Overall our comparison between primary and xenograft samples reveals evidence of tumor heterogeneity and the presence of sub-clonal cell populations. The mutations identified are mostly non-synonymous and some can have direct clinical impact as they are markers of poor prognosis or predictive of drug resistance. Interestingly, the colon primary sample presented the most complex and rich mutational profile, which, with the exception of a few mutations, was faithfully matched in a xenograft after seven successive passages in mouse. While other possible somatic changes, such as copy number alterations, were not assessed, this observation supports the use of xenograft models to reflect the genetics of the primary tumor, in agreement with previous studies [12].
