P1.29: “Real World” Use of Liquid Biopsy in Patients With Lung Adenocarcinoma and Correlation With Tumor Tissue Genetic Profile

Abstract

Several genomic abnormalities have been discovered in adenocarcinoma of the lung (adenoca) in the last year; however, adequate quantity of tumor tissue for molecular analysis is a major handicap to offer most of the patients (pts) a personalized medicine. Reasons such as difficulty to perform tumor biopsy (TBx) either because of tumor location or patient’s co-morbid conditions are among the common factors for lack of molecular profiling results (MPR). Hence, liquid biopsy (LBx) has emerged as a potential alternative to detect these genomic alterations. We analyzed 81 consecutive pts to whom a LBx using Guardant 360 test was ordered in our thoracic oncology clinics at Lynn Cancer Institute, Boca Raton, Florida and Memorial Cancer Institute, Hollywood, Florida. Results from tissue molecular profiling from each subject was obtained or recovered for comparison. MPRs from this cohort was developed by different CLIA laboratories (e.g. Response Genetics, Caris, Foundation Medicine, BioTheranostics, and Genoptix). For LBx analysis, only Guardant 360 test was considered. The Guardant 360 test assays a panel of 70 genes (see Fig. 1) to identify genomic alterations in cancer-associated somatic variants with high sensitivity. Cell-free DNA (cfDNA) is extracted from plasma and genomic alterations are analyzed by massively parallel sequencing of amplified target genes. The distribution by gender was 56 female patients and 25 males; median age 71 (range, 27-99). 65/81 pts (80%) had at least 1 genomic alteration by LBx (range, 1-10). Most common abnormalities found in LBx were: TP53 (32 pts), EGFR (27 pts), NF1 (16 pts), KRAS (11 pts), MET (10 pts). From this 65 pts with + LBx results, 49 pts (75%) had tumor tissue MPRs for comparison. Major reason for lack of tumor tissue MPRs: insufficient tumor (18/81; 22%). For comparison between the 2 modalities, we considered all pts with available results in both tests; hence, 63 pts were used to compare TBx with LBx. 33 pts out of 63 (52%) had at least 1 similar genomic abnormality or MPRs found in both TBx and LBx. Most of the concordance was in EGFR alterations (17/22; 77%). LBx caught 10 additional EGFR genomic aberrations not being identified by TBx (a total of 27 EGFR genomics aberrations were identified in LBx). 14 EGFR found in LBx were actionable; 3/10 of mutant EGFR found only in LBx were actionable. LBx using Guardant 360 evaluation offers an alternative to identify genomic alterations including actionable mutations; still, insufficient tumor is the major reason for lacking of tumor MPRs and more advances to obtain TBx are needed. Our cohort had 48% concordance between LBx and TBx.