Most cells in a tumor are, according to many cancer biologists, just along for the ride. The real instigators are, supposedly, rare cancer stem cells, and so selectively destroying these cells could keep a cancer from growing, or spreading. But a recent study hints that some evidence for rare, dangerous cells is actually experimental artifact, and that the cells capable of propagating a tumor are, in fact, widespread.1

Some of the strongest evidence for cancer stem cells comes from the observation that only one in a million cells from leukemic patients can initiate tumor growth in immune-compromised mice. Publishing in Science, Kelly et al. argue that what's rare about these human cells is not so much their ability to initiate cancer but rather their ability to initiate cancer in a mouse.

In fact, when the research team transplanted mouse tumor cells into histocompatible mice, they found that mice injected with more cells got sicker more quickly and that mice were always dead of lymphoma within 5 weeks, even if only 10 B lymphoma cells had been injected. Moreover, fewer than perhaps 1 in 20 cells expressed markers (Sca-1 or AA4.1) characteristic of stem cells. The researchers tested three kinds of lymphomas and found that the transplanted tumors always mirrored the primary tumor.

Thus, the researchers speculate, cancers need not always be driven by a population of rare cancer stem cells. Instead, certain malignancies, particularly those not marked by a wide variety of cell types, could be sustained by many, maybe even most, of the tumor cells. James Trosko of Michigan State University says he doesn't doubt the experiments Kelly et al. performed, but says the work does not offer an alternative hypothesis as to how cells of different phenotypes and genotypes can sustain a tumor. He thinks cancer stem cells may be less rare than assumed.

Although Kelly et al. studied only leukemia, the authors note that other cancers—including brain, colon, and breast—test for tumor-sustaining cells in human cancers by putting them into mice. But to truly test the hypothesis that a rare subpopulation can sustain a tumor, researchers must create a better mouse model, perhaps transplanting tumor cells along with so-called accessory cells necessary to sustain growth. Of course, if those cells can be properly identified, those seeking to cure cancer by squashing its stem cells may find that they have another worthy target.