This paper compared the percentage of circulating tumor DNA (ctDNA) in plasma obtained after delayed centrifugation of EDTA blood from non-small cell lung cancer (NSCLC) patients. The percentage of ctDNA recovered and the mutational profiles were compared between formalin-fixed paraffin-embedded (FFPE) biopsy specimens and matched ctDNA before and after size selection.
Rescue of Non-Informative Circulating Tumor DNA to Monitor the Mutational Landscape in NSCLC.
Mayer S, Schmidtke-Schrezenmeier G, Buske C, Rücker FG, Barth TFE, Möller P, Marienfeld R
Purpose
Conclusion
High molecular weight DNA indicative of genomic contamination was present in all specimens, but the percentage of genomic DNA was higher and percentage of circulating tumor DNA (ctDNA) lower in specimens where blood was processed >3 days after collection than in specimens processed within 24 h. Size selection using magnetic beads significantly increased the mean percentage of ctDNA independent of pre-selection percentage ctDNA and NSCLC stage. Sequencing found that all KRAS mutations identified in the non-size selected ctDNA were also found in the tumor and size selected ctDNA but additional KRAS and EGFR mutations were identified in size selected ctDNA and/or biopsy specimens. The overall concordance with the biopsy was similar for non-size selected and size selected ctDNA (70% and 71%, respectively) but the mutation specific concordance with the biopsy specimen increased from 39% without size selection to 79% after size selection. Importantly, size selection also tended to increase the variant allele frequency.
Studies (1)
Purpose
This study compared the percentage of ctDNA in plasma obtained after delayed centrifugation of EDTA blood from NSCLC patients. The percentage of ctDNA recovered and the mutational profiles were compared between FFPE biopsy specimens and matched ctDNA before and after size-selection. EDTA blood and FFPE biopsies were obtained from 51 patients (mean age 63.3 years) with NSCLC adenocarcinoma. Blood was stored at room temperature for <24 h (11 patients), 1 day (14 patients), 2 days (11 patients), 3 days (3 patients), 4 days (8 patients), or 7 days (1 patients), before isolation of plasma by centrifugation at 2500 x g for 10 min at room temperature. For the remaining three patients, the delay before plasma separation was not specified. Plasma was aliquoted and stored frozen at -80°C until DNA extraction. Details of biopsy procurement and processing were not provided but the authors state pathologists evaluated each biopsy. DNA was extracted from one to three 5 µm FFPE sections using the Qiagen FFPE DNA Mini Kit and from plasma using the QIAamp Circulating Nucleic Acid Kit. Extracted DNA was stored at -20°C. DNA was quantified using the Qubit dsDNA HS Assay Kit and fragment size evaluated using the Agilent High Sensitivity Kit on a bioanalyzer. To eliminate genomic contamination, DNA size selection was performed using Agencourt AMPure XP magnetic beads. Sequencing libraries were constructed using the Tumor Actionable Mutations Gene Read Kit and sequenced on a MiSeq.
Summary of Findings
High molecular weight DNA indicative of genomic contamination was present in all specimens (mean 69% of DNA, range 14-99%). The fraction of genomic DNA was higher and ctDNA lower in specimens where blood was processed >3 days after collection than in specimens processed within 24 h (77% versus 58%, P=0.042). After size selection using magnetic beads, the mean percentage of ctDNA was increased from 33% to 77% (P<0.0001). Importantly, the post-selection percentage was independent of pre-selection percentage ctDNA and NSCLC stage. For 47 of the 51 specimens, the post-selection percentage ctDNA was ≥58% and, for those with a low percentage of ctDNA, the procedure was repeated with a mean loss of only 11 ng ctDNA/mL plasma. Sequencing of 42 tumor specimens identified oncogenic EGFR mutations in five specimens, oncogenic KRAS mutations in 14 specimens, and zero mutations in the remaining 23 tumor specimens. Sequencing of the non-size selected ctDNA was only able to find six of the KRAS mutations and none of the EGFR mutations. However, after size selection, 10 specimens with KRAS mutations and four specimens with EGFR mutations were found by sequencing ctDNA. All six KRAS mutations identified in the non-size selected ctDNA were also identified in both the matched biopsy and the size selected ctDNA. The remaining KRAS mutations included two identified in both the biopsy and the size selected ctDNA (but not non-size selected), six found only in the biopsy specimen, and three found only in the size selected ctDNA. The EGFR mutations included three found in both the biopsy and the size selected ctDNA (but not non-size selected ctDNA), two found only in the biopsy, and one found only in the size selected ctDNA. The overall concordance with the biopsy was similar for non-size selected and size selected ctDNA (70% and 71%, respectively), but the mutation specific concordance increased from 39% without size selection to 79% after size selection. Importantly, size selection also tended to increase the VAF.
Biospecimens
FluidPlasmaFluidBloodTissueLungPreservation Types
FrozenFormalinDiagnoses
NeoplasticCarcinomaPlatforms
DNAFluorometryDNAAutomated electrophoresis/BioanalyzerDNANext generation sequencingPre-analytical Factors
Classification Pre-Analytical Factor Value(s) Preaquisition Prognostic factor Stage IAStage IIAStage IIBStage IIIAStage IIIBStage IVBiospecimen Aliquots and Components Centrifugation Centrifugation delays investigatedStorage Storage duration <24 h>3 daysNext generation sequencing Specific Template modification Magnetic bead size-selectionNo size selection