Just one cell may be the root of a tumor. Of course, that one cell has properties that it shouldn't have that allow it to make more of itself and differentiate into the cells of a tumor. It's also resistant to chemotherapy and can seed metastases and recurrences. In essence, the cell is acting like a stem cell. And that cell is at the heart of the cancer stem cell hypothesis.
"There's an intense interest because these cells certainly behave differently than many of the other cells in the tumor, and it appears that these are the major cells required to repopulate the tumor and to initiate metastasis," says William Farrar, head of the Cancer Stem Cell Section at the National Cancer Institute.
The cancer stem cell hypothesis has come a long way from being limited to leukemia. In the past few years, it has exploded into the solid tumor arena and has been taken up by more and more researchers for study. While it has its detractors, research based on cancer stem cells is going strong in the lab, particularly in the search to find drugs to shut down those cells. Some of that work is even moving into clinical trials.
"The idea originated really in leukemia and lymphoma literature. That literature started linking what we call stem cells to the oncogenic process," Farrar says. "The big leap was that the phenomena would probably exist in epithelial or solid tumors, and that breakthrough came from the work of Max Wicha and Michael Clarke."
Beginning about 50 years ago, researchers studying leukemia noticed that some leukemia cells were different. By morphology, Memorial Sloan-Kettering's Bayard Clarkson says, certain cancer cells stood out. He and his colleagues also saw that these cells were quiescent, stuck in a Go-like state, and not affected by treatment. "There was always this population of primitive, blast-like quiescent cells," Clarkson says.
It was not until the 1990s that the University of Toronto's John Dick identified the first cancer stem cell in AML. Dick also developed an immunodeficient mouse model in which he could xenograft human leukemia cancer stem cells into mice and watch the disease develop.
Since then, the work has moved into solid tumors. The University of Michigan's Max Wicha and his colleague Michael Clarke, now at Stanford University, applied Dick's xenograft technique to solid tumors. "[We] decided we'd see if we could find similar cancer stem cells in solid tumors," Wicha says.
Using fresh tumors from breast cancer patients, Wicha and Clarke used flow sorting to selectively enrich tumor-initiating cells. "That led to the finding of breast cancer stem cells and that kind of opened up the whole field after that," he says. They also found that the breast cancer stem cells had a particular CD44+/CD24- phenotype that later was found on other tumor types.
Since then, researchers have been scrambling to find cancer stem cells. Farrar identified them in prostate cancer and others have found them in pancreatic and brain tumors.
"In fact, the cancer stems cells have been described in every conceivable [lineage]," adds Harvard's Kevin Struhl.
But what are cancer stem cells? These cells go by many names, including "tumor-initiating cell" and "tumor progenitor cell," Farrar says. "Cancer stem cell" is an operational definition, adds Wicha. It's a "cell that can self-renew, so it perpetuates the tumor and also generates other cells in the tumor that lost the capacity to self-renew," he says.
According to the cancer stem cell hypothesis, one property that cancer stem cells have that other cancer cells lack is the ability to form tumors, including metastases. "They have a very high invasive capacity and they metastasize," Wicha says. "Other cells may get into the blood and lodge in a distant organs, but because they don't have the capacity to self-renew, they are very limited and can't really form a clinically significant metastasis."
Furthermore, cancer stem cells are resistant to chemotherapy and help seed a recurrence of cancer. Clarkson at Sloan-Kettering found that in leukemia, the quiescent cells weren't killed by traditional treatments. In fact, Wicha's lab has found that as tumors shrink during chemotherapy, the percentage of cancer stem cells increases, and not just because the proportion is thrown off. New data from Wicha suggests that when breast cancer cells are dying, they make cytokines that stimulate cancer stem cells. "It's almost like a damage response," he says. "The cancer stem cells are not only resistant to chemotherapy, but they are actually stimulated by the chemotherapy indirectly."
Not too much, however, is known about the biology of cancer stem cells, such as where they originate. They're not necessarily stem cells that went bad, but could be progenitor cells that, through a mutation, reverted to a stem-like state. "At this point, given that much of their biology is still a black box, it really isn't known," says the Broad Institute's Piyush Gupta. "It's very hard to make any sort of statement on how cancer stem cells relate to normal stem cells."
The hypothesis is not without its detractors. "I'm pretty convinced there's nothing there," says Scott Kern at Johns Hopkins University.
However, Kern says that cancer stem cells are well established in liquid tumors like leukemia — it's just the solid tumor claims about which he's skeptical. Instead of there being cancer stem cells that initiate tumor growth, he says there could be inhibitors that sometimes block tumor-initiating activity when the cancer cells are xenografted into mice. Rather than seeing stem cells when tumors form, he contends, researchers are seeing the absence of those inhibitors.
In other words, he says, biologists need to think more like chemists. "If we relate this to stem cells, what we find is one population of cells can xenograft into mice and form tumors at an easier rate than another. The stem cell proponents say, 'Therefore, the population had more stem cells.' A chemist would never do that. A chemist faced with two solutions, one of which seemed to have more activity than the other, would first start wondering if there was an inhibitor," Kern says. "If the two solutions could be shown to not have an inhibitor, then the chemist would start thinking that properties were actually different and that one may be more active." That experiment, Kern says, hasn't been done.
In any event, controversy is good, Wicha says. "A lot of the controversy is over which mouse model to use and those kinds of things. It's very healthy," he says.
A lot of new studies coming out in this field focus on drugs that specifically target those cells. "The major concern about these is that almost all conventional therapies and all conventional drug screens were based on screening and obtaining results against whole populations of tumors," Farrar says. "Technologies were never developed to screen against a very small population of cells that may be what you might call the evil seed of the whole tumor."
Using a new inducible cancer stem cell model, Harvard's Struhl recently found that a diabetes drug, metformin, selectively kills cancer stem cells. Over the years, researchers had noted that people taking this diabetes drug had lower rates of developing cancer than expected.
Struhl and his colleagues first developed a way to transform cells into cancer stem cells using an inducible oncogene. Within 24 to 36 hours of inducing SRC, the team would have cancer stem cells. When Struhl then treated the model mice with both metformin and standard chemotherapy treatment, he saw a better response than either alone, as he reports in the October Cancer Research. "The tumor goes down faster, but more importantly, it doesn't come back because you've actually killed not only the cancer, which the chemo does, but also the cancer stem cells, which the metformin does," he says. "The whole idea of our experiment is that the effect of metformin is really much more potent when combined with chemotherapy because it's actually killing a different kind of cell than the chemotherapy is killing. So if you're killing them together, you're getting the best of both worlds."
The Broad's Gupta and his colleagues took a different approach to find another drug, salinomycin, that also selectively targets cancer stem cells. The Weinberg lab at the Whitehead Institute had shown that cancer stem cells have properties similar to those of cells undergoing an epithelial-mesenchymal transition. During that transition, cells undergo gene expression, protein expression, and functional changes. "We started wondering wonder whether or not the drug sensitivity might also be shared between cancer stem cells and cells that were induced at the EMT," Gupta says.
First they took immortalized epithelial cells — but not cancer cells — and induced them do go through a mesenchymal transdifferentiation. "In so doing, we got sort of a significant enrichment in the proportion of cells that had stem-like properties," he says.
Then they used a high-throughput approach to screen about 16,000 compounds to search for ones that would selectively kill those induced cells. Of those, 35 showed some toxicity against the EMT cells, but only a few — about eight or nine — were available in sufficient quantity to follow up on. Three of those had strong selectivity, and salinomycin had the highest. "It has a 10-fold selectivity for cancer stem cells. When you put it on cells in culture, essentially you see complete loss of stem cell activity," Gupta says. His team is looking for additional compounds that might also target cancer stem cells.
He and his colleagues also plan to work out how these compounds are having their effect on cancer stem cells. "If we can identify some of the intracellular targets, the proteins actually being targeted by some of the compound hits from our screen, that will tell us a little bit about the network of signaling that is operative in cancer stem cells," he says.
Michigan's Wicha is already at work on both those pathways and inhibitors for them. He and Clarke started a company called OncoMed Pharmaceuticals to translate their basic science findings into drugs. A few of their produts are just beginning clinical trials.
The first of OncoMed's products is a monoclonal antibody against the Notch pathway ligand DLL4. Onco-Med says that blocking this ligand disrupts angiogenesis, inhibits cancer stem cell growth, and promotes cell differentiation. Other potential drugs in the company's pipeline include gamma secretase inhibitors that also affect Notch and a Hedgehog inhibitor. "Both of these [pathways] are very involved in stem cell self-renewal," Wicha says.
However, Wicha says clinical trials aren't designed to test drugs targeting cancer stem cells. Trials are set up to look at tumor shrinkage. "Tumor regression is really the marker of the bulk of the population, not of the stem cells, so it's not a good model," he says. "That's why we've developed many compounds that can shrink tumors but patients don't necessarily live longer."
Instead, Wicha says they are using a neoadjuvant trial design. "Our approach is going to be, at the beginning, usually combining stem cell therapy with chemotherapy because we want to hit both cell populations. We are going to hit the bulk population; we can do that with chemotherapy, and we then hit the stem cell population," he says.
In the end, the goal of studying cancer stem cells is to develop better treatments for cancer patients. "We hope that some sort of compound that can target the cancer stem cells when used together with standard chemotherapy or radiation treatment [can] help eliminate some of those cells that will give rise to recurrence and be a problem down the line," says Gupta.