OBJECTIVES: This study sought to evaluate the diagnostic performance of optical frequency domain imaging (OFDI) for recognition of coronary plaque morphologies and to assess additive values of integrated backscatter intravascular ultrasound (IB-IVUS) in detection of vulnerable plaque. BACKGROUND: Precise diagnosis of coronary lesions susceptible to plaque rupture and thrombosis may serve to stratify the risk of future coronary events and to make decisions for appropriate treatment of choice. METHODS: Twenty-seven coronary arteries from 14 human autopsy hearts were interrogated ex vivo by OFDI and IB-IVUS. Imaged segments were sectioned at 3 mm intervals where a total of 360 pairs of cross-sectional images coregistered to histology were investigated. RESULTS: Overall, OFDI could depict various plaque components and structures such as fibrous tissue, sheet and nodular calcification, lipid, cholesterol crystals, and healed plaque rupture. OFDI could detect 14 of 18 thin-cap fibroatheroma (TCFA), however, the diagnostic accuracy was not high (positive predictive value [PPV] = 60.9%, κ = 0.664; area under the curve [AUC]: 0.88) mainly because of signal interference from macrophages. Further, we defined IB-IVUS-derived TCFA by recursive partitioning analysis as: 1) cross-sectional % lipid area >65.1%; 2) % lipid area >32.3 but <65.1% with plaque area >10.5 mm(2), where TCFA detection by IB-IVUS alone was marginal (PPV = 50.0%, κ = 0.545; AUC: 0.82). However, when IB-IVUS was combined with OFDI, all pseudo OFDI-derived TCFA (non-TCFA on histology) were excluded. Accordingly, PPV of TCFAs diagnosed by both OFDI and IB-IVUS was improved to 100.0% (κ = 0.704; AUC: 0.93). CONCLUSIONS: OFDI could recognize detailed morphologies of human coronary plaque. However, diagnostic accuracy of both OFDI alone and IB-IVUS alone to identify TCFA is limited. Combination of IB-IVUS with OFDI improved the accuracy for TCFA detection, suggesting hybrid imaging or further development of novel devices will be required to identify coronary lesions responsible for future events.
OBJECTIVES: This study sought to evaluate the diagnostic performance of optical frequency domain imaging (OFDI) for recognition of coronary plaque morphologies and to assess additive values of integrated backscatter intravascular ultrasound (IB-IVUS) in detection of vulnerable plaque. BACKGROUND: Precise diagnosis of coronary lesions susceptible to plaque rupture and thrombosis may serve to stratify the risk of future coronary events and to make decisions for appropriate treatment of choice. METHODS: Twenty-seven coronary arteries from 14 human autopsy hearts were interrogated ex vivo by OFDI and IB-IVUS. Imaged segments were sectioned at 3 mm intervals where a total of 360 pairs of cross-sectional images coregistered to histology were investigated. RESULTS: Overall, OFDI could depict various plaque components and structures such as fibrous tissue, sheet and nodular calcification, lipid, cholesterol crystals, and healed plaque rupture. OFDI could detect 14 of 18 thin-cap fibroatheroma (TCFA), however, the diagnostic accuracy was not high (positive predictive value [PPV] = 60.9%, κ = 0.664; area under the curve [AUC]: 0.88) mainly because of signal interference from macrophages. Further, we defined IB-IVUS-derived TCFA by recursive partitioning analysis as: 1) cross-sectional % lipid area >65.1%; 2) % lipid area >32.3 but <65.1% with plaque area >10.5 mm(2), where TCFA detection by IB-IVUS alone was marginal (PPV = 50.0%, κ = 0.545; AUC: 0.82). However, when IB-IVUS was combined with OFDI, all pseudo OFDI-derived TCFA (non-TCFA on histology) were excluded. Accordingly, PPV of TCFAs diagnosed by both OFDI and IB-IVUS was improved to 100.0% (κ = 0.704; AUC: 0.93). CONCLUSIONS: OFDI could recognize detailed morphologies of human coronary plaque. However, diagnostic accuracy of both OFDI alone and IB-IVUS alone to identify TCFA is limited. Combination of IB-IVUS with OFDI improved the accuracy for TCFA detection, suggesting hybrid imaging or further development of novel devices will be required to identify coronary lesions responsible for future events.