Validation of ctDNA Quality Control Materials Through a Precompetitive Collaboration of the Foundation for the National Institutes of Health. (2021)

• Record Type: Journal Article
• Authors: Williams, P. Mickey; Forbes, Thomas; P. Lund, Steven; Cole, Kenneth D.; He, Hua-Jun; Karlovich, Chris; Paweletz, Cloud P.; Stetson, Daniel; Yee, Laura M.; Connors, Dana E.; Keating, Susan M.; Destenaves, Benoit; Cleveland, Megan H.; Lau, Christie J.; Barrett, J. Carl; Kelloff, Gary J.; McCormack, Robert T.
• Is Part Of: JCO precision oncology. Volume 5(2021)
• Publication Date: 2021
• Subjects: Precision Medicine; Neoplasms; Pharmacogenetics; Molecular Targeted Therapy; Personalized medicine; Oncology; Pharmacogenomics; Periodical; Periodicals; 616.994
• DOI: 10.1200/PO.20.00528
• Languages: English
• ISSNs: 2473-4284

Abstract

Abstract : PURPOSE: We report the results from a Foundation for the National Institutes of Health Biomarkers Consortium project to address the absence of well-validated quality control materials (QCMs) for circulating tumor DNA (ctDNA) testing. This absence is considered a cause of variance and inconsistencies in translating ctDNA results into clinical actions. METHODS: In this phase I study, QCMs with 14 clinically relevant mutations representing single nucleotide variants, insertions or deletions (indels), translocations, and copy number variants were sourced from three commercial manufacturers with variant allele frequencies (VAFs) of 5%, 2.5%, 1%, 0.1%, and 0%. Four laboratories tested samples in quadruplicate using two allele-specific droplet digital polymerase chain reaction and three (amplicon and hybrid capture) next-generation sequencing (NGS) panels. RESULTS: The two droplet digital polymerase chain reaction assays reported VAF values very close to the manufacturers' claimed concentrations for all QCMs. NGS assays reported most single nucleotide variants and indels, but not translocations, close to the expected VAF values. Notably, two NGS assays reported lower VAF than expected for all translocations in all QCM mixtures, possibly related to technical challenges detecting these variants. The ability to call ERBB2 copy number amplifications varied across assays. All three QCMs provided valuable insight into assay precision. Each assay across all variant types demonstrated dropouts at 0.1%, suggesting that the QCM can serve for testing of an assay's limit of detection with confidence claims for specific variants. CONCLUSION: These results support the utility of the QCM in testing ctDNA assay analytical performance. However, unique designs and manufacturing methods for the QCM, and variations in a laboratory's testing configuration, may require testing of multiple QCMs to find the best reagents for accurate result interpretation.