Objective To analyze the genetic variation characteristics of circulating tumor DNA (ctDNA) in patients with advanced lung cancer, aiming to provide valuable information for clinical decision-making. Methods In this retrospective study, a total of 468 patients with advanced-stage lung cancer were enrolled at West China Hospital, Sichuan University, from February 2017 to November 2019. Peripheral blood from each patient was collected for next generation sequencing (NGS) based on Illumina NextSeq550 platform, using a 23-gene targeted panel. Results Totally, 74.57% of patients were found to have at least one genetic alteration in plasma ctDNA, and drug-sensitive or drug-resistance information was obtained in 68.19% of patients. Frequently mutated genes included EGFR (47.01%), ROS1 (7.91%), ALK (7.69%), and HER2 (6.41%). EGFR mutation was more common in non-smokers (P<0.05), while KRAS mutation tended to occur more frequently in males (P<0.01) and smokers (P<0.01). EGFR 19DEL (28.88%), T790M (25.07%), and L858R (23.43%) were the most common types of EGFR mutations, and some rare potential drug-resistant mutations like K806R, S768I and L861Q, were also detected. Patients with brain metastasis had a relatively lower positive rate (69.23%). EGFR mutations were associated with bone metastasis (P<0.001), PIK3CA and KRAS mutations were enriched in liver metastasis (P<0.01 and P<0.001, respectively), while FGFR1 and HER2 mutations were more common in adrenal metastasis (P<0.01 and P<0.05, respectively). Third-generation EGFR tyrosine kinase inhibitor significantly increased the detection rate of FGFR1 genetic alterations in patient ctDNA when administered as second-line therapy (P<0.01). Conclusions The mutation profile and positive rate of plasma ctDNA are closely associated with smoking history, sex and metastatic sites. Comprehensive plasma ctDNA profiling using NGS technology demonstrates clinical utility by revealing therapeutically sensitive or resistant alterations in over half of the advanced-stage lung cancer patients, thereby informing personalized treatment strategies.
