Little is known about the intravascular ultrasound (IVUS) minimal lumen area (MLA) criteria and their accuracy in lesion subsets assorted according to vessel type, lesion location, vessel size, and clinical settings. We therefore assessed the accuracy of subgroup-specific cut-off values in predicting fractional flow reserve (FFR) <0.80. In total 692 consecutive patients with 784 coronary lesions were assessed by IVUS and FFR before intervention. All patients had ≥1 target vessel with a de novo lesion (30% to 90% diameter stenosis). For prediction of FFR <0.80 in the group overall, the best cut-off value of MLA was 2.4 mm(2) (sensitivity 84% and specificity 63%). Overall diagnostic accuracy was only 69%. In the subgroup analysis, the MLA cutoff was 2.4 mm(2) for the left anterior descending coronary artery, 1.6 mm(2) for the left circumflex coronary artery, and 2.4 mm(2) for the right coronary artery. By lesion location, the optimal cutoff was 2.6 mm(2) for proximal, 2.3 mm(2) for mid, and 1.9 mm(2) for distal segments. Furthermore, the cutoffs were 3.2 mm(2) in lesions with a larger RLD >3.5 mm and 1.9 mm(2) in lesions with a smaller RLD <2.75 mm. Nevertheless, diagnostic accuracies of all subgroup-specific criteria were <80%. In conclusion, because IVUS-measured MLA is only 1 of many factors affecting coronary flow hemodynamics, even subgroup-specific criteria were inaccurate in identifying ischemia-inducible stenosis. In conclusion, direct functional assessment is therefore essential in guiding treatment strategies for coronary lesions.
Copyright © 2012 Elsevier Inc. All rights reserved.