Abstract

Lung cancer is one of the most prevalent and lethal malignancies worldwide. Despite recent advancements in precision medicine, early detection and therapeutic monitoring of lung cancer remain challenging. Here, we present an Endonuclease-Assisted Selective Exponential Amplification (ESEA) platform that selectively depletes wild-type alleles through programmable endonuclease digestion (CRISPR-Cas or restriction enzymes) while simultaneously amplifying mutant alleles, enabling robust and cost-effective detection at mutant allele frequencies (MAF) as low as 0.00625%. To address the PAM-site limitations inherent in CRISPR-based enrichment strategies, we introduce a primer-directed restriction site programming approach that expands the theoretical coverage to over 94% of mutations listed in COSMIC. The ESEA system offers high sensitivity and cost-effectiveness compared with conventional methods, and enables multiplexed detection of key lung cancer hotspot mutations, including EGFR L858R, EGFR exon 19 deletions, EGFR T790M, BRAF V600E, as well as hotspots in KRAS and PIK3CA. In a small-sample-size test, the ESEA system achieved 100% sensitivity and specificity in pleural effusion samples (n=5) and a 100% ctDNA detection rate in patients with extracranial lesions and disease progression (n=6). These results highlight its potential as a cost-effective, highly sensitive, and robust platform for dynamic, real-time companion diagnostics, as well as non-invasive monitoring of treatment response and tumor evolution.