Share
PrintGenome Advance of the Month
Teaching Old Drugs New Tricks
August 2011
By Jonathan Gitlin, Ph.D.
Science Policy Analyst
At some point, most of us rely on prescription drugs to treat or cure a symptom or condition, but few of us realize what's involved in drug development.
It currently takes around 12 years and $1.2 billion to develop a new drug and bring it to market, according to the U.S. Food and Drug Administration (FDA), which monitors and regulates new drug development. What's more, only one in 10,000 new chemicals makes it all the way through the development pipeline. The August Genome Advance of the Month spotlights two studies by Stanford University's Atul Butte, M.D., Ph.D., that repurposed old drugs as new treatments for lung cancer and inflammatory bowel disease.
Dr. Butte and his group relied on the databases at the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus, and the Connectivity Map maintained by the Broad Institute of MIT and Harvard. Both databases maintain gene expression profiles — experiments that measure which genes are being turned on or turned off in a particular cell or tissue.
From the Gene Expression Omnibus, the team constructed a set of gene expression profiles for 100 different diseases. From the Connectivity Map, they constructed gene expression profiles for 164 different drugs currently approved for use in humans. Then, they compared the data sets.
If a disease and a drug both had similar profiles — that is, if they both caused the same genes to turn on and off — then they were a bad match. In fact, the drug might worsen the disease. But when the profiles were opposite — the drug turned off genes that the disease activated and vice versa — then that drug might be used to treat that disease.
This proved true for prednisolone, a drug used to treat a wide range of inflammatory conditions such as ulcerative colitis and Crohn's disease. Using the gene expression profiles, Dr. Butte found that an epilepsy drug called topiramate (Topamax) might also be an effective treatment for ulcerative colitis and Crohn's disease.
Next, researchers looked at cimetidine, commonly known as Tagamet, which is used to treat stomach ulcers. Sir James Black won the 1988 Nobel Prize in Medicine for discovering this drug, considered by many to be the first pharmaceutical blockbuster. The gene expression profile for cimetidine indicated that it might be useful in the treatment of lung cancer. In fact, cimetidine received a better score than gefitinib, a drug now commonly used to treat lung cancer. The group then proved the treatments might work through a series of experiments in cultured human cells and mice.
Physicians are not yet rushing to prescribe cimetidine for lung cancer or topiramate for ulcerative colitis and Crohn's disease — these uses still require confirmation in the lab and in clinical trials. But Dr. Butte's studies, published in the journal Science Translational Medicine, at the very least, validate this approach to repurposing established drugs. And, with this new drug development approach, who knows what the future will bring?
Read the Studies:
Sirota M, Dudley JT, Kim J, Chiang AP, Morgan AA, Sweet-Cordero A, Sage J, Butte AJ. Discovery and preclinical validation of drug indications using compendia of public gene expression data, Science Translational Medicine, 3(96):96ra77. 2011. [PubMed]
Dudley, J. T., Sirota, M., Shenoy, M., Pai, R. K., Roedder, S., Chiang, A. P., Morgan, A. A., Sarwal, M.M., Pasricha, P.J., and Butte, A.J. Computational repositioning of the anticonvulsant topiramate for inflammatory bowel disease. Science Translational Medicine, 3(96). 2011. [PubMed]
Last Updated: May 23, 2012
Leave a Comment!
We encourage you to share your thoughts as they relate to the topics being discussed on genome.gov, other websites or discussion forums owned or administered by the National Human Genome Research Institute (NHGRI). We review all comments for posting. We do not discriminate against any views, but we reserve the right not to post comments, or to remove comments, that are inconsistent with our Comment Policy, at our complete discretion.
Interesting idea and write up. I will now go read the source papers! I am curious, though, how generalizable this approach is to all drugs and how much mechanism matters. For example, say a drug works by blocking the conversion of B to C in some pathway A to B to C, where the disease state results from an overabundance of C. If there is negative feedback regulation of A from C, then blocking C would remove this feedback, thereby inducing expression if A...which might match the disease state GEO profile instead of being opposite it. Thus, the drug that does indeed treat the condition looks to match it under this approach. Perhaps best to restrict the analysis, then, to chemicals known to be involved mechanistically in gene expression? Just some ignorant musings from one who hasn't read the paper yet.