Pericardial fat volume measured by non-contrast tomography improves risk prediction of subclinical coronary artery disease estimated by coronary artery calcium score and Framinghan risk score

Abstract

Background: Although coronary artery calcium (CAC) score has high specificity for predicting subclinical coronary artery diasease (CAD), it lacks sensitivity. Therefore, we aimed to evaluate if the measurement of pericardial fat volume (PFv) by non-contrast computed tomography (CT) in combination with CAC score would improve discrimination of coronary atherosclerotic plaques. Methods: 132 consecutive patients with suspected CAD were addressed to perform 64-slice CT both without contrast for CAC score and coronary angiography. During CT we evaluated both CAC score and PFv (in mm3); and during angiography we evaluated PFv and the presence of coronary stenosis higher than 30%. In addition, we measured visceral abdominal adipose tissue (VAT) by abdominal CT and calculated Framinghan Heart Risk (FHR) score. Results: We found a strong association between PFv measured by non-contrast CT and the presence of coronary plaques with multivariate odds of 5.24 (95% CI 1.2–22, p=0.026). On sensitivity (sn), specificity (sp) and negative predictive value (npv) analysis for discrimination of coronary plaques, CAC score<10 + PFv measured by non-contrast CT presented the best predictive capacity (sn=93.3%; sp=100%; npv=99%) in comparison to CAC score<10 + PFv by angiography (sn=92.0%; sp=100%; npv=97%) or CAC score<10 + VAT (sn=90.7%; sp=100%; npv=95%); or even CAC score = 0 alone (sn=80%; sp=98%; npv=93%) or CAC score<10 alone (sn=80%; sp=100%; npv=93%). Furthermore, while in the entire cohort CAC score alone was better than the combination of CAC and PFv, when considering only patients with CAC score < 10, CAC score alone attained an area under the curve (AUC) of 0.808 (0.64 – 0.97; p=0.001) versus 0.853 (0.74 – 0.97; p<0.001) in CAC plus PFv model. Consistently, adding PFv to FHR score was better both in the entire cohort and in patients with CAC score<10, but tended to be more accurate in lower risk patients. Conclusions: This is the first study to shown PFv evaluation by non-contrast CT in combination with CAC score improves risk prediction of subclinical coronary artery disease in comparison with CAC score alone, especially in lower risk patients, without additional costs and risks of contrast infusion of coronary angiography.