Cognitive Computing Gives Precision Cancer Care a Boost

Researchers report in The Oncologist that molecular tumor boards coupled with cognitive computing could potentially improve patient care by providing a rapid, comprehensive approach for data analysis. They theorize that this type of analysis could help better guide treatment and funnel patients into the most appropriate clinical trials.

The researchers used IBM’s Watson for Genomics to assess whether cognitive computing was more effective than a panel of cancer experts in identifying therapeutic options for tumors with specific genetic abnormalities. They compared Watson’s ability to identify possible therapeutic options tied to potentially clinically significant genetic mutations with a molecular tumor board.

In a retrospective analysis of 1,018 cancer cases, the molecular tumor board identified actionable genetic alterations in 703 cases, which Watson also confirmed. In addition, the team identified additional genomic events of potential significance in 323 (32%) patients, which had not been considered actionable by the molecular tumor board. In most of those cases, Watson identified a new clinical trial. The new analysis turned up 96 tumors that were not previously identified as having an actionable mutation.

“We were surprised by the ability of Watson to keep up to date with the current compendium of clinical trials. For example, one of the biomarker-selected trials it identified for a patient had just opened approximately 1 to 2 weeks prior to the case being run through Watson for Genomics,” said study investigator William Kim, MD, from the University of North Carolina (UNC) Lineberger Comprehensive Cancer, Chapel Hill, North Carolina.

The study drew on data from a clinical trial that used next-generation sequencing to analyze a participant’s tumor with the goal of matching tumor abnormalities with a targeted therapeutic. The current findings suggest that cognitive computing may augment the molecular tumor board process for the interpretation and collection of information regarding a patient’s genomic profile.

“I believe our study points out that despite our best intentions as physicians, it is next to impossible to keep up with the deluge of biomedical literature constantly being published as well as clinical trials that are suitable for a patient. Cognitive computing allows us to do this in a comprehensive manner and therefore in some cases is able to present more potential treatment options for patients,” Dr. Kim told Oncotherapy Network.

Dr. Kim, who is also an associate professor of medicine and genetics at the UNC School of Medicine, noted that this analysis was not designed to examine whether cognitive computing affects patient outcomes or quality of life.  He said further studies are warranted.