Comparison of angiography-derived index of microcirculatory resistance measurements between low and high frame-rate acquisitions: the prospective ADMIRAL-study

Abstract Background Angiography-derived index of microcirculatory resistance (angio-IMR) is a novel, wire-free, angiography-based method (QFR-IMR®) to assess coronary microvascular dysfunction. If a frame-rate of 15 frames per second (fps) is recommended, there is an interest in decreasing this frame-rate to reduce X-ray exposure. However, the impact of frame-rate reduction on angio-IMR quantification has not been investigated. Purpose To compare angio-IMR at 7.5 fps (angio-IMRlow) to angio-IMR at 15 fps (angio-IMRhigh) in a multicentric study. Methods From October 2023 to February 2024, we conducted a multicentre prospective study including consecutive patients referred for an invasive coronary angiogram with a final diagnosis of ischemia with no obstructive coronaries arteries (INOCA) (positive stress test with normal coronary angiogram) or acute coronary syndrome (ACS). Main exclusion criteria were creatinine clearance<45ml/kg/1.73m2, factors that might substantially impact the coronary angiograms quality (e.g: aortic stenosis, supraventricular tachycardia), and TIMI flow <3. In ACS patients, angio-IMR was measured after revascularisation. The same two acquisitions with different angles (≥25°) were recorded at a rate of 7.5 fps and a rate of 15 fps to measure angio-IMR in the selected vessel (culprit coronary vessel in ACS patients or left anterior descending artery in patients with INOCA). Angio-IMR was measured with dedicated software (QFR-IMR® RE, Medis Medical Imaging). Correlation between angio-IMRlow and angio-IMRhigh was assessed by linear regression and Bland-Altman test. The diagnostic accuracy of angio-IMRlow to predict microcirculatory dysfunction (defined as angio-IMRhigh≥ 25 for INOCA and ≥ 40 for ACS) was also assessed. Statistical significance was awarded by p≤0.05. Results Of 87 consecutive eligible patients, 75 were included in the study, comprising 57 with ACS and 18 with INOCA (mean age 61±12 years, 72% male). Angio-IMRlow (median=35; [IQR:28-48]) was slightly but significantly higher (p=0.04) than angio-IMRhigh (median=34; [IQR:25-45]). Angio-IMRlow and angio-IMRhigh were significantly correlated (r=0.90; p<0.0001) (Figure 1). Using Bland-Altman analysis, mean difference between the two measurements was 2.2 (95% CI 0.1-4.4) with a significant higher value with angio-IMRlow (p=0.04). Moreover, as value of Angio-IMRhigh increases, the difference between angio-IMRlow and angio-IMRhigh becomes more significant (Figure 2). Using angio-IMRhigh to diagnose microcirculatory dysfunction as reference, angio-IMRlow achieved positive and negative predictive values of 83% and 85% respectively with 6 false positives, and 6 false negatives. Conclusion Correlation between angio-IMRhigh and angio-IMRlow is acceptable. Lowering frame acquisition rate reduces radiation exposure without relevantly impact on the diagnostic accuracy of angio-IMR, which can contribute to a widespread use of angio-IMR.Correlation between angioIMRlow and highBland-Altman plot