Brain tumour cell hierarchy

Cancers are heterogeneous between patients and between tumour cells. It is still difficult to identify the subpopulations of cells that most contribute to tumour growth and those that are targeted by therapy. Xiaoyang Lan et al. now explore the dynamics of human glioblastoma (GBM) subpopulations using barcodes for tumour cells. They suggest that a proliferative hierarchy emerges through stochastic cell fate decision. In this model, slow-cycling stem cells give rise to rapidly proliferative progenitors that fuel tumour growth and which in turn generate cells that are short-lived and do not proliferate. This is in contrast to a clonal evolution model based on the different fitness of cells that are selected for. The authors also identify a rare subpopulation of GBM cells that is resistant to TMZ treatment (the common treatment for GBM) but can be targeted by drug combinations.