Combination of coronary CT angiography, FFRCT, and risk factors in the prediction of major adverse cardiovascular events in patients suspected CAD

Abstract Background To examine the utility of fractional flow reserve by coronary computed tomography (CT) angiography (FFRCT) for predicting major adverse cardiovascular events (MACE) in patients with suspected coronary artery disease (CAD). Methods This was a nationwide multicenter prospective cohort study including consecutive 1187 patients aged 50−74 years with suspected CAD and had available coronary CT angiography (CCTA). In patients with ≥50% coronary artery stenosis (CAS), FFRCT was further analyzed. The Cox proportional hazards model was used to examine the association of FFRCT and cardiovascular risk factors with incident MACE within 2 years. Results Among 933 patients with available information on MACE within 2 years after enrollment, the incidence rate of MACE was higher in 281 patients with CAS than in those without CAS (6.11 vs. 1.16 per 100 patient‐year). In 241 patients with CAS, the Cox proportional hazards analysis showed that FFRCT as well as diabetes mellitus and low high‐density lipoprotein cholesterol level were independently associated with incident MACE. Moreover, the hazard ratio was significantly higher in patients harboring all three factors compared to those harboring 0−2 of the three factors (6.01; 95% confidence interval: 2.77−13.03). Conclusions Combinatorial assessment using CCTA for stenosis, FFRCT, and risk factors was useful for more accurate prediction of MACE in patients with suspected CAD. Among patients with CAS, those with lower FFRCT, diabetes mellitus, and low high‐density lipoprotein cholesterol level were at highest risk for MACE during the 2‐year period following enrollment.

K E Y W O R D S cardiovascular risk factor, computed tomography, diabetes mellitus, fractional flow reserve, major adverse cardiovascular event

| INTRODUCTION
Coronary artery stenosis (CAS) is associated with the future development of cardiovascular events and is therefore important in the evaluation of patients with suspected or known coronary artery disease (CAD). 1 Coronary computed tomography (CT) angiography (CCTA) is a commonly used high-utility tool for the noninvasive evaluation of CAS, and CCTA findings inform treatment approaches including percutaneous coronary intervention. 2 Fractional flow reserve (FFR) determined with CCTA (FFR CT ), which reflects coronary artery blood flow and reserve capacity, has been recently approved for the clinical evaluation of cardiac ischemia in patients with CAS. [3][4][5][6] A cutoff FFR CT value of ≤0.8 was demonstrated to indicate reduced capacity of coronary artery flow, which suggested therapeutic intervention and revascularization. 7 Moreover, a recent meta-analysis suggested that negative FFR CT was associated with low incidence of adverse events at 12 months compared to positive FFR CT in patients with stable CAD. 8 However, more accurate and valuable prediction of major adverse cardiovascular events (MACE) using FFR CT is open to consideration. The predictive ability of FFR CT in combination with cardiovascular risk factors for future MACE, which has not yet been elucidated, might be an effective strategy for the management of patients with suspected and known CAD. Therefore, we investigated the utility of a combinatorial approach, including cardiovascular risk factors and CCTA to evaluate stenosis and FFR, in predicting future MACE in patients with suspected CAD. To this end, we used data from the Nationwide Gender-Specific Atherosclerosis Determinants Estimation and Ischemic Cardiovascular Disease Prospective Cohort (NADESICO) study. 9 2 | METHODS

| Study design
This study was a subanalysis of the NADESICO study, a prospective, multicenter cohort study that was designed to evaluate sex differences in the association of coronary atherosclerosis including coronary artery calcification (CAC) with MACE. The study protocol was registered with the UMIN Clinical Trials Registry (UMIN-CTR ID: UMIN000001577) before the release of data to the lead author of the present study. The protocol was approved by the institutional review boards of all participating centers including the National Cerebral and Cardiovascular Center (NCVC) (M20-029-9), and written informed consent was obtained from all patients before participation.

| Participants
The detailed study protocol was described elsewhere. 9

| Data collection
Clinical data on diagnostic and therapeutic measures were collected by investigators at each hospital and sent to the NCVC. Hypertension was defined as current use of antihypertensive agents with systolic blood pressure >140 mmHg or diastolic blood pressure >90 mmHg while resting. Diabetes mellitus was defined as self-reported history of adult-onset fasting glucose ≥126 mg/dL or use of insulin or oral glucose-lowering medications. Dyslipidemia was defined as the current use of any lipid-lowering agents, triglyceride ≥150 mg/dL, low-density lipoprotein cholesterol ≥140 mg/dL, or high-density lipoprotein (HDL) cholesterol ≤40 mg/dL in men and ≤50 mg/dL in women. Data on smoking habits and medical history were collected at enrollment by a questionnaire. Laboratory examination included complete blood count, lipid profile, and plasma glucose level.
CT was performed using 64 or more channels with electrocardiography gating according to the Japanese Circulation Society guidelines and institutional protocols. The CT images were digitally transferred to the NCVC and interpreted in a blind fashion by an independent imaging core laboratory using SYNAPSE VINCENT (FUJIFILM Medical Co., Ltd).
A Japan Radiological Society Board-certified radiologist with extensive experience in coronary CT blinded to all clinical data interpreted the plain CT images using the Agatston CAC scoring method. In CCTA images, all >1.5 mm vessels were assessed for the presence of stenosis and the severity was determined by visual estimation using a percentage of the vessel diameter. The stenosis severity with ≥50% was defined as CAS in the present study. Additionally, FFR CT analysis was performed in patients with CAS. All FFR CT analyses were performed in HeartFlow, Inc. In each patient, the lowest FFR CT value of major coronary arteries, such as right coronary artery, left anterior descending coronary artery, and circumflex branch artery, was defined as the FFR CT value used for all analyses. All FFR CT values lower than 0.50 were defined as an FFR CT value of 0.50.

| Follow-up and outcomes
All patients were evaluated for the presence of MACE. Attending physicians contacted patients who did not visit the hospital more than once a year via telephone or mail. If a patient visited another hospital with CAD, the attending physician inquired about the onset with the hospital. MACE included cardiovascular death, myocardial infarction, late revascularization (>3 months after the indexed CCTA), stroke, hospitalization for unstable angina, heart failure, and aortic disease. Cardiovascular death was defined as death induced by myocardial infarction, heart failure, cardiac arrhythmia, sudden cardiac death, aortic disease, or stroke.
The primary outcome was the ability of FFR CT to predict MACE within 2 years after enrollment based on a previous report. 11 The secondary outcomes were the ability of CAS alone and in combination with FFR CT and cardiovascular risk factors to predict MACE within 2 years.

| Statistical analysis
Continuous data were presented as means ± standard deviation or medians with interquartile ranges, and categorical data were presented as numbers with percentages. Continuous variables between two groups were compared using Student's t or the Mann−Whitney U test, and categorical variables between two groups were compared using the for all). Additionally, lower FFR CT values were associated with higher MACE incidence within 2 years (p for trend, <.01; Figure 1).
Moreover, an FFR CT value of ≤0.71 was only associated with male sex and lower HDL cholesterol levels among the cardiovascular risk factors included in the Framingham risk score (Supporting Information: Table 3).

| Risk stratification for MACE in the total and CAS cohorts
In the total cohort, the Kaplan−Meier analysis indicated that the incidence of MACE was significantly higher in patients with CAS than in those without CAS (6.11 vs. diabetes mellitus, and low HDL cholesterol (≤50 mg/dL) were independently associated with MACE within 2 years (Model 2, Table 2). In the Kaplan−Meier analysis, the incidence of MACE was significantly higher in patients with FFR CT ≤0.71 than in those with FFR CT >0.71 ( Figure 2B). The incidence rate of MACE was higher in patients with FFR CT ≤0.71 than in those with FFR CT >0.71 (9.78 vs. 2.68 per 100 patient-year), and the HR was 3.31 (95% CI: 1.32−8.32) after adjustment for age and sex (Table 3). In the combined model including the three independent factors in Model 2, patients harboring all three factors had the highest incident rate of MACE (23.41 per 100 patient-year), and the HR was 6.01 (95% CI: 2.77−13.03) compared with the reference group of patients harboring 0−2 of the three factors, after adjustment for age and sex ( Figure 2 and Table 3). In contrast, patients harboring none of the three factors did not develop MACE during the 2 years after enrollment for suspected CAD.

| DISCUSSION
In the present study, the prognostic utility of FFR CT was examined in 933 patients with suspected CAD and CCTA. Our analyses revealed that CAS (stenosis severity with ≥50%) based on CCTA was a predictive factor for the development of MACE, in agreement with previous reports. 12 We also found that an FFR CT value of ≤0.71 was also a predicting factor for MACE within 2 years in patients with CAS.
Furthermore, the impact was enhanced following combination with other risk factors including history of diabetes mellitus and lower HDL cholesterol level.

| CCTA in patients suspected CAD
CCTA is recommended as a noninvasive approach for patients with symptomatic chest pain and intermediate CAD risk. Several randomized trials have demonstrated that CCTA has similar or better diagnostic ability and prognostic outcomes compared with standardof-care noninvasive testing. 13 However, there are limitations in specificity and physician agreement with CCTA in patients with moderate-severe coronary atherosclerosis with >50% stenosis. 14 For example, CAC significantly reduces the diagnostic specificity and overall accuracy of CCTA. 15 Thus, further stratification and refinement in diagnostic and prognostic assessments are necessary, and invasive coronary angiography and invasive FFR are often performed.
As a noninvasive test following CCTA, FFR CT assesses functional severity by utilizing computational fluid dynamics to calculate coronary blood flow, and exhibits good correlation with invasive FFR. 16 The utility of FFR CT has been demonstrated in several studies of patients with suspected CAD and CAS, and FFR CT ≤0.80 was used as a predictive indicator with functional significance. 16