Abstract 5017: Analytical performance of contrived samples for validation of liquid biopsy assays

Amanda E. Harvey,Cindy K. Maddox,Sonomi O. Dillon,Omar T. Khawar,Ellen L. Verner,Amy Greer,Kenneth C. Valkenburg,Brian J. Caveney,Shakti Ramkissoon

DOI: https://doi.org/10.1158/1538-7445.am2024-5017

IF: 11.2

2024-03-24

Cancer Research

Abstract:Introduction: Cell-free DNA (cfDNA) extracted from plasma is critical for the analytical validation of liquid biopsy assays. cfDNA is comprised of genetic material from tumor and non-tumor cells, which can be highly variable across tumor types and disease stage. Relative to tumor tissue, cfDNA yield is lower and may not always capture all clinically relevant variants (e.g. gene fusions). Supplementing analytical validation studies with contrived samples prepared from cell line derived DNA to mimic the features of cfDNA allows for more robust testing of rare variants in clinical samples. To determine the commutability of contrived sample DNA to clinical plasma sample cfDNA, a retrospective analysis of sensitivity studies using two methods of contrived sample preparation was performed. Methods: The PGDx elio Plasma Focus assay was used to compare sample base size post library preparation and limit of detection (LoD) of clinically relevant variants. Both methods used fragmented cell line DNA with a variant of interest and for method #1, cell line DNA was spiked into noncancerous donor plasma, which was then extracted resulting in a cfDNA and cell line DNA mixture. This contrived sample was diluted using wild-type (WT) DNA extracted from noncancerous donor plasma to obtain a target variant allele frequency (VAF). Method #2 used fragmented cell line DNA diluted with fragmented WT cell line DNA to obtain a target VAF. This retrospective data analysis will include a minimum of 28 samples run in replicate (3-20) across multiple studies. Results: Data obtained from a DNA fragment analyzer post library preparation show similarities between fragmented cell line DNA spiked into plasma, fragmented cell line DNA, and clinical plasma cfDNA, with fragment peaks ranging from 290-320 base pair (bp). The median LoDs across the three sample types (n=28) were also similar for single nucleotide variants (SNVs) at VAFs of 0.70%, 0.76%, 0.72%, indels at VAFs of 0.70%, 1.0%, 0.78%, amplifications at folds of 1.4, 1.3, 1.5, and translocations at fusion read fractions of 0.5%, 0.61%, 0.55%. When comparing the same variants in a contrived sample functional characterization study, the call rates of various dilution levels were similar. A Fisher's Exact Test of the Hit Rate showed no significant difference for KRAS G12V SNV, ERBB2 amplification, and RET translocation between cell line DNA and clinical cfDNA. Conclusions: This data demonstrates that contrived samples perform with general equivalence to and may be an appropriate substitute for clinical plasma samples, though certain DNA sources may influence overall performance. Overall, fragmented cell line DNA spiked into plasma and fragmented cell line DNA diluted with WT fragmented cell line DNA performed similarly to clinical cfDNA samples representative of the intended use population. Citation Format: Amanda E. Harvey, Cindy K. Maddox, Sonomi O. Dillon, Omar T. Khawar, Ellen L. Verner, Amy Greer, Kenneth C. Valkenburg, Brian J. Caveney, Shakti Ramkissoon. Analytical performance of contrived samples for validation of liquid biopsy assays [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular s); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl) nr 5017.

oncology