Abstract A17: Pan-solid tumor comparison of variant detection in paired liquid and tissue biopsies

Abstract

Abstract Comprehensive genomic profiling (CGP) of circulating tumor DNA (ctDNA) provides an opportunity to noninvasively monitor a patient’s tumor burden via liquid biopsy. Liquid biopsies can assess a patient’s mutational landscape through time and provide information on treatment response and relapse. It has become increasingly common to have paired genomic data analysis between liquid and tissue biopsies from the same patient. However, the impact of clinical, temporal, and biologic factors on percentage of positive agreement (PPA) of variant detection between these biopsies is unclear. For our study, we leveraged two databases containing CGP data from paired liquid and tissue biopsy specimens tested by Foundation Medicine (FMI): the FMI database (>1700 paired samples) and the Flatiron Health-Foundation Medicine Clinico-Genomic Database (CGDB). The CGDB is a subset of the FMI database linked with the Flatiron Health nationwide de-identified EHR-derived database, which includes demographic, treatment, and clinical outcomes information (>500 paired samples). We report pan-solid tumor and per-indication prevalence and PPA for variants commonly assayed in both liquid and tissue tests. We found that PPA depended on the tumor DNA concentration in plasma and tissue biopsies, as well as the amount of time between tests. We also investigated the effect of treatments administered in the time between liquid and tissue tests on the detected variants and observed some well-described resistance alterations at a higher frequency in liquid samples, including a higher prevalence of ESR1 point mutations in breast cancer, more EGFR T790M alterations in lung cancer, and frequent KRAS Q61H alterations in colorectal cancer. Liquid biopsies can also contain DNA shed from multiple metastatic sites. We observed evidence of this in the form of polyclonal resistance alterations, which may also account for differences in PPA over time. Our findings indicate that while PPA is generally high between samples, it may be influenced by factors such as intervening therapies, resistance, therapy efficacy, and polyclonality of liquid samples. Citation Format: Zoe June F. Assaf, Smruthy Sivakumar, Dexter X. Jin, Sophia L. Maund, Svetlana Lyalina, Guneet Walia, Ethan S. Sokol, Sally E. Trabucco. Pan-solid tumor comparison of variant detection in paired liquid and tissue biopsies [abstract]. In: Proceedings of the AACR Special Conference on Advances in Liquid Biopsies; Jan 13-16, 2020; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2020;26(11_Suppl):Abstract nr A17.