on the aneuploidy test (Fig. 2).

2. Cloud-based clinical aneuploidy test

In total, 221 maternal blood samples were obtained from voluntary participants, and three different cell lines of chromosome 13, 18, and 21 trisomies were used as positive controls. Among the pregnant participants, 207 cases were categorized as high-risk based on previous tests and/or classifications including 137 cases of positive serum screening, 41 cases of advanced maternal age, five cases of abnormal ultrasonography, four cases of genetic history, and 20 cases of multiple indications of the above. Mean maternal age was 33.2 years, ranging from 26 to 40 years, and most samples (80.5%) were collected during the second trimester (Table 1). This study was approved in September 2013 by the Eone Laboratories Institutional Review Board (IRB #128477-201309-BR-001-01). Informed consent was obtained from the participants for the study based on IRB guidelines.

The 221 cfDNA and six genomic DNA reference samples were extracted from the participants’ blood and cultured three positive control cells, respectively. The Illumina HiSeq NGS platform was used to generate 100-bp PE or 50-bp SR sequences from 50 and 20 cfDNA samples, respectively, while 151 cfDNA and six control DNA samples were processed with the Proton Life Technology NGS platform to generate SR sequencing reads. Each sequencing dataset was transferred to the cloud to perform the NICE aneuploidy test (Fig. 3).

Of the 221 samples, 29 exhibited aneuploidy based on our Z-score criteria: 22 for trisomy 21, 6 for trisomy 18, and 1 for trisomy 13 (Table 1 and Fig. 3). After the genetic consultation, additional tests for 21 samples including amniocentesis, quantitative fluorescent polymerase chain reaction (QF PCR), and/or fluorescence in situ hybridization (FISH), were performed to validate the aneuploidy results. All were consistent with the NICE results except for one trisomy 21 sample (#25) (Table 2), which was confirmed as normal by amniocentesis, QF PCR, and FISH. However, interestingly, an Illumina Verifi test for the sample reported the same result as the NICE test. All positive controls were clearly detected in the aneuploidy Z-score region (Fig. 4). Additionally, among the 192 women classified as NIPT-negative, 64 underwent further aneuploidy testing, including amniocentesis, QF PCR, and/or FISH; all were confirmed to be negative for aneuploidy. As a result, we compared 81 NICE validation tests with amniocentesis results, as amniocentesis has been accepted as the gold standard for diagnosing aneuploidy (Table 3). There were 20 true positives and one false positive case in this test, resulting in a test sensitivity of 100% and specificity of 98.36%, and the incidences of trisomies 13, 18, and 21 in the sample were 1.23%, 6.17%, and 17.28%, respectively.

We also performed a proficiency test to assess how well our data is analyzed compared to the Verifi NIPT results for 129 samples. Among 14 aneuploid samples based on Verifi results, 13 were detected as aneuploid based on the NICE test, all of which were consistent with amniocentesis results. All of the results for the negative-aneuploidy samples using NICE were also consistent with Verifi.

Discussion

Due to its noninvasiveness, superior sensitivity and specificity, and higher success rate compared to conventional aneuploidy screening methods, NGS-based NIPT has been widely accepted for prenatal aneuploidy tests. We developed the NICE procedure and implemented the pipeline on a cloud-based system. We successfully identified trisomy 13 (Patau), 18 (Edward), and 21 (Down) syndromes in 221 clinical samples and three trisomy positive controls using the cloud-based pipeline.

In total, we found 22 cases of trisomy 21, six cases of trisomy 18, and one case of trisomy 13. These results were verified with other conventional aneuploidy tests. Only one of the positive samples was found to be a false-positive upon comparison with the amniocentesis result. Of the 81 clinical samples that were compared to the amniocentesis results, we were able to successfully classify 80 NIPT samples (98.77%), including 59 normal and 21 aneuploidy cases, due to the high performance of our CLIA-certified NICE pipeline. Positive controls were also clearly detected using the same pipeline. An LOQ test showed the importance of the proportion of the fetal fraction in the sample, indicating that an insufficient amount of fetal DNA would lead to a false interpretation of aneuploidy. Further efforts are needed to develop a method of examining the amount of fetal DNA in maternal plasma in order to avoid misinterpretation.

The NICE analysis pipeline and reporting system were implemented on a secure Amazon cloud-based system, and the whole pipeline was certified under the CLIA in the United States. The clinically approved NICE cloud system offers an opportunity to perform clinical sample testing under the CLIA guidance to the labs which want to provide the NIPT service.

Acknowledgements

The authors would like to express appreciation to the patients who donated blood and to Eone reference laboratory for collecting samples for the study. We also thank the team at Maverix Biomics for helping with cloud preparation and ResearchDx for CLIA consultation.

Figures
Fig. 1.

Noninvasive chromosome examination (NICE) work flow. CLIA, Clinical Laboratory Improvement Amendments.


Fig. 2.

Limit of quantitation test with negative and positive aneuploidy samples. Trisomy genomic DNA samples were mixed with the genomic DNA solution of the non-fetal mother in concentrations of: 0.5%, 2%, 4%, 6%, 8%, and 10%. Z-scores of each 13, 18 and 21 chromosome are shown. (A) Normal brother of T21. (B) Trisomy 13. (C) Trisomy 18. (D) Trisomy 21.


Fig. 3.

High-level cloud architecture of noninvasive chromosome examination (NICE) test. AWS, Amazon web service.


Fig. 4.

Z-scores of the 221 clinical and six positive control samples.


TABLES

Table 1

Demographic information for the pregnant participants of the NICE test

VariableSample size (n)Number of samples with aneuploidy

Trisomy 21Trisomy 18Trisomy 13
Maternal age (yr)
?268000
?2714210
?2816210
?2916000
?3017100
?3115211
?3211100
?3316200
?3417100
?3515000
?3616010
?3716220
?3816100
?3916200
?4012600
Gestational age at test
?First trimester (10-13 weeks)29310
?Second trimester (14-27 weeks)1861951
?Third trimester (≥28 weeks)5000
?Unknown1000
Test platform
?Illumina (PE or SR)a701050
?Proton (SR)1511211
?Additional test for NICE-positive samplesa292261

NICE, noninvasive chromosome examination; PE, paired-end; SR, single-read.

a

Amniocentesis, quantitative fluorescent polymerase chain reaction (QF PCR), fluorescence in situ hybridization (FISH) and/or Illumina Verifi.


Table 2

Additional tests performed for the samples showing NICE aneuploidy results

SampleNICEAmniocentesisQF PCRFISHVerifi
1T13T13-a-T13
2T18T18T18T18-
3T18T18--Hemolysis
4T18T18T18--
5T18T18---
6T18T18T18--
7T18NT--T18
8T21T21T21--
9T21T21-T21-
10T21T21---
11T21T21T21--
12T21T21T21T21-
13T21T21---
14T21T21---
15T21T21---
16T21T21T21--
17T21T21T21--
18T21T21T21--
19T21T21T21--
20T21---T21
21T21---T21
22T21---T21
23T21---T21
24T21---T21
25T21NegativeNegativeNegativeT21
26T21T21--T21
27T21T21--T21
28T21---T21
29T21---T21

NICE, noninvasive chromosome examination; QF PCR, quantitative fluorescent polymerase chain reaction; FISH, fluorescence in situ hybridization.

Blank cells indicate ‘not-tested’.


Table 3

Validation test comparing NICE results with amniocentesis

TrisomyNSensitivity (%)95% CISpecificity (%)95% CIIncidence (%)
T1381100 (1/1)2.50-100100 (80/80)95.49-1001.23
T1881100 (5/5)47.82-100100 (76/76)95.26-1006.17
T2181100 (14/14)76.84-10098.51 (66/67)91.96-99.9617.28
Total81100 (20/20)83.16-10098.36 (60/61)91.20-99.9624.69

NICE, noninvasive chromosome examination; CI, confidence interval.


References
  1. Zimmermann B, Hill M, Gemelos G, Demko Z, Banjevic M, and Baner J et al. Noninvasive prenatal aneuploidy testing of chromosomes 13, 18, 21, X, and Y, using targeted sequencing of polymorphic loci. Prenat Diagn 2012;32:1233-41.
    Pubmed KoreaMed CrossRef
  2. Curnow KJ, Wilkins-Haug L, Ryan A, Kırkızlar E, Stosic M, and Hall MP et al. Detection of triploid, molar, and vanishing twin pregnancies by a single-nucleotide polymorphism-based noninvasive prenatal test. Am J Obstet Gynecol 2015;212:79.e1-9.
    Pubmed CrossRef
  3. Xu Y, Li X, Ge HJ, Xiao B, Zhang YY, and Ying XM et al. Haplotype-based approach for noninvasive prenatal tests of Duchenne muscular dystrophy using cell-free fetal DNA in maternal plasma. Genet Med 2015;17:889-96.
    Pubmed CrossRef
  4. Fan HC, Blumenfeld YJ, Chitkara U, Hudgins L, and Quake SR. Noninvasive diagnosis of fetal aneuploidy by shotgun sequencing DNA from maternal blood. Proc Natl Acad Sci U S A 2008;105:16266-71.
    Pubmed KoreaMed CrossRef
  5. Chiu RW, Chan KC, Gao Y, Lau VY, Zheng W, and Leung TY et al. Noninvasive prenatal diagnosis of fetal chromosomal aneuploidy by massively parallel genomic sequencing of DNA in maternal plasma. Proc Natl Acad Sci U S A 2008;105:20458-63.
    Pubmed KoreaMed CrossRef
  6. Li H, and Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 2009;25:1754-60.
    Pubmed KoreaMed CrossRef
  7. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, and Homer N et al. 1000 Genome Project Data Processing Subgroup. The Sequence Alignment/Map format and SAMtools. Bioinformatics 2009;25:2078-9.
    Pubmed KoreaMed CrossRef
  8. Miller CA, Hampton O, Coarfa C, and Milosavljevic A. ReadDepth: a parallel R package for detecting copy number alterations from short sequencing reads. PLoS One 2011;6:e16327.
    Pubmed KoreaMed CrossRef
  9. Vora NL, Johnson KL, Basu S, Catalano PM, Hauguel-De Mouzon S, and Bianchi DW. A multifactorial relationship exists between total circulating cell-free DNA levels and maternal BMI. Prenat Diagn 2012;32:912-4.
    CrossRef
  10. Levine RJ, Qian C, Leshane ES, Yu KF, England LJ, and Schisterman EF et al. Two-stage elevation of cell-free fetal DNA in maternal sera before onset of preeclampsia. Am J Obstet Gynecol 2004;190:707-13.
    Pubmed CrossRef


December 2016, 13 (2)
  • Crossref Similarity Check logo