SAN FRANCISCO – A noninvasive test that computes fractional flow reserve from coronary CT angiography images was highly accurate in detecting ischemia, compared with anatomic interpretation from CT angiography or invasive coronary angiography, in a study of 254 patients and 484 vessels.
The primary endpoint was per-patient diagnostic performance as assessed by the area under the receiver operating characteristic curve (AUC) of the test, compared with coronary CT angiography, for the diagnosis of ischemia. The AUC for the new test was 0.82, significantly better than 0.63 for coronary CT angiography, Dr. Bjarne L. Nørgaard reported at the Transcatheter Cardiovascular Therapeutics annual meeting.
The specificity nearly doubled when the HeartFlow test was used to compute fractional flow reserve from coronary CT angiography images (FFRCT), compared with coronary CT angiography assessment. FFRCT correctly reclassified 68% of false positives from CT angiography to true negatives, said Dr. Nørgaard of Aarhus (Denmark) University.
The video associated with this article is no longer available on this site. Please view all of our videos on the MDedge YouTube channel.
Invasive assessment of FFR is considered the gold standard for diagnosis of lesion-specific functional ischemic disease, but it carries more risk than a noninvasive test. Coronary CT angiography detects anatomic stenosis but is not good at determining the physiologic significance of lesions, he said.
The FFRCT technology builds a quantitative model using data from conventional coronary CT images, and develops a physiological model using left ventricular and coronary anatomy and established form-function principles, Dr. Nørgaard said. A fluid model calculates flow and pressure under simulated hyperemic conditions.
In the 254 patients, FFRCT had an accuracy of 81%, compared with 64% for anatomic assessment using invasive coronary angiography and 53% with CT angiography. The specificity was 79% with FFRCT, 51% with invasive angiography, and 34% with CT angiography. Positive predictive values were 65% with FFRCT, 46% with invasive angiography, and 40% with CT angiography. In each of these categories, FFRCT performed significantly better than CT angiography.
Sherry Boschert/IMNG Medical Media
A panel of experts said the noninvasive coronary test may change clinical practice.
The sensitivity in per-patient diagnosis was 86% with FFRCT, 91% with invasive angiography, and 94% with CT angiography. The negative predictive values were 92% with FFRCT, 93% with invasive angiography, and 92% with CT angiography. Differences between groups were not significant for sensitivity and negative predictive values.
Similar trends were seen in results for the 484 vessels in the study. FFRCT had an accuracy of 86%, compared with 71% for invasive angiography and 65% for CT angiography. The per-vessel specificities were 86%, 66%, and 60%, respectively, and the positive predictive value was 61% with FFRCT, 40% with invasive angiography, and 33% with CT angiography. Again, there was no significant loss in sensitivity (84%, 84%, and 83%, respectively) or in negative predictive value (95%, 94%, and 92%).
The accuracy of FFRCT and invasive assessments of FFR compares favorably with the accuracy of other tests, Dr. Nørgaard said, including stress echo, coronary CT angiography (cCTA), cCTA with transluminal attenuation gradient, single-photon emission CT, and intravenous ultrasound.
"The diagnostic performance of other tests is not impressive," he added. "I think the FFR is a major breakthrough."
The study enrolled patients at 10 centers on three continents who underwent CT and invasive coronary angiography with no more than 60 days between tests.
"I think this will be incorporated into practice," Dr. James B. Hermiller Jr. commented in a panel discussion of the study during a press briefing. A cost analysis is needed, added Dr. Hermiller of St. Vincent Heart Center of Indiana, Indianapolis.
Dr. Philippe Généreux, of Hôpital du Sacré-Coeur de Montréal, called the trial "a brilliant study" and "a breath of fresh air" in the area of noninvasive testing.
Dr. Bernard J. Gersh of the Mayo Clinic, Rochester, Minn., said, "This is a really important trial." He predicted that over the next 2-3 years, great strides will be made in noninvasive assessments of ischemia. "Stay tuned. A number of other methods for evaluating FFR" are being studied, he noted.
The meeting was cosponsored by the American College of Cardiology.
HeartFlow, which markets the FFRCT test, funded the study. Dr. Nørgaard reported having no other financial disclosures. Dr. Hermiller, Dr. Généreux, and Dr. Gersh reported financial associations with multiple companies, but not with HeartFlow.
sboschert@frontlinemedcom.com On Twitter @sherryboschert