Aengus Stewart, Aron Charles Eklund, Graham Clark, Lisa Pickering, Nicholas Matthews, Keiran Raine, Gordon Stamp, Martin Gore, Patrick S. Tarpey, Stuart Horswell, David Endesfelder, Andrew Rowan, Bradley Spencer-Dene, Marco Gerlinger, Calli Latimer, Eva Grönroos, James Larkin, Adam Butler, Charles Swanton, Zoltan Szallasi, Neil Q. McDonald, P. Andrew Futreal, Claudio R. Santos, Sharmin Begum, Julian Downward, Ignacio Varela, David T. Jones, Pierre Martinez, Benjamin Phillimore, and Mahrokh Nohadani
Background: Genetic intra-tumor heterogeneity (ITH) may foster tumor adaptation by providing selectable phenotypes for Darwinian evolution. Extensive genetic heterogeneity may also hinder personalized medicine strategies that rely on the portrayal of the mutational landscape from single tumor biopsies. Methods: To examine ITH, we have subjected multiple biopsies from primary renal cell carcinomas and associated metastatic sites to exome capture sequencing (n=2), SNP-array based chromosomal aberration and ploidy profiling analysis (n=4). Phylogenetic relationships of tumor regions were reconstructed by clonal ordering of non-synonymous somatic mutations. The phenotypic consequences of genetic ITH were characterised by immunohistochemistry, mutation functional analysis and mRNA expression profiling. Results: Phylogenetic reconstruction revealed branched evolutionary tumor growth with 63-69% of somatic mutations identified from single biopsies not detectable across all sequenced tumor regions. Independent tumor suppressor gene loss of function mutations with distinct regional distributions were detected within individual tumors: SETD2 harbored 5 different mutations in 2 tumors and PTEN and KDM5C two different mutations in one tumor, each. Thus, despite genetic divergence during tumor progression, phenotypic convergent evolution occurred, indicating a high degree of early mutational diversity. ITH was observed for a mutation in the kinase domain of mTOR, correlating with S6 and 4EBP phosphorylation specifically in cancer regions carrying the mutation and constitutive activation of mTOR kinase activity in vitro. Expression of a renal cancer-specific prognostic signature differed between tumor regions. Chromosomal aberration analysis revealed extensive ITH with 26/30 tumor biopsies from four tumors harboring divergent allelic imbalance profiles. Ploidy profiling revealed heterogeneity in two out of four tumors and identified an aneuploid tumor cell population in a metastasis that probably evolved from a tetraploid intermediate detectable in the primary tumor. Conclusions: Genetic ITH was present in all tumors and occurs through spatially separated heterogeneous somatic mutations and chromosomal and ploidy aberrations leading to both, phenotypic intra-tumor diversity (mTOR activating mutation) and convergent loss of function(SETD2, PTEN and KDM5C). ITH can lead to underestimation of the tumor genomic landscape portrayed from single biopsies and may present significant challenges to personalized medicine and biomarker development. ITH, associated with heterogeneous protein function, may foster tumor adaptation and therapeutic failure through Darwinian selection. Targeting mutations commonly located on the trunks, rather than the branches of phylogenetic trees, may improve therapeutic outcomes. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 964. doi:1538-7445.AM2012-964