MS02.02 Liquid Biopsy: Who, When, What and How

Liquid biopsy (LB) refers to a multitude of biomarkers that can be isolated with minimally invasive methods from human body fluids (blood, saliva, urine, ascites, pleural effusion, etc.) that include cell free DNA (cfDNA), circulating tumor cells (CTCs), microRNAs (miRNAs), long non coding RNAs (lncRNAs), and exosomes (Fig. 1). Plasma genotyping of cfDNA entered clinical practice in non-small cell lung cancer (NSCLC) for detection of EGFRmutations in both treatment-naïve and EGFR-mutated NSCLC after progression to EGFR tyrosine kinase inhibitors (TKIs) as an alternative source to tissue when histological samples are insufficient or when a biopsy is not feasible. Indeed according to the LB IASLC statement1, plasma cfDNA analysis should be offered to the same population candidate for molecular testing using DNA isolated from tissue (all non-squamous NSCLC subjects, or adenosquamous or in patients with clinical features suggestive of the presence of a molecular driver) in cases with insufficient tumor tissue specimens or where tissue specimens are not obtainable. Moreover, LB is indicated for the identification of acquired resistance mutations in EGFR-mutated NSCLCs progressing during treatment with first- or second-generation EGFR TKIs, reserving tissue re-biopsy in case of negative or inconclusive results. The PCR-based cobas EGFR Mutation Test v2 was the first FDA approved LB test for NSCLC, although several limitations on sensitivity of this technique. Fortunately, LB is a rapidly evolving field and several commercial and “in house”NGS platforms have been developed, allowing a more comprehensive plasma genotyping that include genetic rearrangements (such as ALK, ROS1, RET, and NTRK) as well as other relevant oncogene drivers, including mutations of BRAF, HER2, MET, and KRAS. Some of the vendors have a Medicare reimbursement in United States. Recently, the multicenter prospective Noninvasive versus Invasive Lung Evaluation (NILE) study demonstrated that a validated and highly sensitive plasma 73-gene NGS test (Guardant360) used at the time of diagnosis of NSCLC was non-inferior to standard-of-care tissue genotyping in identifying guideline-recommended genomic biomarkers, allowing a guideline-complete genotyping in a higher proportion of patients with a shorter median turnaround time2. These results support the rationale for a “blood-first” approach, reserving tissue for PD-L1 immunohistochemistry or in case of negative liquid biopsy testing. In addition to cfDNA, cfRNA is a novel approach to enhance the comprehensive analysis of circulating biormakers. Furthermore, the increasing use of more sensitive detection methods, such as NGS, poses novel technical and biological challenges in our clinical practice, including the identification of non tumor-related mutations due to clonal hematopoiesis (CH), increased risk of false positive in presence of low variation allelic fraction (VAF), need for standardization and validation of the analytical methods, definition of requirements for appropriate report and interpretation of the results. Novel technologies such as CH-filtering ultradeep NGS have been tested with promising results3. Moreover, the adoption of molecular tumor board is essential for helping oncologists in interpreting NGS testing results4, using evidence-based scales, such as OncoKB and ESCAT tiers5,6. Recently, LB use has been advocated in immunotherapy-treated patients as a minimally invasive method that can allow a real time monitoring of treatment response and interpretation of challenging radiographic situations, overcoming the limits of conventional radiological assessment methods7,8. Further prospective studies are needed to better clarify the role of cfDNA as a predictive biomarker for immune checkpoint blockage in NSCLC. LB can also allow the estimation of tumor mutational burden (TMB) in plasma, proving a valuable alternative to tissue. Exploratory analyses of two large randomized phase III studies explored the potential utility of blood TMB (bTMB) as predictive biomarker for immunotherapy, using two different platforms (Foundation Medicine and GuardantOMNI)9,10. The results of these studies support bTMB as potential biomarker for immunotherapy in NSCLC. Further prospective studies will clarify the role of bTMB in treatment selection of patient candidate for immunotherapy, as well as the optimal cut-off value, the minimum number of genes necessary for TMB estimation, and the specific mutations useful. References: Rolfo C, et al. J Thorac Oncol 2018 Leighl N, et al. Clin Cancer Res 2019 Li BT, et al. Ann Oncol 2019 Rolfo C, et al. ESMO Open 2018 Chakravarty D, et al. JCO Precis Oncol 2017 Mateo J, et al. Ann Oncol 2018 Anagnostou V, et al. Cancer Res. 2019 Guibert N, et al. ASCO-SITC 2019 Peters S, et al. AACR Annual meeting 2019 Gandara DR, et al. Nat Med 2018 [Figure presented] Keywords: Liquid Biopsy, cfDNA, NGS