Computational method predicts new uses for existing medicines
For the first time ever, scientists are using computers and genomic information to predict new uses for existing medicines.
Balintfy: An NIH-funded computational study has analyzed genomic and drug data to find new uses for medicines that are already on the market. Dr. Atul Butte, an NIH grantee at Stanford University says a computational study is when computers are used to address a biomedical need; and in this case:
Butte: Not just using computers, but also using publicly available data.
Balintfy: Dr. Butte explains that a computer program searched through publicly available data on drugs and diseases. An algorithm compared thousands of possible combinations to find drugs and diseases whose gene expression patterns essentially cancelled each other out — a genetic “opposites attract” search on a molecular level.
Butte: If we could find 5 diseases that were similar to each other now in a molecular level — you know never mind what the doctors say, if they look like each other at a molecular level — and along came a 6th disease, but those 5 diseases are treated the same way, maybe that 6th disease could be treated the same way too.
Balintfy: Dr. Butte’s group focused on 100 diseases and 164 drugs. While many of the drug-disease matches were known, and already in clinical use adding to the credibility of this computational study, others were surprises. Dr. Butte says out of the hundreds of predicted matches, his group chased down two.
Butte: One prediction was that the anti-seizure drug Topiramate – it’s a seizure drug, we thought that should have efficacy against Crohn’s Disease which is a kind of inflammatory bowel disease. The second prediction we had was even more astounding in some ways. The over-the-counter drug Tagamet or Cimetidine, we thought should have efficacy against lung cancer, specifically lung adenocarcinoma.
Balintfy: He says animal models show the drugs work, but cautions that more research is needed.
Butte: We still have a while to go before we can show this actually works in human clinical trials.
Balintfy: But researchers are optimistic that this method of finding new ways to repurpose drugs that are already approved could improve treatments and save both time and money. Experts point out that bringing a new drug to market typically takes about $1 billion and many years of research development.
Butte: The beauty of using existing drugs with known safety profiles is that the potential for getting these to patients is that much quicker.
Balintfy: Dr. Butte adds there is another potential.
Butte: If we didn’t find these uses for these drugs I would argue that nobody would ever find these uses for these drugs.
Balintfy: He explains that biotech or pharmaceutical companies might not pursue existing drugs for other treatments because of low financial incentive. That could be because the drugs are already off-patent, meaning low-cost generics exists; or the treatment might be for a so-called orphan disease that has relatively few patients.
Butte: Those are businesses decisions more than scientific ones.
Balintfy: Dr. Butte notes that this research has since expanded to 300 diseases and over 1,500 different drugs; and in addition to these two medications, described in the online issue of Science Translational Medicine, a third and fourth are also showing promise. For more information on this computational study and this kind of research, visit www.nigms.nih.gov. This is Joe Balintfy, National Institutes of Health, Bethesda, Maryland.
About This Audio Report
Reporter: Joe Balintfy
Sound Bite: Dr. Atul Butte
Topic: computers, computational study, medicine, drug, existing medicine, existing drug, disease, match, molecule, clinical trial
Additional Info: Computational method predicts new uses for existing medicines