New Software Simulates How Blood Clots Dissolve, Will Help Tailor Drug

BUFFALO, N.Y. -- If physicians knew precisely how blood clots dissolve, they could tailor drug therapies and reduce patient risks.

The answers they need to accomplish those goals are being provided by new computer software developed by researchers at the University at Buffalo that simulates clot dissolution using the same techniques employed to analyze chemical reactors.

The computer model, called CLOTSIM, is described in a paper by Scott Diamond, Ph.D., associate professor of chemical engineering at UB, and Sriram Anand, Ph.D., UB postdoctoral fellow, published in the August issue of Circulation.

The UB research is the first successful attempt to capture the physics of clot-dissolving therapy in a single, accurate model.

"Every hundredth of a second, the program provides a 'snapshot' of a blood clot undergoing dissolution," said Diamond.

He recently was awarded a $575,000 National Institutes of Health grant to enhance CLOTSIM and to use it to predict more clinical outcomes.

By simulating reaction and transport processes at and near the clot, CLOTSIM provides critical information about the function of tissue plasminogen-activator (tPA) and other thrombolytic agents used to dissolve blood clots in coronary arteries and elsewhere in the body.

The program will help researchers and clinicians discover why some blood clots are harder to treat than others and eventually will allow doctors to tailor thrombolytic therapies that significantly reduce patient risks.

It also is expected to shorten the amount of time it takes to bring a new thrombolytic agent to clinical trials.

"Although tPA has been on the market since the 1980s, there are still controversies about how best to use it," said Diamond.

"The goal is to dissolve the clot and re-establish flow through the blood vessels as quickly as possible," he explained. "If you use too much drug, you risk bleeding complications. If you use too little, you don't dissolve the clot. It can be very difficult to predict how best to use these thrombolytic agents for each patient."

Under the best circumstances, the clot will dissolve quickly and uniformly. However, many variables, such as the type of clot, its age, where it is located and the initial disease that caused it, can affect the safety and efficacy of the treatment for an individual.

"Our program takes the biochemistry and clot structure into account," said Diamond. "Based on specific drug concentrations and methods of administration, it calculates what happens when the drug is administered, how quickly the clot dissolves and the chances of it suddenly breaking apart, which is hazardous."

As chemical engineers, the researchers developed the computer model from an unusual perspective. They analyzed the blood clot the same way they would a chemical reactor.

For example, Diamond said, as solid rocket fuels burn, a combustion front moves across the fuel, obliterating the fuel as it is consumed.

"In the same way," he added, "the enzymes in the thrombolytic drugs move across the blood clot in a dissolution front, obliterating the clot's solid structure."

The approach has provided fundamental insights into the dynamics of clot dissolution.

"The model predicts how the blood chemistry is changing in real-time in response to what the clinician is doing, based on any thrombolytic agent used or infusion regimen," Diamond said. "No one could predict these dynamics and outcomes before."

For that reason, it is expected to help streamline the design of clinical trials for new, clot-dissolving drugs.

By computationally testing scenarios that involve different agents, administration regimens, doses and rates of infusion, CLOTSIM can predict valuable information about how new thrombolytic compounds will perform, even before the first human or animal study is done.

"Hopefully, this program will promote the safe use of new compounds by taking advantage of all we have learned with the first generation of compounds that took us a decade to understand," Diamond said.