An unusual collection: A brain tumor tissue bank

long read

Physician researchers at Dana-Farber on a race to personalize treatment for glioblastoma patients

Five years ago, as she was walking into Caritas
Holy Family Hospital and Medical Center in Methuen, Mass., Patricia Fay saw a
priest she knew and cornered him. “I’m like ‘Oh, Father Peter! And I sort of
grabbed him by his arm,” she recounts.“I said, ‘What are you doing here? Father
Peter! I could use a prayer right now. He asked me what was going on and I told
him, “They found a brain tumor and I’m about to get set up for radiation. It’s

“Father Peter put his hand on the top of my head,
closed his eyes, and started saying a prayer,” Fay continues. But all she could
think was, “Oh no! He’s blessing the wrong side!”

Today, a piece of the tumor the priest prayed
over, along with many like it from other patients, is living in a laboratory at
the Dana-Farber Cancer Institute where researchers are studying
the tissue samples, hoping to use  them to tailor  treatments for

Harvard Medical School (HMS) and Dana-Farber Cancer Institute neuro-oncologist Santosh Kesari, M.D., Ph.D. and colleagues are
beginning — on an experimental basis — to venture into the challenging and
possibility-filled world of personalized medicine. “Studying (the living tumor)
outside of the body could be more fruitful scientifically and directly
applicable to each patient instead of working with a static dead sample,” says
Kesari’s colleague, Keith L. Ligon, M.D., Ph.D. assistant professor
of pathology at Harvard Medical School, the Brigham and Women’s Hospital and a
consultant in neuropathology at Children’s Hospital Boston.

Kesari, an assistant professor of neurology at
HMS, and Ligon have established a tumor bank at Dana-Farber, called the
DFCI/BWH Living Tissue Bank, which contains living tumors cultured from
patients’ tumor cells.

“Part of the reason we started working on this
tissue bank was just the obvious potential for using such cells for both
research and potentially clinical care further down the line,” Ligon says.

“It’s directly translating science to the
patient,” Kesari adds.

Kesari and Ligon first met six years ago when
they were both postdoctoral fellows at the Dana-Farber Cancer Institute. 
Back then, they studied the development of glioblastoma
brain tumors
in mice. The work soon evolved into the analysis of
this type of tumor in humans, assisted by new discoveries on ways to culture
glioblastoma-type tumor cells. Researchers found a way to effectively isolate
cells from these malignant masses, a discovery that led to the Living Tissue
Bank project for Kesari and Ligon.

For Ligon, “the type of person who likes to
collects things,” and who used to collect stamps and comic books when he was a
boy, it was only natural. For Kesari, the project has recently taken the
undertone of a personal crusade where his most important, though not the only,
contender is time – a family member has been diagnosed with glioblastoma.

For years, Kesari had remained a bench
researcher, focused solely on doing basic science far removed from patient
care. “But then you know patients are dying…weekly,” he says. “I realized that
with the recent development of new technologies: genomics, proteomics- there’s
a possibility of bringing these to inform decisions in the clinic. We see
patients every day, and there are things we can learn from them individually
and from translational studies that you can do just from the blood of patients
undergoing treatment with specific drugs or the specimens that we get at

Researchers have learned over the years that a
brain tumor is not “one thing,” that every patient has a different “flavor,” as
Kesari says, even in the way the illness presents itself.

Take Patricia Fay, for example. 

In 2005, while participating of her daughter’s
eighth-grade graduation, she began experiencing an intense headache. “I had
been getting these headaches, lots of headaches,” she explains. “They didn’t go
away. That day I couldn’t stand it. “It was like something pressing on my

Fay was diagnosed with glioblastoma, one of the
most common malignant brain tumors. Hers was a grade 4, the most lethal.
Although the prognosis for a patient with glioblastoma usually is grim (12
months), it’s been five years since Fay’s diagnosis.

Ephraim Friedman, a physician from Beverly Farms,
Mass., also became a patient five years ago when he was diagnosed with a brain
tumor following a seizure that left him unconscious. Today, with slightly
slurred speech and some issues with his short-term memory, Friedman is well
enough to work on the sculptures he loves to create and says he is doing well.

These long-term survivals are great news for both
Kesari and his patients. But Kesari and Ligon want to better understand the
science behind these success stories. “They probably have a particular kind of
glioblastoma that we don’t understand well enough,” Kesari says.

The problem, according to Kesari is
“lumping” patients into one disease when “… the reality is, you look
at the genetic profile of individual glioblastomas, they are very different
from one another,” says the investigator. “Each person’s tumor is actually
different even though it’s called the same thing. And each person responds
differently to different drugs. So if we understood the connection between the
genetic makeup of the tumor and the response to specific treatments that we
have at hand, then we would be able to make more rational, personalized
treatments for patients.”

Glioblastoma is also as nondiscriminatory as it
is varied.

“This is one of the cancers that span all ages,
from childhood to old age,” Kesari says, “That’s why it’s an important disease
to understand, because it affects everyone.”

“The common concept in the past was that a tumor
was basically just a clone of itself,” Ligon says. “Yet in certain kinds of
cancer, for example glioblastoma, when you look at the patient’s tumor, you
notice cells don’t all look the same. It’s a heterogeneous mix when it starts.”

What researchers have found is that different
cell types exist within an individual tumor. These cells may be highly
genetically related, they may look different, or they may have different
genetic makeup as well. It is this heterogeneity what makes “killing” the tumor
so difficult. Drugs used in therapy can partially shrink the tumor by
terminating a certain group of cells while other types of cells remain behind and
replicate, thus maintaining the possibility that the tumor will propagate. One
drug doesn’t target all cell types at once.

“The bottom line is that not every tumor cell is
the same as its neighboring cell. That’s where the cancer stem cells come in,” Ligon says,
referring to a controversial theory holding that not all tumor cell types are
important for tumor growth.

 “The question is what are the most
important ones to kill within the tumor?” Ligon says.

The cancer stem cell theory, as it is known, is
that there is a small population of  “mother”cells with the potential to
survive therapy. Much like the body’s normal stem cells, capable of generating
the different tissues that comprise the body, these “mother” cells, or cancer
stem cells, would reproduce infinitely turning into powerful, malignant
“daughter” cells. The daughter cells might die with standard treatment but stem
cells may survive since such standard treatments are not yet designed to
destroy the stem cell population.

The adult brain retains a certain number of normal
stem cells over a person’s lifetime. These cells help regenerate the brain, and
are possibly involved in memory and cognition, but some scientists think it is
possible that, when lacking, they might also be causing degenerative diseases
such as Alzheimer’s, and Parkinson’s. “And the other flipside is,” Kesari
says, “if they go wild, stem cells could also be involved in the formation of
brain tumors. That’s something that’s not proven, though.”

And it’s a hard task because, if they exist,
these sought-after cancer stem cells could be rare, making up approximately one
in 100,000 tumor cells. Currently, scientists have two preliminary ways to
identify, isolate, and grow the suspected cancer stem cells.

The first involves dissociating the tumor in a
culture dish. “The ones [cells] that continue to grow and make spheres, tumor neurospheres
as we call them, are the ones that functionally are defined as cancer stem
cells,” Ligon says.

The second involves inserting the cells inside
the brains of mice with immunological systems that have been previously
compromised to prevent them from rejecting the inserted human cells. “You put
those cells in the mouse, and presumably the cells that grow out are the cancer
stem cells,” says Ligon.

Kesari and Ligon have been growing their
patients’ cells in the same media other stem cell scientists use to culture
‘traditional’ stem cells. So far, it only works about 50% of the time, and when
it does, the medium keeps the suspected tumor stem cells alive and in the same
state as they would be found inside the patient’s brain.

“Many people,” says Ligon, “are doing research to
try to study (the cultured cells) based on their genetic characteristics or the
markers on their cell surface with the idea of separating the different kind of
cells and purifying them.”

Instead of giving the patient an ‘off the shelf’
treatment, a more specific ‘lumping’ of the disease would allow a physician
scientist to take part of the tumor to a laboratory and test all available
treatments on it in order to identify the one that would be most

“Conceptually, we should be able to develop sensitivity
tests to different drugs, such as we routinely do for infectious agents to pick
the right drugs that will work for each patient” Kesari says.

A simple concept that the team of researchers is
currently embarking upon by initiating a series of clinical
with the potential to help, at least, a certain population of
patients with glioblastomas made up of cells that exhibit a specific kind of
surface marker.

Inside his office space, an excited Kesari points
at his computer monitor showing a brain image of a patient’s glioblastoma
tumor. “See this mass in the brain,” he says, “both sides of the brain should
be the same but clearly there’s a big mass here causing a lot of problems for this
patient. The patient had several surgeries, radiation, chemotherapy, and several other clinical trial
drugs. They failed every time.”

Having no other choice Kesari gave his patient a
drug called Imatinib, which has the effect of inhibiting the expression of a gene
called PDGFR. “This drug surprisingly had a dramatic effect. You can see the
shrinkage of the tumor in one month, which is very unusual.”

The researchers were surprised to look at other
patients who in the past had been treated with the same drug. “This drug was
thought not to work,” Kesari says, “but in fact the few patients did respond.” 

What Ligon and Kesari found was not a miracle
drug, but one possible way to predict whether a tumor would respond or not to a
specific treatment. “… because when we looked at this further we found a
biomarker that predicted response of this patient’s tumor.”

Kesari’s and Ligon’s inspiration for the upcoming
clinical trial that will use the biomarker to identify a specific kind of
glioblastoma, began with their interaction with another of their patients. “We
were treating this patient who responded well to the drug,” Kesari says, “we
tried of figure out why, and in the process found the biomarker.”  

 “Now, we expect not to be saying ‘oh, the
drug failed when given to all brain tumor patients’” Kesari says. Knowing that
only patients with specific markers would respond to the drug would allow
Kesari and his team to be more rational in their treatment decisions. “…every
patient [would] be screened [for the marker] so if only 1 in 10 patients have
the marker, we’re going to only enroll that subset onto the trial and we would
predict that 100 percent of those selected patents will now respond … It’s a
concrete example of personalized medicine,” Kesari says. A clinical trial will
be starting in the near future to test this hypothesis.

These days Kesari’s project has turned more
personal. When he started in the field six years ago, an uncle of his died of a
glioblastoma. And now a very close aunt has been diagnosed with something that
at least by “name” is the same illness. “She’s one of my favorite aunts. Six
years ago it was personal, but now this is very important for me to do,” he
says. “So I’m doing everything I can to figure this out.”

Kesari is going “full steam ahead to personalize
her treatment,” as he says. The culturing of tumor cells, growing tumors in
mice brains, molecular and genetic testing – it will all get done. “Although
she’ll get the standard treatment of radiation and chemotherapy, we are going
to try to do as much testing as we can to see whether there are any of the
markers that can help us predict response and the best drugs for her,” he says.

“We are trying to do it fast [despite] barriers
that are always in the way,” says Kesari. “Everyone knows the right thing to do
but you need support-institutional, financial, etc.-to do it, you need
infrastructure, and right now, we’re just doing it ourselves. We really have no
choice but to do it ourselves.”

And if he and his colleagues find something that
really works? “The clinical trial will be one proof of concept. If the trial
works, it will be big news. Then if we can find something for her [his aunt],
we’ll certainly publish it and make it known for other patients. If we can find
out what makes her tumor tick, what drug will work, that’s going be the basis
for another trial to treat another subset of patients with that specific
glioblastoma,” Kesari says. “That’s the only way to do it. You have to look at
each patient, each tumor, because each one is different. It’s not just one
disease. We’ve learned that from a lot of other cancers. You just need to do it
in a brute force fashion, I think.”