Are baseline conditions of coronary arteries sufficient for calculating angio-based index of microcirculatory resistance and fractional flow reserve?

Chenguang Li,Yumeng Hu,Jingpu Wang,Congcong Pan,Hao Lu,Yizhe Wu,Zhangwei Chen,Zhiqiang Pei,Li Shen,Jingsong He,Xiaochang Leng,Jianping Xiang,Junbo Ge

DOI: https://doi.org/10.21037/qims-23-72

2023-08-31

Quantitative Imaging in Medicine and Surgery

Abstract:Background: Angio-based index of microcirculatory resistance (IMR) and fractional flow reserve (FFR) have been developed, however, the differences between baseline and hyperemic data and their effects on their computation have not yet been discussed. This study aimed to compare the diagnostic performance of a novel method for calculating IMR and FFR from coronary angiography under baseline and hyperemic conditions.Methods: We performed a retrospective study to investigate the diagnostic performance of angiography-derived IMR (AccuIMR) and FFR (AccuFFRangio) computed from the hyperemic condition (AccuIMRhyp, AccuFFRangiohyp) and baseline condition (AccuIMRbase, AccuFFRangiobase) in 101 consecutive patients with chronic coronary syndrome (CCS) who underwent measurements of IMR and FFR at a single center, using wire-based IMR and FFR as the reference standard.Results: AccuIMRhyp showed much better correlation with IMR than AccuIMRbase (r=0.77 vs. 0.47, P<0.001). The diagnostic accuracy and area under the curve (AUC) for identifying significant microvascular dysfunction was higher for AccuIMRhyp than AccuIMRbase [92.1% (95% CI: 85.0-96.5%) vs. 83.2% (95% CI: 74.4-89.9%), P=0.012; 0.942 (95% CI: 0.877-0.979) vs. 0.815 (95% CI: 0.726-0.886), P=0.003]. The computed AccuFFRangio showed good correlations with FFR and good diagnostic performance under both hyperemic and baseline conditions [r=0.68 vs. 0.68, P>0.99; diagnostic accuracy =95.9% (95% CI: 89.8-98.9%) vs. 94.9% (95% CI: 88.4-98.3%), P=0.728; AUC =0.989 (95% CI: 0.942-1.000) vs. 0.973 (95% CI: 0.919-0.995), P=0.381]. The net reclassification index (NRI) demonstrated that hyperemic group had improved reclassification ability compared to the baseline group in identification of IMR >25 (NRI =0.20, P<0.001) and FFR ≤0.8 (NRI =0.11, P<0.001).Conclusions: By comparing the calculated angio-derived IMR and FFR under the baseline and hyperemic conditions, this study demonstrates that AccuIMR calculation is more accurate using the hyperemic condition, while AccuFFRangio calculation is accurate under both conditions.

radiology, nuclear medicine & medical imaging