STRENGTH
IN NUMBERS
UF Is Home To National Statistical Data Center
For Children’s Cancer Research
By Melanie Fridl Ross
Like
most parents, Lisa and Keith Cannady can proudly recite all the important
facts and figures about their daughter. Marissa is 6 years old. She was
born on the 4th of July. She weighed 7 pounds, 4 ounces at birth, stood
39.5 inches tall four years later. And so on.
But the Cannadys’ data bank delves even deeper. They also monitor
an entirelydifferent set of digits and dates, numbers that signal how
Marissa is faring in the battle against cancer. The size of the tumor
nestled in her right thighbone. Her latest blood counts. The drug doses
that comprise her chemotherapy cocktail. The odds she’ll survive
the battle.
While these statistics merit careful review by Marissa’s parents
and doctors, they also receive attention somewhere else: the Children’s
Oncology Group Research Data Center at the University of Florida International
Center for Childhood Cancer Research, the first stop for research data
from virtually every pediatric cancer patient in the United States and
beyond.
The Children’s Oncology Group is a newly created, federally supported
consortium that pools resources to design, conduct and analyze studies
of pediatric cancer therapies worldwide. Clinicians throughout the United
States, Canada, Switzerland, Australia and parts of Europe funnel their
data to UF for analysis.
The reason is simple: During the past 25 years, many of the major treatment
breakthroughs have arisen out of the collaborative efforts of research
groups around the world. Such 
cooperation,
UF officials say, helps produce faster research results and hastens the
discovery of cures.
The information pours into their offices
at a dizzying rate. Using sophisticated computational tools, they scrutinize
data from about 240 institutions participating in randomized clinical
trials to help determine what is working and what isn’t. Is a once-incurable
cancer becoming vulnerable to a new drug regimen? Are kids responding
poorly to a certain medication? Is a therapy enabling them to spend more
time in school or avoid side effects?
“Back in the ‘50s it was recognized
that cancer in childhood is rare. And cancer in children often has many
different manifestations we don’t see in adults,” says pediatric
oncologist John Graham-Pole, a professor of pediatrics at UF’s College
of Medicine and the UF Shands Cancer Center, and principal investigator
for the UF clinical membership of the Children’s Oncology Group.
“To learn anything about this disease, you need the numbers. One
institution alone could not possibly have enough patients. It became evident
that the only way to work was to network nationally and, increasingly,
throughout the world.”
The cooperative approach has been scientifically
proven to increase survival rates in all forms of leukemia, lymphomas
and solid tumors. In addition, the research has helped practitioners identify
improvements in the treatment of childhood cancers, and contributed to
a better understanding of cancer cause and development.
UF statisticians design research protocols
and aid in the development of research questions. They also monitor the
performance of member institutions and conduct statistical analyses as
data are gathered. On average, 100 clinical trials are under way on any
given day, covering all major childhood malignancies, according to Brad
Pollock, an associate professor and interim chair of the College of Medicine’s
Department of Health Policy and Epidemiology at UF, and associate director
for cancer control at the UF Shands Cancer Center. Pollock also has been
appointed to the Medical Scientific Advisory Board for the National Childhood
Cancer Foundation, the organization that sponsors the new group.
“This puts us on the map in terms
of being a major institute for pediatric cancer research,” Pollock
says. “Harvard, Johns Hopkins, Yale, USC, Stanford ... every major
medical center with a childhood cancer research program belongs to this
consortium. Collectively, about 65 percent of all children with cancer
get placed on one of these treatment protocols. This is the standard of
care for childhood cancer.”
The National Cancer Institute gives
UF about $2.5 million a year for its pediatric cancer research programs.
Approximately $2 million of that supports direct costs of the data center,
headed by statistics professor Jonathan Shuster and staffed by 32 statisticians,
systems analysts, data managers and administrators. The other half million
is earmarked for research on pediatric cancer prevention and control.
The data center,
part of the Department of Statistics in UF’s College of Liberal Arts
and Sciences, serves as the repository for the research data submitted
by member institutions and by centralized laboratories and tumor banks
that analyze cancer specimens.
Center officials track timeliness of data submission and police quality
of the institutions’ involvement in the studies. They leave no stone
unturned, analyzing whether appropriate laboratory tests have been done,
how drugs were administered and monitored, and how various malignancies
were categorized based on pathological data.
To aid in these efforts, UF has developed a new Web-based system that
researchers use for remote data entry.
The Children’s Oncology Group includes pediatric oncologists, surgeons,
pathologists, psychologists, radiation oncologists, nurses, pharmacists
and clinical research associates who pool clinical case material and laboratory
resources, and serve on numerous disease, discipline and scientific support
committees.
The Pediatric Oncology Group, the Children’s Cancer Group, the National
Wilms Tumor Study Group and the Intergroup Rhabdomyosarcoma Study Group
united last spring to create the organization, now the largest pediatric
cooperative group in existence, consisting of about 240 hospitals, cancer
centers and academic medical centers.
UF has been a member of the Pediatric Oncology Group since it was formed
in 1980 and was the statistical center for that organization as well.
The merger reduces the number of NCI-supported cooperative research groups
from 12 to nine; the others are dedicated to the treatment of adult cancers.
A Bump On The Knee
Marissa’s ordeal started when she was in kindergarten, after she
bumped her knee on the playground.
“She told us about it when she came home,” says her father,
Keith. “She said her knee was bothering her a little bit, but after
that she seemed fine. Two weeks later she bumped her knee again on the
ottoman in our house, and my wife took her to have her knee checked. We
saw a little swelling, and she was favoring it a bit. The doctor where
we lived said it was a bruised muscle.”
After her discomfort persisted, the Cannadys took their daughter to two
other doctors. After one raised the specter of cancer, they rushed Marissa
to the UF Shands Cancer Center from their home in the Florida Panhandle.
Graham-Pole confirmed their daughter had osteosarcoma, a form of bone
cancer, and within days she began chemotherapy, enrolling in a national
trial designed to improve treatment for these patients.
Previous attempts to increase the amount of chemotherapy given raised
the risk of toxic effects. So Marissa and other study participants also
received a drug designed to protect the heart.
“He said, ‘This is what’s available: We can give her something
that has worked in the past or we can try her on something using different
doses, and she’d probably spend less time on the treatment,’”
says Marissa’s mother, Lisa. “We looked at that and thought
that would be good because we wanted to get this over with. We wanted
it to be effective, but we wanted to get this over with.”
A few months later, after the drugs had begun to shrink the tumor, Marissa
underwent an above-the-knee amputation using a technique to preserve mobility.
Once her lower leg and knee joint were removed, the ankle joint was rotated
and attached to become the new knee joint. Marissa wears a short prosthetic
limb to complete the lower portion of her leg. Her prognosis is good.
Behind The Numbers
Shuster will likely never meet Marissa. But he never forgets who’s
behind his work.
“As statisticians, we analyze numbers, but we’re never unaware
of the children,” he says. “You never lose sight of that. That’s
who those numbers represent.”
Shuster and his colleagues act as watchdogs, studying the data as they
come in to see which treatments are working … and which aren’t.
“You just want to know what the truth is out there,” he says.
“In a nutshell, we’re there to make sure that the study is not
doing something that is unexpected.”
Shuster, also a professor of statistics at UF, thinks back to 1994, to
a study of about 300 children with leukemia. Researchers were evaluating
whether outcomes could be improved by altering when certain drugs were
administered.
“Patients started having seizures in fairly large numbers, and because
we had a finger on the pulse of the study — we watch our studies
so closely — we found it really quickly,” Shuster says. “That’s
where our collective experience helps. If everyone is feeding their data
into the system in a timely way you can find out about the problem and
you can correct it much faster, and when you make a correction it goes
national.”
Statisticians also have helped uncover drug interactions and evaluated
findings that ultimately revised the standard of care. Thanks to seven
years’ worth of data, about 20 percent of children with neuroblastoma
now undergo successful surgery and are spared chemotherapy or radiation,
which used to be the traditional therapy, Shuster says. About 95 percent
of them are cured.
In addition, findings from a study dating to the early 1980s showed a
combination of three drugs placed directly into the spinal fluid was much
more effective than a single drug for patients whose leukemia recurred
in the central nervous system.
“We had to curtail the study and give the patients who were on the
other treatment the triple drugs, which just did a much better job of
protecting the central nervous system,” Shuster says. “If you’re
watching a study carefully and you have large numbers, when one treatment
turns out a lot better you find out soon enough so patients on the other
treatment can benefit.”
Since the 1950s, cancer mortality in children has dropped 55 percent,
and more than 70 percent of children with cancer are cured and grow up
to be healthy adults, says Shuster. In comparison, the death rate has
declined just 10 percent in adults, only 1 to 2 percent of whom enroll
in group trials.
As cure rates continue to rise for children with cancer, research efforts
have been broadened to look at the long-term effects of cancer therapy
and how it affects the lives of childhood cancer survivors.
“Some of the drugs, although they cure children of cancer, leave
them with heart problems, learning disabilities and other difficulties,”
Shuster says. “This is another area we’re very concerned with.
Curing them is not enough.”
Says Lisa Cannady, Marissa’s mother, “You don’t want anybody
to have to go through what your child had to go through. You definitely
hope that whatever the findings, good or bad, they’re going to help
another child along the way.”
John Graham-Pole
Professor, Department of Pediatrics
(352) 392-5633
grahajr@peds.ufl.edu
Brad Pollock
Associate Professor and Interim Chair, Department of Health Policy and
Epidemiology
(352) 395-8040 x8-6538
brad@hpe.ufl.edu
www.cog.ufl.edu