The idea that a subpopulation of tumor cells, dubbed cancer stem cells (CSCs), drive tumor survival and metastasis has gained momentum in recent years, but it is not without skeptics. Last week, researchers published three studies that may silence a few of those critics -- or at least answer some of their critical questions, including whether CSCs might be relevant in patient care.
In each of the studies, the investigators used lineage-tracing techniques to show that a small population of long-lived, generally quiescent cells drive tumor formation and growth in different tumor models: glioblastoma in a genetic mouse model, intestinal adenomas in a genetic mouse model, and squamous cell skin cancer in a chemical-induction model.
“It is not like there is a major scientific leap that is made with any of these papers, it is more a validation. But it is an important one, because they are using a different general approach than we’ve all used to study the problem,” says William Matsui, MD, a hematologist at the Sidney Kimmel Comprehensive Cancer Johns Hopkins University School of Medicine in Baltimore, who studies CSCs in blood cancers and pancreatic cancer.
Previous CSC experiments relied on grinding up tumors, segregating subpopulations of cells in vitro based on cell surface proteins, and then watching their growth capacity after xenograft transplantation or in in vitro cultures. Some critics have expressed concern that the putative CSCs identified in this way are not really relevant to tumor biology, but rather are an artifact of the stressful purification methods or xenograft techniques.
Two of the new papers, though, take advantage of genetic mouse models, which seem to closely resemble the cognate human disease. In each case, when the investigators used molecular tricks to label subsets of cells in the animals, they found that a small subpopulation of cells drove tumor formation.
In the skin cancer model, the researchers, led by Cédric Blanpain of the Université Libre de Bruxelles, Belgium, saw that most of the cells in a benign papilloma have limited proliferation capacity. By contrast, a small fraction of cells have the capacity for long-term survival such that their progeny comprise a substantial portion of the tumor.
In the glioblastoma work, Luis Parada, at the University of Texas Southwestern Medical Center in Dallas, and colleagues took the experiments one step further and showed that the small subpopulation of stem cells in the tumor was indifferent to temozolomide treatment. The mouse tumors initially stopped growing in response to temozolomide treatment, similar to what is seen in patients. However, a small group of neural-derived stem cells (genetically labeled with green fluorescent protein, GFP) slowly repopulated the tumor.
Parada’s team still had one more trick up their sleeve though: The GFP-labeled stem cells were also genetically engineered to die in the presence of the antiviral drug ganciclovir. When the researchers treated the animals with temozolomide and ganciclovir they saw improved survival, compared with temozolomide-only treatment. That observation suggests the subpopulation of GFP-labeled stem cells are responsible for tumor survival and treatment resistance.
“All three of the papers are very nice pieces of work to show that the [CSC] concept may be valid,” Matsui says. “There are caveats to it. They are mouse tumors. Mouse tumors are not human tumors. But nothing is perfect and the more levels of evidence you have, the more believable it becomes.”
For me, as tantalizing as the CSC hypothesis is, the million dollar question remains: Whether treatments built on the CSC hypothesis will benefit patients and improve their outcomes. For that answer, we still all have a while to wait.