"This work highlights the importance of both biological and computational model systems to unravel the complexities and heterogeneity of human cancer," said Daniel Gallahan, Ph.D., program director for the Integrative Cancer Biology Program at the National Cancer Institute. "This type of analysis can be exploited to better align a therapeutic strategy with an individual's specific cancer."

Running parallel to human trials, the mouse trials will show what works well and what doesn't in the trial methods, data collection, analysis and other aspects of the trials. Researchers can then translate these findings immediately to keep the human clinical trials advancing as effectively as possible.

With so much mouse model research happening around the globe, why weren't these mouse tumor differences noted before? The gene expression analyses performed on mouse tumors simply haven't been large enough, Nevins said.

"We examined a large number, up to 80 samples of mouse tumors. And in the same way that a picture gets clearer when you add more pixels, the information about the tumors became clearer as we examined more samples," he said. "In effect, we went to a higher resolution and could begin to see patterns more clearly."

The study was funded by the National Institutes of Health and the V Foundation for Cancer Research, named in honor of the late North Carolina State basketball coach Jim Valvano.

Other authors include Eran R. Andrechek, Jeffrey T. Chang, Michael L. Gatza, Chaitanya R. Acharya, and Anil Potti of the Duke Institute for Genome Sciences and Policy, and Robert D. Cardiff of the Center for Comparative Medicine at the University of California at Davis.

Source: Duke University Medical Center