Assessing diagnostic accuracy of intravascular imaging for functionally significant coronary stenosis: updated findings from a meta-analysis

Background Accurate discrimination of functionally significant coronary stenosis using intravascular imaging remains uncertain, particularly with regard to vessel size. This meta-analysis evaluates the diagnostic performance of intravascular ultrasound (IVUS) and optical coherence tomography (OCT) for identifying functionally significant coronary stenosis as confirmed by fractional flow reserve (FFR). Methods A systematic search of PubMed, Scopus and Google Scholar identified studies that assessed the diagnostic accuracy of IVUS and OCT by minimal luminal area (MLA) with FFR as the reference standard. Sensitivity and specificity were analysed across different vessel diameters including left main coronary artery (LM) lesions. Hierarchical models estimated the summary receiver operating characteristic curve, sensitivity and specificity. Results 31 studies involving 4039 patients and 4413 lesions were analysed. For IVUS, a median MLA threshold of 2.9 mm² (IQR: 2.6–3.2) predicted significant lesions, yielding an area under the curve (AUC) of 0.76. In vessels≥3 mm, this threshold increased to 3.0 mm² (IQR: 2.7–3.1) with an AUC of 0.76 while in smaller vessels it decreased to 2.6 mm² (IQR: 2.4–2.7) with an AUC of 0.79. For LM lesions, the median threshold was 6.0 mm² (IQR: 4.9–6.2) with an AUC of 0.88. OCT demonstrated a median threshold of 2.0 mm² (IQR: 1.7–2.3) and an AUC of 0.82 with better performance in larger vessels (≥3 mm, median 3.0 mm², AUC 0.87) than in smaller ones (<3 mm, median 1.8 mm², AUC 0.75). Conclusions IVUS and OCT show moderate diagnostic accuracy for identifying functionally significant coronary stenosis with OCT providing improved accuracy in vessels≥3 mm. IVUS is more accurate in assessing LM lesions, suggesting that vessel size should guide modality selection. PROSPERO registration number CRD 42024514538.