Diagnostic performance of fractional flow reserve derived from computed tomography in the stented coronary arteries

Abstract

Abstract Background The current system of HeartFlow fractional flow reserve derived from computed tomography (FFRCT) is not available for stented coronary arteries. However, the latest version of HeartFlow FFRCT has the potential of rendering an accurate assessment for stented coronary arteries. Objectives To evaluate the feasibility of FFRCT for the stented coronary arteries. Methods We retrospectively enrolled patients with a history of coronary stent implantation who underwent invasive fractional flow reserve (FFR) for the stented coronary arteries within 3 months after coronary computed tomographic angiography (cCTA). As a subgroup analysis, we analyzed optical coherence tomography (OCT) images in patients who underwent OCT for stented vessels. OCT was performed as a part of the Kobe University Hospital OCT registry, which is a single-centre registry of consecutive patients who underwent OCT for the coronary arteries. The diagnostic performance of HeartFlow FFRCT for stented vessels was evaluated by comparing with that of cCTA alone. Minimum lumen area (MLA) within stent segments derived from cCTA was also compared with MLA derived from optical coherence tomography (OCT). Results A total of 30 vessels in 23 patients were studied. The diagnostic accuracy, sensitivity, specificity, positive and negative predictive value of FFRCT for myocardial ischemia using invasive FFR as a reference standard was 73.3%, 88.9%, 67.7%, 53.3% and 93.3%, respectively. FFRCT provided superior diagnostic performance than cCTA alone (p=0.016). FFRCT was significantly correlated with the invasive FFR (r=0.620, p<0.001), and Bland-Altman plot showed a mean bias of −0.041 with 95% limit of agreement of −0.173 to 0.092 (Figure 1). Among a total of 30 studied vessels, OCT was available for 25 vessels. MLA derived from cCTA correlated significantly with that derived from OCT (r=0.695, p<0.001), and Bland-Altman plot showed a mean bias of −0.32 mm2 with 95% limit of agreement of −2.89 to 2.25 mm2 (Figure 2). Conclusions The latest version of the HeartFlow FFRCT has the potential to be available for stented coronary arteries. Further investigation is required to elaborate our results. Funding Acknowledgement Type of funding sources: None. Figure 1Figure 2