The cancer stem cell hypothesis states that a tumour consists of a subpopulation of cells that give rise to all the heterogeneity found in the cancerous tissue. These cells have some common markers and characteristics of normal stem cells. These, the hypothesis suggests, give rise to the tumour and are therefore the best target for cancer therapeutics.
A recent paper in Nature seems to have caused a bit of a stir in the cancer stem cell field. (first born child or expensive subscription required). Part of the evidence for the cancer stem cell hypothesis is that when human tumour cells are implanted into an immunocompromised host mouse only a small percentage of these cells are capable of reproducing a tumour. This new paper demonstrates that if the host is more immunocompromised then a larger number of cells are capable of reproducing a tumour, instead of only as low as one in a million cells to as many as one in four.
A wired artcle on the work really tries to stir the pot:(free)
"The controversial idea that all tumors are created by cancer stem cells received a setback Wednesday."
There are also some great summaries of what the impact of this paper may be in this field of research in Nature. News and Views. Nature News. Again you will have to sell organs or have an institutional subscription.
I can't say that I know much about cancer stem cells, only that I find the hypothesis interesting. As the summaries suggest, I would not find it surprising that some cancers do indeed consist of a subpopulation of cancer stem cells wereas others do not. I also don't know if specifically targetting cancer stem cells is going to cure a patient since these cells, while capable of causing a recurrance don't cause the symptoms of cancer.
I also don't follow the logic that the cancer stem cells are the reason that chemoresistance occurs, and other hypothesis that have come out of the cancer stem cell hypothesis.
That being said I certainly don't find this study to be blow to the cancer stem cell hypothesis.