Application of ctDNA sequencing in patients with metastasized lung cancer.

e20033 Background: Circulating tumor DNA (ctDNA) is a form of cell-free DNA (cfDNA) found in blood that originates from tumor cells, which has emerged as a promising non-invasive biomarker in monitoring cancer prognosis and precision medicine with the advancement of deep sequencing. Here we compared the mutational landscape between solid tumor and cfDNA in metastasis-free and metastasized lung cancer patients using targeted NGS to investigate the feasibility of ctDNA sequencing in monitoring cancer prognosis. Methods: Blood-derived cfDNA samples and formalin-fixed paraffin embedded tissue were collected from 217 Chinese patients with metastatic lung cancer (n = 39) or metastasis-free lung cancer (n = 178). Blood samples were collected after surgery. Panel sequencing targeting 680 cancer-related genes was performed on both samples. Somatic single nucleotide variations and indels were analyzed. Results: Among the 178 metastasis-free patients, 89. 3% of the patients have at least one mutation both detected in tumor and blood samples. On average, 50. 4% of tumor mutations were detected in blood for each patients. EGFR, TP53, CSMD3, LRP1B and SYNE1 were the top five most frequently mutated genes, with the proportion of tumor mutation found in blood of 89. 1%, 80. 2%, 70. 0%, 61. 3% and 64. 3% respectively. In the metastasized cohort, all 39 individuals have at least one tumor mutations found in blood, achieving a concordance of 100%. On average, 59. 9% of tumor mutations were detected in blood for each patients. Notably, the proportion of tumor mutations detected in blood for each patients is significantly higher than the metastasis-free cohort (unpaired T-test, p = 0. 0167). EGFR, TP53, RBM10, PIK3CA and CSMD3 were the top five most frequently mutated genes, with the proportion of tumor mutation found in blood of 93. 5%, 80. 8%, 57. 1%, 83. 3% and 42. 9% respectively. Moreover, the proportion of patients having EGFR p. L858R mutation in metastasized cohort was higher than the metastasis-free cohort (54. 3% vs 32. 8%). While within the metastasized cohort, blood samples detected a higher proportion of EGFR p. T790M (12. 9% vs 7. 69%) but a lower proportion of EGFR exon19 deletion (p. E746A750del, 19. 4% vs 27. 7%) comparing to tumor sample, which can indicate drug resistance. Conclusions: We demonstrated a high mutation detection concordance between tumor and blood using targeted NGS, with an exceptional performance especially in driver oncogenes like EGFR and TP53. Our findings showed that metastasized lung cancer patients exhibit a higher positive concordance rate compared to metastasis-free cohort using ctDNA. We also showed a varied mutational landscape between ctDNA and tumor sequencing in metastasized lung cancer patients. Overall, this study not only reaffirms the viability of using blood as an alternative of tissue biopsy, but also provides new insights on application of ctDNA monitoring lung cancer metastasis and its clinically targetable mutations.