April 08, 2014

Should we abandon mouse models of human disease?

Dr. Ross Summer, Associate Professor, Division of Pulmonary & Critical Care Medicine at Thomas Jefferson University reviews a recent high profile and provocative paper published in PNAS 2013 by Seok, et al, Genomic responses in mouse models poorly mimic human inflammatory diseases.

Do mouse models accurately mimic human diseases? Apparently not, according to a recent study by Seok et. al. published in Proceedings of the National Academy of Sciences. This study, which involved researchers from many institutions across the U.S., compared gene expression profiles in circulating white blood cells from patients with severe burns, sepsis or trauma to those in white blood cells from mice with similar injuries.  Results demonstrated some interesting findings, but most notably, researchers found a high correlation in gene expression responses in humans, regardless of the disease, but importantly these changes did not correlate with gene expression profiles in cells from mice. While the reflexive response to this study is to conclude that mouse models should be immediately abandoned, or at the very least, recognized as providing only misleading data for studying human disease — we strongly suggest caution against making such a conclusion. First and foremost, this conclusion would be against what decades of pre-clinical studies have already shown us.  Animal studies have a proven track record for contributing to major advances in our ability to understand and to treat diseases (e.g. leukemia, obesity, diabetes, cancer). Second, while mice may look and behave differently, they share 99% of their genes with humans. Finally, mice have many well-documented advantages over working with larger species including relatively low cost, short generation times, and acceptability from an ethical standpoint. However, this study does reinforce what many scientists may actually already know; mouse models of human disease are not perfect. But instead of abandoning or discounting mouse models, we suggest this study may be sending a very different message. This message speaks to the importance of holding scientists to higher standards and demanding that findings from pre-clinical studies be robust, reproducible and well-validated before moving on to clinical studies.   

Click Here to read the article

Jefferson