Skip to main content

Liquid biopsy in colon cancer: comparison of different circulating DNA extraction systems following absolute quantification of <i>KRAS</i> mutations using Intplex allele-specific PCR.

Kloten V, Rüchel N, Brüchle NO, Gasthaus J, Freudenmacher N, Steib F, Mijnes J, Eschenbruch J, Binnebösel M, Knüchel R, Dahl E

Publication: Oncotarget (2017) • Vol 8 • Page 86253-86263PMID: 29156792View on PubMed

Purpose

This paper investigated the effects of using serum rather than plasma, DNA extraction method, and inclusion of carrier RNA on circulating cell-free DNA (cfDNA) yield, fragment size, and allele frequencies. Additionally, the KRAS mutational status of matched serum and formalin-fixed and paraffin-embedded (FFPE) specimens was compared and it was investigated if cfDNA or circulating tumor DNA (ctDNA) from serum could distinguish between healthy patients and those with colorectal carcinoma (CRC). The effects of analytical method on quantification of cfDNA and determination of copy number were also investigated.

Conclusion

Although serum yielded more cfDNA than plasma, it also resulted in higher quantification of KRAS than expected and lower calculated allele frequencies. Use of magnetic bead-based kits rather than silica membrane-based kits for cfDNA extraction from serum resulted in a significantly higher yield of small cfDNA fragments compared to large fragments and a higher percentage of fragments <600 bp. cfDNA yields using either kit type were unaffected by the use of carrier RNA.

The concordance of KRAS status between tumor and serum specimens was 66%. The specificity of serum for KRAS status in tumors was 70% but the sensitivity was only 50%. cfDNA concentrations were higher in serum from patients with CRC than healthy controls and appeared to rise further at occurrence of a second metastasis. ctDNA yield and allele frequency were not capable of discriminating among CRC stages.

cfDNA amounts and KRAS wildtype copy numbers were very strongly correlated between Intplex PCR and Qubit and digital PCR, respectively, but only 50% of KRAS G12D mutations identified with Intplex were found at similar copy numbers by digital PCR.

Studies (2)

Purpose

The purpose of this study was to investigate the effect of extraction kit type, inclusion of carrier RNA, and using plasma versus serum on the isolation of cell-free DNA and circulating tumor DNA from healthy patients and those with CRC. Blood was obtained by venipuncture from eight healthy individuals and 38 CRC patients in S-Monovette tubes containing clot activator (serum) or K3EDTA (plasma) and centrifuged within 3 h of collection. Serum and plasma were obtained by centrifugation at 2000 x g for 10 min and transferred to 1.5 mL tubes. Specimens were centrifuged again at 16000 x g for 10 min, transferred to new tubes, and frozen at -80˚C. DNA was extracted from plasma and serum using the magnetic bead based Maxwell RSC ccfDNA Plasma kit and the silica-membrane based QIAamp Free Circulating Nucleic Acid kit with or without 1 µg carrier RNA. DNA carrying KRAS G12S or G12D mutation was spiked in at various concentrations to investigate the effects on allelic frequency. DNA was stored at -80˚C until analysis. DNA was analyzed by Qubit, Intplex allele-specific PCR, and droplet digital PCR amplification of KRAS.

Summary of Findings

No differences in total cfDNA yields from serum were found between silica membrane-based and magnetic bead-based kits by Qubit. While a slightly higher yield of fragments <600 bp was found using the silica-membrane-based rather than magnetic bead-based kits, there was also a higher yield of fragments >600 bp, indicating additional lymphocyte contamination. Consequently, a significantly higher yield of small cfDNA fragments compared to large (>600 bp) was only observed when magnetic bead-based kits were used and use of magnetic beads resulted in a higher percentage of fragments <600 bp. cfDNA yields using either kit type were unaffected by the use of carrier RNA.

As expected, use of serum resulted in 6.7-fold higher cfDNA yields than plasma when extracted using the magnetic bead-based kits and 3.7-fold higher using silica membrane-based kits. While quantified copies of G12D and G12S in serum were slightly higher than expected for both kit types, they were closer to expected values in plasma specimens but only when the magnetic bead-based systems were used for extraction. Quantification of spiked-in ctDNA by Intplex PCR showed lower than expected allele frequencies in serum, likely due to increased cfDNA.

  • Biospecimens

    FluidPlasma
    FluidSerum
  • Preservation Types

    Frozen
  • Diagnoses

    Normal
    NeoplasticCarcinoma
  • Platforms

    DNADigital PCR
    DNAFluorometry
    DNAReal-time qPCR
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen Aliquots and ComponentsBlood and blood products
    Plasma
    Serum
    Analyte Extraction and PurificationAnalyte isolation method
    Magnetic bead based extraction kit
    Silica-membrane based extraction kit
    With carrier RNA
    Without carrier RNA
    Real-time qPCR SpecificTechnology platform
    Digital PCR
    Intplex PCR
    Qubit