Leukaemia researchers have known about cancer stem cells for many years, but solid tumours have recently been found to contain populations of highly proliferative cells that have a lot in common with stem cells. Like normal stem cells, they have the capacity not only to self-renew, but also to undergo differentiation. These are likely to be the most dangerous cells in the tumour, as they have the greatest potential to form metastases. Two articles in this issue explain how defects in self-renewal pathways lead to transformation, and discuss ways of blocking this process.

On page 895, Ricardo Pardal, Michael Clarke and Sean Morrison discuss how cancer stem cells were first discovered, and have been recently identified in tumour types such as breast cancer and glioblastoma. Cancer stem-cell development is a two-way street — mutations in pathways that control self-renewal can not only cause normal stem cells to undergo transformation, but also cause differentiated cells to take on stem-cell-like traits. Researchers are now developing tools to determine which cells in a tumour are benign and post-mitotic and which ones have self-renewal capacity, and are investigating ways to specifically target these cells.

On page 903, Marina Pasca di Magliano and Matthias Hebrok explain how activation of one of the pathways described by Pardal et al. — the Hedgehog (HH) signalling pathway — initiates cancer formation, and how continued HH signalling is required for survival of pancreatic adenocarcinomas and small-cell lung carcinomas. Various inhibitors already exist that block HH signalling at several levels and these are being developed as anticancer agents. But targeting stem cells is not without risks — normal stem cells exist in the adult bone marrow, gut and brain, so this approach will have to be undertaken with great caution.