DISCOVER-FLOW Published: Noninvasive FFR Shows High Diagnostic Accuracy
A novel method of calculating fractional flow reserve (FFR) noninvasively based on coronary computed tomographic angiography (CTA) yields high diagnostic accuracy comparable to that of the gold standard invasive FFR, according to a study published in the November 1, 2011, issue of the Journal of the American College of Cardiology. It also provides superior prediction of functionally significant disease compared with CTA alone.
Results of the DISCOVER-FLOW (Diagnosis of Ischemia-Causing Stenoses Obtained Via Noninvasive Fractional Flow Reserve) trial were initially presented in May 2011 at EuroPCR in Paris, France.
For the prospective, multicenter study, investigators led by James K. Min, MD, of Cedars-Sinai Medical Center (Los Angeles, CA), evaluated the new technology in 103 stable patients (159 vessels) with known or suspected CAD who also underwent coronary CTA, conventional angiography, and FFR.
Obstructive CAD was defined as the presence of at least 50% stenosis on CTA, while lesion-specific ischemia was defined as a measurement of 0.80 or below on both conventional and noninvasive FFR. Invasive FFR was used as the reference standard.
FFR Without a Pressure Wire
For the comparison, computational fluid dynamics was applied to CTA imaging data to produce a 3-D model of blood flow and pressure fields in coronary arteries. Diagnostic results from this noninvasive version of FFR correlated well with conventional FFR (R = 0.72; P < 0.0001), with a slight underestimation (0.022 ± 0.116; P= 0.016) at the per-vessel level but no difference on a per-patient basis (P = 0.131). Use of CT-derived FFR yielded a false-positive rate of 11.3% and a false negative rate of 4.4%.
On a per-vessel basis, the noninvasive FFR method provided markedly superior diagnostic accuracy, the primary endpoint, and, greater specificity and positive predictive value than CTA. Both methods, meanwhile, yielded strong sensitivity and negative predictive value (table 1).
Table 1. Diagnostic Performance of Noninvasive FFR vs. Coronary CTA: Per-Vessel Analysis
|
Noninvasive FFR |
Coronary CTA |
Sensitivity |
87.9% |
91.4% |
Specificity |
82.2% |
39.6% |
Positive Predictive Value |
73.9% |
46.5% |
Negative Predictive Value |
92.2% |
88.9% |
Diagnostic Accuracy |
84.3% |
58.5% |
No improvement in accuracy was provided by addition of coronary CTA information to noninvasive FFR findings on either a per-vessel or per-patient level (P = 0.50 and P = 0.64, respectively).
The authors suggest that a CT-derived FFR method “might permit an ‘all-in-one’ approach, whereby patient-specific stenosis can be assessed for lesion-specific ischemia.”
Not Ready for Prime Time
Although Dr. Min and colleagues should be congratulated for this novel research, the method is not ready for widespread use, according to Ron Waksman, MD, of Washington Hospital Center (Washington, DC).
“The research demonstrates that there is unlimited potential for software to tell us more about the physiology of the anatomy and to be more accurate in detecting patients with significant coronary artery disease,” Dr. Waksman told TCTMD in a telephone interview. “However, we have to put things in perspective, and it is not ready to be applied to day-to-day practice.”
Instead, the results should be viewed as proof of concept, and more research should be conducted to support the findings, Dr. Waksman said.
This assessment was echoed by Stephan Achenbach, MD, of the University of Giessen (Giessen Germany), in an editorial accompanying the study. “It is an impressive first step into what needs to follow, the painstakingly detailed workup of whether this method will translate into clinical benefit when applied to a broader scale and which patient groups are the ones to most likely benefit from this additional analysis,” he writes.
‘Real-World’ Applicability Questioned
Drs. Waksman and Achenbach agreed that the study had a major limitation that may, for the time being, restrict the application of this new technology in a real-world setting. Accuracy may have been distorted by exclusion of 6 very-high-grade lesions, which reduced the number of ‘true positive’ findings, and inclusion of only patients with “at least 1 stenosis ≥ 50% in a major coronary artery,” which limited the number of ‘true negative’ cases, Dr. Achenbach explained.
“In a completely nonselected collective of individuals undergoing [coronary] CTA, the incremental value of [CT-based] FFR determination might therefore be less than suggested by the analysis presented here, and if speculations about the best clinical applications were appropriate at this stage at all, applications would probably be most interesting in lesions that visually seem of questionable severity,” he writes.
Study Details
All patients had stenosis in a major epicardial coronary artery as well as diagnostic-quality coronary CTA obtained from a 64-detector row or higher CT scanner. Most lesions were located in the LAD (54.7%), with the remainder in the right coronary (19.5%) and left circumflex arteries (25.8%).
Sources:
1. Koo B-K, Erglis A, Doh J-H, et al Diagnosis of ischemia-causing coronary stenoses by noninvasive fractional flow reserve computed from coronary computed tomographic angiograms. Results from the prospective multicenter DISCOVER-FLOW (Diagnosis of Ischemia-Causing Stenoses Obtained Via Noninvasive Fractional Flow Reserve) study. J Am Coll Cardiol. 2011;58:1989-1997.
2. Achenback S. Anatomy meets function: Modeling coronary flow reserve on the basis of coronary computed tomography angiography. J Am Coll Cardiol. 2011;58:1998-2000.
Related Stories:
Disclosures
- The study was sponsored by HeartFlow (Redwood City, CA).
- Noninvasive FFR analysis, an investigational technology, was performed by HeartFlow.
- Drs. Koo and Waksman report no relevant conflicts of interest.
- Dr. Achenbach reports receiving research grants from Bayer Healthcare and Siemens, lecture honoraria from Siemens, serving as a consultant to Circle, Guerbet, Servier, and being supported by research grant from Bundesministerium fur Bildung and Forschung (BMBF), Bonn, Germany.
Comments