Inter-scan Reproducibility of Cardiovascular Magnetic Resonance Imaging-Derived Myocardial Perfusion Reserve Index in Women with no Obstructive Coronary Artery Disease

Abstract

Background:Cardiovascular magnetic resonance imaging (CMRI) derived myocardial perfusion reserve index (MPRI) has recently been shown to detect coronary microvascular dysfunction (CMD) in women with signs and symptoms of ischemia and no obstructive coronary artery disease (CAD). The aim of this study was to determine the inter-scan reproducibility of MPRI in this patient group in order to assess its diagnostic robustness in serial scans and assess its utility as a marker of potential therapies for CMD.Methods:Rest/stress perfusion CMR was performed at 1.5T using a standardized protocol in 17 women with signs and symptoms of ischemia and no obstructive CAD on two separate days (within 90 days of each other). The same pharmacological stress agent (adenosine/regadenoson) was used for both scans. MPRI was calculated from time-intensity curves of the whole myocardium and blood pool at stress and rest. One experienced observer, blinded to clinical data, performed all measurements. Intra-class correlation coefficients (ICC), coefficient of variation (CoV), and Bland-Altman plots were determined.Results:Mean age was 53±10 years old and BMI 28±7 kg/m2; 47% had hypertension, 4% diabetes, 9% hyperlipidemia and 10% family history of CAD. Mean MPRI for the 17 women was higher for scan 2 compared to scan 1 (1.98±0.3 vs. 1.65±0.78, respectively, p<0.001); and this relationship persisted even when corrected for resting rate pressure product (RPP) (2.42±0.81 vs. 1.97±0.92, respectively, 0.002), The mean bias for MPRI between sequential scans was 0.34 (95% CI: 0.18 to 0.49, limits of agreement: −0.31, 0.98 and when corrected for resting RPP it was 0.45 (95% CI: 0.21 to 0.68, limits of agreement: −0.52, 1.41), ICC and CoV also indicated modest inter-scan reproducibility (ICC 0.57; CoV 20.3%), but both measures were comparable to values seen in prior studies in CAD populations and healthy volunteers.Conclusion:Inter-scan reproducibility of CMRI-derived MPRI in women with suspected CMD is modest, with relatively wide limits of agreement. This variability is similar to that seen in other populations, suggesting that some caution must be exercised when using absolute MPRI cut-offs in isolation for the diagnosis of CMD or repeated measures of MPRI to track response to therapy. Additional work is ongoing to improve reproducibility from both biological and technological standpoints.