Quantitative flow ratio (QFR) identifies functional relevance of non-culprit lesions in coronary angiographies of patients with acute myocardial infarction

Fig. 1. Details of study inclusion. Allowed ischemia tests were staged FFR, stress cardiac MRI or SPECT within 6 months of the initial angiography. Abbreviations: MI myocardial infarction; CAD coronary artery disease; CABG coronary artery bypass graft

Fig. 2. A representative QFR analysis of a non-culprit lesion in a patient with STEMI. In A, a cranial LAO projection shows total acute LCx occlusion (marked with a white asterisk) as well as an intermediate proximal LAD stenosis (indicated by the red arrow) as non-culprit lesion. In B, ECG-supported selection of end-diastolic frames from cranial LAO (right) and caudal RAO (right) views, both with good visualization of the non-culprit LAD stenosis. In C, automatic detection of vessel contours in the two projections (two upper quadrants), three-dimensional vessel reconstruction (right lower quadrant) and diameter profile (left lower quadrant). In D, the resulting QFR value (0.81) is shown, demonstrating no functional relevance of the non-culprit lesion.

Fig. 3. aQFR effectively predicts hemodynamic relevance of non-culprit lesions. In A, ROC curve shows excellent diagnostic efficiency of aQFR and the optimal aQFR cut-off-value for assessing ischemia of non-culprit lesions in patients with MI. In B, superiority of aQFR (blue line) compared to percent area stenosis (%DS, orange line) and minimal luminal diameter (MLD, red line) in the assessment of hemodynamic relevance of non-culprit lesions.

Fig. 4. Head-to-head comparison of aQFR with staged FFR. In A, a strong linear correlation between aQFR and staged FFR of non-culprit lesions is shown. In B, Bland–Altman plot shows a good agreement of aQFR and staged FFR.

Fig. 5. aQFR-based model for direct evaluation non-culprit lesions. Specificity (in red) and sensitivity (in blue) in predicting hemodynamic relevance of non-culprit lesions of patients with AMI are plotted in dependence of aQFR values; based on a (pre-specified) 95% sensitivity and 95% specificity level (which is marked by the black dashed line), we selected aQFR cut-offs. Therefore, we color-coded aQFR values in red (definite hemodynamic relevance), green (definite hemodynamic non-relevance) and grey (uncertain, further testing required)