Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Optimization of sources of circulating cell-free DNA variability for downstream molecular analysis.

Author(s): Till JE, Black TA, Gentile C, Abdalla A, Wang Z, Sangha HK, Roth JJ, Sussman R, Yee SS, O'Hara MH, Thompson JC, Aggarwal C, Hwang WT, Elenitoba-Johnson KSJ, Carpenter EL

Publication: J Mol Diagn, 2021, Vol. , Page

PubMed ID: 34454115 PubMed Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   This study compared cfDNA concentration and KRAS G12/13 VAF among plasma obtained after two-step centrifugation at different speeds, after three-step centrifugation, and among fresh and frozen plasma specimens. cfDNA concentration was determined using three different cfDNA quantification methods (Qubit, real-time PCR, ddPCR) and results were compared for potential correlations. Blood was collected from 26 healthy patients and 24 patients with metastatic pancreatic ductal adenocarcinoma into Streck cell-free DNA BCTs. Unless otherwise specified, plasma was obtained by centrifugation at 1600 g for 15 min followed by a second centrifugation at 4100 g for 15 min and stored frozen at -80°C.  Plasma was thawed at room temperature before cfDNA extraction. Unless otherwise specified, cfDNA was extracted from plasma using the QIAsymphony DSP Circulating DNA Kit. cfDNA was quantified using each of the following methods: the Qubit dsDNA HS Assay Kit, real-time PCR amplification of a 115 bp amplicon of Alu, and ddPCR amplification of the RPP30 locus. KRAS mutation status was detected using the ddPCR KRAS G12/G13 Screening Assay Kit after preamplification with a threshold of 0.04%. To test potential effects of centrifugation speed, plasma from 24 specimens was centrifuged at low speed (4100 g) followed by high speed (16000 g). To investigate potential effects of the number of centrifugation steps, 24 specimens (12 from healthy patients and 12 from pancreatic cancer patients) underwent a third centrifugation at 4100 g for 15 min before storage at -80°C. To investigate potential effects of frozen storage, 24 specimens (12 from healthy patients and 12 from pancreatic cancer patients) were analyzed immediately after isolation and again after frozen storage at -80°C. Concordance between results was evaluated using Lin’s concordance correlation coefficient (CCC) or Spearmans rank correlation coefficient (ρ).

   Summary of Findings:

   The speed of the second centrifugation (4100 g versus 16000 g) and addition of a third centrifugation step (at 4100 g) had no effect on cfDNA concentration in plasma, regardless of quantification method. Importantly, the yield of cfDNA was strongly to very strongly correlated among samples obtained after a second centrifugation at 4100 g or 16000 g (CCC=0.981 for qPCR, CCC=0.982 for Qubit, and CCC=0.876 for ddPCR; P<0.001, all) and when processed by two-step versus three-step centrifugation (CCC=0.959 for qPCR, CCC=0.986 for Qubit, and CCC=0.996 for ddPCR; P<0.001, all).  Further, plasma specimens obtained at the two different centrifugation speeds had a similar VAF in the two cases where the variant was detected in both specimens (CCC=0.973, P<0.001); however, in the remaining two cases a VAF <0.1% was detected in the low-speed, but not the high speed, specimen. Similarly, plasma obtained using two- or three-step centrifugation protocols had similar VAFs in the nine cases where the variant was detectable in differentially processed specimens (CCC=0.998, p<0.001); although, there were two cases where the variant was detected at <0.1% in one specimen but not the other (one specimen for each centrifugation protocol).

   Measured concentrations of cfDNA were very strongly correlated when any two of the three quantification methods evaluated were compared (ρ=0.959 for qPCR versus Qubit, ρ=0.960 for ddPCR versus Qubit, and ρ=0.957 for qPCR versus ddPCR); the CV of each of the three quantification methods evaluated were also comparable. cfDNA concentrations were significantly higher when cfDNA was extracted from fresh rather than frozen plasma if quantified by Qubit (odds ratio, OR=1.206, P<0.001) or real-time PCR (OR=1.134, P=0.001), but cfDNA concentration was lower when fresh specimens were used for extraction if quantified by ddPCR (OR=0.933, P=0.01).  Regardless of measurement method, cfDNA concentrations were strongly correlated between fresh and frozen plasma (CCC=0.987 for qPCR, CCC=0.931 for Qubit, and CCC=0.990 for ddPCR; P<0.001, all). When the variant was detected in both specimens, the VAF ratios were comparable among frozen and fresh plasma (CCC=0.997, P<0.001). In four cases, the variant was detected in one specimen at low frequency (<0.1%) but undetectable in the remaining specimen (in 3 cases VAF was detected only in fresh specimens, and in 1 case it was detected only in frozen specimens).

   Biospecimens

   • Bodily Fluid - Plasma
   • Bodily Fluid - Blood

   Preservative Types

   • Frozen
   • None (Fresh)

   Diagnoses:

   • Normal
   • Neoplastic - Carcinoma

   Platform:

   Analyte	Technology Platform
   DNA	Digital PCR
   DNA	Fluorometry
   DNA	Real-time qPCR

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Biospecimen Aliquots and Components	Centrifugation	Different number of centrifugation steps compared Multiple speeds compared
   Biospecimen Preservation	Type of fixation/preservation	Frozen None (fresh)
   Digital PCR Specific	Targeted nucleic acid	KRAS G12/13
   Digital PCR Specific	Technology platform	Qubit Real-time PCR ddPCR

   View more