[Computer-assisted geometric measuring technic in coronary angiography interval studies: results of initial angiograms of the International Nifedipine Trial of Anti-atherosclerotic Therapy (INTACT) study]

Abstract

In 396 of 423 (93.6%) patients with mild to moderate coronary artery disease participating in a coronary angiographic follow-up trial, diameters of the epicardial coronary arteries were measured with an automatic contour detection system (CAAS). The underlying INTACT study (International Nifedipine Trial on Antiatherosclerotic Therapy), a prospective, placebo-controlled, randomized, double-blind multicenter trial, investigates the influence of the calcium antagonist nifedipine (80 mg/day) on the progression of coronary atherosclerosis over a three-year interval. The study is based on coronary angiograms repeated in identical projections after premedication with 10 mg isosorbide dinitrate sublingually. For quantitative analysis the coronary artery system was, in consideration of all anatomical variations, subdivided into 25 segments. In the first angiograms of the 396 patients evaluated up to April 1, 1988, altogether 5,425 different coronary segments could be analyzed over their entire length in one or more angiographic projections--on the average 13.7 +/- 2.8 segments per patient. Analysis parameters were the mean diameters of the entire segments and of the individual subsegments (about 5 mm in length). The 10 major proximal segments could be evaluated in 76-94% of patients respectively in more than two different angiographic projections on the average; in 0-13% of patients the respective segments were occluded. In 1-4% of patients the evaluation of these segments was prevented by poor film quality and in 1-16% of patients prevented by anatomical abnormalities (e.g., segment too short or too small). 184 segments were found occluded (RCA 42%, LAD 30%, CX 28%) and 909 mostly low-grade to moderate stenoses (RCA 34%, CX 32%, LAD 30%, left main 4%) were analyzed with a special algorithm. The following obstruction parameters were derived: minimal diameter, percentage severity, length, and plaque area. The present data demonstrate that in an angiographical multicenter follow-up study such as INTACT a nearly complete quantitative morphometric analysis of the visualized coronary artery system can indeed be obtained in virtually all angiograms when a computer-assisted contour detection system is applied.