Comparison of assessment of native coronary arteries by standard versus three-dimensional coronary angiography

Vessel foreshortening is a major limitation of standard coronary angiography due to the 2-dimensional representation of 3-dimensional structures. Three-dimensional models may overcome it. The aim of this study was to compare measurements of coronary segments from quantitative coronary angiography (QCA) in an operator-selected working view of standard 2-dimensional coronary angiography with those from 3-dimensional coronary angiography (3D-CA) reconstruction models, which are automatically generated from software applied to rotational coronary angiographic acquisitions. Patients who underwent percutaneous coronary intervention were considered. Two or 3 segments of the artery needing treatment were prespecified, using bifurcations as edges. The operator selected a working view from standard angiography as the view best representing each segment. Rotational angiography was performed, allowing 3-dimensional reconstruction of the selected segments. Additionally a marker guidewire (with 4 markers 10 mm away from one another at the distal tip) was used to further measure segment length, and it was considered the gold standard reference. In 36 patients, 81 segments from 12 left anterior descending, 12 circumflex, and 12 right coronary arteries were evaluated. Three-dimensional coronary angiography was always feasible. Although reference vessel diameter was not different between 3D-CA and QCA (p >0.05), segment length measurements were on average 2.3 +/- 2.5 mm longer with 3D-CA than with QCA (p < 0.001) and 0.4 +/- 1.8 mm longer than with marker guidewire measurement (p = 0.047). Marker guidewire measurements were 1.9 +/- 2.8 mm longer than QCA measurements (p <0.001). According to Bland-Altman plots, 3D-CA and marker guidewire measurements had the best agreement. In conclusion, 3-dimensional coronary modeling is highly feasible and yields more accurate assessments of the lengths of coronary segments than standard QCA. (C) 2008 Elsevier Inc. All rights reserved.