Shared Resource: Biostatistics
In many ways, rocket science has nothing on cancer research. Rocket science is firmly based on the reproducible domain of physics. Rockets and fuels can be replicated en masse, and identical rockets will fly the same distance the first time as they do they do the thousandth.
But most cancer research takes place in a living laboratory. Even if you have a refined research topic in place--such as a new treatment or an at-risk population, this laboratory can present innumerable challenges to a study.
Some of the questions researchers at the DCI face every day include: How many patients of a rare cancer should be studied, and for how long, to estimate how well a treatment works? For rare diseases, how can a feasible design be formulated? What if the treatment is harmful or has side effects? Should fast- and slow-growing cancer types be lumped together or treated as separate categories? How can a researcher recruit underrepresented populations? And how can you assure that the data I collect will provide the information I will need to answer my question?
Fortunately, the Biostatistics Shared Resource exists to answer these (and other) questions. With full-time faculty and staff assigned to work in every program area, the group is the largest shared resource within the Duke Cancer Institute.
“Our shared resource is a bit different than others within the DCI,” said Terry Hyslop, PhD, director of the Biostatistics Shared Resource. We don’t have any machines or equipment. Instead we have a full cadre of people who work with investigators in everything from basic research to clinical trials to preparing adverse event reports. We run the whole gamut.”
Hyslop’s own busy schedule reveals the breadth of the shared resource’s work. A typical day might include regular research meetings with investigators about future clinical trials, abstract deadlines or data analysis, a discussion with Dan George, MD, and Steven Patierno, PhD, about using molecular analysis to better understand cancer disparities by integrating biology with socioeconomic status, and drafting a grant to examine populations at high risk for breast cancer for Dr. Hyslop’s own research.
“People come to us from all over the DCI with promising new research areas or ideas. We work to make that research the best it can be. That means a lot of thinking, designing, analyzing and thinking again during all stages of research,” Hyslop said.
This foresight and analysis will be especially useful for future research about the use of a re-engineered polio virus to treat glioblastoma, a rare form of brain cancer. A clinical trial featuring this technique recently gained national attention when it was covered on the news program 60 Minutes (read the story here).
James Herndon, PhD, a biostatistician who works with the Preston Robert Tisch Brain Tumor Center, and a member of the Biostatistics Shared Resource, has been working with Annick Desjardins, MD, Matthias Gromeier, MD, and other researchers on the conduct of this trial.
“The results so far have been promising, but we have so much more to learn,” Herndon said. “Right now I’m helping the team lay out the study design, clarify logistical issues and study objectives, and determine the number of patients needed to address these objectives. Biostatisticians Frances McSherry, MA and Patrick Healy, MS, are also part of the team that collaborate with Brain Tumor Center investigators and help them interpret their research data.”
“I see the job of our shared resource as a constant collaboration,” said Herndon. “There’s a high level of respect that runs back and forth between all the people we work with. We’re constantly talking back and forth, whether a project’s about to be finished or about future research further down the road.”