Drugs useful in the long-term management of cardiac arrhythmia, which occurs when electrical impulses in the heart become irregular and put patients at risk of sudden death, have eluded researchers for decades. Despite best efforts, most of the medications developed to calm abnormally fast heartbeats, a type of arrhythmia known as tachyarrhythmia, have faltered. Several clinical trials, including the seminal 1986 Cardiac Arrhythmia Suppression Trial(CAST), even showed that the use of certain drugs designed to correct tachyarrhythmia—encainide and flecainide, in particular—actually increased the risk of death.
Arrhythmias cause nearly 250,000 deaths annually in the U.S. Tachyarrhythmia is triggered when an extra heartbeat develops in a person’s normal cardiac cadence. Drugs like flecainide were expected to suppress the trigger that caused more heartbeats per minute but instead created conditions for the tachyarrhythmia to worsen. The problem with medications to treat this condition is their unpredictability—the druglidocaine, for example, has proved beneficial in treating some types of tachyarrhythmia under certain conditions.
“There’s been no way to screen what drugs would be most useful in a given clinical setting,” says Colleen Clancy, an associate professor in the University of California, Davis, Department of Pharmacology. [Read more about cardiac computer modeling: “Using Computers to Model the Heart… Why Bother?”]
With this in mind, Clancy has been working with a team of researchers from Columbia University, Cornell University and The Johns Hopkins University to develop a computer model of a human heart that can help forecast at least some of the side effects of drugs used to treat certain tachyarrhythmias. Their goal is to create a “virtual drug-screening system that models drug-channel interactions and predicts the effects of drugs on emergent electrical activity in the heart,” according to the study that Clancy and her colleagues published in the August 31 issue ofScience Translational Medicine.
Heart cells generate electrical signals through ion channels in the cell membrane that open and close. These signals spread from cell to cell and manifest as electrical waves throughout the heart, telling the heart muscle to contract at regular intervals and pump blood to the brain and vital organs. To predict how antiarrhythmic drugs will alter cardiac rhythms, researchers must first determine how the drugs will interact with these ion channels.
Initially, it was thought that antiarrhythmics reduced electrical activity by plugging the ion channels. With the CAST and other studies calling those assumptions into question, Clancy and her colleagues developed mathematical equations representing the opening and closing of ion channels. They combined these with other mathematical algorithms to create a computer model of a human heart. Using data from experimental measurements of the action of the drugs flecainide and lidocaine, the researchers then introduced virtual representations of these antiarrhythmics to the heart model.
The simulation software was run on multiple high-performance computer clusters consisting of several servers working simultaneously to boost processing power. In one computer experiment, both drugs successfully slowed down heartbeats when the researchers tested the effects on individual heart cells. However, when tested on a virtual model of a whole heart in another computer-based experiment, flecainide created serious side effects by causing an arrhythmia, rather than preventing abnormal rhythms. These results were consistent with the CAST findings, which showed that patients given flecainide were two-to-three times more likely to experience lethal arrhythmias than placebo. “In the absence of those data, we wouldn’t know if the model was behaving properly and we were making accurate predictions,” Clancy says.
The results are significant for a number of reasons. For one, the researchers were able to simulate and test the effects of drugs over a wide range of concentrations, heart rates and arrhythmia triggers much more efficiently than they could if they had performed their study on lab animals. “We were testing those drugs in an environment that simulates a human heart,” Clancy says. “The models aren’t perfect, but they’re more human than a mouse, and it’s well known that drugs exhibit species-dependent effects. Until a time when we can test in humans, and I don’t see that happening anytime soon, this may be the next best thing.”
A future version of the software might also be used to screen out drug compounds that are problematic as well as identify those that have potential well before expensive animal and, later, human trials are conducted. “The model could even be used to identify ideal therapeutics—the properties that a drug needs to have to be useful in a particular clinical setting,” Clancy says.
Within five years the researchers want to create a database documenting the behavior of all prototypical drugs used to treat different types of arrhythmia, including bradyarrhythmia, characterized by an abnormally slow heart rate. The new models that result might also be used to reveal the mechanisms driving common side effects of antiarrhythmic drugs.