Abstract
The identification of epidermal growth factor receptor (EGFR) tyrosine kinase (TK) domain mutations in non-small cell lung cancer (NSCLC) patients is crucial for therapeutic decision-making and monitoring EGFR-tyrosine kinase inhibitor (TKI) resistance. Liquid biopsy has emerged as a promising alternative for patients ineligible for invasive tissue sampling. This study investigated the clinical utility of a novel chip-based digital PCR (dPCR) platform for detecting two important EGFR mutations - exon 19 deletions (19del) and threonine-methionine amino acid substitution at position 790 (T790M) - in serum samples, while exploring potential serum biomarkers for mutation prediction. The collection of 350 serum samples were conducted on patients diagnosed with NSCLC at Huashan Hospital between August 2023 and February 2024. Cell-free deoxyribonucleic acid (cfDNA) was extracted from serum and was analyzed for EGFR mutations using dPCR. The serum tumor marker levels were quantified. The dPCR assay demonstrated positive predictive values of 73.33% for 19del and 28.57% for T790M. Biomarker analysis revealed a carbohydrate antigen (CA) 199 cutoff of 11.75 U/mL (AUC = 0.707, 95% CI: 0.573-0.841, P = 0.005) for 19del detection, while progastrin-releasing peptide (ProGRP) showed a cutoff of 45.15 pg/mL (AUC = 0.628, 95% CI: 0.521-0.735, P = 0.028) for T790M identification. Variant rate exhibited significant positive correlations with biomarker concentrations: 19del variant rates significantly associated with CA125 levels (r = 0.624, P = 0.010), while T790M correlated with both carcinoembryonic antigen (CEA) (r = 0.531, P = 0.004) and ProGRP (r = 0.395, P = 0.041) in mutation-positive cohorts. These findings indicate that serum-based dPCR liquid biopsy demonstrates potential clinical utility as a supplementary approach to tissue biopsy for NSCLC genotyping. Notably, elevated serum tumor marker levels correlate with enhanced mutation detection rates in liquid biopsy, implying their potential supplementary value in prioritizing patients for molecular profiling.