The incremental value of coronary artery calcium in predicting obstructive coronary artery disease in a symptomatic mixed Asian cohort: external validation of RF-CL and CACS-CL

Abstract

Abstract Background Estimating the pre-test probability (PTP) of coronary artery disease (CAD) is an essential step in the evaluation of patients with stable chest pain. Risk factor-weighted clinical likelihood (RF-CL) and coronary artery calcium score-weighted clinical likelihood (CACS-CL) are two simple models recently proposed to estimate the probability of obstructive CAD. However, these tools were predominantly developed and validated on Western populations. There is limited evidence of their predictive value in Asian populations. Purpose We sought to evaluate the performance of RF-CL and CACS-CL in a symptomatic mixed Asian cohort. Methods The study cohort included symptomatic patients referred for coronary computed tomography angiography (CTA) at a tertiary cardiac centre in Singapore. Patients with known CAD were excluded. Obstructive CAD was defined as ≥ 50% stenosis on coronary CTA. RF-CL was calculated with age, sex, symptom characteristics, and the number of risk factors. CACS-CL was calculated from RF-CL by further incorporating the CACS. Discriminative performance was evaluated with the area under the receiver-operating characteristic curves (AUC). Categorical net reclassification index (NRI) was calculated to assess the reclassification improvement of CACS-CL as compared with RF-CL. Results A total of 1293 patients were analysed (mean age 55.8±11.3 years, 45.6% female). The majority of the patients were of Asian descent (73.9% Chinese, 10.6% Indian, 6.4% Malay). The overall prevalence of obstructive CAD was 26.1% (n=338). The RF-CL achieved an AUC of 0.72 (95% CI, 0.69-0.75). The incorporation of CACS in CACS-CL further improved the AUC to 0.89 (95% CI, 0.87-0.91; p<0.001) (Figure 1). Compared to RF-CL, CACS-CL demonstrated an improved reclassification, resulting in a categorical NRI of 0.28 (95% CI, 0.23-0.33) at a 5% cutoff. Among 955 patients without obstructive CAD, the CACS-CL model reclassified a significant proportion (21.4%, n=204) into the low-risk group, in which further testing is not recommended. (Table 1) Conclusions In a symptomatic mixed Asian cohort, the RF-CL model had good discriminative performance in predicting obstructive CAD. Incorporating CACS in the CACS-CL model improved both discrimination and reclassification substantially. These results indicate that the RF-CL and CACS-CL perform similarly in different ethnicities and support their use in optimising the diagnostic workflow in Asian patients with stable chest pain.