Reproducibility of intravascular ultrasound radiofrequency data analysis: implications for the design of longitudinal studies

The purpose of this study was to assess in vivo the reproducibility of tissue characterization using spectral analysis of intravascular ultrasound (IVUS) radiofrequency data (IVUS-VH). Despite the need for reproducibility data to design longitudinal studies, such information remains unexplored. IVUS-VH (Volcano Corp., Rancho Cordova, USA) was performed in patients referred for elective percutaneous intervention and in whom a non-intervened vessel was judged suitable for a safe IVUS interrogation. The IVUS catheters used were commercially available catheters (20 MHz, Volcano Corp., Rancho Cordova, USA). Following IVUS-VH acquisition, and after the disengagement and re-engagement of the guiding catheter, an additional acquisition was performed using a new IVUS catheter. Fifteen patients with 16 non-significant lesions were assessed by 2 independent observers. The relative inter-catheter differences regarding geometrical measurements were negligible for both observers. The inter-catheter relative difference in plaque cross-sectional area (CSA) was 3.2% for observer 1 and 0.5% for observer 2. The limits of agreement for (observer 1 measurements) lumen, vessel, plaque and plaque burden measurements were 0.82, -1.10 mm(2); 0.80, -0.66 mm(2); 1.08, -0.66 mm(2); and 5.83, -3.89%; respectively. Limits of agreement for calcium, fibrous, fibrolipidic and necrotic core CSA measurements were 0.22, -0.25 mm(2); 1.02, -0.71 mm(2); 0.61, -0.65 mm(2); and 0.43, -0.38 mm(2) respectively. Regarding the inter-observer agreement, the limits of agreement for lumen, vessel, plaque and plaque burden measurements were 2.61, -2.09 mm(2); 2.20-3.03 mm(2); 1.70, -3.04 mm(2); and 9.16, -16.41%; respectively, and for calcium, fibrous, fibrolipidic and necrotic core measurements of 0.08, -0.09 mm(2); 0.89, -1.28 mm(2); 0.74, -1.06 mm(2); and 0.16, -0.20 mm(2); respectively. The present study demonstrates that the geometrical and compositional output of IVUS-VH is acceptably reproducible.