Clinicians and researchers in neuro-oncology at Dana-Farber are excited. Experts have spent decades trying to find ways of controlling or killing glioblastomas, the most highly malignant of brain tumors, but progress has been slow.
Now, scientists are discovering a growing list of molecular targets for "smart" drugs that might dent the tumors' resistance that has stymied conventional therapies, and could potentially lengthen survival and improve quality of life.
Until glioblastomas yield to some future breakthrough, the best near-term bet for helping patients is to develop and test these targeted drugs — first one-at-a-time, then in combinations — say the researchers. Rather than aiming to destroy tumor cells by blasting them with radiation and DNA-damaging chemotherapy, the new agents are designed to block runaway growth signals that drive the tumors' aggressive spread.
"There is hope that over the next few years, new cocktails of drugs will be available to target the key molecular pathways involved in these tumors," says Patrick Wen, MD, director of the Center for Neuro-Oncology at Dana-Farber/Brigham and Women's Cancer Center (DF/BWCC).
However, the road from promising candidate to effective treatment is often fraught with detours and delays. Each new compound from pharmaceutical laboratories comes with a list of necessary questions: Can it reach its target within the brain? Will it be effective in disrupting tumor growth? Does it have acceptable risks of side effects? Every new agent must pass these tests on its own, then again when combined with other drugs.
The urgency of getting these answers runs up against the slow and deliberate pace of clinical trials, which require enrolling hundreds of patients — whose tumors have various genetic characteristics — and testing drugs for months or years to see if they are effective.
"If patients in clinical trials don't live longer, we say the drug failed but we don't know why it failed," notes Keith Ligon, MD, PhD, of DF/BWCC's Center for Molecular Oncologic Pathology (CMOP). "Maybe the drug killed off only half the cells, but in fact we did make some progress with the other part of the tumor. But we have no way to tell, other than crude measures that take months to demonstrate results."
Determined to do better, investigators are working on ways to shortcut the testing process, evaluate more quickly whether experimental drugs are reaching targets in the brain, and personalize treatments for patients who are racing against the clock.
What if different drugs could quickly be tried in an exact animal model of a patient's brain tumor, with results guiding the treatment of that patient? Could physicians remove a small number of cells from the tumor during surgery and rapidly screen a panel of drugs for their effectiveness? Are there better ways to predict whether a drug will cross the barrier that protects the brain, so that it can reach the tumor?
Though none of these are ready for prime time, Center for Neuro- Oncology scientists and colleagues are hopeful that these strategies will prove valuable in the fight against stubborn glioblastomas.
One of the newest and most advanced resources is on the second floor of Dana-Farber's Jimmy Fund Building, where incubators and freezers contain cells taken from almost every DF/BWCC patient who undergoes surgery for high-grade brain tumors, such as glioblastoma.
Five years ago, Ligon, with the support of his colleagues in Neuro- Oncology, started this collection of brain tumor cells that survive indefinitely as a resource for the DF/BWCC Brain Tumor Program. He calls it the "Living Tissue Bank" in contrast to conventional tissue banking, which has focused on freezing or storing tissue in a form where the cells are no longer alive. Not all brain tumors can be grown in laboratory dishes, but the bank currently has 50 or 60 permanently established cell cultures, or cell lines, they can mine for valuable information. The bank continually grows as new patients come to DF/BWCC for treatment.
"We take out samples and try to establish a living model of that patient's actual tumor, so that instead of trying to guess from its genetic profile how the tumor will behave, we try to have a model so we can go after it in a more direct fashion," explains Ligon.
"These tumor lines are an invaluable resource for preclinical testing, to make sure the drugs do what they are supposed to do," notes David Reardon, MD, medical director of the Center for Neuro-Oncology. Reardon, Wen, and Ligon are currently designing clinical trials based on evidence from such studies to help pick drugs with the most promise.
With brain tumor cells from the Living Tissue Bank, Ligon has begun work on another type of personalized cancer medicine known as patient "avatars." In online gaming, an avatar is an image that represents a person. In medicine, an avatar is an exact animal model of a single individual's disease — such as a brain tumor — created to enable personalized drug research for that patient. Scientists create avatars by injecting mice with tumor cells from the living tissue bank or even tissue fresh from the operating room. If all goes well, an exact copy of the patient's tumor will develop in the mouse.
Additional mouse models can be created from the initial avatar, so that a variety of drugs can be tested on the tumors growing in the avatars. Ligon notes that this strategy is well-established in microbiology, where doctors take a sample of bacteria from a patient with an infection and grow them in a petri dish, trying different antibiotics to find one that kills the bacteria that is causing disease in the patient.
"The ideal situation would be if we could do this fast enough — if we could test not just the drug the patient was prescribed, but all the drugs he or she could possibly be given" to weed out those that would be ineffective, Ligon says. The model could evaluate new drugs fresh from pharmaceutical company laboratories.
Many scientists are skeptical that mouse avatars will prove valuable and cost-effective; Ligon agrees that the method is a long way from being proven.
One hurdle is that not all tumor samples grow well in mice. Also, the process in its current form may take too long to help patients with rapidly growing brain tumors during their lifetime. In addition, the expense of creating and maintaining the mouse models is high. Ligon notes, however, that patients are increasingly interested in having avatars made of their disease. He is optimistic that this concept, originally tried in the 1960s, has a future. "Our models are more sophisticated and we have a lot more drugs to test than were available back then," he says.
Ligon, who is also affiliated with the Neuro-Oncology Program of the Dana-Farber/Harvard Cancer Center, has begun exploring a futuristic scenario: subjecting one or a few cells isolated from a patient's tumor to precise tests that might predict the tumor's behavior and how it will respond to different drugs. Because tumors often have more than one type of cancer cell — tumor heterogeneity — the ability to test cells from different parts of the cancer could be extremely valuable.
"Potentially we could take tumor cells right out of patients and analyze them almost immediately to determine what kind of drug to give them," Ligon says. Such a shortcut may seem far-fetched, but Ligon is pursuing this vision in collaboration with Scott Manalis, PhD, a bioengineer at Massachusetts Institute of Technology (MIT) and a member of the David H. Koch Institute for Integrative Cancer Research.
Manalis has created a microfluidic system so sensitive that it can measure the mass of a single cell. Within a few minutes, the device can tell, from changes in the mass and density of the cell, whether the cell is dormant, growing, or shrinking. And it can even tell scientists how the cell's growth varies at different times — its "growth profile."
"If the mass goes up, the cell is happy and growing," says Ligon, "so if you apply the drug to a tumor cell and its growth profile doesn't change, you know the drug had no effect." If, however, the cancer cell's growth stalls or shrinks, the drug is likely to be working.
In a 2011 experiment at MIT, the system was tested on leukemia cells that were treated with an antibiotic drug. Less than an hour later, the weighing device detected a change in the cancer cells' density, signaling that they were beginning to die. To the scientists, this suggested the potential value of the system in screening cancer drugs.
Ligon's collaboration with Manalis has been under way only a few months, but he is excited about the possibilities of measuring drug effects in real-time. "If we get the technology working, it would go into clinical trials to validate it; eventually it could be a diagnostic test." Funds for the research come from a special National Institutes of Health program to support scientists who are asking "provocative questions" rather than continuing incremental, "safe" research.
In another slant on analyzing the genome of single brain tumor cells, Dana-Farber's Matthew Meyerson, MD, PhD, and Ligon are collaborating with chemical engineer Christopher Love, PhD, of MIT in a "Bridge Project" supported by the Koch Institute and Dana-Farber/Harvard Cancer Center.
"We now are using technologies which can take a single cell from a patient's tumor, amplify it to make more DNA, and sequence its entire genome," Ligon explains. With such precise information, scientists might be able to determine the genetic makeup of the exact cells within a brain tumor found, in devices like those Manalis has developed, to use a specific mechanism for resisting drugs. And if the tumor is made up of several kinds of tumor cells, this technique could guide physicians in selecting multiple drugs — combination therapy — for the attack.
The most potent anticancer drug is worthless if it can't reach its target. In neuro-oncology, many potentially effective drugs fail because they can't penetrate the "blood-brain barrier" — tight-fitting blood vessel cells evolved to protect the brain against bacteria and poisons. A new tool being developed by DF/BWCC researcher Nathalie Agar, PhD, can track a drug as it enters the body and flows through the brain's circulation, revealing whether molecules of the drug are getting through the barrier.
With the power of mass spectrometry, which detects molecules by their mass, Agar has created a platform that creates images of a drug's distribution in the body and the brain after the compound has been infused into a research animal. What's new about Agar's method is that there is no need for the drug molecules to be tagged with fluorescent chemicals, which can change the drugs' structure, she says. (See Seeing with Mass Spectrometry for details.)
Agar and Dana-Farber colleague Charles Stiles, PhD, are using her mass spectrometry technique to determine if experimental drugs are able to breach the blood-brain barrier and reach tumor targets within the brain.
Stiles, who is co-leader of the Dana-Farber/Harvard Cancer Center Neuro-Oncology Program, remembers the first time he saw Agar demonstrate her technique. "She showed me what she was doing with the mass spec technology, allowing us to visualize the inside of a blood capillary and drugs getting into the interstitial space [of the brain], and I was just blown away," he says.
"An especially powerful feature of Nathalie's technology is that it does not require any radioactive or fluorescent labeling of the drugs being tested," Stiles explains. "She can image the exact forms of drugs that are intended for use in patients."
Stiles and Agar discovered, in one experiment, that a drug did penetrate the barrier and could be seen on the computer display to have accumulated in the tumor. But the image also showed that the drug was confined mainly to the network of blood vessels that the tumor had created. "So the drug was just sitting there," she says, "rather than reaching the tumor tissue itself."
Another interesting finding: The mass spectrometry system took 25 cross-section images of a mouse brain and created a 3-D representation. "We found that there was one region of the brain where the drug crossed the bloodbrain barrier more efficiently," says Agar, which might lead to a strategy for delivering the drug where it is most likely to reach a tumor.
One way of using the method clinically, Agar says, would be to give a drug to a brain tumor patient prior to surgery. The mass spectrometry image could then be used to study the tumor tissue that's been removed "to see if the drug actually made it to the tumor." If so, it would be a good sign for further treatment with the drug.
Agar says she and Stiles, together with Dana-Farber chemist Nathanael Gray, PhD, plan to use the system to test more drugs from pharmaceutical companies.
If they can demonstrate that certain drugs do reach the brain, she says, it "might be a very effective way to convince companies of the value of developing trials that could help us re-purpose existing drugs to treat brain cancer."
Only time will tell what impact these innovations — mouse avatars, single-cell testing, mass spec studies of drug penetration, and others — will have on the treatment of brain tumors. But one thing is certain: They reflect the power of collaboration among DF/BWCC investigators, both basic and clinical, to mount an intense effort against this stubborn form of cancer.
Paths of Progress Spring/Summer 2013 Table of Contents
Dana-Farber Cancer Institute, 450 Brookline Avenue, Boston, MA 02215 | Call us toll-free: