Circulating Tumor DNA Measurement Provides Reliable Mutation Detection in Mice with Human Lung Cancer Xenografts

Ling Wei,Li Xie,Xingwu Wang,Hongxin Ma,Liyan Lv,Lisheng Liu,Xianrang Song
DOI: https://doi.org/10.1038/s41374-018-0041-8
IF: 5.511
2018-01-01
Laboratory Investigation
Abstract:Genotype-directed targeted therapy has become one of the standard treatment options for non-small cell lung cancer (NSCLC). There have been numerous limitations associated with mutation analysis of tissue samples. Consequently, mutational profile analysis of circulating cell-free DNA (cfDNA) by highly sensitive droplet digital PCR (ddPCR) assay has been developed. Possibly due to differences in cfDNA concentrations, previous studies have shown numerous discrepancies in mutation detection consistency between tissue and cfDNA. In order to rigorously analyze the amount of cfDNA needed, we constructed 72 athymic nude mice xenografted with NCI-H1975 (harboring a EGFR T790M mutation) or NCI-H460 (harboring a KRAS Q61H mutation) human NSCLC. We thoroughly investigated the relationship between plasma cfDNA using Q-PCR targeting human long interspersed nuclear element-1 ( LINE-1 ) retrotransposon and the mouse ACTB gene, and the accuracy of mutation detection by ddPCR at different times post-graft. Our results show that the concentration and fragmentation of human (tumor) derived cfDNA (hctDNA) were positively correlated with tumor weight, but not with mouse-derived cfDNA (mcfDNA). Quantification of cfDNA by Q-PCR depends on the amplified target length. Mutation copies in plasma of per milliliter were positively linked to tumor weight, hctDNA level and hctDNA/mcfDNA ratio, respectively. Furthermore, tumor weight, hctDNA level and ratio of hctDNA/mcfDNA were significantly higher in cfDNA mutation-positive mice than in negative mice. Also, our data indicate that when plasma hctDNA level and hctDNA/mcfDNA ratio reach a certain level in xenografted mice, plasma cfDNA mutation can be detected. In summary, the present study suggests that determination of ctDNA levels may be essential for reliable mutation detection by analysis of cfDNA.
What problem does this paper attempt to address?