Abstract 4954: Clinical validation of a cell-free DNA liquid biopsy approach for noninvasive molecular profiling

Abstract

Abstract Clinical molecular profiling of advanced cancers enables the identification of actionable genomic alterations to guide therapeutic decisions. Although profiling of tissue samples is considered the gold standard, specimens may be unavailable or unsuitable for testing due to limited tumor purity or specimen quality. Further, genetically informed treatment decisions are increasingly necessary after disease progression and re-biopsy in this setting may not be feasible. Circulating tumor DNA (ctDNA) approaches for identification of genetic alterations in cancer patients may be more informative as the alterations reflect the current status of the tumor. ctDNA is representative of multiple tumor sites within a patient and may aid in the detection of alterations throughout the course of therapy. However, the fraction of ctDNA obtained from a blood sample is often very low (<1.0%) and difficult to detect. Additionally, many methods to evaluate ctDNA interrogate single hotspot or a few mutations. The next generation of ctDNA assays must identify clinically actionable genetic alterations and novel biomarkers with high precision and accuracy. To address these issues, we have developed and validated PlasmaSelect64, a ctDNA approach to comprehensively detect genetic alterations at low allele frequencies in the circulation of cancer patients. Utilizing digital genomic approaches, we demonstrated robust sensitivity and specificity in our CLIA laboratory in 64 well-established cancer genes that were identified based on clinical actionability. In addition to the evaluation of exons in 58 genes that are frequently mutated in cancer for sequence mutations, we performed a comprehensive genomic analysis of translocations in 18 genes and copy number analyses in 19 genes. We have also developed a novel approach for identification of microsatellite instability (MSI) using these error correction methodologies. To evaluate the PlasmaSelect64 approach, we developed and optimized the pre-analytical conditions for sample collection and processing with K2EDTA and Streck blood collection tubes. Analytical validation studies were performed with clinical samples and contrived cell-line mixtures containing known alterations determined by orthogonal methods. We robustly identified sequence mutations at 0.50% mutant allele frequency (MAF) with a limit of detection of 0.05% MAF, corresponding to a per-base specificity of 99.9997% and a sensitivity of 99.4%. For detection of focal amplifications and translocations, analytical method validation studies demonstrated sensitivity of 97.2% and 94.4% and specificity of >99% at MAFs of ≥20% and ≥0.50%, respectively. PlasmaSelect64 provides a non-invasive platform to enable detection of clinically relevant genetic alterations across a large number of genomic regions to aid in the therapeutic management of cancer patients. Citation Format: Monica Nesselbush, Samuel Angiuoli, Luis A. Diaz, Andrew Georgiadis, Shannon Glynn, Siân Jones, Laurel Keefer, Peter LoVerso, Derek Murphy, Sonya Parpart-Li, David Riley, Naomi Sengamalay, Manish Shukla, John Simmons, Snehal Talati, Rebecca Steinberg, Laura Tucker, Victor E. Velculescu, Ellen Verner, Angela Villarta, Mark Sausen. Clinical validation of a cell-free DNA liquid biopsy approach for noninvasive molecular profiling [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 4954. doi:10.1158/1538-7445.AM2017-4954