PROGNOSTIC VALUE OF CORONARY CALCIFICATION AND ANGIOGRAPHIC STENOSES IN PATIENTS UNDERGOING CORONARY ANGIOGRAPHY

Objectives. This investigation sought to determine the relative progno stic value of coronary calcific deposits and coronary angiographic fin dings for predicting coronary heart disease-related events in patients referred for angiography. Background. The relation among coronary cal cification, coronary stenoses and coronary heart disease-related event s is of interest on a clinical as well as a pathophysiologic basis. Me thods. Four hundred ninety-one symptomatic patients underwent coronary angiography and electron beam computed tomography at five different c enters between April 1989 and December 1993. The electron beam compute d tomograms were interpreted by a cardiologist with no knowledge of th e coronary angiographic and clinical data. Receiver operating characte ristic (ROC) curves were constructed to determine the relation between electron beam computed tomographic and coronary angiographic findings . A follow-up telephone survey was completed in 86% of patients. The r ecords for all patients who died or were admitted to the hospital for chest pain or suspected myocardial infarction were reviewed by three o ther cardiologists with no knowledge of the coronary angiographic and electron beam computed tomographic study results. Results. The mean (/-SE) area under the ROC curve was 0.75 +/- 0.02 for the coronary calc ium score, indicating moderate discriminatory power for this score for predicting angiographic findings. Thirteen coronary heart disease-rel ated deaths and eight nonfatal acute infarctions occurred over 30 +/- 13 months. Scores were sorted in ascending order and divided into quar tiles of equal size. One patient in the first quartile had a fatal myo cardial infarction (coronary calcium score range 0 to 2.1); 2 in the s econd quartile (range 2.1 to 75.3), 8 in the third quartile (range 75. 3 to 397.1) and 10 in the fourth quartile (>397.1) had a coronary hear t disease related event. Application of bivariate logistic regression showed that log score but not number of angiographically dis eased ves sels significantly predicted the probability of a coronary heart disea se-related event occurring during follow up. Conclusions. Electron bea m computed tomographic calcium scores correlate moderately well with a ngiographic findings. These scores predict coronary heart disease-rela ted events in patients undergoing angiography as well as do the number of angiographically affected arteries.