However, in order to avoid cross-contamination, it is essential to initiate the tumor organoid culture from a pure tumor population and/or use selective culture conditions. CRC lesions are generally well defined which

allows the pathologist to exclude potentially contaminating normal SD-208 mouse epithelium. Theoretically, selective culture conditions can be applied for the majority of CRCs given the high penetrance of activating Wnt pathway mutations [31 and 32]. Indeed mouse intestinal organoids with genetically inactivated Apc grow in the absence of Wnt or R-spondin-1, whereas wild-type organoids do not [ 23••, 37 and 38]. Likewise, this selection pressure can be applied to most CRC organoids by withdrawing R-spondin-1 [ 23••] or Wnt. Since EGF is dispensable for growing a different subset of CRC organoids (presumably with KRAS or BRAF mutations) [ 23••], withdrawal of this growth factor or addition SGI-1776 of EGFR inhibitors can enforce the necessary selection pressure. However, standard HISC conditions have to be used in order to grow organoids from adeno(carcino)mas without Wnt or EGFR pathway mutations. In that case, the differentiation

between normal and CRC organoids relies on sample purity and organoid characterization. It is therefore not trivial to generate organoid lines that fully represent the spectrum of CRCs. Given the high success rate of establishing CRC organoids, their unlimited proliferative potential, biological stability, and cryostorage ability it seems, however, to be merely a question of effort to do so. If combined with genetic information and pharmacological profiles, such an organoid collection could aid in identifying CRC specifics that predict a patient’s drug response similar to the Cancer Cell Line Cyclooxygenase (COX) Encyclopedia [ 13••]. Advanced cancers display genomic instability which drives tumor progression by accumulating additional mutations [30]. Assuming random mutability, it is therefore unlikely that

tumor organoids ex vivo undergo the same genetic alterations as their parental tumor in vivo (unless the same selection pressures apply). On the other hand, targeted therapeutic treatment is known to evoke resistance and favor the selection of subclones, potentially also in vitro. To directly compare tumor progression and drug induced selection in vitro and in vivo, multiple organoid lines from the same patient could be established (e.g. early, progressed, and metastasized cancers; pre-treatment and post-treatment) and treated in parallel. A possible disadvantage of organoid culture may be that organoids from progressed cancers counterintuitively grow worse than those from early tumors or normal tissue due to culture conditions (optimized for normal culture) and potential loss of epithelial integrity (epithelial-mesenchymal transition).