부산대학교 도서관

Abstract Background Although the efficacy of molecularly target agents in vitro, their use in routine setting is limited mainly to the use of anti‐HER2 and antiEGFR agents in vivo. Moreover, core biopsy of a single cancer site may not be representative of the whole expanding clones and cancer molecular profile at relapse may differ with respect to the primary tumor. Methods We assessed the status of a large panel of cancer driver genes by cell‐free DNA (cfDNA) analysis in a cohort of 68 patients with 13 different solid tumors at disease progression. Whenever possible, a second cfDNA analysis was performed after a mean of 2.5 months, in order to confirm the identified clone(s) and to check the correlation with clinical evolution. Results The approach was able to identify clones plausibly involved in the disease progression mechanism in about 65% of cases. A mean of 1.4 mutated genes (range 1‐3) for each tumor was found. Point mutations in TP53, PIK3CA, and KRAS and copy number variations in FGFR3 were the gene alterations more commonly observed, with a rate of 48%, 20%, 16%, and 20%, respectively. Two‐points‐Next‐Generation Sequencing (NGS) analysis demonstrated statistically significant correlation between allele frequency variation and clinical outcome (P = .026). Conclusions Irrespective of the primary tumor mutational burden, few mutated genes are present at disease progression. Clinical outcome is consistent with variation of allele frequency of specific clones indicating that cfDNA two‐point‐NGS analysis of cancer driver genes could be an efficacy tool for precision oncology.