Reliability of mechanical and phased-array designs for serial intravascular ultrasound examinations--animal and clinical studies in stented and non-stented coronary arteries.

Abstract

Both mechanical and multi-element intravascular ultrasound designs have potential advantages and limitations that may impact on their value for clinical and research purposes. Determination of the reproducibility of measurements is critical before a given system can be used in studies such as regression of atherosclerosis trials. We performed serial intravascular ultrasound imaging with catheters using mechanical and phased-array designs in stented and non-stented coronary arteries in dogs and in patients. Both systems correlated well for areas (r > or = 0.90, p < 0.0001) and diameters (r > or = 0.84. p < 0.0001) in dogs and in patients. There was a slight difference between multi-element and mechanical designs for measurements of area (mean difference in dogs and in patients: -0.24 and 0.96 mm2, p < 0.055) and diameter (-0.08 and 0.16 mm, p < 0.0001). The reproducibility of the multi-element system for reanalysis of the same frames and for analysis of serial pullbacks was similar to the same measurements with the mechanical system (r > or = 0.96 for all measurements). The differences in absolute and relative variability between the mechanical and phased-array designs, both for reanalysis of same frames and serial pullbacks, were very small. Although multi-element and mechanical intravascular ultrasound designs are not strictly interchangeable, their similar reproducibility and the small differences in measurements demonstrate that both designs are acceptable alternatives for trials of regression of atherosclerosis. Determination of the variability for serial pullbacks of both designs was also important to assess the statistical power of such trials.