Evaluation of the Idylla system to detect the EGFR T790M mutation using extracted DNA

During the last few years, detection of epidermal growth-factor-receptor (EGFR)-activating mutations has become a routine part of clinical practice because of their importance in choosing the optimal treatment strategy for non-small-cell lung cancers (NSCLCs). The emergence of third-generation EGFR-tyrosine-kinase inhibitors required the implementation of sensitive methods to detect the subclonal EGFR mutation. Clinical implications make it essential to rapidly search for the T790M mutation, which is a real challenge for laboratories. The aim of this study was to compare performances of next-generation sequencing (NGS), one of the most frequently used molecular biology methods, and Idylla EGFR-Mutation Assay (henceforth Idylla), a fully automated real-time polymerase chain reaction (PCR) that is increasingly used in pathology laboratories, to detect the EGFR mutation using DNA. This retrospective study used 47 DNA samples extracted from NSCLC biopsies that previous NGS identified as: 29 harboring EGFR and T790M resistance mutations, 11 EGFR-activating mutation without T790 M and 7 wild-type EGFR. EGFR limit-of-detection (LOD) experiments used a commercial DNA known to harbor that mutation. Idylla detected primary EGFR-activating mutations and the T790 M mutation in 97.5 % and 65.5 % of the cases, respectively. The results of this retrospective analysis and LOD experiments showed that the Idylla should only be used to detect EGFR mutations in samples with > 25 ng of DNA and > 10 % tumor cells. Idylla was able to rapidly detect EGFR-activating mutations but detecting subclone mutations, like T790M, with < 25 ng of good-quality DNA or < 10 % tumor cells (variant allele frequency below the assay's validated LOD) was not always reliable.