Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Study of Preanalytic and Analytic Variables for Clinical Next-Generation Sequencing of Circulating Cell-Free Nucleic Acid.

Author(s): Mehrotra M, Singh RR, Chen W, Huang RSP, Almohammedsalim AA, Barkoh BA, Simien CM, Hernandez M, Behrens C, Patel KP, Routbort MJ, Broaddus RR, Medeiros LJ, Wistuba II, Kopetz S, Luthra R

Publication: J Mol Diagn, 2017, Vol. 19(4), Page 514-24

PubMed ID: 28506684 PubMed Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   This study investigated the effects of blood collection tube type and delayed centrifugation on the yield and size of cfDNA, as well as the detection of mutations using next-generation sequencing. Two methods of DNA quantification were also compared. The study included peripheral blood from 31 patients with known mutational status including eight patients with endometrial adenocarcinomas, seven with colon adenocarcinomas, three with breast adenocarcinomas, three with prostate adenocarcinomas, one with an ovarian carcinoma, one with a lung adenocarcinoma, one with a liver adenocarcinoma, one with an esophagus signet ring adenocarcinoma, one with an appendiceal mucinous adenocarcinoma, one with a squamous cell carcinoma of the tongue, one with a parathyroid carcinoma, and three with astrocytoma or ganglioneuroblastoma. Blood was drawn into both a regular K3-EDTA tube and a cell-free DNA BCT tube. Both tubes were inverted 10 times, stored for 0, 2, 4, 16 and >24 h and centrifuged at 2000 x g for 10 min at room temperature. Plasma was transferred to a new tube and centrifuged again at 2000 x g for 10 min at room temperature before extraction of cfDNA using the QIAamp Circulating Nucleic Acid Kit. cfDNA was stored at -20˚C until analysis. DNA was quantified using the Qubit dsDNA HS Assay and by amplifying 115 and 247 bp Alu repeats. Four cfDNA specimens were also sequenced using the Ion PGM.

   Summary of Findings:

   High molecular weight DNA was not present in plasma specimens as determined by bioanalyzer when centrifuged within 16 h regardless of the collection tube type. cfDNA yield was strongly correlated between blood collection tube types (r=0.820, P<0.0001) and values were comparable when plasma was obtained within 16 h of blood collection. However, significantly more cfDNA was obtained from specimens collected in K3-EDTA than BCT tubes when centrifugation was delayed by more than 24 h, (37.03 ng/mL versus 22 ng/mL , P<0.0001). cfDNA isolated from specimens collected in K3-EDTA tubes also had a higher ratio of real-time PCR amplicons (Alu 247 to Alu 115) than specimens collected in BCT tubes (0.39-0.55 versus 0.35-0.43, P<0.01). Importantly, when plasma specimens collected in Streck and K3-EDTA tubes were sequenced by Ion PGM mutant allele frequencies were 100% concordance and the variant allele frequencies (VAFS) were very strongly correlated (r=0.95, P<0.0001). Calculated DNA yields were slightly higher when determined by Qubit then by Alu real-time PCR.

   Biospecimens

   • Bodily Fluid - Blood
   • Bodily Fluid - Plasma

   Preservative Types

   • Streck/BCT
   • None (Fresh)

   Diagnoses:

   • Neoplastic - Other
   • Neoplastic - Carcinoma

   Platform:

   Analyte	Technology Platform
   DNA	Next generation sequencing
   DNA	Real-time qPCR
   DNA	Automated electrophoresis/Bioanalyzer
   DNA	Fluorometry

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Biospecimen Acquisition	Type of collection container/solution	Streck BCT K3-EDTA Vacutainer
   Biospecimen Aliquots and Components	Centrifugation	Centrifugation delays investigated
   Biospecimen Preservation	Type of fixation/preservation	Blood collection tube additive None (fresh)
   Real-time qPCR Specific	Targeted nucleic acid	Alu
   Fluorometry Specific	Technology platform	Real-time PCR
   Real-time qPCR Specific	Length of gene fragment	115 bp 247 bp
   Storage	Time at room temperature	0 h 2 h 4 h 16 h >24 h

   View more

2. Study Purpose

   This study compared the mutations detected in cfDNA isolated from plasma specimens using two platforms with those detected previously in FFPE tissue specimens from the same patient. The study included peripheral blood from 31 patients with known mutational status including eight with endometrial adenocarcinomas, seven with colon adenocarcinomas, three with breast adenocarcinomas, three with prostate adenocarcinomas, one with an ovarian carcinoma, one with a lung adenocarcinoma, one with a liver adenocarcinoma, one with an esophagus signet ring adenocarcinoma, one with an appendiceal mucinous adenocarcinoma, one with a squamous cell carcinoma of the tongue, one with a parathyroid carcinoma, and three with astrocytoma or ganglioneuroblastoma. Blood was drawn into both a regular K3-EDTA tube and a cell-free DNA BCT tube. Both tubes were inverted 10 times, stored for 0 h and centrifuged at 2000 x g for 10 min at room temperature. Plasma was transferred to a new tube and centrifuged again at 2000 x g for 10 min at room temperature before extraction of cfDNA using the QIAamp Circulating Nucleic Acid Kit. cfDNA was stored at -20˚C until analysis. DNA was quantified using the Qubit dsDNA HS Assay and by amplifying 115 and 247 bp Alu repeats. cfDNA specimens from all patients were pooled and sequenced using the Ampliseq Cancer Hot Spot Panel v2 on the Ion proton platform and using the Accel amplicon 56G oncology targeted panel on an Illumina MiSeq 300 Cycles v2 kit.

   Summary of Findings:

   Of the 54 mutations that were expected based on sequencing data obtained with FFPE tissue, 45 (87%) were identified in case-matched plasma specimens. Ion proton and MiSeq generated discordant results for 2 of these 45 mutations, with each platform detecting a mutation that was missed by the other. Upon manual inspection, all mutations identified in the tissue specimens were confirmed in plasma specimens, but they occurred at an allele frequency below the 1% allele frequency threshold. In addition, two mutations were identified in plasma but not tissue specimens using the two sequencing technologies, but were not confirmed by manual inspection of reads obtained from tissue specimens. As the time between tissue and plasma collection ranged between 736 and 530 days, the authors postulate that these mutations may have arisen in the interim between tissue and plasma collection. 

   Biospecimens

   • Bodily Fluid - Plasma
   • Tissue - Brain
   • Tissue - Breast
   • Tissue - Endometrium
   • Tissue - Esophagus
   • Tissue - Appendix
   • Tissue - Lung
   • Tissue - Ovary
   • Tissue - Prostate
   • Tissue - Tongue
   • Tissue - Liver

   Preservative Types

   • Streck/BCT
   • None (Fresh)
   • Formalin

   Diagnoses:

   • Neoplastic - Other
   • Neoplastic - Carcinoma

   Platform:

   Analyte	Technology Platform
   DNA	Next generation sequencing

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Next generation sequencing Specific	Technology platform	Ion Proton MiSeq
   Biospecimen Acquisition	Biospecimen location	Tumor Plasma

   View more

3. Study Purpose

   This study investigated potential effects associated with DNA input amount and next generation sequencing technology on the number of reads, mean sequencing depth and accuracy of mutation detection. Potential effects of bioinformatic pipeline on mutation discovery using MiSeq were also investigated. The study included peripheral blood from 28 patients with carcinomas and three patients with brain cancer. Blood was drawn into both a regular K3-EDTA tube and a cell-free DNA BCT tube. Both tubes were inverted 10 times before being centrifuged at 2000 x g for 10 min at room temperature. Plasma was transferred to a new tube and centrifuged again at 2000 x g for 10 min at room temperature before extraction of cfDNA using the QIAamp Circulating Nucleic Acid Kit. cfDNA was stored at -20˚C until analysis. DNA was quantified using the Qubit dsDNA HS Assay and size distribution was determined by amplifying 115 and 247 bp Alu repeats. DNA was pyrosequenced using the Ampliseq Cancer Hot Spot Panel v2 on the Ion PGM and proton platforms and sequenced using the Accel amplicon 56G oncology targeted panel on an Illumina MiSeq 300 Cycles v2 kit. The effect of DNA input amount was investigated by pooling and diluting libraries created from 4 specimens and the limit of detection was determined by dilution of a single specimen with two mutations. cfDNA libraries from four plasma specimens collected in K3-EDTA and Streck tubes were pooled, respectively, and sequenced using both ION platforms. Twelve cfDNA libraries were sequenced using MiSeq.

   Summary of Findings:

   When 5 pmol of DNA was used for library construction rather than 20 pmol, the number of pyrosequencing reads was higher (542 million versus 272 million) and the percentage of polyclonal reads was lower (32% versus 67%) , but the variant allele frequency was unaffected by input amount. Pyrosequencing using the Ion Proton platform as opposed to the Ion PGM platform resulted in 15.7-fold more reads (78.1 million versus 4.25 million) and 2.3-fold greater mean depth (8860x versus 3893x), but allelic calls were 100% concordant and very strongly correlated between the two platforms (R=0.9972, P<0.0001). The ION PGM and ION Proton were able to detect mutations even when DNA was diluted to 1.5% but further dilution required manual inspection of the sequencing reads for mutation detection. The average fold coverage using MiSeq was 3000x. Concordance between MiSeq and Ion Proton results was observed for 50 of the 52 variants detected (96%) and results were very strongly correlated (R=0.9213, P<0.0001). Analysis of MiSeq data using in-house and Swift Biosciences bioinformatics pipelines identified 3569 and 3876 variant calls, respectively. After filtering based on variant coverage and frequency and sequencing coverage, 949 and 910 variants remained, of these 899 variants were identified using both bioinformatics pipelines. 

   Biospecimens

   • Bodily Fluid - Plasma

   Preservative Types

   • None (Fresh)
   • Streck/BCT

   Diagnoses:

   • Neoplastic - Other
   • Neoplastic - Carcinoma

   Platform:

   Analyte	Technology Platform
   DNA	Next generation sequencing

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Next generation sequencing Specific	Template/input amount	5 pmol 10 pmol 12 pmol 20 pmol 50% mutated DNA 25% mutated DNA 12.5% mutated DNA 6.25% mutated DNA 3.15% mutated DNA 1.5% mutated DNA 0.75% mutated DNA
   Next generation sequencing Specific	Data handling	In-house bioinformatics pipeline Swift Biosciences bioinformatics pipelines
   Next generation sequencing Specific	Technology platform	Ion proton Ion PGM MiSeq

   View more