Functional Evaluation of CAD Using Angiographic Images Alone

February 20, 2013

SHEFFIELD, UK — British researchers have developed a computerized tool for measuring myocardial fractional flow reserve (FFR) using angiographic images alone, without the need for a pressure-sensitive intracoronary wire [1]. The model of intracoronary physiology is based on a rotational coronary angiogram and fluid dynamics analysis and is able to identify significant coronary lesions with 97% accuracy.

"The starting point for our work is the well-documented disconnect between what we see during angiography and the functional significance of coronary lesions," lead author Paul Morris (University of Sheffield, UK) told heartwire . "There is substantial evidence demonstrating that visual interpretation of coronary lesions, especially moderate lesions, can be misleading. Hence the development, use, and impact of fractional flow reserve, which introduces objectivity and reduces subjectivity in these assessments. We know from DEFER, FAME, and FAME II that if we use physiology to guide PCI, patients do better and costs are reduced in the long term."

Despite the known benefits of using FFR, it is still an underutilized technology, with less than 10% of patients in the UK having FFR assessed prior to PCI, said Morris. There are likely multiple factors for the slow uptake of FFR in clinical practice, not least is the time required to perform the extra step and the increased procedure cost. If it is possible to develop an accurate estimate of FFR from angiography alone, the use of physiologically guided PCI might be available to more of the population.

The results of the study are published in the February 2013 issue of the Journal of the American College of Cardiology: Cardiovascular Interventions.

FFR Not Being Used Enough in the UK

For their "virtual FFR," the researchers obtained 3D images of coronary lesions in patients with established coronary disease using a rotational angiogram. They then applied computational fluid dynamics within the virtual reconstructed artery to estimate the coronary pressures in a total of 10 right coronary arteries and 12 left coronary arteries.

We would like to create a system that you don't have to be a PhD student to operate.

Next, researchers compared results of the virtual FFR with traditional FFR measurements taken in patients at the time of the angiography. The lesions were stratified into two groups: those with an invasively measured FFR >0.80 and those with an FFR <0.80. The virtual FFR assessments were stratified in the same way. The authors report a high level of agreement between the invasively measured FFR and the virtual FFR assessments: sensitivity, specificity, positive predictive value, and negative predictive value were 86%, 100%, 100%, and 97%, respectively. The overall diagnostic accuracy was 97%

"We were able to successfully predict the pressures within the artery. However, we had to make sure that these pressures were reliable," said Morris. "We therefore compared our results with the invasively measured FFRs to see how well we performed. The results were not perfect, but we were all pleased with the reliability. I think this is the first step in the development of virtual FFR. We don't have the finished system yet."

Some Tinkering Still Needed

To heartwire , Morris said there are three key areas they would like to improve upon with virtual FFR. First, they would like to make the assessment more accurate. Second, they would like to make the process faster. At present, it takes several hours to process the pressures inside the diseased artery using computational fluid dynamics. In contrast, it takes just a few minutes to perform FFR in the cath lab. And finally, the researchers would like to make the system more accessible to the general cardiologist.

"We would like to create a system that you don't have to be a PhD student to operate," said Morris. "We believe we have demonstrated the feasibility of the approach. We wouldn't suggest we have a system we can roll out tomorrow, but it does appear feasible, and hopefully, by the next time we publish, we'll have been able to optimize its accuracy, speed, and ease of use. It's the first step in a promising direction."

As previously reported by heartwire , other researchers are also seeking ways to noninvasively combine the benefits of FFR with computed tomography (CT). The hope, not unlike virtual FFR, is that FFR-CT can improv e the diagnostic accuracy of CT imaging by providing functional information, without the need for an invasive test, adenosine, or additional radiation exposure.