Each of these things by itself may be harmless enough, but together they could be a recipe for disaster.
While we know disturbingly little about the specifics of drug interactions, UNM professor and Chief of the Translational Informatics Division Dr. Tudor Oprea and his team are developing a system that may help inform patients and physicians about the potentially dangerous effects of mixing drugs.
Oprea works in a field known as “cheminformatics,” which involves taking available data and assembling and examining them to find useful knowledge that can be used to solve a problem in a chemical field. For example, cheminformatics is often used in drug discovery, an application that Oprea has worked on for many years.
The idea behind this platform is that interactions between drug molecules and their targets, the proteins, could be modeled much like a social network, said Gergely Zahoranszky-Kohalmi, a Ph.D. student working in Oprea’s lab.
“If something goes wrong we need to fix the broken communication link between the players, and using small-molecule drugs is one way to do so,” he said. “However, the result of any network analysis is only as good as the quality of the underlying network.”
While his research in drug discovery is still ongoing, lately Oprea is also focusing on tackling the problem of drug-drug interactions (DDIs).
When multiple drugs are taken together, they can have potentially dramatic effects on each other’s activities, Oprea said. One drug can increase the effect of the other, decrease the effect, or cause the drug to act in entirely new way.
“If you look at celebrities like Anna Nicole Smith, Philip Seymour Hoffman, Michael Jackson, Heath Ledger — they died because of lethal effects of drug-drug interactions,” Oprea said. “We have tried to develop a system that would be computationally advanced enough to handle multiple active pharmaceutical ingredients with the dosage included into that system.”
However, prescription drug use isn’t just a celebrity problem. A study performed by the researchers at the Mayo Clinic last year found that nearly 70 percent of Americans are taking at least one prescription drug, and the number only increases in older populations.
According to research conducted at the Centers for Disease Control, 114 people die every day from drug overdose and thousands of people are hospitalized — almost 80 percent of which are accidental. In the last few decades overdoses from prescription drugs have seen a drastic increase, becoming far more common than deaths from illicit drugs such as heroin and cocaine. A huge risk factor for drug overdose: combining multiple drugs.
“When it comes to drug interactions, partly because too much funding would be required and no one wants to put in the effort, most DDIs are studied pairwise,” Oprea said. “Nobody has the ability to study five drugs at the same time because it is too complex, so we try to do that with computers.”
Get content from The Daily Lobo delivered to your inbox
To do this, his group harvests information from nearly every available source, taking into account data from published scientific research, drug labels and clinical trials and assembling them into a chemical library.
“We try to integrate all of that knowledge into a system that is both chemically cognizant as well as clinically cognizant, and those are not very simple things to do,” he said.
He used these data to help update an in-house database, DrugsDB, that assembles information on indications and side effects of all prescription drugs, Oprea said.
He said he believes his system is making headway on providing the useful information that both patients and advising physicians need.
“We think we have a good grip on it,” he said.
Lauren Topper is a freelance writer at the Daily Lobo. She can be reached at news@dailylobo.com or on Twitter @DailyLobo.
