Statistical dynamics of tumor initiation and progression

Cancer cells are generally not well defined molecularly or phenotypically. How  cellular heterogeneity of tumor initiating cells (TICs) fuels tumor growth  remains an open question. To address this challenge we analyzed cellular  dynamics in human colorectal cancer samples by following the fate of distinct  tumor cells in vivo. Consistent with similar  recent experiments we find clonal dominance with a broad distribution of clonal  contributions (more than 100-fold difference) across the successful lineages, which has been suggested to constitute evidence for cellular heterogeneity  among TICs, i.e. some cells are better than others at forming tumors. Surprisingly, outcomes of mixing experiments between dominant monoclonal  tumors and polyclonal tumors showed a pattern where the fractional contribution  from the monoclonal tumor was below its injected fraction. To address this  puzzle, we have suggested a model with stochastic fate for TICs, which not only  explains differences in clonal composition among different lineages in tumors, but  also the surprising patterns found in mixing experiments.