Pulling Cancer’s Roots
July 11, 2008
By Leigh MacMillan (from the Spring 08 Momentum)
Every gardener knows that to truly eliminate a weed, you have to pull out the root.
Nip the weed off at the surface only, and in time it will grow back. Such might also be the case with cancer – treatments aimed at shrinking tumors may be leaving behind the “roots,” a core of cells with the unique capacity to regenerate the tumor. Proponents of the idea call it “the dandelion phenomenon,” and they argue that new treatments need to target these “cancer stem cells,” which appear to be present in a wide variety of tumor types. Clinical trials that aim to extend patient survival by killing these cells are under way.
“Within two to three years, we’re going to have clinical trials to treat almost every kind of common cancer with an agent that we think targets cancer stem cells,” predicts Max Wicha, M.D., director of the University of Michigan Comprehensive Cancer Center and a leading expert on cancer stem cells. “The real proof that cancer stem cells are clinically important will be in the results: do the patients do better than with our current therapies?”
The cancer cell, stem cell connection
In 1875, the pathologist Julius Cohnheim suggested that stem cells misplaced during embryonic development give rise to tumors in adult life. Over the course of the next century, scientists increasingly recognized that cancer cells and normal stem cells share certain properties, chief among these their seeming immortality.
Normal stem cells populate the tissues of the developing organism and maintain and regenerate tissues throughout life. They have two defining characteristics: they can divide nearly indefinitely to produce more copies of themselves – the immortality that scientists call self-renewal – and they can produce daughter cells that mature into various cell types.
Embryonic stem cells, which are perhaps the most versatile of stem cells, have made news headlines and sparked political controversy since they were first isolated from human embryos a decade ago. Because embryonic stem cells normally generate all of the diverse cell types in the organism, they are a tantalizing source of healthy cells for repairing diseased tissues. And while it might be possible in the future to direct their maturation for cell-based therapies, clinical “stem cell” applications may come first from killing their dark twins – the cancer stem cells that replenish and renew tumors.
“I believe there will be benefits on both sides, but I think we’ll know sooner if targeting cancer stem cells has a clinical impact than we’ll know how to direct stem cells to replace damaged tissues,” Wicha says.
Mark Magnuson, M.D., director of Vanderbilt’s Center for Stem Cell Biology, says that the two areas of stem cell research are cross-fertilizing and informing each other.
“Everything we’re learning about normal stem cells – what they are, how they grow, what genes confer ‘stemness’ – these are all interesting findings that are transforming our understanding of stem cells in general and that have relevance to both normal and cancer stem cells,” he says.
All cancer cells are not created equal
Although the idea that a small population of cancer cells has stem cell-like properties is more than a century old, the technologies for identifying these rare cells were only recently developed.
Using flow cytometry – a method for sorting living cell populations based on cell surface proteins – and a mouse model for growing human blood stem cells in mice, John Dick, Ph.D., and colleagues at the University of Toronto began to identify cancer stem cells in leukemia in the 1990s.
The investigators reported in a widely cited 1997 Nature Medicine paper that only a fraction of human leukemia cells could reproduce leukemia in a mouse. These cancer stem cells were selected based on certain cell surface proteins, and they represented less than one in 10,000 of the human leukemia cells. The leukemia that was produced in the mice shared the diversity of cells present in the original leukemia, supporting the idea that the cancer stem cells could both reproduce themselves and give rise to various mature cell types.
Evidence that solid tumors also contain cancer stem cells was first reported in 2003. Michael Clarke, M.D., Wicha, and colleagues at the University of Michigan used surface proteins to sort cells from human breast tumors. They showed that only one sub-population of cells was able to re-create the original tumor in mice. As few as 200 of these breast cancer stem cells, which represented between 1 percent and 10 percent of the original tumor, could form tumors, whereas 20,000 cells isolated from the same tumor but without the same cell surface characteristics did not form tumors.
Since then, cancer stem cells have been identified in a range of tumors including brain, colon, head and neck, prostate and pancreas.
“It appears that virtually all cancers have only a small component of cells that is capable of transferring the cancer in a mouse model; they’re likely the only cells that are really tumorigenic, that are driving the cancers,” Wicha says.
Even though many types of tumors appear to have sub-populations of cells that can regenerate the tumor with all of its diverse cell types, the cancer stem cell hypothesis is still debatable.
“Currently, there’s no clear definition of what a cancer stem cell is,” says Susan Kasper, Ph.D., assistant professor of Urologic Surgery and Cancer Biology at Vanderbilt. “Cells identified as cancer stem cells appear to have a few cell surface markers in common, but one of the key challenges in the field is to define the characteristics of the cancer stem cell.”
Magnuson agrees. “The research related to cancer stem cells is all tumor-based. And the problem is you really don’t know what the stem cell in the tumor is,” he says.
Cancer stem cells might come from normal stem cells, when mutations dismantle the normally tight controls on their self-renewal properties. They might also come from mutations that restore the power of self-renewal to so-called progenitor cells, the offspring of stem cells that mature into certain cell types.
Both normal stem cells and progenitor cells, because of their long-lived natures – stem cells could be around for an entire lifetime – have the potential to accumulate the multiple mutations required for carcinogenesis. The hypothesis is appealing as an explanation for how tissues with very short-lived mature cells, like the blood, skin, and lining of the gut, can accumulate enough mutations to give rise to a tumor: the mutations happen in the long-lived stem/progenitor cell population.
A cancer stem cell can be thought of as lurking in the general stem cell population, Kasper explains. Once it has accumulated a number of mutations, it’s there “waiting for the right stimulus to activate it so that it begins proliferating.
“No one knows what those signals are – we talk about cancer stem cells and how they might work, but very little is known about the biology of cancer stem cells,” she says. “For example, how do these cells arise? How do they survive and proliferate? Is the core of a metastatic lesion a cancer stem cell?”
Kasper and colleagues have developed human prostate cancer stem cell lines (cells that can be grown in the laboratory indefinitely) that they will use to address these kinds of questions.
Are cancer treatments off target?
The cancer stem cell hypothesis, if correct, could explain why many cancer treatments don’t improve long-term patient survival. Treatments are selected for their ability to cause tumor shrinkage, which doesn’t necessarily predict improved survival.
“We’ve basically designed a lot of treatments that kill the wrong cells in the tumor,” Wicha says. “The treatments leave the cancer stem cells behind, and those cause recurrence.”
Wicha cites evidence from animal models and from ongoing studies in patients with breast and pancreatic cancer. He and colleagues have examined the cells that remain after chemotherapy and radiation therapy shrink the tumors.
“If you transfer those cells that are left to a mouse, they grow like crazy,” he says.
Cancer stem cells may be more resistant to cancer treatments because of the properties they share with normal stem cells: slow cell division cycles (cancer therapies often target rapidly dividing cells) and high levels of proteins that protect against DNA damage and cell death.
These shared properties may also lead to the development of novel therapies that act across many different tumor types. Signaling pathways that are important for normal stem cells during development, such as the Wnt, Hedgehog and Notch pathways, also appear to be important regulators of cancer cell growth.
“What we learn about one cancer stem cell in one kind of tumor is informing us about what’s going on in another kind of cancer stem cell,” Wicha says. “If we develop an effective therapy for one kind of cancer – and that’s a big if – it might be effective in killing stem cells in another cancer too.”
Clinical trials of existing drugs or unique combinations of drugs that target surface proteins on cancer stem cells are already ongoing for multiple myeloma and leukemias.
And Wicha and colleagues at the University of Michigan, along with investigators at the Dana-Farber Cancer Center and Baylor College of Medicine, are gearing up for the first clinical trial targeting cancer stem cells in a solid cancer. The trial will test an inhibitor of the Notch signaling pathway in breast cancer. The drug, developed by Merck, kills breast cancer stem cells in laboratory studies.
The ultimate test: survival
Clinical trials of cancer stem cell-directed therapies face challenges. Will the treatments kill normal stem cells that are important for tissue maintenance and regeneration? What are the measures of success for a treatment that’s directed against a tumor’s slow-growing “roots” rather than its visible “weed?”
The potential for killing normal stem cells is perhaps the biggest challenge to the field, Wicha says. There are data now being published that support the notion that cancer stem cells may have different sensitivities to certain drugs, even though the drugs target pathways that are also active in normal stem cells.
“The extra mutations in the cancer stem cell may make it particularly vulnerable to certain kinds of treatment that don’t affect a normal stem cell,” Wicha says. “That remains to be proven. The real test will be in giving these agents to patients; we’ll be watching very carefully for side effects.”
Tumor shrinkage is a traditional measure of success for cancer therapies. But cancer stem cell-targeted treatments may not have any visible effects on the bulk of a tumor. One notion is to first use another agent to “de-bulk” the tumor and induce remission, and then follow with a cancer stem cell-targeted agent, using duration of remission as a measure of success. Investigators are also exploring alternate laboratory-based tests.
Ultimately, the question is – are cancer stem cells truly the cells that regenerate the tumor, and if they are killed, does that eliminate the cancer and improve survival?
Measuring survival takes a long time, which is why investigators are designing alternate tests. They will eventually have to prove that these quicker measures “correlate with patients living longer – that’s the ultimate test for any therapy,” Wicha says.
“I hope that people won’t get discouraged if the first trials don’t work,” he adds. “We’re really just at the beginning of this, and I think the idea is right.”