October 26, 2007

In pursuit of personalized medicine

Featured Image

Dan Roden, M.D., left, and Daniel Masys, M.D., are leading a multi-center study of how electronic medical records systems can be used for large-scale genetic research. (photo by Neil Brake)

In pursuit of personalized medicine

Vanderbilt University Medical Center has been selected to participate in and coordinate a national network that will pursue the ideal of personalized medicine — health care “tailored” to each individual's genetic makeup.

The network of five sites, collectively awarded $20 million by the National Institutes of Health, will evaluate how electronic medical record systems can be used for large-scale genetic research.

Principal investigator Dan Roden, M.D., said Vanderbilt was well-positioned to compete for the new grant because of the institution's investments over the last few years in the DNA Databank Resource, an anonymous collection of genetic and clinical information.

“The idea was to create a large collection of DNA samples that are associated with clinical information, and to do that in a way that preserves patient privacy and yet allows investigators to access the very rich information that we have in the electronic medical record,” said Roden, assistant vice chancellor for Personalized Medicine.

“It is gratifying that the NIH has recognized our efforts in conceptualizing and building the DNA Databank,” said Harry Jacobson, M.D., vice chancellor for Health Affairs. “We expect that this resource will enable researchers to discover genetic links to disease susceptibility and drug response that will ultimately personalize and improve patient care. Our participation in this new consortium will move us toward that goal more quickly.”

Sample collection for the DNA Databank resource, which is directed by Jill Pulley, M.B.A., launched in February, and Roden said he expects the resource to hold 50,000 samples by next May.

Each center in the new consortium, organized by the National Human Genome Research Institute (NHGRI) with additional funding from the National Institute of General Medical Sciences, will study the genetic variation underlying a particular human trait using a technique called genome-wide association analysis.

“The NHGRI essentially challenged institutions to come up with proposals that will test how useful our collections of DNA samples and electronic records really are for genome science,” Roden said.

The Vanderbilt team, which includes experts in genomics, statistics, informatics, ethics and clinical medicine, will probe the genetics underlying heart rhythm variation, while exploring issues of privacy, confidentiality and interactions with the broader community. Bradley Malin, Ph.D., assistant professor of Biomedical Informatics, and Ellen Clayton, M.D., J.D., professor of Pediatrics and Law, will lead efforts related to privacy and community involvement, respectively.

Daniel Masys, M.D., chair of Biomedical Informatics and co-principal investigator of the project, will lead the development of “natural language processing tools” — tools that will allow investigators to query the electronic medical records and find cases and controls that fit certain clinical parameters. For this project, the focus will be to find records that include a normal electrocardiogram (ECG), a recording of the heart's electrical activity, without other conditions or medications that affect the ECG.

The investigators will compile a set of 1,500 samples from patients who have slower conduction of electrical impulses through the heart and a set of 1,500 samples from patients with faster conduction. Slower conduction translates into an increased risk for abnormal heart rhythms, Roden said. The NHGRI will analyze the samples for genetic variation at hundreds of thousands of sites in the genome and return the data to Vanderbilt, where a team of genetic epidemiologists headed by Jonathan Haines, Ph.D., will analyze the data for genetic variation linked to regulation of cardiac conduction.

“This grant gives us an amazing opportunity to participate with these other centers in moving forward our concept of the DNA Databank Resource,” Roden said.

Vanderbilt's selection as the Administrative Coordinating Center for the network means that it will orchestrate the research efforts of all five groups as they address common challenges including the development of methods to extract information from medical records, protection of privacy, and standards for submitting data to national databases.

“Vanderbilt's Department of Biomedical Informatics is the largest in the country, and home to experts in medical language processing and data de-identification,” Masys said. “This expertise will help advance the state of the art for all participants in the research consortium.”

Other sites participating in the network are Marshfield Clinic, Mayo Clinic, Northwestern University and the University of Washington.