IBM Watson Advances Medical Research and Data Analysis
As oncology researchers turn to IBM’s Watson supercomputer to facilitate their work, patients, clinicians and other researchers are beginning to benefit. At the same time, the pathway from Watson’s analysis to practical clinical information becomes more clear.
When a team of clinician oncologists from the Lineberger Comprehensive Cancer Center at the University of North Carolina at Chapel Hill trained Watson to annotate research, the supercomputer came up with a set of clinically actionable genes above the 90 or so identified in the course of a large-scale study. The additional genes located by Watson showed up in 30 percent of the 1,000 patients who participated in the study, according to Nirali M. Patel, M.D., assistant professor of pathology and laboratory medicine at UNC.
“That’s important, especially in these highly treated cancer patients,” she says. In other words, Watson had effectively located new treatment options for a patient population rapidly running out of them.
Getting this type of practical, actionable information into clinics that don’t operate at the size and scale of a research institution represents another challenge. A three-way partnership among Quest Diagnostics, Memorial Sloan-Kettering Cancer Center, and IBM Watson provides a channel through which to deliver such insights to community oncologists, from whom 70 percent of cancer patients receive their care.
“Watson is doing the heavy lifting that formerly only major cancer centers could do, and then we’re working with them to deliver that to the community doc in a way that they can digest,” says Jay G. Wohlgemuth, M.D., chief medical officer and senior vice president of research, development and medical at Quest. By giving physicians something to do with the information, Wohlgemuth believes Watson will help to spur broader adoption of mutational profiling.
Quest also adds its test results to Watson’s knowledge base in order to help yield the next generation of clinical advancements. Initiatives like a partnership between the Cambridge, Mass.-based Broad Institute and Watson on a large-scale study of drug-resistant genes also help to strengthen platform’s analysis.
“We have been able to do this comprehensive genomic profiling for at least five years, but because it was being done in a very one-off, localized fashion, we tended to get information on patients who were at a very advanced stage or happened to be at an academic medical center,” Patel says. She expects the broader application of these tools to lead to better outcomes data, though she stresses that the process of finding long-term endpoints will still take time.
“I believe Watson will have a significant impact over the next two to five years,” says Alan Louie, research director of life sciences for IDC Health Insights. In addition to the advancements it continues to facilitate in oncology, he sees potential in Watson’s use as “the next extension beyond clinical support systems.” Given the appropriate training, he says, Watson could then give primary care physicians in remote situations access to high-level, high-quality ways of looking at disease and suggested modes of moving forward, with demonstrated patient outcomes.
