Premium
Diagnostic accuracy of intravascular ultrasound‐derived minimal lumen area compared with fractional flow reserve—Meta‐analysis: Pooled accuracy of IVUS luminal area versus FFR
Author(s) -
Nascimento Bruno R.,
Sousa Marcos R.,
Koo BonKwon,
Samady Habib,
Bezerra Hiram G.,
Ribeiro Antônio L.P.,
Costa Marco A.
Publication year - 2014
Publication title -
catheterization and cardiovascular interventions
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.988
H-Index - 116
eISSN - 1522-726X
pISSN - 1522-1946
DOI - 10.1002/ccd.25047
Subject(s) - medicine , fractional flow reserve , intravascular ultrasound , cutoff , receiver operating characteristic , meta analysis , coronary artery disease , cardiology , gold standard (test) , area under the curve , clinical trial , radiology , stenosis , lumen (anatomy) , confidence interval , coronary angiography , myocardial infarction , physics , quantum mechanics
Although intravascular ultrasound minimal luminal area (IVUS‐MLA) is one of many anatomic determinants of lesion severity, it has been proposed as an alternative to fractional flow reserve (FFR) to assess severity of coronary artery disease. Objective Pool the diagnostic performance of IVUS‐MLA and determine its overall accuracy to predict the functional significance of coronary disease using FFR (0.75 or 0.80) as the gold standard. Methods Studies comparing IVUS and FFR to establish the best MLA cutoff value that correlates with significant coronary stenosis were reviewed from a Medline search using the terms “ fractional flow reserve ” and “ ultrasound .” DerSimonian Laird method was applied to obtain pooled accuracy. Results Eleven clinical trials, including two left main (LM) trials (total N = 1,759 patients, 1,953 lesions) were included. The weighted overall mean MLA cutoff was 2.61 mm 2 in non‐LM trials and 5.35 mm 2 in LM trials. For non‐LM lesions, the pooled sensitivity of MLA was 0.79 (95% CI = 0.76–0.83) and specificity was 0.65 (95% CI = 0.62–0.67). Positive likelihood ratio (LR) was 2.26 (95% CI = 1.98–2.57) and LR− was 0.32 (95% CI = 0.24–0.44). Area under the summary receiver operator curve for all trials was 0.848. Pooled LM trials had better accuracy: sensitivity = 0.90, specificity = 0.90, LR+ = 8.79, and LR− = 0.120. Conclusion Given its limited pooled accuracy, IVUS‐MLA's impact on clinical decision in this scenario is low and may lead to misclassification in up to 20% of the lesions. Pooled analysis points toward lower MLA cutoffs than the ones used in current practice. © 2013 Wiley Periodicals, Inc.