POSTER Presented at 2020 AMP Annual Meeting & Expo, November 2020
Introduction: Liquid biopsies, especially those that use plasma ctDNA,
are emerging as a powerful complement, and, in some cases, alternative
to solid tumor biopsies for the molecular characterization of cancer. Nonsmall
cell lung cancer (NSCLC) has proven particularly amenable to liquid
biopsies due to the availability of an FDA-approved tyrosine kinase
inhibitor for patients whose tumors have EGFR exon 19 deletions or exon
21 L858R mutations, as well as the need to monitor disease progression
and treatment response. However, collection of ctDNA is challenged by
the low abundance of ctDNA in blood, the potential for contamination by
genomic DNA, as well as other factors. These challenges can lead to false
negative results that limit utility of liquid biopsy. Methods: We compared
the recovery of ctDNA using a commercial column-based approach
(Roche) to a novel magnetic hydrogel particle (Ceres Nanosciences)
method. We created contrived liquid biopsy specimens by spiking
fragmented EGFR wild-type and mutant DNA sequences into pooled
donor plasma at varying concentrations. Recovered DNA was quantitated
and tested for EGFR mutations using a semi-quantitative multiplex PCR
assay (COBAS 4800). The performance of these 2 ctDNA extraction
methods were further compared using plasma derived from NSCLC
patients with known mutational profiles. Results: In the contrived
specimens, the column-based approach recovered 48% to 63% of spiked
DNA, whereas the magnetic hydrogel particles recovered 79% to 84% of
spiked DNA. Importantly, the hydrogel particle-based approach prevented
genomic DNA contamination for up to 24 hours post-collection. Extracted
ctDNA from both approaches demonstrated comparable performance in
the detection of EGFR mutations, with EGFR exon19del, L858R, and
T790M mutation detection possible at plasma DNA concentrations of 0.5,
4, and 20 ng/mL, respectively. Patient data demonstrated concordance
between ctDNA and tissue analysis in most cases, regardless of ctDNA
extraction method. ctDNA analysis revealed an EGFR L858R mutation not
detected by solid tumor testing. Conclusions: The magnetic hydrogel
particle-based ctDNA extraction technique was highly efficient in
recovering ctDNA from plasma. Moreover, the hydrogel particle-based
technique required less starting material (1 mL plasma) than column based
approaches with similar performance in downstream applications.
These results demonstrate the potential benefit of a magnetic hydrogel
particle-based ctDNA extraction technique in liquid biopsy testing of
NSCLC patients. On-going work is focused on integrating this extraction
technique into a multigene liquid biopsy assay.
