Taking informatics approach to lung cancer treatment
Massive data crunching effort may help clinicians pick best treatment pathsMay 9th, 2013
Experts from The Ohio State University and University of Kentucky are collaborating to create what they claim is a first-of-its-kind biomedical informatics-based approach to help improve survival rates in people with lung cancer, which is the leading cancer killer among men and women in the United States.
The approach combines two different areas of biomedical informaticsdigital imaging and genomic sequencingto match patients with treatments most likely to extend their survival.
TheAmerican Lung Association says that more than half of people with lung cancer die within a year of being diagnosed.
"Lung cancer treatment choice hangs critically on how a pathologist classifies certain cancer cell traitsthe phenotype. But this interpretation can be highly subjective," says Dr.Kun Huang, associate professor ofbiomedical informaticsat Ohio State's College of Medicine.
"We have genomic data that tells us what treatments might work best for a specific person, but that doesn't tell us how aggressive the cancer type may be," Huang says. "So clinicians today are making decisions on the best available data, but it's an incomplete set of information."
The data from one individual's genome and lung cancer phenotype equals a little more than four gigabytes, about as much data as is contained on a full-length DVD movie. Huang and his research partner, Dr. Lin Yang, of University of Kentucky, will be working with computers and programs that are able to crunch and compare more than 15 terabytes of data, the equivalent of about 3,000 DVDs.
"A pathology report gives you one set of insights, a person's genome another. This model will bring the power of both together to help clinicians select the right treatment for the best outcome," asserts Yang.
Huang and Yang initially will create their model using a repository of tissue from more than 4,000 lung cancer patients from Appalachia, and genomic data from a National Institutes of Health database,The Cancer Genome Atlas.
The team hopes to start sharing results from their work in one year.