The coronary calcium score is a more accurate predictor of significant coronary stenosis than conventional risk factors in symptomatic patients

Aims In this retrospective study we assessed the predictive value of the coronary calcium score for significant (> 50%) stenosis relative to conventional risk factors.
Methods and Results We investigated 5515 symptomatic patients from Denmark, France, Germany, Italy, Spain and the USA. All had risk factor assessment, computed tomographic coronary angiogram (CTCA) or conventional angiography and a CT scan for coronary artery calcium (CAC) scoring. 1539 (27.9%) patients had significant stenosis, 5.5% of whom had zero CAC. In 5074 patients, multiple binary regression showed the most important predictor of significant stenosis to be male gender (B = 1.07) followed by diabetes mellitus (B = 0.70) smoking, hypercholesterolaemia, hypertension, family history of CAD and age but not obesity. When the log transformed CAC score was included, it became the most powerful predictor (B = 1.25), followed by male gender (B = 0.48), diabetes, smoking, family history and age but hypercholesterolaemia and hypertension lost significance. The CAC score is a more accurate predictor of > 50% stenosis than risk factors regardless of the means of assessment of stenosis. The sensitivity of risk factors, CAC score and the combination for prediction of > 50% stenosis when measured by conventional angiogram was considerably higher than when assessed by CTCA but the specificity was considerably higher when assessed by CTCA. The accuracy of CTCA for predicting > 50% stenosis using the CAC score alone was higher (AUC = 0.85) than using a combination of the CAC score and risk factors with conventional angiography (AUC = 0.81).
Conclusion In symptomatic patients, the CAC score is a more accurate predictor of significant coronary stenosis than conventional risk factors.