The Brain Tumor Center at Duke University
runs more clinical trials of new treatments for
brain tumors than virtually any other center
in the world. For patients not in these trials,
the Center tries drugs approved for other cancers
and promising against brain cancer cells in
the laboratory. The Center’s surgeons operate on
brain tumors that others have pronounced
inoperable. It is a battle that buys about 10 to
15 percent of their patients another three to ﬁve
years of life. The Center’s motto is “At Duke,
there is hope,” but is the hope real? Is it worth the
struggle? And what is “hope” when no treat
ment is working? For the sake of his patients,
the oncologist who co-directs one of America’s
largest clinical programs for patients with brain
tumors is aggressive, experimental, indignant,
and compassionate—but always hopeful.
Brain cancer is one of our most terrifying diseases. All too often, it kills with appalling speed; the most common primary brain cancer in adults, glioblastoma multiforme, is also the deadliest. In the United States, only half of patients receiving the standard treatments survive for a year after diagnosis. Fewer than one in ten are alive ﬁve years later. In adults, there are about 17,000 cases of primary brain cancer each year (in addition, cancer can spread to the brain from some other part of the body), resulting in about 14,000 deaths. Brain cancer is the third leading cause of cancer deaths among adults aged 20 to 39, and the biggest killer among cancers of children and young people up to age 20.
The fear that brain cancer brings runs deep. This cancer assaults the organ of the self—the seat of our mind, emotions, and basic sense of control over our bodies—and so threatens to alter and destroy who we are. Symptoms vary depending on what part of the brain is affected. The frontal lobes, the part of the brain right over the eyes, are involved with decision making, mood, and judgment. Tumors that inﬁltrate both frontal lobes can induce personality changes (for example, in levels of inhibition), lack of initiative, irritability, depression, and paralysis.
Tumors located in, or pressing on, the language areas of the brain are usually in the posterior part of the frontal lobe, in the temporal lobe behind the eye and above the ear, and usually on the side opposite to the person’s handedness. They may bring problems with speaking, understanding speech, thinking, and writing. Tumors hitting the visual areas, which are at the back of the brain, may cause distortion of sight, including blindness. Brain-stem tumors are deep in the brain within its stalk, which extends from the top of the spinal cord. These tumors can cause vomiting, swallowing and speech problems, lack of coordination, clumsiness, and muscle weakness in the face that may cause a person’s smile to appear crooked and cause loss of coordination of eye movements. All brain tumors can lead to seizures, which may be manifested as subtle alterations of mood or levels of consciousness. Interestingly, brain tumors are only rarely a cause of headaches.
Obviously, the entire brain cannot be cut out, but malignant brain tumors are not neatly circumscribed; they have edges that inﬁltrate normal brain tissue. Nor does the brain tolerate well the slashing, burning, and poisoning that still characterize much of the arsenal of modern cancer medicine. The body’s need to protect the brain at all costs means that even reaching a brain tumor with medication, radiation, or surgery is much harder than reaching cancers in other areas. No other organ is tightly enclosed in its own protective bone case; and no other type of cancer treatment must breach a defense like the blood-brain barrier—the tightly knit system of blood vessels and cells that keeps out or immediately expels most foreign substances, including medicines, from the brain.
The comparatively few cases of primary brain cancer each year (in comparison to, for example, lung or breast cancer) make it what is called an orphan condition. Studying it offers little potential proﬁt to pharmaceutical companies and little promise of wide impact to the supporters of biomedical research, so these groups invest very little time and money in it. Indeed, most brain tumors are secondary, originating in other organs (frequently the lungs or breasts) or in skin melanomas and spreading to the brain only late in the course of illness. About 170,000 of these brain metastases are diagnosed each year. The bitter consequence is that by the time cancer is discovered in the brain, it usually has spread widely from its original site in the body. If treatment is already under way for the primary cancer, its appearance in the brain means that the therapy is failing and the cancer is out of control.
As a result, when physicians see brain tumors, they tend to give up. Either they have found a voracious primary tumor, for which they can merely relieve the symptoms, or they have discovered a sign that treatment is failing. They often tell patients that “there’s nothing more we can do.” Indeed, the majority of patients I see as co-director of the clinical neuro-oncology program, the clinical arm of the Brain Tumor Center at Duke University, have been told that there is no hope. Scott Beckert, the husband of one such patient, said simply: “The doctors at home said she had six months to live, that the cancer was inoperable, incurable, hopeless.” Like Eileen Beckert, most of my patients arrive here having been sent away from somewhere else to die.
The overriding goal of the Brain Tumor Center at Duke is to give these patients hope and keep that hope alive as long as any chance remains for ﬁnding a treatment that prolongs life. It is a philosophy that raises difﬁcult ethical questions: What constitutes legitimate hope? How aggressive should treatment be when the unambiguous alternative is death? When treatment options run out, what can and should we do?
In the early 1990s, Duke modiﬁed its approach to brain cancer. The distinctive combination of specialties and interests represented in the center’s current interdisciplinary leadership emerged at that time. I had been trained as a pediatric oncologist, but became interested in helping adults as well; Allan Friedman, M.D., co-director of the clinical neuro-oncology program (no relation, though we are often asked), was an adult neurosurgeon who wanted to learn more about pediatrics and neurooncology. I had trained in the laboratory of Darell Bigner, Ph.D., who directs the Brain Tumor Center and its research, so I knew of the challenges in coordinating the laboratory and clinic. We merged the adult and pediatric clinical programs for brain tumor patients and sought the tightest possible link between research and practice.
MORE CLINICAL TRIALS FOR MORE PATIENTS
The Brain Tumor Center runs more clinical trials of new brain cancer treatments (speciﬁcally, trials generated by a single institution) than virtually any other center in the world. At any given time, about 40 different trials are underway, with roughly 40 percent of our patients enrolled in them. Other research centers also enroll large percentages of their patients in trials, of course, but overall only 1 to 2 percent of adult cancer patients in the United States participate in research. For childhood cancer, the proportion is closer to 60 to 80 percent of patients enrolled in clinical trials. This approach has helped produce dramatic advances in survival in cancers like childhood leukemia.
Clinical trials make medicine a science. Before they were introduced in the mid-twentieth century, medical treatment was based on instinct, guesswork, and what a particular doctor had been taught at a particular school. Because even without treatment many conditions improve over time, and because the belief that a medication will work occasionally produces improvement or even a cure, totally ineffective treatments were often deemed successful and used for generations.
To discover real weapons in the battle against disease, scientists evolved techniques that, in combination, constitute today’s Phase III clinical trial. In a Phase III trial, to separate out any possible effect of a patient’s simply believing in a treatment, some patients are given the new treatment and others a comparison treatment. The comparison treatment may be a placebo or, in the case of brain tumors, the standard therapy, which means that any difference between the two treatment groups should show true effectiveness of the new therapy.
Patients do not know which treatment they are getting, so the study is characterized as “blind.” Furthermore, assignment of a patient to one or the other treatment is random, so that the researchers cannot unconsciously select the most promising candidates for the new treatment and assign the sicker ones to the comparison group, thus biasing the results. Finally, to keep the researcher’s potentially misleading hope or optimism at bay when he is assessing how well the patients have done under each treatment, researchers do not know which medication has been given to which patients. This makes the trial “double blind.”
At Duke, the wellspring of new treatment strategies for clinical trials is a research organization of 100 or more people with interests that range from trying to understand the molecular and genetic processes by which a normal cell becomes cancerous to seeking new radiation therapies, chemical therapies, and immune-based treatments like anti-cancer vaccines and viruses. When our researchers ﬁnd a therapy that seems promising, we quickly test the therapy for efﬁcacy and toxicity in animal models and get it into human trials; and when researchers need tumor samples our surgeons provide them without delay.
And yet, these scientiﬁc considerations, however compelling, cannot be our sole guide. Every clinical trial to get a new drug or therapy approved by the U.S. Food and Drug Administration (FDA) has three phases. Phase I tests ask whether the drug is safe for human use and what is the optimal dose. Here, both patient and scientist know that the real drug (not a control treatment or placebo) is being given, and escalating doses are used to ascertain how much can be used to attack the cancer without producing unacceptable side effects. Patients in Phase I trials are medical heroes, frequently the ﬁrst humans to try a drug previously tested only on animals—a drug that may have unknown and even deadly effects. Their chances of seeing improvement are impossible to quantify, although, of course, treatments are not tested unless there is good evidence from animal experiments to believe they are promising. If the drug is effective, these patients are ﬁrst to beneﬁt.
Treatments established as safe in Phase I move to Phase II, which asks what effect the drug or other therapy has on the disease or disorder. In Phase II, a single group of patients gets the new treatment and is closely followed to see whether it is working. Phase III is frequently a randomized trial of the kind described earlier involving many patients and comparing the new treatment to the present “standard of care”—the best treatment already in use. For conditions less serious than brain cancer, half of the patients in a trial may be randomly selected to receive a placebo—not the standard treatment—as the comparison, but this is rightly deemed unethical where a life-threatening illness is involved.
Since standard-of-care treatment for glioblastoma multiforme is only modestly effective, we view randomizing patients into that treatment as handing them a death sentence, which we refuse to do.
At the Brain Tumor Center, we take those ethical considerations further. Since standard-of-care treatment for glioblastoma multiforme is only modestly effective, we view randomizing patients into that treatment as handing them a death sentence, which we refuse to do. Instead, we may compare two promising new treatments— for example, a combination of radiotherapy, surgery, and a single newer drug versus a combination of radiotherapy, surgery, the newer drug, and a new pharmaceutical agent being tested. If we run these two treatments concurrently it is easier to compare the data and a success may stand out more clearly. For an additional comparison, we can use the historical rate of success in disease control achieved in patients who received the standard of care. We know that there are investigators who maintain that a trial is scientiﬁcally more rigorous when standard of care is used as a control group in Phase III, but we do not think that any extra precision gained is worth the cost to the patients with this disease. Of course, this risk-to-beneﬁt calculation would be different in the case of diseases not imminently fatal or for which an existing treatment was at least somewhat effective. Not so in many patients with brain cancer.
GOING “OFF LABEL”
Another challenge for an aggressive approach to treatment is what to do with patients not eligible for clinical trials. At the Brain Tumor Center we employ drugs approved for other cancers and that have shown efﬁcacy in treating brain cancer cells in the laboratory. This is called “off label” use. The drug is an FDA-approved medication, recognized to be safe and effective in treating some other cancer, but has not been tested and approved to treat brain cancer. The situation is not uncommon, because, as we have seen, the potential sales and proﬁts do not justify a drug company’s huge expense in gaining FDA approval for a new medication to treat an orphan disease. We go ahead and use off-label drugs when we have some evidence that they may help in a particular case. For example, we use a drug approved for colon cancer, CPT-11. It has certainly helped some brain tumor patients, but it does not score a knockout. We hope to build on the potency of CPT-II by combining it with other agents.
Using cancer drugs off label puts us at odds with insurance companies, who consider the treatment experimental and often refuse to pay for it. They claim that they can control medical costs better when they cover only treatments already proven effective for a particular condition in the large clinical trials necessary to win FDA approval. We might agree, if there were alternatives for our patients. But many patients cannot participate in clinical trials. For example, someone might have already had another treatment that would confound our analysis of results if we included that patient in a study. With standard treatment producing such an abysmal success rate, we think it is unfair not to offer that patient some other chance.
OPERATING ON THE “INOPERABLE”
Eileen Beckert’s tumor was located directly on the speech area of her brain. She had been told that it was inoperable. Because of the poor prognosis for brain cancer in general, and the propensity of tumors to grow back after surgery, many surgeons do not even try operations in cases like this. One result is that they do not develop the skill that comes with having done hundreds of them.
We believe in giving patients their best chance. As Allan Friedman puts it, time is precious. No one lives forever, but he sees his job as getting his patients as much quality time as possible. Many patients he sees have already been turned away by other surgeons.
Allan was an unlikely candidate for excelling at neurosurgery. The ﬁrst time he observed an operation, he fainted when the surgeon made the ﬁrst cut. He now says that, while you may never get used to it, the surgery eventually becomes fairly routine. Now, he does 12 operations in a typical week, each lasting about 4 hours.
To perform a difﬁcult operation like Eileen’s, a surgeon must visualize the procedure in his head before he begins. He takes every precaution to avoid mistakes and so goes in with conﬁdence. “You don’t think in terms of what you are going to do. You think in terms of what pitfalls you are likely to encounter and how to avoid or deal with them,” says Allan.
For example, bleeding is a major problem in brain surgery. With other organs, loss of blood or exposure to excess blood is not necessarily a problem, but loss of blood ﬂow to an area of the brain can quickly produce damage. Any chance that a surgical move will cause unexpected bleeding must be anticipated and, if possible, prevented. Here good preparation can make the difference between life and death, disability and normal functioning.
To achieve the most intense concentration during surgery, Allan has to distance himself from the person on the operating table. He says that there are still times during an operation when he catches himself thinking about the person and has to stop for a second to dissociate himself from the human attachment, which distracts him from the mechanical process. The patient, on the other hand, should be distracted if awake during the surgery. There is always a technician talking to the patient. Allan says: “I don’t want the patient sitting there thinking, ‘Oh, my God, my head’s open! What’s going on?’ They shouldn’t be left to their own thoughts.”
To avoid damaging the language centers of the brain, neurosurgeons electrically probe areas that they consider cutting to see if disrupting the signaling at that location interferes with speech. If a man speaks several languages, he must be tested in each one because each may be localized in a different brain region. Probing one area may show no effects on the ability to speak English, for example, but cutting it might bar the person from speaking Spanish. One patient of ours was a rabbi who spoke English, Hebrew, and Yiddish, each localized to a different area of his brain. You do not want a rabbi waking up no longer able to speak Hebrew!
Because we deal with the brain, the fear factor among patients is high. Many feel sure they will not be able to tolerate being conscious during the operation. Yet, during hundreds of surgeries he has done on conscious patients, Allan has never had someone become unbearably distraught or agitated. For the most traumatic parts, such as opening the skull, the patient is under general anesthesia; but any time a patient is conscious there is someone in the operating room whose sole task is to talk to the patient.
Eileen Beckert, who was a 47-year-old mother of four children when she was Allan’s patient, said: “The day before, I was really nervous. I thought: What if I don’t wake up? I wrote a letter to my kids, just in case.” Though she was conscious for parts of her operation, Eileen cannot remember much of the operation itself: “I do remember lying in pre-op and just praying. And I remember at the end, I couldn’t talk for a while.”
In fact, for several weeks after the operation, as she recovered in the intensive care unit, Eileen could not initiate speech. Her husband, Scott, says: “I would say ‘Simon’ and she could say ‘Garfunkel,’ but she couldn’t start.” This is probably because the areas for planning speech were temporarily disconnected from the areas that execute speech—the result either of the surgery itself or of postoperative swelling that interfered with the connections. When she was cued by external speech, Eileen’s speech-planning areas did not have to come into play.
Eileen also endured three terrifying days when she entirely lost the ability to speak, probably as a the result of brain swelling during recovery. Her speech is entirely normal now, nearly two years after the surgery, but she has some problems with concentration and focus that prevent her from returning to her job (ironically, as an oncology nurse). Treatment for cancer, and sometimes the tumor itself, in any region of the brain, can cause this problem. Eileen has completed her radiation and chemotherapy treatments and at her latest checkup showed no sign of tumor regrowth.
In Eileen’s case, an aggressive approach to treatment succeeded where to some there had appeared to be no hope. Unfortunately, the news is not often as good as it was for Eileen. We recently had to tell a mother that her two-year old child’s brain tumor had failed to respond to treatment. Even with consistently aggressive treatment, only 10 to 15 percent of our patients with glioblastoma multiforme survive three to ﬁve years, although we have some survivors in the sixth, seventh, or eighth year after initial diagnosis.
We are asked why we do not run clinical trials of the drugs and methods we are using for patients who are not already in a clinical trial program. If we began formal testing protocols with these patients, we would be locked into a particular treatment for the duration of the testing. That treatment might not always be the best option, which really matters if you are one of the roughly 10 percent of patients who survive much longer than you would have with traditional care.
Darell Bigner’s research team is attacking brain cancer on several fronts. One is immunotherapy. How do you get the immune system to recognize and attack the cancer? This would spare healthy tissue and avoid the many toxic side effects associated with current chemical, radiation, and other therapies. We are exploring the use of monoclonal antibodies, which are proteins our bodies produce that attach themselves to an invader as one step in the immune system’s recognition of its enemies. At ﬁrst, we made the antibodies by injecting mice with material from a patient’s brain tumor, collecting the antibodies the mouse made in response, and infusing them into the patient. We have advanced now to developing human antibodies and attaching them to radioactive molecules or toxins that harm the tumor when the antibodies home in on it. Collaborating with the National Cancer Institute, our Center has treated about 400 patients this way—more than any other medical center. We see promising results for longer survival and better quality of life, but we are not at the ﬁnal stages of testing. With several different antibodies and different groups of patients we are in Phase II trials. This year, we will also start a new Phase I study.
Another line of research employs dendritic cells, which are like commanders of the immune system, telling the soldiers what to attack and what to leave alone. Again, the complexity of the brain makes this vaccine-like approach difﬁcult. When researchers are trying to teach dendritic cells to recognize tumors, one common technique is to grind up the tumor and present it to dendritic cells taken from the patient. Animal experiments suggest, however, that brain tumors have antigens on them that, if targeted by the immune system, could lead to a disease similar to multiple sclerosis (MS). In MS, the immune system attacks healthy areas of the central nervous system.
To avoid that problem, we have identiﬁed a protein, EGFR variant III, that is found only in brain tumors, not in healthy tissue. This protein does not induce the MS-like condition in animals when it is used as a target for dendritic cells. We take a patient’s dendritic cells, prepare them so that they will attack this protein (by exposing them to a synthetic version of EGFR variant III) and return the cells to the patient.
So far, we have done this with 10 patients, who have shown no signs of toxicity or MS-like disease. Preliminary tests indicate that they have developed immunity to the protein and have T cells active against it. It is too early to determine what effect this has had on their tumors.
We are also testing new radiation therapy that makes use of a radioisotope called astatine 211, which is 1,000 times more active against tumors than iodine 131, now used for radiotherapy. We have used this new isotope to treat 17 patients with glioblastoma multiforme and seen lengthened survival—in fact, some of these patients are our long-term survivors—and better quality of life.
A fourth key area is genetics. What genes are responsible for turning a normal cell into a cancerous one? Some genes prevent cells from going wrong; others promote tumors. We are looking at one gene, called MRP3, that is involved with drug resistance. If a tumor cell has this gene, and expresses it, then as soon as a toxic agent like a chemotherapy drug enters that cell, the cell is able to pump it back out. Obviously, we want to ﬁgure out how to block this gene.
Another gene we are studying allows the cancer to spread from one cell to another. In other words, it makes malignant cells migrate, so that surgery often fails to cure the cancer. Metastasis is a terrifying word to the cancer patient. We often speak of whether surgery “got all of the cancer,” but with brain tumors you cannot strive for what are called “clean surgical margins.” If you cut away extra tissue around the cancer to ensure that you have removed all the dangerous cells, the surgery itself can cause paralysis, speech loss, or other problems. You do not want to take out one more brain cell than necessary. If we can kill cells that express the gene that causes cancer to spread, or if we can inactivate the gene within those cells, we might prevent metastasis or a recurrence of the primary cancer.
A ﬁnal research area is signal transduction, which, in brain cancer, is how internal cell signals go awry and turn a normal cell into a cancerous one. Some of our scientists are studying brain cells called astrocytes, which ordinarily support neurons but can become cancerous astrocytomas. What genes induce that change? So far, our researchers have been able to turn normal brain cells into tumor cells in a test tube, and it turns out that at least three molecular abnormalities must be present before that transformation occurs. Now we can test each gene involved in the transformation to discover what it contributes to the process of cancer development and then how to interfere with that process.
LIVING AND DYING WITH HOPE
Research will soon bring us better treatments. We are immensely optimistic. But the question we are asked most often is: Are you offering false hope to patients now, today, by treating their tumors so aggressively?
I have tried to characterize in scientific and ethical terms why we favor aggressive treatment, but how does it make patients and families feel about their chances? We have one good yardstick. The families of patients who have died should be our harshest critics, but rarely has a family said we gave them false hope. They are glad that their loved one preserved a sense of optimism about the future and got the best shot at prolonging it.
When we say that “At Duke, there is hope”—the motto of our center—we mean more than the potential for physical survival, although we ﬁght ﬁercely for that. We also mean there is hope that the patient’s remaining time, long or short, will be worth living. We believe that a patient’s psyche can affect this disease. When a doctor at home tells a patient that nothing can be done, the patient and his family believe that he is half-dead already.
If a tumor is not responding to treatment and we must conclude that further attempts will fail to improve the patient’s quality of life, we say so. But our team is devoted to quality-of-life concerns and does everything possible to ensure psychological and spiritual support, as desired, for both patient and family. Our family support center is the model for dozens of similar centers in hospitals across the country. We conduct research on our quality-of-life services to see which ones help patients most, and so we can use our resources most effectively. What I am saying is that “hope” changes as situations and expectations change. Like a quilt that keeps you warm and comfortable when it is dark and cold, hope may be patchwork, but there are distinct patterns.
As for us, the Center’s staff, we often battle death, giving no quarter, and then must turn and accept it. Cancer medicine is a ﬁeld with tremendous burnout. You can absorb only so much sadness and keep responding. I can tap into the feelings of loss, but—like many doctors who treat often-fatal illnesses—I also have a protective mechanism that enables me to go on providing care and not lapse into sadness that would get in the way of the medicine. Because I love sports, I believe in the mind-set of a team ﬁxed on victory. If you play or watch basketball, you know that a team can change the game’s outcome right up till the very end. We want every remaining moment to count for a patient, and every patient has a right to the best that science and spirit can bring to the ﬁght.