were shared by most of the primary tumor regions, 28 were shared by most of the metastatic regions and 29 were ‘private mutations’ unique to specific regions. Only 34% of the mutations were present in all regions of the nephrectomy sample. For a second patient, six different areas of the nephrectomy and a liver metastasis were exome sequenced. A total of 119 somatic mutations were detected across the samples, but only 31% of the mutations were ubiquitous in the primary tumor and metastases (37% were detected in all regions of the primary renal cancer). These findings suggest that in such advanced renal cancers, multiple biopsies will be required to fully characterize all subclones in order to make therapeutic decisions.
The data from the two patients were used to construct phylogenetic trees of the tumor regions by clonal ordering, revealing in both cases branching tumor evolution and a common trunk (the original clone) containing a mutation in the von Hippel–Lindau tumor-suppressor gene (VHL). The first report of clonal evolution in an epithelial human malignancy7 had already shown that, of 30 mutations present in a metastasis, only 10 could be found in the primary tumor diagnosed 9.5 years earlier. Similar patterns of branched clonal evolution based on sequencing data