Real-World Data of the Correlation between EGFR Determination by Liquid Biopsy in Non-squamous Non-small Cell Lung Cancer (NSCLC) and the EGFR Profile in Tumor Biopsy.

Abstract

EGFR-mutated non-small cell lung cancer (NSCLC) has significant improved outcomes when treated with EGFR-tyrosine kinase inhibitors (TKI). Thus, EGFR-mutational status should be assessed at diagnosis and in the course of treatment with TKI. However, tissue samples are not always evaluable, and molecular profiling has been increasingly performed in cell-free tumor DNA (ctDNA) from blood samples. Our objective is to evaluate the reliability of ctDNA profiling in plasma samples in a real-world setting. We retrospectively analyzed the patients diagnosed with non-squamous NSCLC from May 2016 to December 2017 at Hospital Universitario Doctor Peset who had been tested for EGFR mutations in tissue and plasma samples. Both samples were sent to an external laboratory to perform the analysis by the cobas® EGFR assay. Percentage of agreement and concordance were calculated by kappa statistic. Of 102 patients reviewed, 89 were eligible. The overall EGFR mutation frequency was 18.6% for the evaluable tissue samples and 19.6% for evaluable plasma samples. Mutation status concordance between matched samples was 87.4%. Cohen's kappa index (κ) = 0.6 (sensitivity 70.6%, specificity 91.7%, positive predictive value 66.7%, negative predictive value 93%). When concordance was stablished only in stage IV tumors κ = 0.7, suggesting a higher agreement in advanced disease. This real-world data suggest that plasma is a feasible sample for ctDNA EGFR mutation assessment. Results of ctDNA molecular profiling are reliable when using a validated technique such as the cobas® EGFR assay, especially in patients that cannot undergo a tissue biopsy.