Coronary angiography-derived index for assessing microcirculatory resistance in patients with non-obstructed vessels: The FLASH IMR study

Abstract

Assessing index of microcirculatory resistance (IMR) is customarily performed using intracoronary wires fitted with sensors by at least 3 intracoronary injections of 3 to 4 mL of room-temperature saline during sustained hyperemia, which is time- and cost-consuming. The FLASH IMR study is a prospective, multicenter, randomized study to assess the diagnostic performance of coronary angiography-derived IMR (caIMR) in patients with suspected myocardial ischemia with nonobstructive coronary arteries using wire-based IMR as a reference. The caIMR was calculated by an optimized computational fluid dynamics model simulating hemodynamics during diastole based on coronary angiograms. TIMI frame count and aortic pressure were included in computation. caIMR was determined onsite in real time and compared blind to wire-based IMR by an independent core laboratory, using wire-based IMR ≥25 units as indicative of abnormal coronary microcirculatory resistance. The primary endpoint was the diagnostic accuracy of caIMR, using wire-based IMR as a reference, with a pre-specified performance goal of 82%. A total of 113 patients underwent paired caIMR and wire-based IMR measurements. Order of performance of tests was based on randomization. Diagnostic accuracy, sensitivity, specificity, positive and negative predictive values of caIMR were 93.8% (95% CI: 87.7%-97.5%), 95.1% (95% CI: 83.5%- 99.4%), 93.1% (95% CI: 84.5%-97.7%), 88.6% (95% CI: 75.4%-96.2%) and 97.1% (95% CI: 89.9%-99.7%). The receiver-operating curve for caIMR to diagnose abnormal coronary microcirculatory resistance had area under the curve of 0.963 (95% CI: 0.928-0.999). Angiography-based caIMR has a good diagnostic yield with wire-based IMR. NCT05009667.