Evaluating the Coronary Artery Disease Consortium Model and the Coronary Artery Calcium Score in Predicting Obstructive Coronary Artery Disease in a Symptomatic Mixed Asian Cohort

Abstract

Background The utility of a given pretest probability score in predicting obstructive coronary artery disease (CAD) is population dependent. Previous studies investigating the additive value of coronary artery calcium (CAC) on pretest probability scores were predominantly limited to Western populations. This retrospective study seeks to evaluate the CAD Consortium (CAD2) model in a mixed Asian cohort within Singapore with stable chest pain and to evaluate the incremental value of CAC in predicting obstructive CAD. Methods and Results Patients who underwent cardiac computed tomography and had chest pain were included. The CAD2 clinical model comprised of age, sex, symptom typicality, diabetes, hypertension, hyperlipidemia, and smoking status and was compared with the CAD2 extended model that added CAC to assess the incremental value of CAC scoring, as well as to the corresponding locally calibrated local assessment of the heart models. A total of 522 patients were analyzed (mean age 54±11 years, 43.1% female). The CAD2 clinical model obtained an area under the curve of 0.718 (95% CI, 0.668–0.767). The inclusion of CAC score improved the area under the curve to 0.896 (95% CI, 0.867–0.925) in the CAD2 models and from 0.767 (95% CI, 0.721–0.814) to 0.926 (95% CI, 0.900–0.951) in the local assessment of the heart models. The locally calibrated local assessment of the heart models showed better discriminative performance than the corresponding CAD2 models ( P <0.05 for all). Conclusions The CAD2 model was validated in a symptomatic mixed Asian cohort and local calibration further improved performance. CAC scoring provided significant incremental value in predicting obstructive CAD.

Deep learning‐based prediction of coronary artery calcium scoring in hemodialysis patients using radial artery calcification

“…In high CACS and CAD, 21,22 and it has now become a reliable method to detect and quantitatively evaluate coronary artery calcification in patients to screen for CAD risk. 23,24 Furthermore, CACS can provide a long-term and independent prognosis for the clinical risk of CAD events.…”

“…In 1991, Agatston developed a calculation method to evaluate the burden of coronary atherosclerosis 20 by evaluation of CACS in coronary arteries, called the Agatston score, which is measured by CACT. Previous studies have demonstrated a correlation between high CACS and CAD, 21,22 and it has now become a reliable method to detect and quantitatively evaluate coronary artery calcification in patients to screen for CAD risk 23,24 . Furthermore, CACS can provide a long‐term and independent prognosis for the clinical risk of CAD events 25,26 .…”

Evaluating the American Heart Association/American College of Cardiology Guideline—Recommended and Contemporary Pretest Probability Models in a Mixed Asian Cohort: The Contribution of Coronary Artery Calcium

“… 3 , 10 Given the clinical utility of the Agatston coronary artery calcium score (CACS) as a gatekeeper to further testing, 11 it has been incorporated in PTP tools such as the CAD2 (CACS) model and the CACS‐weighted clinical likelihood model (CACS‐CL), resulting in its guideline recommendation. 1 , 8 , 12 , 13 …”

“…Tools derived and initially well validated in Western cohorts have performed less well in Asian contexts. 13 , 14 , 15 Singapore has a population comprising 3 major Asian ethnicities (Chinese [74.3%], Malay [13.5%], and Indian [9.0%]) that provide a unique snapshot of the genetic diversity across East, Southeast, and South Asia. 16 , 17 There is a paucity of evidence comparing contemporary Western tools to a model in the Asian context.…”

“…The recently developed local assessment of the heart (LAH) score is a PTP tool that was calibrated to a Singapore‐based mixed Asian population with the optional incorporation of CACS in estimating obstructive CAD defined by coronary computed tomography angiography (CCTA). 13 Although providing superior discrimination compared with CAD2, the validity of the LAH models has not been fully assessed.…”