Download this article's Factoid (PDF & PPT for Gold Subscribers)

An alternate method of measuring fractional flow reserve (FFR) may be more sensitive than the conventional approach in identifying ischemia-causing lesions, according to a small animal study published online February 4, 2014, ahead of print in Circulation: Cardiovascular Interventions. Instantaneous hyperemic electrocardiogram (ECG)-gated FFR acquired at end-diastole shows a stronger correlation with directly measured blood flow than does FFR measured during the full cardiac cycle.

Researchers led by Sabee Molloi, PhD, of the University of California, Irvine (Irvine, CA), compared FFR measurements from different stages of the cardiac cycle in 5 hyperemic swine after simulating stenosis in epicardial arteries. In addition, coronary flow was measured directly using an ultrasound probe; this value served as a reference standard for FFR comparisons.

A total of 17 FFR measurements were made during each of 4 selected cardiac cycles (end-diastole, mid-diastole, full diastole, and full cardiac cycle), 13 in the LAD and 4 in the LCX. For conventional FFR, distal and aortic pressures were recorded as an interval value averaged over 3 full cardiac cycles. For end-diastolic FFR, instantaneous values obtained at 60 milliseconds before the ECG R-wave were averaged.

The hypothesis behind the study, Dr. Molloi and colleagues explain, is that “decreased residual effects of contraction on intracoronary pressure measurement, as well as lower and more constant zero-flow pressure at end-diastole, would result in reduced variability in combined subject data during comparison of pressure-based FFR with actual reductions in coronary blood flow.”

End-Diastolic FFR More Closely Tied to Flow

Hyperemic FFR in the presence of an epicardial stenosis was 0.81 ± 0.13 for conventional FFR and 0.72 ± 0.11 for instantaneous end-diastolic FFR. Direct flow reserve, a ratio of stenosed to normal flow, was 0.73 ± 0.15.

Instantaneous end-diastolic FFR showed a closer correlation with direct flow reserve than did conventional full cardiac cycle FFR (0.941 ± 0.050 vs 0.876 ± 0.069), although both correlations were significant (P < 0.0001). Full-diastole and mid-diastole FFR also performed well compared with the full cardiac cycle (0.943 ± 0.053 and 0.916 ± 0.069, respectively), but despite a higher correlation with direct flow, mid-diastole FFR did not reduce the standard error compared with the conventional strategy.

Progress in Minimizing FFR Variability

In an e-mail communication with TCTMD, Justin E. Davies, MD, PhD, of Imperial College London (London, United Kingdom), noted that this is the first study to try to verify the original research that provided the foundation for FFR, and it shows a looser correlation of pressure with flow. The current results are “more in keeping with the intrinsic variability of the parameters being measured,” he asserted.

“This is a small physiologic study showing that if you eliminate the systolic component, you get a purer flow-pressure relationship,” Morton J. Kern, MD, of the University of California, Irvine (Irvine, CA), said in a telephone interview with TCTMD. “The standard error is smaller and the correlation is tighter. If investigators have the capacity to revamp their software, this would be an easy acquisition.”

Current data suggest that “we should probably start to cut out some of the variances in FFR measurements,” Dr. Kern continued. He noted, however, that another, perhaps more important, source of variability is patients’ heterogeneous responses to adenosine, a phenomenon that has been highlighted in several recent papers.

Drawbacks of Preclinical Models

While calling the current study a “solid proof of concept,” Juan F. Granada, MD, of the CRF Skirball Research Center (Orangeburg, NY), drew attention to the limitations inherent in any research using a healthy-animal model. These swine have normal endothelium and microcirculatory response and no significant collateral circulation, he noted, “so all the variables in humans that are responsible for the ischemic response beyond the hemodynamic obstruction are absent.”

Nonetheless, Dr. Davies observed that “this is the second study to highlight the variability of FFR between the 0.6 and 0.9 range,” which is “mainly driven by the large and extremely variable reduction of resistance during systole. By eliminating the systolic resistance component from the analysis, . . . this variability can be minimized.”

These are “exciting times,” he continued, as these results “suggest that we may be able to achieve significant improvement in accuracy in the critical intermediate zone [of stenosis] using these more refined approaches.”

However, Dr. Kern cautioned that whether the greater physiological precision of the end-diastolic method would have much impact on clinical decision making remains unknown. For now, “there is no correlation of these data with anything clinical . . . to see whether it is worth the effort to rebuild our FFR systems,” he noted.

Study Details

An IV drip of adenosine (400 µg/kg/min) by a syringe pump was used to induce maximum hyperemia. Intracoronary pressure measurements were performed using a pressure wire (Radi Medical Systems, Uppsala, Sweden).

Epicardial stenoses were produced by an external vascular occluder filled with saline. Direct flow was measured via an ultrasound probe placed proximal to the occluder.

Chalyan DA, Zhang Z, Takarada S, et al. End-diastolic fractional flow reserve: Comparison with conventional full-cardiac cycle fractional flow reserve. Circ Cardiovasc Interv. 2014;Epub ahead of print.

Related Stories:

• Novel Ischemia Measurement Strategies Show Potential vs. FFR
• Study Seeks to CLARIFY Accuracy of Adenosine-Free Pressure Index vs. FFR
• Adenosine-Free Pressure Measurement Found Equivalent to FFR