The purpose of this paper was to compare DNA yield and circulating tumor DNA (ctDNA) variant allele frequency (VAF) among plasma from blood stored in EDTA or Streck BCT for up to 14 days and between different cell-free DNA (cfDNA) extraction kits using blood from colorectal carcinoma patients. The ability to detect colorectal cancer relapse using ctDNA was also investigated.
Sensitive Quantification of Cell-Free Tumor DNA for Early Detection of Recurrence in Colorectal Cancer.
Stasik S, Mende M, Schuster C, Mahler S, Aust D, Tannapfel A, Reinacher-Schick A, Baretton G, Krippendorf C, Bornhäuser M, Ehninger G, Folprecht G, Thiede C
Purpose
Conclusion
When blood was stored in EDTA tubes DNA yield increased and VAF decreased, which resulted in KRAS mutations falling below the limit of detection after 14 days of storage. In contrast, when blood from the same patient was stored in Streck BCT only slight increases in DNA yield and decreases in the VAF were observed. The highest DNA yield and most accurate VAF were obtained with manual extraction using the Zymo Research Kit, followed by the Qiagen Kit, although a single specimen was used in the experiment. When specimens from 15 patients were examined, yield and mutation detection were not significantly different between extraction methods (Zymo Research automated method and the QIAsymphony automated method). In a serially diluted specimen, detection of the KRAS c.35G>A mutation by NGS was linear with a false positive rate of 0.0016 ± 0.0007%, but VAF increased slightly when 20 ng of DNA was used as input compared to 5 ng.
Tumor-specific mutations were detected in 13 of 14 plasma specimens obtained during imaging-based detection of relapse, with higher levels of ctDNA observed when cfDNA levels increased. In the prospective cohort, 9% of patients had a post-operative mutation detected at a median VAF of 0.042%. The false positive rate was below 0.01% for each mutation, but the actual level differed between types of mutations.
Studies (2)
Purpose
This study investigated the effects of storing blood in different types of collection tube on KRAS genotyping results using a single specimen. Blood from a KRAS mutation positive (c.38G>A VAF 35%) colorectal carcinoma patient was diluted with the blood of healthy donors to achieve a VAF of 20%, then transfer to S-Monovette EDTA tubes and Cell-Free DNA BCT tubes and stored at room temperature for 0, 1, 2, 3, 4, or 14 days before plasma separation. Plasma was separated by centrifugation at 300 g for 20 min followed by 5000 g for 10 min and stored at -20°C until cfDNA extraction with either Jena PME free-circulating DNA extraction kit QIAamp Circulating NA Kit Zymo Quick cfDNA serum and plasma kit. DNA (cfDNA and genomic) was quantified by real-time PCR amplification of β-globin. KRAS mutations were detected by mutation-specific real-time PCR.
Summary of Findings
DNA yield increased 10-fold after 3 days and 100-fold after 14 days when blood was stored in EDTA tubes, indicating a high level of genomic contamination; consequently, the VAF for the KRAS mutation declined rapidly when blood was stored in EDTA tubes and was below the limit of detection (0.01%) by day 14. In contrast, when blood was stored in Streck BCT DNA yield only increased 1.5-fold and the VAF only declined slightly (from a technical replicate mean of 20% to 14.64%) by day 14.
Biospecimens
FluidBloodFluidPlasmaPreservation Types
Streck/BCTFrozenDiagnoses
NeoplasticCarcinomaNormalPlatforms
DNAReal-time qPCRPre-analytical Factors
Classification Pre-Analytical Factor Value(s) Biospecimen Acquisition Type of collection container/solution EDTA tubeStreck BCTBiospecimen Aliquots and Components Centrifugation Centrifugation delays investigatedBiospecimen Preservation Type of fixation/preservation Blood collection tube additiveEDTAStorage Storage duration 0 days1 days14 days2 days3 days4 days
Purpose
This study investigated the effects of cfDNA isolation method and input amount on KRAS mutation detection and investigated the feasibility of using a cfDNA-based assay to detect a relapse of colorectal cancer. Plasma was isolated from blood from a KRAS mutation positive (c.38G>A VAF 10%) colorectal carcinoma patient (methods not specified). DNA was extracted from plasma using Jena PME free-circulating DNA extraction kit, QIAamp Circulating NA Kit and the Zymo Quick cfDNA serum and plasma kit. Additionally, cfDNA was isolated from the plasma of 15 colorectal carcinoma patients (details of collection and processing not provided) using the Zymo Research kit and the automated PAXcircDNA_STA_2400 protocol on a QIAsymphony instrument. cfDNA and ctDNA were quantified by real-time PCR amplification of β-globin and mutation-specific assays and stored at -20°C until NGS. Fusion libraries were cleaned by two rounds of Agencourt AMPure XP Reagent and sequenced on an Ion S5 XL NGS system with a target of 100,000 X. To determine linearity of SNV detection, DNA extracted from a patient (VAF 45%) was serially diluted with wildtype cfDNA and the mutation was detected by NGS. To test if a colorectal cancer relapse could be detected using cfDNA, blood was obtained from colorectal carcinoma 14 patients at the time of imaging-based detection for a relapse and VAF was compared with the tumor and with a blood specimens collected 1 month post-initial tumor resection. To determine if ctDNA levels are predictive of a cancer relapse, a blood specimen was collected from 67 colorectal cancer patients 5-35 days after surgical resection of a tumor and deep sequencing was conducted to identify mutations.
Summary of Findings
The highest DNA yields from plasma were obtained with manual extraction with the Zymo Research Kit (median 7.01 ng/mL), followed by the Qiagen Kit (median 6.47 ng/mL) and the Analytik Jena Kit (median 2.42 ng/mL). VAF detection of the KRAS c.38G > A mutation using cfDNA was influenced by cfDNA isolation kit: Zymo Research Kit (median 8.78%) followed by the Qiagen kit (median 5.86%) and the Analytik Jena Kit (3.96%).Among blood specimens with a VAF of 10%, However, yield and mutation detection were not significantly different between specimens extracted using the Zymo Research automated method and the QIAsymphony method in blood specimens from 15 patients. In a serially diluted specimen, detection of the KRAS c.35G>A by NGS was linear (r2 =0.9946) between 0.0045% and 45%, with a false positive rate (detection in wildtype specimen) of 0.0016 ± 0.0007%. Importantly, specificity for detection of KRAS mutations at a VAF >0.1% was 100%, but detection levels increased when 20 ng of DNA was used as input instead of 5 ng (median VAF of 0.12 versus 0.08%).
Tumor-specific mutations were detected in 13 of 14 plasma specimens obtained during imaging-based detection of a cancer relapse; specimens for which a mutation was not detected by cfDNA analysis had a VAF of 9% in the primary tumor.
Biospecimens
FluidBloodFluidPlasmaPreservation Types
FrozenDiagnoses
NeoplasticCarcinomaNormalPlatforms
DNAReal-time qPCRDNANext generation sequencingPre-analytical Factors
Classification Pre-Analytical Factor Value(s) Biospecimen Acquisition Time of biospecimen collection <1 month post-resectionAt image detected relapseAnalyte Extraction and Purification Analyte isolation method Automated PAXcircDNA_STA_2400 protocol on a QIAsymphony instrumentAutomated Zymo Research methodManual Jena PME free-circulating DNA extraction kitManual QIAamp Circulating NA KitManual Zymo Quick cfDNA serum and plasma kitNext generation sequencing Specific Template/input amount 0.0045%0.045%0.45%20 ng4.5%45%5 ng