High Incidence of False Positives in EGFR S768I Mutation Detection Using the Idylla qPCR System in Non-Small Cell...

Background/Objectives: The accurate detection of EGFR mutations, particularly the rare S768I variant, is crucial for guiding treatment decisions in non-small cell lung cancer (NSCLC) patients. This study investigated the incidence of false positives in S768I mutation detection using the Idylla^TM qPCR system and compared results with next-generation sequencing (NGS). Methods: A prospective observational study was conducted at the Dr. Negrín University Hospital between July 2023 and July 2024. Six NSCLC patient samples with S768I variant detection by Idylla^TM were analyzed from all NSCLC cases tested during the study period. Initial testing was performed on tissue samples (Idylla1), followed by replicate analysis using extracted DNA (Idylla2). Results were compared with NGS as the reference method. Statistical analysis included the calculation of sensitivity, specificity, accuracy, and Kappa concordance index. Results: Initial Idylla testing showed an 80% false positive rate, with only one of five positive results confirmed by NGS. The first analysis demonstrated high sensitivity (100%) but low specificity (20%), with an accuracy of 0.333 and poor concordance with NGS (Kappa = 0.077). Repeat testing using extracted DNA showed improved performance, with increased accuracy (0.833) and better agreement with NGS (Kappa = 0.571). Analysis of amplification curves revealed that false positives typically showed normalized fluorescence values below 12 points, with no clear correlation between false positives and factors such as sample quantity or tumor content. Conclusions: While the Idylla^TM system shows high sensitivity for S768I detection, its initial specificity is problematic, leading to frequent false positives. These findings emphasize the importance of confirming positive S768I results through alternative methods like NGS, particularly when these results could influence therapeutic decisions. Results suggest the need to refine the system's interpretation algorithms to improve specificity.