Non-invasive fractional flow reserve still out of reach
medwireNews: A novel, non-invasive approach to determining fractional flow reserve (FFR) failed to show sufficient diagnostic accuracy to meet the primary endpoint of the DeFACTO trial, but was more accurate for diagnosing coronary artery ischemia than computed tomography (CT), reported the trial's lead investigator.
The approach applies computational fluid dynamics to calculate FFR from CT (FFRCT) without the need for additional imaging, modification of CT acquisition protocols, or administration of additional medicines.
Speaking at a press conference ahead of his presentation at the European Society of Cardiology Congress 2012, James Min (Cedars-Sinai Heart Institute, Los Angeles, California, USA) stressed: "DeFACTO is the first large-scale demonstration of patient-specific computational models to calculate physiologic pressure and velocity fields from CT images."
All patients in the DeFACTO (Determination of Fractional Flow Reserve by Anatomic Computed Tomographic Angiography) study underwent CT, invasive coronary angiography, FFR, and FFRCT between October 2010 and October 2011. The accuracy of FFRCT and CT for diagnosis of ischemia was compared with an invasive FFR reference standard.
Min and colleagues found that of 252 patients, 137 (54.4%) had a diagnosis of ischemia on FFR (defined as mean distal coronary pressure/mean aortic pressure ≤0.80).
An FFRCT of 0.80 or less was also taken to indicate ischemia, while anatomically obstructive CAD was defined by a stenosis of 50% or larger on CT and invasive coronary angiography.
On a per-patient basis, diagnostic accuracy of FFRCT was 73% (95% confidence interval [CI]: 67-78%), meaning that it did not meet the prespecified primary endpoint of the study, requiring the lower boundary of the 95% CI to exceed 70%. Nonetheless, "FFRCT had increased discriminatory power," remarked Min.
The diagnostic accuracy of CT alone was 64%, and FFRCT demonstrated significantly better discrimination compared with CT alone, at an area under the receiver operating characteristic curve value of 0.81 versus 0.68.
Discussant Jean-Pierre Bassand (University of Besançon, France) remarked that the computation of FFRCT is achieved through an "elegant but complex methodology." However, he added that "its diagnostic accuracy is far from being optimal and as presented in this report is more or less tautology."
Correlation with the gold standard that is invasive FFR is poor, he added, and it is premature to envisage clinical evaluation.
"Therefore, my conclusion is significantly different from that of Doctor Min," he finished.
Full results of the study are published online in JAMA.
By Piriya Mahendra, medwireNews Reporter