Their effort represents the first attempt to use ensemble learning to predict colorectal cancer recurrence. This will likely lead to better molecular staging tests for colorectal cancer in the future.

— Kelly M. McMasters, MD, PhD

Tweet this quote

“Several gene-expression profiles have been evaluated to predict prognosis in colorectal cancer, but none have become widely accepted or U.S. Food and Drug Administration–approved. This has important implications for deciding which patients may benefit from adjuvant chemotherapy,” commented session moderator Kelly M. McMasters, MD, PhD, a surgical oncologist with the University of Louisville School of Medicine, in Kentucky.

“Dr. Castellanos and colleagues from Vanderbilt evaluated the complex bioinformatics approach of ensemble learning to evaluate gene-expression data from six publicly available microarray data sets and develop more robust prognostic models,” he said.

“Their effort represents the first attempt to use ensemble learning to predict colorectal cancer recurrence. This will likely lead to better molecular staging tests for colorectal cancer in the future, which will allow more personalized treatment decisions of patients with this disease,” Dr. McMasters concluded. ■

Disclosure: Dr. McMasters reported no potential conflicts of interest.