Establishing comprehensive, quantitative criteria for detection of restenosis and remodeling after percutaneous transluminal coronary angioplasty.

To establish comprehensive criteria for detecting restenosis and remodeling, inter- and intraobserver reproducibility of quantitative arteriography in the analysis of 20 lesions immediately after and 6 months after percutaneous transluminal coronary angioplasty (PTCA) were assessed. Geometric single-plane (minimum, maximum, mean diameter and percent diameter stenosis), biplane (absolute and relative cross-sectional area stenosis), relative densitometric area stenosis and the average of densitometric area stenosis in orthogonal views were compared. A high intra- and interobserver reproducibility of all absolute measurements was found, with the highest correlations for minimum diameter and cross-sectional area (interobserver, r = 0.85 and 0.85; intraobserver, r = 0.93, and 0.95 for minimum diameter and cross-sectional area, respectively). Of the relative measurements, biplane geometric percent cross-sectional area stenosis was the most reliable and percent densitometric area stenosis was the most variable (interobserver, r = 0.67; intraobserver, r = 0.71). Only small differences were demonstrated for the absolute measurements between the analysis of lesions immediately after PTCA and after follow-up, whereas a greater variability was found for relative measurements, especially videodensitometry. In both circumstances, a poor correlation between relative densitometric cross-sectional area from orthogonal views was found, whereas geometric elliptical cross-sectional area correlated quite well with the average of densitometric percent cross-sectional area in orthogonal views (interobserver, r = 0.86; intraobserver, r = 0.84). Thus, data in this study support the suitability of geometric quantitative analysis for the assessment of PTCA results. Densitometry was the least reliable quantitative parameter.