CT Angiography (CTA) and Diagnostic Performance of Noninvasive Fractional Flow Reserve: Results From the Determination of Fractional Flow Reserve by Anatomic CTA (DeFACTO) Study

OBJECTIVE: Fractional flow reserve (FFR) computed from standard coronary CT scans (FFRCT) is a novel noninvasive method for determining the functional significance of coronary artery lesions. Compared with CT alone, FFRCT significantly improves diagnostic accuracy and discrimination for patients with and without hemodynamically significant coronary artery stenoses. To date, the impact of CT image quality on diagnostic performance of FFRCT is unknown. We evaluated the impact of patient preparation, CT scan protocol, and factors related to image quality on the diagnostic accuracy of FFRCT. SUBJECTS AND METHODS: We studied stable patients with suspected coronary artery disease (CAD), enrolled from 17 centers, who underwent CT, invasive coronary angiography, FFR, and FFRCT. The accuracy of CT and FFRCT for diagnosis of ischemia was compared against an invasive FFR reference standard. Anatomically obstructive CAD was defined by a stenosis value of at least 50 by CT or invasive coronary angiography, whereas ischemia was defined by an FFR or FFRCT of up to 0.80. Ischemia was assessed at the per-patient and per-vessel levels. Diagnostic performance of FFRCT was then evaluated in relation to patient preparation, including administration before CT of a β-blocker or nitroglycerin, as well as in relation to imaging characteristics, including misalignment, noise, motion, and coronary artery calcium. RESULTS: Among 252 study participants, 137 (54.0%) had an abnormal FFR. Administration of a β-blocker increased FFRCT specificity (51.0% vs 66.0%; p = 0.03) with lower bias (-0.084 vs -0.048; p = 0.008), whereas nitroglycerin pretreatment within 30 minutes of CT was associated with improved specificity (54.0% vs 75.0%; p = 0.013). Misalignment artifacts resulted in impaired sensitivity (43.0% vs 86.0%; p = 0.001) with resultant reductions in overall accuracy (56.0% vs 71.0%; p = 0.03). No differences in diagnostic performance of FFRCT were noted in the presence of coronary motion or increasing coronary artery calcium score. CONCLUSION: Use of β-blockade and nitroglycerin administration before CT improve diagnostic performance of FFRCT. Diagnostic accuracy of FFRCT is significantly reduced in the setting of misalignment artifacts.