Cancer Institute Network Aims To Personalize Cancer Treatments
Cancer Biomedical Informatics Grid matches up patients with the right experimental treatments.
The Cancer Biomedical Informatics Grid, or caBig, research initiative is helping researchers at leading U.S. cancer centers share clinical data and research results to develop more personalized cancer treatments.
Built five years ago, caBig provides shared applications, data standards, and data models to its 100 members--government and academic research centers, plus national and community cancer centers involved in research and clinical trials of experimental treatments. The underlying infrastructure is a service-oriented architecture with more than 100 grid nodes connecting the participating organizations. Researchers use the network to manage repositories of biological specimens and medical images, conduct molecular research, oversee clinical trials, and develop analytical tools to work with all the data they're collecting.
"We want to get to a point where all raw data is connected and then put business analysis-type tools on it," says Ken Buetow, director of the center of bioinformatics and IT at the National Cancer Institute, which is part of the National Institutes of Health.
The University of California San Francisco Medical Center is one caBIG participant providing raw data. It launched the Athena Breast Health Network in September in which it plans to collect data from 150,000 women who will be screened for breast cancer and tracked for several decades.
Data will come from medical records, regular surveys of participants, and a bio-specimen repository that will track mammograms, tumor specimens, and blood. The project aims to identify what types of women are most at risk for specific kinds of breast cancer and what treatments work best in women with specific medical and molecular profiles, says Dr. Laura Esserman, a UCSF physician and a professor of surgery and radiology.
The goal is to speed up advancements in breast cancer prevention, screening, and treatment, and ultimately help women survive the disease by "compressing the time it takes to implement innovations into clinical practice," Esserman says.
Other caBig participants are using the network's data and analytics software to identify the best patients to participate in clinical trials of experimental cancer treatments. Researchers assess patients' medical histories, genomic data, age, and other factors to weed out people unlikely to have good outcomes, so they don't waste time on a trial that isn't suited to their needs, Buetow says.
Drugs can have different effects on patients even if they have the same disease. For instance, the cause of a person's cancer--smoking versus other environmental or genetic factors--can affect how well the patient responds to a treatment. The caBig network lets researchers more easily weigh all the complex factors involved. The clinical decision-support tools help medical researchers tap into "rich data on subtypes of cancers and survival outcomes," Buetow says.
If doctors can quickly identify the best candidates for trials, it could help increase participation in research. Only 5% of adult cancer patients participate in clinical trials, compared with 67% of childhood cancer patients, Buetow says. That level of involvement can lead to faster advancements. "Thirty years ago, leukemia for kids was a death sentence," Buetow says. "Now survival rates are 80% to 85%."