Coronary Artery Calcium Score–Weighted Clinical Likelihood Model Performance in Patients with Stable Chest Pain and Coronary Artery Calcium Scores of Zero

TL;DR

This study evaluates a risk model for predicting heart disease in patients with chest pain and zero coronary artery calcium scores.

Contribution

The study shows that a 15% cutoff in the CACS-CL model is more effective than a 5% cutoff for predicting outcomes in these patients.

Findings

Higher CACS-CL values were linked to increased coronary artery disease prevalence.

The 15% cutoff showed stronger associations with major adverse cardiovascular events than the 5% cutoff.

Using the 15% cutoff improved risk classification accuracy compared to the 5% cutoff.

Abstract

For individuals with persistent stable chest pain (SCP) and
a coronary artery calcium score (CACS) of 0, it might be challenging to establish
the best risk assessment method for determining the individuals who will not
benefit from further cardiovascular imaging testing (CIT). Thus, we investigated
the CACS-weighted clinical likelihood (CACS-CL) model in SCP patients with a CACS
of 0.

Thus, to assess SCP, we originally enrolled 14,232
individuals for CACS and coronary computed tomography angiography (CCTA) scans
between January 2016 and January 2018. Finally, patients with a CACS of 0 were
included and followed up ​until January 2022. According to the established
CACS-CL cutoffs of 15% and 5%, the associations between coronary artery disease
(CAD) and major adverse cardiovascular events (MACEs) in risk groups were
evaluated, alongside the net reclassification improvement (NRI).

Of the 6689 patients with a CACS of 0, the prevalence of CAD
increased significantly (p
< 0.0001) in patients with higher CACS-CL.
However, there was no significant difference in the CAD distribution (p
= 0.0637) between patients with CACS-CL <5% and 5–15%. The association
between the CACS-CL = 15%-determined risk groups and the occurrence of MACEs was
stronger than for a CACS-CL = 5% (adjusted hazard ratio (HR): 7.24 (95% CI: 1.93–16.42)
versus 3.68 (95% CI: 1.50–8.26)). Compared with the cutoff for CACS-CL = 5%,
the NRI was 10.61% when using a cutoff for CACS-CL = 15%.

Among patients with an SCP and CACS of 0, the CACS-CL model provided accurate
predictions of CAD and MACEs. Compared to the cutoff for CACS-CL = 5%, the
cutoff for CACS-CL = 15% seemed to be more effective and safer for deferring
further CIT.

NCT04691037.

Figures