Analytic and Clinical Validation of an Ultrasensitive, Quantitative Polymerase Chain Reaction Assay for EGFR Mutation Analysis With Circulating Tumor DNA

Abstract

Context.— The mutation analysis of epidermal growth factor receptor (EGFR) has become a common test to guide therapeutic decision making for lung cancer. Molecular testing with circulating tumor DNA in plasma allows diagnosis of mutations when tumor tissue is not available as well as monitoring treatment response with repeat biopsies. Objectives.— To develop a timely and cost-effective assay that can accurately detect EGFR mutations in circulating tumor DNA and to evaluate the analytic and clinical performance of the assay. Design.— Analytic assessment was conducted with a set of reference materials carrying classic EGFR mutations. A recently developed Poisson distribution–based approach was employed to understand the assay sensitivity. Clinical evaluation was performed with 224 pairs of plasma and matched tissues from patients with stage I to IV disease. EGFR mutation rates of 390 consecutive plasma samples processed in the central service laboratory were compared with previously reported prevalence in an Asian population. Results.— Our results suggested that limit of detection for the EGFR quantitative polymerase chain reaction assay was 10 mutation copies, and the lowest detectable copy numbers could be extended to a single-digit level. The clinical sensitivity was 53.3% for all stages combined and 81.4% for late stages, with a high specificity of 100%. Clinical observations showed an overall positive finding rate of 32.5% and 41.4% for stage IV disease, which is consistent with previously reported EGFR mutation prevalence in an Asian population. Conclusions.— Our results supported the clinical utility of the ultrasensitive, quantitative polymerase chain reaction assay for EGFR mutation analysis with circulating tumor DNA.