Redefining the normal angiogram using population-derived ranges for coronary size and shape: validation using intravascular ultrasound and applications in diverse patient cohorts

To develop a method for quantitating coronary angiographic abnormalities of segmental size and shape (tapering) in comparison to gender- and segment-specific, population derived, normal values. In the absence of obvious focal stenoses, remodeling renders the angiogram insensitive to the presence of atherosclerosis and invalidates use of a "normal reference segment" for calculation of percent diameter stenosis. Equations were created for detection of size/shape abnormalities of coronary angiographic segments. After validation using intravascular ultrasound (IVUS), the equations were applied to a cohort of segments judged to be completely normal by a panel of highly experienced, core laboratory technicians; and a cohort of patients judged by an experienced interventionalist to have completely normal coronaries. In patients assessed by core technicians, 53% (162/303) of males, 39% (209/538) of normal segments in males, 60% (56/94) of females, and 40% (81/205) of normal segments in females had quantifiable abnormalities. In patients with normal coronaries as judged by an experienced interventionalist, 100% of males (n = 14) and females (n = 19), 37% (67/182) of segments in males and 43% (105/247) of segments in females had abnormalities. The left main segment was most commonly abnormal. We propose a set of equations validated using IVUS and based on gender- and segment-specific normal values for coronary angiographic size and shape that markedly improves the sensitivity of the coronary angiogram for detection of abnormalities. The method should replace the unfounded practice of labeling coronary angiograms as "normal" based solely on the failure to detect focal stenoses.