Briefly, the CCTA Improves Clinical Management of Stable Chest Pain (CICM-SCP) registry is an ongoing cohort of patients who were referred to CCTA as the first-line CIT to assess their SCP in relation to CCS (ClinicalTrials.gov Identifier: NCT04691037). As shown in Fig. 1, during a period of 24 months, between January 2016 and January 2018, 14,232 patients were finally enrolled in this registry, while details on the registry have been previously described [8, 17]. In the present analysis, patients with a CACS of 0 were included and followed up until January 2022. This study was approved by the Ethics Committees at local institutions and all participants provided informed consent.

Fig. 1.

Flow diagram. CCTA, coronary computed tomography angiography; ACS, acute coronary syndrome; CAD, coronary artery disease; CACS, coronary artery calcium score; NYHA, New York Heart Association.

The scanning, as well as interpretation of CACS and CCTA, were conducted as described previously [8, 17]. Patients with a CACS of 0 were included in the present study. Based on the Coronary Artery Disease–Reporting and Data System [20], the maximal degree of coronary diameter stenosis was defined as no CAD (0%), nonobstructive CAD (1–49%), and obstructive CAD ($\ge$50%).

Baseline clinical data, including age, sex, diabetes, hypertension, hyperlipidemia, smoking, family history of CAD, and symptoms were collected and defined as described previously [8, 17]. Hypertension was defined as blood pressure $\ge$140/90 mmHg or the use of anti-hypertension medication. Hyperlipidemia was defined as total cholesterol levels of $\ge$220 mg/dL, low-density lipoprotein cholesterol of $\ge$140 mg/dL, fasting triglycerides of $\ge$150 mm/dL, or receiving treatment with oral lipid-lowering agents. Diabetes was defined as fasting glucose levels over 7 mmol/L or treatment currently with either diet, oral glucose-lowering agents, or insulin. Smoking was defined as currently smoking or having smoking in the past 6 months. A family history of CAD was defined as a diagnosis of the disease in a male first-degree relative before 55 years of age or in a female first-degree relative before 65 years of age. Typical angina was defined as having 3 characteristics [1]: (1) substernal discomfort of characteristic quality, (2) precipitated by physical exertion or emotion, and (3) relieved with rest or nitroglycerin within 10 min. Atypical angina was defined as having 2 of the 3 defined characteristics. Nonanginal chest pain was characterized as chest pain or discomfort that met 1 or 0 of the 3 defined characteristics.

CACS-CL was estimated for each patient using the variables mentioned above (age, sex, diabetes, hypertension, hyperlipidemia, smoking, family history of CAD, and symptoms) as well as CACS based on plugins to the statistical software, packaged from the original study by Winther et al. [18]. The CADPTP models for implementation in R were available at https://github.com/CardioLab/cadptp/tree/master/R. Based on the recommendations of recent guidelines, CIT is not recommended in low-risk patients and should be referred only to high-risk patients [1, 2]. Thus, we selected two different cutoffs for CACS-CL (5% and 15%) to classify patients into either the low (CACS-CL 5% and $\le$15%) and high (CACS-CL $\ge$5% and $>$15%) risk group, respectively.

The research endpoint and the collection of follow-up data were previously defined in detail [8, 17]. Briefly, all patients were followed at the 6th, 12th, 24th, 36th, 48th, 60th, and 72nd months after CCTA. MACEs were defined as a composite of all-cause death and nonfatal myocardial infarction (MI). All-cause death was used rather than cardiovascular death to eliminate the competing events of cardiovascular and noncardiovascular mortality, as well as the need for possibly difficult adjudication of causes of death, especially given the relatively low mortality. All MACEs were adjudicated via a review of the follow-up information and medical records by an independent clinical event committee who were blinded to other study data.

All data analyses were performed using MedCalc (version 15.2.2; MedCalc Software, Mariakerke, Belgium) and R (version 3.2.4; R Foundation for Statistical Computing, Vienna, Austria). A p value 0.05 (two-tailed) was considered significant. The ANOVA or Mann–Whitney U test was used to evaluate the differences in continuous variables as appropriate. The $\chi$${}^2$ test or Fisher’s exact test was used to evaluate the differences in the categorical variables as appropriate. The discrimination and calibration of the CACS-CL model were assessed by the area under receiver operating characteristic curve (AUC) and Hosmer–Lemeshow goodness-of-fit statistic (H-L $\chi$${}^2$) [21]. Net reclassification improvement (NRI) was assessed in a reclassification table and used to determine how strategies using different CACS-CL cutoffs reclassified patients into various risk groups compared with each other [21]. The cumulative MACE-free survivals were estimated using Kaplan–Meier curves and were compared by the log-rank test. We used Cox proportional hazard regression models to calculate the adjusted hazard ratios (HRs) and 95% confidence intervals (CIs), which assessed the association of risk groups to the onset of the first MACE. The models were adjusted for age, sex, cardiovascular risk factors (diabetes, hypertension, hyperlipidemia, smoking, family history of CAD), symptoms, and CACS.

Baseline characteristics according to CACS-CL (5%, 5–15%, and $>$15%) are shown in Table 1. The CCIM-SCP cohort contained 14,232 symptomatic patients, 6689 (47%) of whom had a CACS of 0 and were included in the current investigation. The average age was 57.26 years, and 57% of the participants were men. Of the 6689 patients with a CACS of 0, 81% were classified as CACS-CL 5%, 15% as CACS-CL 5–15%, and 4% as CACS-CL $>$15%. All baseline characteristics were significantly different across the three CACS-CL categories (p 0.0001 for all comparisons).

The receiver operating characteristic curve for the CACS-CL model is presented in Fig. 2. The discrimination of the CACS-CL model was excellent, with an AUC of 0.805 (95% CI: 0.790–0.819, p 0.0001). The calibration plot for the CACS-CL model is illustrated in Fig. 3. Graphically, in the group with negative CACS-CL, the probability of the obstructive CAD was overestimated, resulting in a moderate calibration (H-L $\chi$${}^2$ = 13.19, p 0.0001).

Fig. 2.

ROC curve for the CACS-CL model to predict obstructive CAD. ROC, receiver operating characteristic; CACS-CL, coronary artery calcium score-weighted clinical likelihood; CAD, coronary artery disease.

Fig. 3.

Comparison of CACS-CL and proportion of obstructive CAD on CCTA by deciles of CACS-CL. CACS-CL, coronary artery calcium score-weighted clinical likelihood; CAD, coronary artery disease; CCTA, coronary computed tomography angiography; CL, clinical likelihood.

In total, 247 (4%), 1023 (15%), and 5419 (81%) patients had obstructive, nonobstructive, and no CAD, respectively, based on CCTA. As illustrated in Fig. 4, compared to patients with CACS-CL 5% and 5–15%, patients with CACS-CL $>$15% had more (p 0.0001 for both comparisons) obstructive (36% (219/612) versus 0.4% (21/5204) and 0.8% (7/873)) and nonobstructive CADs (49% (297/612) versus 12% (610/5204) and 13% (116/873)), respectively. However, there was no significant (p = 0.0637) difference between patients with CACS-CL 5% and 5–15% in terms of CAD distribution.

Fig. 4.

Distribution of CAD according to CACS-CL (5%, 5–15%, and $\mathrm{>}$15%). CACS-CL, coronary artery calcium score-weighted clinical likelihood; CAD, coronary artery disease; CCTA, coronary computed tomography angiography.

During the 5-year follow-up (median: 61 months, interquartile range: 54 to 68 months), 1% (73/6689) of patients experienced MACEs: 19 patients died and 54 patients suffered from nonfatal MI. As shown in Table 2, the rates of the MACEs were 0.6%, 1% (10/863), and 13% in patients with CACS-CL 5%, 5–15%, and $>$15%, respectively. Among the 32, 95, and 542 typical anginal patients with CACS-CL 5%, 5–15%, and $>$15%, 23, 5, and 19 experienced MACEs, respectively. Thus, most (64%, 47/73) MACEs occurred in typical anginal patients. Moreover, as shown in Table 2, typical anginal patients with a CACS-CL 15% had similar a rate of MACEs as patients with CACS-CL $>$15% who have not yet developed typical angina. This suggests that three distinct risk cohorts can be further redefined. Cohort 1 (very low risk) included patients with CACS-CL 15% but without typical angina. The rate of MACEs in this cohort was 0.3% (17/5805). Cohort 2 (moderate risk) included patients with CACS-CL 15% and typical angina or patients with CACS-CL $>$15% but without typical angina. The rate of MACEs in this cohort was 4.3% (37/852). Cohort 3 (high risk) included patients with CACS-CL $>$15% and typical angina. The rate of MACEs in this cohort was 59% (19/32).

Fig. 5 illustrates the Kaplan–Meier estimates for patients surviving free from MACEs. The patients in the high-risk group according to the cutoff for CACS-CL = 5% and 15% had a significantly higher (log-rank p 0.0001 for both) risk of MACEs, respectively, although the association between CACS-CL = 15%-determined risk groups and MACEs was stronger than for CACS-CL = 5% (adjusted HR for CACS-CL $>$15% versus $\le$15%: 7.24, 95% CI 1.93–16.42; adjusted HR for CACS-CL $\ge$5% versus 5%: 3.68, 95% CI 1.50–8.26).

Fig. 5.

Cumulative survival probability from MACEs according to different cutoffs for CACS-CL. (A) CACS-CL $\ge$5% versus 5%. (B) CACS-CL $>$15% versus $\le$15%. CACS-CL, coronary artery calcium score-weighted clinical likelihood; MACE, major adverse cardiovascular event; CCTA, coronary computed tomography angiography.

Table 3 is the reclassification table for obstructive CAD. Compared to the strategy, which used the cutoff for CACS-CL = 5%, among the 6442 patients without obstructive CAD, 901 patients were correctly reclassified from the high-risk group to the low-risk group. However, 35 patients were incorrectly reclassified from the low-risk group to the high-risk group by the strategy that used the cutoff for CACS-CL = 15%. Among the 247 positive patients, the strategy that used the cutoff for CACS-CL = 15% correctly reclassified one patient from the low-risk group to the high-risk group, although eight patients were incorrectly reclassified from the high-risk group to the low-risk group. Therefore, the NRI of the strategy that used the cutoff for CACS-CL = 15% compared with the strategy that used the cutoff for CACS-CL = 5% was 13.45% for negative patients, –2.84% for positive patients, and 10.61% for all patients. Similarly, when regarding any CAD in Table 4, the NRI was 13.84% for negative patients, –9.68% for positive patients, and 4.16% for all patients.