Abstract
Background
Risk stratification of patients with symptomatic nonobstructive coronary artery disease remains uncertain. Our study assessed the clinical value of single‐vessel, multivessel, and 3‐vessel computational angiography–derived fractional flow reserve (caFFR) measurement in patients with nonobstructive coronary artery disease.
Methods and Results
We enrolled patients with ≤50% stenosis with a caFFR value ≥0.8 in all 3 coronary arteries on coronary angiography. The sum of caFFR values in the 3 vessels was computed for each patient. Patient stratification was based on the median value of the following criteria: single‐vessel analysis, multivessel analysis, and 3‐vessel analysis. The primary end point of this study was major adverse cardiac events at 5 years, defined as a composite of cardiac death, myocardial infarction, and ischemia‐driven revascularization. A total of 490 patients were included. The 5‐year major adverse cardiac event rates in single‐vessel analysis were statistically insignificant between low‐ and high‐caFFR groups (left anterior descending artery [P=0.163]; left circumflex artery [P=0.797]; right coronary artery [P=0.127]). In multivessel analysis, patients in the multiple‐vessel low‐caFFR group (with 2–3 vessels lower than median value of all coronary arteries) showed an increased risk of 5‐year major adverse cardiac events compared with patients in the single‐vessel low‐caFFR group (0–1 vessel) (hazard ratio [HR], 2.648 [95% CI, 1.141–6.145]; P=0.023). In 3‐vessel analysis, patients in the low 3‐vessel caFFR group demonstrated a greater 5‐year major adverse cardiac event risk than the high 3‐vessel caFFR group (HR, 2.43 [95% CI, 1.087–5.433]; P=0.031).
Conclusions
We demonstrated that both multiple‐vessel and 3‐vessel caFFR measurements serve as valuable prognostic indicators for risk assessment in patients with nonobstructive coronary artery disease.
Keywords: computational angiography–derived fractional flow reserve, major adverse cardiac events, nonobstructive coronary artery disease
Subject Categories: Coronary Artery Disease, Percutaneous Coronary Intervention, Imaging, Angiography, Coronary Circulation
Nonstandard Abbreviations and Acronyms
- 3V‐caFFR
3‐vessel computational angiography–derived fractional flow reserve
- caFFR
computational angiography–derived fractional flow reserveFAME‐2 FFR
Fractional Flow Reserve Versus Angiography for Multivessel Evaluation 2
- FFR
fractional flow reserve
- FFR‐FRIENDS
Fractional Flow Reserve–Guided Revascularization in All Comers Versus FRIED in Non‐STEMI Patients
- MACE
major adverse cardiovascular events
- NOCAD
nonobstructive coronary artery disease
- PROMISE
Prospective Multicenter Imaging Study for Evaluation of Chest Pain
Clinical Perspective.
What Is New?
Introduction of the 3‐vessel caFFR cutoff value (2.78) for risk stratification in patients with nonobstructive coronary artery disease.
Computational angiography–derived fractional flow reserve (caFFR) is an innovative, faster, and less invasive imaging technique that evaluates multiple coronary arteries, providing a comprehensive functional assessment of the heart by measuring overall physiological atherosclerotic plaque and total atherosclerotic burden.
What Are the Clinical Implications?
The use of multivessel caFFR and 3‐vessel caFFR for risk stratification in patients with nonobstructive coronary artery disease has shown prognostic value, as lower caFFR values in multiple vessels and a reduced 3‐vessel caFFR are associated with higher major adverse cardiac event risk, allowing clinicians to better identify high‐risk patients and implement aggressive pharmacotherapy and potential early revascularization.
Nonobstructive coronary artery disease (NOCAD) has emerged as a significant health concern, leading to worsening cardiovascular outcomes and imposing substantial financial burdens on individuals. 1 , 2 , 3 , 4 , 5 Studies have shown that >40% of patients with angina undergoing invasive coronary angiography are diagnosed with NOCAD. 6 , 7 , 8 , 9 While prior research characterized individuals with NOCAD as having a relatively benign prognosis, recent studies have unveiled a disconcerting trend, indicating a high incidence of major adverse cardiovascular events (MACEs) within this patient cohort. 4 , 8 , 9 , 10 , 11 , 12 In addition, research based on angiography and coronary computed tomography angiography (CCTA) demonstrated that the severity of nonobstructive disease carries significant prognostic implications. 4 , 13 , 14 The incidence of MACEs increases in proportion to the extent of NOCAD, and individuals with NOCAD affecting all 3 coronary arteries suffer from an equivalent risk and symptom burden compared with patients with obstructive coronary artery disease (CAD). 3 , 4 , 8 Despite the established importance of atherosclerotic severity in NOCAD, there is currently a gap in risk assessment and guideline‐based interventions to better manage the functional ischemic burden associated with the condition. 15 , 16
In a study that evaluates patients with CAD including those who had significant obstruction that requires intervention, 3‐vessel fractional flow reserve (FFR) is indicated as a prognostic marker, providing valuable insights into the total physiological atherosclerotic burden. 17 Notably, wire‐based FFR presents various drawbacks, such as the requirement of administering adenosine, which could potentially induce myocardial steal in cases of severe stenosis and in a myocardium dependent on collateral circulation. 18 The use of angiography‐derived FFR overcomes these limitations, and several models have demonstrated favorable correlations with traditional FFR. 19 , 20 , 21 The advent of computational angiography–derived fractional flow reserve (caFFR) combining invasive aortic pressure with computational pressure‐flow dynamics during angiography provides excellent diagnostic accuracy of 95.7% compared with wire‐based FFR. 22 , 23 The objective of this study is to evaluate the implications of single‐vessel, multivessel, and 3‐vessel caFFR measurements on the occurrence of MACEs (a composite of cardiac death, myocardial infarction [MI], and ischemia‐driven revascularization), as well as on all‐cause death in patients diagnosed with NOCAD.
Methods
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Study Design and Patient Population
Patients aged ≥18 years who have symptoms indicative of heart disease and underwent coronary angiography at Queen Mary Hospital in Hong Kong between January 1, 2014, and December 31, 2017, were retrospectively included and followed up for 5 years for caFFR analysis. Patients in this study underwent coronary angiography on the basis of various clinical indications, including angina, dyspnea, abnormalities detected in ECGs, treadmill stress test results, or as per their physician's recommendation. The definition of NOCAD has been described previously. 14 Patients excluded from the study were those with any of 3 major coronary arteries’ caFFR value <0.8, with >50% stenosis by visual stimulation in major epicardial coronary arteries, acute coronary syndrome within 72 hours (including both ST‐segment–elevation MI and non–ST‐segment–elevation MI), acute decompensated heart failure with ejection fraction <50%, congenital heart disease, severe valvular heart disease, and underwent staged percutaneous coronary intervention (PCI) and coronary artery bypass grafting. Patients' angiography data that fulfilled both the including and excluding criteria were retrieved for caFFR analysis. Baseline characteristics and prescribed medication were obtained through the Clinical Data Analysis and Reporting System, 24 which is an electronic health record database managed by the Hong Kong Hospital Authority. Diagnoses were determined using the International Classification of Diseases, Ninth Revision (ICD‐9), and previous studies have shown high coding accuracy. This study was approved by the ethics committee of the West Cluster Hospital Authority of Hong Kong (UW—19‐575).
Quantitative Coronary Angiography and Measurement of caFFR
Coronary angiography was performed using standard techniques, and angiographic views were obtained after intracoronary administration of nitrate (100 or 200 μg).
The technical requirements for conducting caFFR analysis are as follows 25 : ensuring contrast opacification of the entire vessel by manually applying standard force without any table movement during the contrast injection, and acquiring ≥2 coronary angiograms with projections separated by at least 30 degrees. The coronary angiograms were recorded at a standard frame rate of 15 frames per second. Moreover, a simulated 3‐dimensional mesh reconstruction of the specific coronary artery of interest was generated with the mean aortic pressure, which was performed using computational pressure‐flow dynamics methods. In our study, only the main branch of all 3 major coronary arteries is used to generate caFFR value in the patients in line with the included criteria, using the FLASH software (Rainmed Ltd, Suzhou, China). The 3‐dimensional mesh reconstruction encompassed the vessel from the inlet until a point at least 10 mm downstream of the most distal stenosis. If the distal segment of the vessel contained >2 daughter branches (eg, the posterolateral branch and posterior descending artery in the right coronary artery [RCA]), the branch with a larger diameter was considered part of the main vessel. The mean aortic pressure data required for caFFR analysis were computed on the basis of arterial tracing records obtained during coronary angiography sessions. An example of caFFR analysis is illustrated in Figure S1. All caFFR analyses were conducted by an independent investigator (Y.F.) at Rainmed Ltd, Suzhou, China, who was blinded to patient data and outcomes. 25
Cohort and Grouping
Although there are limited data available on the ideal cutoff value of caFFR for predicting adverse events, caFFR has demonstrated a strong correlation with FFR. 22 caFFR was used as a substitute for FFR in our study; we hence established a threshold of 0.80 as the cutoff value for patients with NOCAD in caFFR measurements.
To assess the physiological burden of atherosclerosis in patients with NOCAD, different groupings were employed for the study. As cohort 1 for 3‐vessel caFFR (3V‐caFFR) analysis, patients were categorized on the basis of the median value (2.78) of the total 3V‐caFFR to assess the overall physiological burden of atherosclerosis (Figure 1). As cohort 2 for multivessel caFFR analysis, patients are classified according to the severity of NOCAD on the basis of the median value of all coronary arteries (0.93) into 2 groups as follows: single–low‐caFFR group: patients with 0 to 1 vessel <0.93; multi–low‐caFFR group: patients with 2 to 3 vessels <0.93 (Figure 1). As cohort 3 for single‐vessel caFFR analysis, patients were divided into groups on the basis of the median values of lesions in the anterior descending (LAD) artery (0.92), left circumflex (LCx) artery (0.94), and RCA (0.94) (Figure 1). Patients exhibiting diminutive LCx or RCA were excluded from the single‐vessel analysis involving the LCx and RCA. During the analysis of a total of 1470 vessels, 48 of 490 (9.80%) RCA and 43 of 490 (8.78%) LCx were excluded from the single‐vessel analysis. Through these groupings, we were able to examine the clinical implications of single‐vessel assessment, multivessel analysis, and 3‐vessel analysis in patients with NOCAD.
Figure 1. Study design.
3V‐caFFR indicates 3‐vessel computational angiography‐derived fractional flow reserve; CABG, coronary artery bypass grafting; caFFR, computational angiography‐derived fractional flow reserve; CKD, chronic kidney disease; LAD, left anterior descending artery; LCX, left coronary artery; NSTEMI, non–ST‐segment–elevation myocardial infarction; PCI, percutaneous coronary intervention; RCA, right coronary artery; and STEMI, ST‐segment–elevation myocardial infarction.
Patient Follow‐Up and End Point Measurement
The primary end point of the study was to assess MACEs at 5 years, cardiac death, any instance of MI, and ischemia‐driven revascularization (Table S1). Secondary end points included all‐cause death (Table S1). Unless a clear cardiac cause was evident, all deaths were considered cardiac nonrelated. The definition of MI was based on the fourth universal definition of MI, any significant increase in cardiac troponin, symptoms, new electrocardiographic changes, and imaging findings of myocardial viability loss. 26 Ischemia‐driven revascularization was defined as the occurrence of either PCI or coronary artery bypass grafting performed on the same vessel following the initial admission. All clinical outcomes were determined on the basis of comprehensive clinical data obtained from interhospital centralized electronic medical records. These records contain detailed information such as clinical notes, procedural records, and results from radiological and laboratory investigations. These clinical data were used to evaluate the end points of the study, and the follow‐up rate was complete in our cohort of patients.
Statistical Analysis
The primary hypothesis of the current study was that patients with low 3V‐caFFR would show an increase in frequency of 5‐year MACEs compared with those with high 3V‐caFFR. Categorical variables were expressed as proportions and compared using the χ2 test or Fisher exact test. Continuous variables were expressed as mean±SD or median with interquartile range (IQR; quartiles 1–3) based on their distribution, checked by the Kolmogorov–Smirnov test. Differences between groups were compared using the t test for continuous variables and the χ2 test for categorical variables. Differences between groups for nonnormally distributed data were further compared using the Mann–Whitney U test.
To compare differences between groups at risk of MACEs and all‐cause death, Cox proportional hazard regression was used to calculate the hazard ratio (HR) along with a 95% CI. 27 The parameters that were considered clinically relevant or that showed a univariate relationship with outcome (P<0.05) were entered into multivariate Cox models (Table S2). Survival curves were plotted using Kaplan–Meier estimates, and differences between groups were tested using the log‐rank test. 28 To select the best cutoff value of 3V‐caFFR, a method using maximally selected log‐rank statistics was used as previously described. 29
All probabilities were 2‐sided, and statistical significance was determined by a P value <0.05. All statistical analyses were performed with R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria) and SPSS version 28.0 (SPSS Inc., Chicago, IL) statistical packages.
Patients and Public Involvement
Patients and the public were not involved in the design, conduct, outcomes, or dissemination of this study. All procedures were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000. This study was approved by the ethics committee of the West Cluster Hospital Authority of Hong Kong. Considering the retrospective study design, the need to obtain informed consent was waived.
Results
Patient Characteristics
We identified 490 NOCAD patients between 2014 and 2017, the mean age was 64.39 years, 272 (55.5%) were men, nearly one‐third had diabetes (n=160 [32.7%]), and more than one‐third had hypertension (n=190 [38.8%]). Patients were included in 3 models for analysis, namely 3V‐caFFR analysis, multivessel analysis, and single‐vessel analysis. Baseline characteristics of the entire cohort population are summarized in Table 1.
Table 1.
Baseline Characteristics
| Total | 3V‐caFFR analysis (cohort 1) | Multivessel analysis (cohort 2) | |||
|---|---|---|---|---|---|
| Low caFFR | High caFFR | Single‐vessel low‐caFFR | Multivessel low‐caFFR | ||
| (≤2.78) | (>2.78) | 0 to 1 vessel (≤0.93) | 2 to 3 vessel (≤0.93) | ||
| No. of patients | 490 | 268 (56.7) | 222 (45.3) | 211 (43.1) | 279 (56.9) |
| General characteristics | |||||
| Age, median (quartiles 1–3) | 66 (57–72) | 65 (57–72) | 66 (55–72) | 65 (57–72) | 65 (57–72) |
| Male sex, n (%) | 272 (55.5) | 148 (55.2) | 124 (55.9) | 114 (54.0) | 158 (56.6) |
| Smoking, n (%) | 98 (20.0) | 56 (20.5) | 43 (19.4) | 41 (19.4) | 57 (20.4) |
| Cardiovascular risk factors, n (%) | |||||
| History of MI | 61 (12.4) | 37 (13.8) | 24 (10.8) | 19 (9.0) | 42 (15.1) |
| History of PC | 64 (13.1) | 36 (13.4) | 28 (12.6) | 22 (10.4) | 42 (15.1) |
| Peripheral vascular disease | 13 (2.7) | 6 (2.2) | 7 (3.2) | 5 (2.4) | 8 (2.9) |
| Stroke (transient ischemic attack) | 54 (11.0) | 26 (9.7) | 28 (12.6) | 25 (11.8) | 29 (10.4) |
| Diabetes | 160 (32.7) | 91 (34.0) | 69 (31.1) | 66 (31.3) | 94 (33.7) |
| Hypertension | 190 (38.8) | 102 (38.1) | 88 (39.6) | 86 (40.8) | 104 (37.3) |
| Elevated lipid level | 169 (34.5) | 95 (35.4) | 74 (33.3) | 66 (31.3) | 103 (36.9) |
| Family history of CAD | 8 (1.6) | 7 (2.6) | 1 (0.5) | 3 (1.4) | 5 (1.8) |
| Atrial fibrillation | 34 (6.9) | 17 (6.3) | 17 (7.7) | 15 (7.1) | 19 (6.8) |
| Heart failure | 13 (2.7) | 7 (2.6) | 6 (2.7) | 3 (1.4) | 10 (3.6) |
| Chronic coronary syndrome classification, n (%) | |||||
| Class I | 353 (72.0) | 198 (73.9) | 155 (69.8) | 150 (71.1) | 203 (72.8) |
| Class II | 130 (26.5) | 65 (24.3) | 65 (29.3) | 59 (28.0) | 71 (25.4) |
| Class III | 7 (1.4) | 5 (1.9) | 2 (0.9) | 2 (0.9) | 5 (1.8) |
| Estimated glomerular filtration rate, median (quartiles 1–3) | 87.26 (74.44–97.58) | 89.16 (73.78–98.42) | 85.12 (75.18–95.92) | 87.06 (75.23–98.03) | 87.47 (74.26–97.30) |
| Baseline medications, n (%) | |||||
| β‐blocker | 214 (43.7) | 124 (46.3) | 90 (40.5) | 86 (40.8) | 128 (45.9) |
| Calcium channel blocker | 159 (32.4) | 87 (32.5) | 72 (32.4) | 67 (31.8) | 92 (33.0) |
| Angiotensin‐converting enzyme inhibitors | 103 (21.0) | 56 (20.9) | 47 (21.2) | 41 (19.4) | 62 (22.2) |
| Angiotensin receptor blockers | 81 (16.5) | 46 (17.2) | 35 (15.8) | 34 (16.1) | 47 (16.8) |
| Diuretics | 46 (9.4) | 25 (9.3) | 21 (9.5) | 18 (8.5) | 28 (10.0) |
| Digoxin | 5 (1.0) | 1 (0.4) | 4 (1.8) | 4 (1.9) | 1 (0.4) |
| Nitrates | 122 (24.9) | 58 (21.6) | 64 (28.8) | 59 (28.0) | 63 (22.6) |
| Statin* | 361 (73.7) | 207 (77.2) | 154 (69.4) | 145 (68.7) | 216 (77.4) |
| Insulin injection | 17 (3.5) | 8 (3.0) | 9 (4.1) | 7 (3.3) | 10 (3.6) |
| Antidiabetic drugs | 93 (19.0) | 54 (20.1) | 39 (17.6) | 40 (19.0) | 53 (19.0) |
| Aspirin† | 328 (66.9) | 192 (71.6) | 136 (61.3) | 118 (55.9) | 210 (75.3) |
| Clopidogrel/ticagrelor/prasugrel, n (%) | 58 (11.8) | 32 (11.9) | 26 (11.7) | 26 (12.3) | 32 (11.5) |
| Oral anticoagulant, n (%) | 26 (5.3) | 10 (3.7) | 16 (7.2) | 15 (7.1) | 11 (3.9) |
| Index of CAD burden | |||||
| 3V‐caFFR median (quartiles 1–3)‡ | 2.78 (2.72–2.82) | 2.73 (2.68–2.76) | 2.83 (2.81–2.86) | 2.79 (2.75–2.83) | 2.78 (2.71–2.79) |
3V‐caFFR indicates 3‐vessel computational angiography‐derived fractional flow reserve; CAD, coronary artery disease; caFFR, computational angiography‐derived fractional flow reserve; MI, myocardial infarction; and PCI, percutaneous coronary intervention.
P‐value of cohort 2 <0.05.
P‐value of cohort 1 <0.05, and cohort 2 <0.001.
P‐value of both cohort 1 and cohort 2 <0.001.
Baseline Characteristics in 3V‐caFFR Analysis: Cohort 1
The median value of 3V‐caFFR was 2.78 (IQR, 2.72–2.82; Figure 2). The best cutoff value of 3V‐caFFR to predict 5‐year MACEs was 2.78 on the basis of the maximum log‐rank statistics. In accordance with the median and best cutoff value, 268 (54.7%) patients were classified into the low–3V‐caFFR group, and 222 (45.3%) were classified into the high–3V‐caFFR group. While age and sex were similarly distributed across groups, the low–3V‐caFFR group had a higher proportion of patients using aspirin and lower per‐vessel mean caFFR values (Table 1).
Figure 2. Per‐vessel and 3‐vessel caFFR.
3V‐caFFR indicates 3‐vessel computational angiography‐derived fractional flow reserve; caFFR, computational angiography‐derived fractional flow reserve; and IQR, interquartile range.
Clinical Outcome of Cohort 1
Table 2 illustrates the difference in 5‐year MACE rate between the low–3V‐caFFR group and the high–3V‐caFFR group. Patients in the low–3V‐caFFR group indicated a higher MACE rate than those in the high–3V‐caFFR group (9.49% versus 3.6%, HR, 2.551 [95% CI, 1.146–5.679]; P=0.022). This increase in 5‐year MACE rate in the low–3V‐caFFR group compared with the high–3V‐caFFR group is attributed to a higher rate of ischemia‐driven revascularization (5.14% versus 1.35%; HR, 3.638 [95% CI, 1.037–12.77]; P=0.044). After multivariate adjustment, the low–3V‐caFFR group was depicted as an independent predictor of 3‐year MACEs (Table 3, Table S3). Figure 3 presents the cumulative incident curve of rate of 5‐year MACEs between low– and high–3V‐caFFR groups. Patients in the low–3V‐caFFR group show a higher 5‐year MACE rate than high–3V‐caFFR groups (P=0.017). Table 2 indicates the difference in 5‐year all‐cause mortality rate between the low–3V‐caFFR group and the high–3V‐caFFR group. The low–3V‐caFFR group exhibited an elevated risk of all‐cause death compared with the high–3V‐caFFR group (7.09% versus 3.15%; HR, 2.807 [95% CI, 1.127–6.99]; P=0.027). Various multivariate Cox regression models showed that low–3V‐caFFR was an independent predictor of 5‐year rate of all‐cause death.
Table 2.
Univariate Cox Regression for 3V‐caFFR and Multivessel caFFR
| 5‐year clinical outcome | Groups | Events rate, n (%) | HR (95% CI) | P value |
|---|---|---|---|---|
| MACE | ||||
| 3V‐caFFR analysis | Low 3V‐caFFR (n=268) | 24 (9.49) | 2.55 (1.15–5.68) | 0.02 |
| High 3V‐caFFR (n=222) | 8 (3.60) | |||
| Multivessel analysis | Single‐vessel low caFFR (n=211) | 7 (3.32) | 2.77 (1.20–6.40) | 0.02 |
| Multivessel low caFFR (n=279) | 25 (8.96) | |||
| Cardiovascular death | ||||
| 3V‐caFFR analysis | Low 3V‐caFFR (n=268) | 6 (2.24) | 5.02 (0.60–41.70) | 0.1 |
| High 3V‐caFFR (n=222) | 1 (0.45) | |||
| Multivessel analysis | Single vessel low caFFR (n=211) | 2 (0.95) | 1.90 (0.37–9.80) | 0.4 |
| Multi‐vessel low caFFR (n=279) | 5 (1.79) | |||
| Myocardial infarction | ||||
| 3V‐caFFR analysis | Low 3V‐caFFR (n=268) | 12 (4.48) | 1.67 (0.63–4.45) | 0.3 |
| High 3V‐caFFR (n=222) | 6 (2.70) | |||
| Multivessel analysis | Single‐vessel low caFFR (n=211) | 5 (2.37) | 1.98 (0.71–5.56) | 0.2 |
| Multivessel low caFFR (n=279) | 13 (4.66) | |||
| Ischemia‐driven revascularization | ||||
| 3V‐caFFR analysis | Low 3V‐caFFR (n=268) | 13 (5.14) | 3.64 (1.04–12.77) | 0.04 |
| High 3V‐caFFR (n=222) | 3 (1.35) | |||
| Multivessel analysis | Single‐vessel low caFFR (n=211) | 3 (1.42) | 3.31 (0.94–11.61) | 0.06 |
| Multivessel low caFFR (n=279) | 13 (4.66) | |||
| All‐cause death | ||||
| 3V‐caFFR analysis | Low 3V‐caFFR (n=268) | 19 (7.09) | 2.81 (1.13–6.99) | 0.03 |
| High 3V‐caFFR (n=222) | 7 (3.15) | |||
| Multivessel analysis | Single‐vessel low caFFR (n=211) | 7 (3.32) | 2.07 (0.87–4.92) | 0.1 |
| Multivessel low caFFR (n=279) | 19 (6.81) | |||
3V‐caFFR indicates 3‐vessel computational angiography‐derived fractional flow reserve; caFFR, computational angiography‐derived fractional flow reserve; HR, hazard ratios; and MACE, major adverse cardiovascular event.
Table 3.
Multivariate Cox Regression for 3V‐caFFR and Multivessel caFFR
| 5‐Year clinical outcome | Comparator | HR (95% CI) | P value |
|---|---|---|---|
| MACE | |||
| 3V‐caFFR analysis* | Low vs high 3V‐caFFR | 2.49 (1.12–5.55) | 0.03 |
| Multivessel analysis* | 0–1 vs 2–3 low‐caFFR vessels | 2.55 (1.10–5.91) | 0.03 |
| Ischemia‐driven revascularization | |||
| 3V‐caFFR analysis* | Low vs high 3V‐caFFR | 3.56 (1.01–12.49) | 0.04 |
| Multivessel analysis* | 0–1 vs 2–3 low‐caFFR vessels | 3.20 (0.91–11.26) | 0.07 |
| All‐cause death | |||
| 3V‐caFFR analysis* | Low vs high 3V‐caFFR | 2.86 (1.15–7.14) | 0.02 |
3V‐caFFR indicates 3‐vessel computational angiography‐derived fractional flow reserve; caFFR, computational angiography‐derived fractional flow reserve; and MACE, major adverse cardiovascular event.
Adjusted for age, sex, heart failure, and history of myocardial infarction.
Figure 3. Cumulative incidence curves.
3V‐caFFR indicates 3‐vessel computational angiography‐derived fractional flow reserve; caFFR, computational angiography‐derived fractional flow reserve; and MACEs, major adverse cardiovascular events.
Baseline Characteristics in Multi‐Vessels Group: (Cohort 2)
The median value for all coronary arteries caFFR, including LAD, LCx, and RCA, was 0.93 (IQR: 0.9–0.95) (Figure 2). Among the study population, 211 (43.1%) patients were categorized as the single‐vessel low‐caFFR group (with 0–1 vessels ≤0.93) and 279 (56.9) as the multi‐vessel low‐caFFR group (with 2–3 vessels ≤0.93). Table 1 presents the baseline characteristics of these 2 groups, indicating that the multivessel low‐caFFR group had a greater percentage of patients using aspirin and statin medications, as well as lower per‐vessel mean caFFR values. Although age and sex were similarly distributed across groups, a higher percentage of patients in the multi‐vessel low‐caFFR group (2–3 vessels) demonstrated an elevated risk profile, encompassing conditions such as MI, peripheral vascular disease, and diabetes. Consequently, this group had a greater proportion of patients on aspirin and statin therapy, along with lower mean caFFR values per vessel at baseline.
Clinical Outcome of Cohort 2
Figure 3 presents the comparison of 5‐year MACE rate between the single low caFFR group (patients with 0–1 vessels less than the median value of all vessels) and the multi–low‐caFFR group (patients with 2–3 vessels less than the median value of all vessels). The 5‐year MACE rate was higher in the multi–low‐caFFR group compared with the single–low‐caFFR group (8.96% versus 3.32%; HR, 2.766 [95% CI, 1.196–6.395]; P=0.017). This is primarily driven by the higher rate of ischemia‐driven revascularization in the multi–low‐caFFR group (4.66% versus 1.42%; HR, 3.307 [95% CI, 0.943–11.61]; P=0.062; Table 2). Multiple Cox regression models that were adjusted for various variables revealed that the multi–low‐caFFR group emerged as a significant independent predictor of 5‐year MACEs (Table 3, Table S3). Regarding the secondary end point, the association between multi–low‐caFFR group and the rate of 5‐year all‐cause death is insignificant, with the HR showing there is an increasing risk of all‐cause death with the multi–low‐caFFR group (HR, 2.069 [95% CI, 0.870–4.922]; P=0.100).
Baseline Characteristics in Single‐Vessel Analysis
Baseline characteristics of the 3 groups, stratified according to individual coronary arteries, are outlined in Tables S4 through S6. The median caFFR values of LAD were 0.92 (IQR, 0.89–0.94), LCx was 0.94 (IQR, 0.91–0.96), and RCA was 0.94 (IQR, 0.92–0.95). In the caFFR analysis of the LAD, patients in the low‐LAD caFFR group were found to be older, with elevated lipid levels, a higher prevalence of diabetes, and a higher use of statins. Conversely, the differences in baseline characteristics between the low‐ and high‐caFFR groups in the LCx were not statistically significant. In the analysis of the RCA caFFR, patients in the low‐RCA caFFR group exhibited a higher prevalence of previous PCI, along with increased use of aspirin and oral anticoagulants compared with the high‐RCA caFFR group. Notably, a higher proportion of patients in the high‐RCA caFFR group reported the use of nitrates.
Clinical Outcome of Cohort 3
Table S7 demonstrates the comparison of 5‐year MACEs and all‐cause mortality rates between high‐ and low‐caFFR groups for LAD, LCx, and RCA, respectively. The 5‐year MACE rates between the low‐ and high‐caFFR groups were found to be statistically insignificant (LAD, P=0.114; LCx, P=0.66; RCA, P=0.126). The comparison of 5‐year all‐cause mortality rates between the high‐ and low‐caFFR groups revealed statistically insignificant differences in the LAD and LCx (LAD, P=0.053; LCx, P=0.706). However, the low‐caFFR group indicated a higher risk of death than the high‐caFFR group in the RCA (HR, 3.485 [95% CI, 1.186–10.24]; P=0.023).
Discussion
Our study in patients with NOCAD unveiled the comprehensive use of caFFR as a useful prognostic indicator for predicting adverse outcomes. The key findings indicated that patients with low–3V‐caFFR, when categorized on the basis of the median value, faced a significantly elevated risk of MACEs compared with their counterparts with high–3V‐caFFR. Additionally, patients with multiple vessels exhibiting low caFFR demonstrated a higher rate of MACEs when compared with those with a single vessel displaying low caFFR. These results were consistently observed across various multivariate Cox regression models. Nonetheless, the prognostic value of single‐vessel analysis in patients with NOCAD remains neutral.
The prevalence of NOCAD, a condition that is often overlooked, escalated from 19% to 41% between 2000 and 2009 in the symptomatic population, with patients with diffuse NOCAD exhibiting a comparable susceptibility to MACEs as their counterparts with obstructive CAD. 5 Similarly, a comprehensive meta‐analysis involving 41 960 symptomatic patients who underwent coronary angiography revealed that 33.8% of the population were diagnosed with NOCAD, which carries a 6‐fold increased risk of MI or cardiac death compared with individuals without CAD. 30 These findings aligned with prior research demonstrating a higher 1‐year risk of MI and all‐cause death in patients with NOCAD, particularly as the extent of CAD involvement increases. 4 Concomitantly, the rate of MACE and all‐cause death in our NOCAD population is 6.3% and 5.3% in 5 years of follow‐up, respectively. The prevalence and challenges of NOCAD highlight the urgent need for improved risk stratification and clinical awareness to prevent undiagnosed patients from experiencing adverse events. 31
Recent studies have unveiled a notable disparity in the overall burden of atherosclerosis between patients with NOCAD and those who exhibit normal myocardial perfusion (indicated by a single‐photon emission computed tomography summed stress score of <1). 32 This observation has led to the hypothesis proposing that the presence of early CAD is associated with the development of substantial atherosclerosis, which induces endothelial dysfunction and increases the deposition of lipid content within the vascular wall. 33 The recent 3‐vessel FFR‐FRIENDS (Fractional Flow Reserve–Guided Revascularization in All Comers Versus FRIED in Non‐STEMI Patients) trial was performed to investigate the utility of FFR imaging in patients diagnosed with CAD (>30% stenosis) over a 2‐year follow‐up period and discovered that the total sum of FFR values obtained from the 3 main coronary arteries, including those who required PCI, could serve as an indicator of the overall physiological burden of atherosclerosis. The trial demonstrated a negative correlation between 3V‐FFR and the total physiological atherosclerotic burden, in which patients with low–3V‐FFR showed a higher risk of 2‐year clinical events. 17 However, this study focused solely on using this functional and anatomic imaging technique in patients with obstructive CAD, without specifically examining its application in patients with NOCAD. Our study exclusively enrolled patients with NOCAD with coronary artery stenosis of ≤50% and a caFFR value of ≥0.8 in each coronary artery. 34 Building upon prior research, we established a cutoff value for 3V‐caFFR (2.78) based on both median and best cutoff value, enabling the stratification of patients diagnosed with NOCAD into high– and low–3V‐caFFR groups in cohort 1. Upon stratifying patients on the basis of their 3V‐caFFR values, it was observed that the low–3V‐caFFR group exhibited a nearly 2‐fold increased risk of MACEs compared with the high–3V‐caFFR group. The higher risk of MACEs was due to a heightened likelihood of ischemia‐driven revascularization, which was in line with the 3‐vessel FFR‐FRIENDS trial investigation. 17 Although no significant difference in cardiovascular death was observed between the low– and high–3V‐caFFR groups, all‐cause death was notably higher in the low–3V‐caFFR group. This disparity could be partly explained by the small sample size in our study, and future studies with larger populations should verify this finding.
The unfavorable MACEs and ischemia‐driven revascularization outcome observed in the low–3V‐caFFR group can be attributed to the pathological progression of CAD. Initially, these patients exhibit a lower atherosclerotic burden in individual vessels, leading to similar ischemia‐driven revascularization rates at 3 years in both the low– and high–3V‐caFFR groups. However, with the gradual increase in total atherosclerotic volume, encompassing both the extent and severity of stenosis, the severity of stenosis progresses. 31 This explains the divergence in ischemia‐driven revascularization rates beyond the 3‐year mark, underscoring the importance of 3V‐caFFR in predicting long‐term outcomes in patients with NOCAD. This finding aligns with previous investigations in patients with NOCAD with CCTA, demonstrating that patients with NOCAD with a higher total coronary artery plaque burden experienced poorer outcomes. 35 Therefore, the evaluation of 3V‐caFFR enables the comprehensive assessment of the total physiological impact of atherosclerotic plaque, which correlates well with total atherosclerotic burden, serving as a holistic functional evaluation of the heart to assess the risk in patients with NOCAD.
A substudy of the FAME‐2 FFR 2, 36 demonstrated that patients with moderate CAD (FFR, 0.78–0.86) experienced a significantly elevated rate of MACE over a 2‐year period compared with patients with normal FFR values (FFR, 0.87–1.00). Similarly, a post hoc analysis of the 3‐vessel FFR‐FRIENDS study 37 examined FFR in patients with NOCAD, indicating that individuals with moderate CAD (FFR, 0.81–0.87) affecting multiple vessels exhibited a poorer MACE outcome compared with patients with moderate CAD limited to a single vessel. However, there are constraints associated with the use of FFR, as the clinical application of intravenous adenosine during FFR measurement can be burdensome and associated with side effects and high costs. 38 Furthermore, despite being strongly recommended as a class IA indication for guiding PCI on the basis of the severity of stenosis, the use of wire‐based FFR in clinical practice is restricted, with a significant majority (>80%) of patients with intermediate obstructive coronary lesions not undergoing FFR assessment. 39 Consequently, caFFR has garnered significant interest. Not only does it circumvent technical inadequacies linked to wires, but more importantly, it can eliminate the requirement for wire manipulation and hyperemia stimulation while demonstrating comparable diagnostic accuracy to conventional FFR. 40 , 41 , 42 In our study, we have implemented computational pressure‐flow dynamics–derived caFFR, a method that incorporates invasive aortic pressure alongside computational flow modeling. In addition to the superior correlation with wire‐based FFR, caFFR offers a clinical advantage of a brief operating time, with the total procedure completed in <5 minutes and computational analysis taking <1 minute. 22 Based on caFFR‐derived moderate CAD (caFFR, 0.80–0.93), we have categorized patients into single vessel (0–1 vessel) and multiple vessels (2–3 vessels) affected by moderate CAD in cohort 2. Notably, we observed that moderate CAD affecting multiple vessels suffered from a higher 5‐year MACE rate than patients with moderate CAD affecting a single vessel. This observation is consistent with previous studies investigating mortality risk in patients with NOCAD, where the highest risk was observed in cases involving NOCAD affecting all 3 epicardial vessels. 12 With prior clinical studies showing comparable rates of adverse events between patients with moderate CAD affecting multiple vessels and those with single‐vessel obstructive disease, 4 , 5 the extent of moderate CAD affecting multiple vessels is indicative of risk prediction.
A recent study conducted on patients with NOCAD used CCTA‐derived FFR in the single‐vessel RCA. The findings suggested that the diagnostic accuracy of computed tomography‐derived fractional flow reserve (CT‐FFR) alone in patients with NOCAD is limited due to inconsistent vessel morphology. 43 Similarly in our study, we categorized patients into low‐ and high‐caFFR groups on the basis of the median value of individual coronary arteries, namely, LAD, LCx, and RCA. Interestingly, we observed no significant association between single‐vessel caFFR and the rate of MACEs during the 5‐year follow‐up period, but a significant association exists between low caFFR in the RCA and all‐cause death (Table S4). However, it is important to note that the wide confidence interval in RCA (1.186–10.24) suggests statistical instability, which could impact the interpretation of this association. The lack of significance in a single vessel through imaging may be attributed to various factors, including the influence of hemodynamics in single‐vessel cases by factors such as energy loss due to turbulence caused by the bifurcation angle into the LAD and LCx, 44 , 45 as well as the variable morphology of the RCA. 43 Thus, the location‐based assessment of single‐vessel caFFR alone does not provide prognostic indications or substantial impact on patient outcomes.
Clinical Implications
Prior studies have focused on employing noninvasive CCTA anatomic imaging in risk stratification of patients with NOCAD. 43 , 46 , 47 , 48 Through coronary CCTA anatomic imaging, the PROMISE (Prospective Multicenter Imaging Study for Evaluation of Chest Pain) minimal risk score identifies patients with a low likelihood of having obstructive CAD, allowing for the deferral of further diagnostic testing and reducing unnecessary procedures. 49 , 50 However, the specificity of CCTA imaging compared with the gold standard of FFR is only 76%. 51 The sole reliance on anatomic CCTA imaging is not enough in symptomatic patients 52 ; the use of both anatomic and functional imaging hence holds greater significance in the risk assessment of patients. In our study, we have demonstrated the prognostic implication of the use of a 3V‐caFFR cutoff value (2.78) in stratifying patients with NOCAD into high– and low–3V‐caFFR groups for risk stratification. This stratification remained evident even after accounting for multiple factors in the multivariate regression models. By using this global functional assessment of the heart, this approach enables the measurement of both the overall physiological atherosclerotic plaque and atherosclerotic burden, aiding the identification of high‐risk patients with NOCAD. Beyond merely 3V‐caFFR analysis, multivessel caFFR provided prognostic value in showing that an increased number of vessels with low caFFR values is associated with poorer outcomes. It is also noteworthy that a novel technique of caFFR is used clinically in patients with symptomatic NOCAD specifically. During coronary angiography, the caFFR measurement stands out for its short and efficient operating time, as well as its less invasive nature when compared with FFR. Additionally, caFFR exhibits a high level of diagnostic accuracy comparable to FFR, while avoiding the side effects associated with FFR. The use of 3V‐caFFR might therefore guide future studies in determining the necessity of more intensive cardiac treatments for high‐risk patients with NOCAD, ranging from a single medication such as a statin to combinations of statins and aspirin, or even early revascularization treatments. Conducting further analysis through randomized controlled trials in high‐risk patients with NOCAD may help in controlling and reducing adverse outcomes such as MACEs and ischemia‐driven revascularizations, as observed in the present study. Further analysis is also warranted to comprehensively investigate the prognostic impact of 3V‐caFFR specifically in patients without stent placement.
Limitations
Our study has certain limitations that should be acknowledged. First, the overall size of the NOCAD population in our study was relatively small, and the observed differences in MACE rates were not primarily due to direct cardiac causes but were mainly driven by ischemia‐driven revascularization. Nonetheless, we observed that patients with low 3V‐caFFR had a higher HR for MI and cardiac death compared with those in the high–3V‐caFFR group. Furthermore, due to the limited size of the current study population, we were not able to subdivide the cohort into 4 groups of low caFFR. Second, the event rates in our study were generally lower compared with previous studies, possibly due to the inclusion of a relatively healthy population. Third, invasive imaging studies and precise measurement of stenosis <50% were not conducted; the interindividual difference in coronary anatomy is hence not considered, preventing the investigation of the direct relationship between anatomic atherosclerotic burden and clinical outcomes. Fourth, an inherent limitation of our study is the absence of invasive imaging modalities such as optical coherence tomography and intravascular ultrasound to validate total atherosclerotic burden. Implementing these techniques across all 3 vessels would have prolonged the intervention duration, increasing procedural risks. Fifth, it should be noted that although all patients meet the minimum left ventricular ejection fraction requirement of ≥50%, the precise measurements necessary for inclusion in the multivariable analysis are not available. In addition, despite the absence of 5‐year follow‐up data on statin use for all patients, 71.9% of patients with MACEs and 76.9% of those with all‐cause death were on statin therapy before angiography. Finally, it should be noted that although caFFR does not require the use of a pressure wire or the induction of hyperemia, it remains an invasive procedure.
Conclusions
In patients with NOCAD, the higher count of vessels exhibiting low caFFR values and the low 3V‐caFFR values have been found to be associated with an increased rate of MACEs over a 5‐year period. The difference was mainly driven by ischemia‐driven revascularization. This highlights the importance of evaluating both the presence of multiple vessels and the overall physiological atherosclerotic burden using 3V‐caFFR as prognostic assessments in patients with NOCAD.
Sources of Funding
None.
Disclosures
None.
Supporting information
Tables S1–S7
Figure S1
This manuscript was sent to Amgad Mentias, MD, Associate Editor, for review by expert referees, editorial decision, and final disposition.
Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.124.036717
For Sources of Funding and Disclosures, see page 11.
References
- 1. Widmer Robert J, Samuels B, Samady H, Price JM, Jeremias A, Anderson RD, Jaffer AF, Escaned J, Davies EJ, Prasad M, et al. The functional assessment of patients with non‐obstructive coronary artery disease: expert review from an international microcirculation working group. EuroIntervention. 2019;14:1694–1702. doi: 10.4244/EIJ-D-18-00982 [DOI] [PubMed] [Google Scholar]
- 2. Shaw LJ, Merz CN, Pepine CJ, Reis SE, Bittner V, Kip KE, Kelsey SF, Olson M, Johnson BD, Mankad S, et al. The economic burden of angina in women with suspected ischemic heart disease: results from the National Institutes of Health–National Heart, Lung, and Blood Institute–sponsored Women's Ischemia Syndrome evaluation. Circulation. 2006;114:894–904. doi: 10.1161/CIRCULATIONAHA.105.609990 [DOI] [PubMed] [Google Scholar]
- 3. Kissel CK, Chen G, Southern DA, Galbraith PD, Anderson TJ. 3210 Long‐term follow‐up of patients with non‐obstructive coronary artery disease presenting with stable angina. Eur Heart J. 2017;38(Suppl 1):3210. doi: 10.1093/eurheartj/ehx504.3210 [DOI] [Google Scholar]
- 4. Maddox TM, Stanislawski MA, Grunwald GK, Bradley SM, Ho PM, Tsai TT, Patel MR, Sandhu A, Valle J, Magid DJ, et al. Nonobstructive coronary artery disease and risk of myocardial infarction. JAMA. 2014;312:1754–1763. doi: 10.1001/jama.2014.14681 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Jespersen L, Hvelplund A, Abildstrøm SZ, Pedersen F, Galatius S, Madsen JK, Jørgensen E, Kelbæk H, Prescott E. Stable angina pectoris with no obstructive coronary artery disease is associated with increased risks of major adverse cardiovascular events. Eur Heart J. 2012;33:734–744. doi: 10.1093/eurheartj/ehr331 [DOI] [PubMed] [Google Scholar]
- 6. Patel MR, Peterson ED, Dai D, Brennan JM, Redberg RF, Anderson HV, Brindis RG, Douglas PS. Low diagnostic yield of elective coronary angiography. N Engl J Med. 2010;362:886–895. doi: 10.1056/NEJMoa0907272 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Merz CN, Kelsey SF, Pepine CJ, Reichek N, Reis SE, Rogers WJ, Sharaf BL, Sopko G. The Women's ischemia syndrome evaluation (WISE) study: protocol design, methodology and feasibility report. J Am Coll Cardiol. 1999;33:1453–1461. doi: 10.1016/S0735-1097(99)00082-0 [DOI] [PubMed] [Google Scholar]
- 8. Ostrom MP, Gopal A, Ahmadi N, Nasir K, Yang E, Kakadiaris I, Flores F, Mao SS, Budoff MJ. Mortality incidence and the severity of coronary atherosclerosis assessed by computed tomography angiography. J Am Coll Cardiol. 2008;52:1335–1343. doi: 10.1016/j.jacc.2008.07.027 [DOI] [PubMed] [Google Scholar]
- 9. Nakazato R, Arsanjani R, Achenbach S, Gransar H, Cheng VY, Dunning A, Lin FY, Al‐Mallah M, Budoff MJ, Callister TQ, et al. Age‐related risk of major adverse cardiac event risk and coronary artery disease extent and severity by coronary CT angiography: results from 15 187 patients from the international multisite CONFIRM study. Eur Heart J Cardiovasc Imaging. 2014;15:586–594. doi: 10.1093/ehjci/jet132 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. De Bruyne B, Pijls NH, Kalesan B, Barbato E, Tonino PA, Piroth Z, Jagic N, Möbius‐Winkler S, Rioufol G, Witt N, et al. Fractional flow reserve‐guided PCI versus medical therapy in stable coronary disease. N Engl J Med. 2012;367:991–1001. doi: 10.1056/NEJMoa1205361 [DOI] [PubMed] [Google Scholar]
- 11. Hoffmann U, Ferencik M, Udelson JE, Picard MH, Truong QA, Patel MR, Huang M, Pencina M, Mark DB, Heitner JF, et al. Prognostic value of noninvasive cardiovascular testing in patients with stable chest pain: insights from the PROMISE trial (prospective multicenter imaging study for evaluation of chest pain). Circulation. 2017;135:2320–2332. doi: 10.1161/CIRCULATIONAHA.116.024360 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Lin FY, Shaw LJ, Dunning AM, Labounty TM, Choi JH, Weinsaft JW, Koduru S, Gomez MJ, Delago AJ, Callister TQ, et al. Mortality risk in symptomatic patients with nonobstructive coronary artery disease: a prospective 2‐center study of 2,583 patients undergoing 64‐detector row coronary computed tomographic angiography. J Am Coll Cardiol. 2011;58:510–519. doi: 10.1016/j.jacc.2010.11.078 [DOI] [PubMed] [Google Scholar]
- 13. Adjedj J, De Bruyne B, Floré V, Di Gioia G, Ferrara A, Pellicano M, Toth GG, Bartunek J, Vanderheyden M, Heyndrickx GR, et al. Significance of intermediate values of fractional flow reserve in patients with coronary artery disease. Circulation. 2016;133:502–508. doi: 10.1161/CIRCULATIONAHA.115.018747 [DOI] [PubMed] [Google Scholar]
- 14. LaMonte MJ, FitzGerald SJ, Church TS, Barlow CE, Radford NB, Levine BD, Pippin JJ, Gibbons LW, Blair SN, Nichaman MZ. Coronary artery calcium score and coronary heart disease events in a large cohort of asymptomatic men and women. Am J Epidemiol. 2005;162:421–429. doi: 10.1093/aje/kwi228 [DOI] [PubMed] [Google Scholar]
- 15. Herscovici R, Sedlak T, Wei J, Pepine CJ, Handberg E, Bairey Merz CN. Ischemia and no obstructive coronary artery disease (INOCA): what is the risk? J Am Heart Assoc. 2018;7:e008868. doi: 10.1161/JAHA.118.008868 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Sarah Payne EF, Michel J, Widmer RJ. Medical management and outcomes of non‐obstructive coronary artery disease. Circulation. 2020;142(Suppl 3). doi: 10.1161/circ.142.suppl_3.16054 [DOI] [Google Scholar]
- 17. Lee JM, Koo BK, Shin ES, Nam CW, Doh JH, Hwang D, Park J, Kim KJ, Zhang J, Hu X, et al. Clinical implications of three‐vessel fractional flow reserve measurement in patients with coronary artery disease. Eur Heart J. 2018;39:945–951. doi: 10.1093/eurheartj/ehx458 [DOI] [PubMed] [Google Scholar]
- 18. Pijls NH, Van Gelder B, Van der Voort P, Peels K, Bracke FA, Bonnier HJ, el Gamal MI. Fractional flow reserve. A useful index to evaluate the influence of an epicardial coronary stenosis on myocardial blood flow. Circulation. 1995;92:3183–3193. doi: 10.1161/01.CIR.92.11.3183 [DOI] [PubMed] [Google Scholar]
- 19. Masdjedi K, van Zandvoort LJC, Balbi MM, Gijsen FJH, Ligthart JMR, Rutten MCM, Lemmert ME, Wilschut JM, Diletti R, de Jaegere P, et al. Validation of a three‐dimensional quantitative coronary angiography‐based software to calculate fractional flow reserve: the FAST study. EuroIntervention. 2020;16:591–599. doi: 10.4244/EIJ-D-19-00466 [DOI] [PubMed] [Google Scholar]
- 20. Fearon WF, Achenbach S, Engstrom T, Assali A, Shlofmitz R, Jeremias A, Fournier S, Kirtane AJ, Kornowski R, Greenberg G, et al. Accuracy of fractional flow reserve derived from coronary angiography. Circulation. 2019;139:477–484. doi: 10.1161/CIRCULATIONAHA.118.037350 [DOI] [PubMed] [Google Scholar]
- 21. Xu B, Tu S, Qiao S, Qu X, Chen Y, Yang J, Guo L, Sun Z, Li Z, Tian F, et al. Diagnostic accuracy of angiography‐based quantitative flow ratio measurements for online assessment of coronary stenosis. J Am Coll Cardiol. 2017;70:3077–3087. doi: 10.1016/j.jacc.2017.10.035 [DOI] [PubMed] [Google Scholar]
- 22. Li J, Gong Y, Wang W, Yang Q, Liu B, Lu Y, Xu Y, Huo Y, Yi T, Liu J, et al. Accuracy of computational pressure‐fluid dynamics applied to coronary angiography to derive fractional flow reserve: FLASH FFR. Cardiovasc Res. 2020;116:1349–1356. doi: 10.1093/cvr/cvz289 [DOI] [PubMed] [Google Scholar]
- 23. Ai H, Zheng N, Li L, Yang G, Li H, Tang G, Zhou Q, Zhang H, Yu X, Xu F, et al. Agreement of angiography‐derived and wire‐based fractional flow reserves in percutaneous coronary intervention. Front Cardiovasc Med. 2021;8:654392. doi: 10.3389/fcvm.2021.654392 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Kwok WC, Tam TCC, Sing CW, Chan EWY, Cheung CL. Validation of diagnostic coding for asthma in an electronic health record system in Hong Kong. J Asthma Allergy. 2023;16:315–321. doi: 10.2147/JAA.S405297 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Leung CK, Lam LY, Li KY, Feng Y, Cao G, Wu M, Wang R, Wu MZ, Ren QW, Yu SY, et al. Clinical value of computational angiography‐derived fractional flow reserve in stable coronary artery disease. J Cardiovasc Transl Res. 2023;16:1166–1176. doi: 10.1007/s12265-023-10381-x [DOI] [PubMed] [Google Scholar]
- 26. Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax JJ, Morrow DA, White HD. Fourth universal definition of myocardial infarction (2018). Circulation. 2018;138:e618–e651. doi: 10.1161/CIR.0000000000000617 [DOI] [PubMed] [Google Scholar]
- 27. Prentice RL, Zhao S. Regression models and multivariate life tables. J Am Stat Assoc. 2021;116:1330–1345. doi: 10.1080/01621459.2020.1713792 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Xie J, Liu C. Adjusted Kaplan‐Meier estimator and log‐rank test with inverse probability of treatment weighting for survival data. Stat Med. 2005;24:3089–3110. doi: 10.1002/sim.2174 [DOI] [PubMed] [Google Scholar]
- 29. Hothorn T, Lausen B. On the exact distribution of maximally selected rank statistics. Comput Stat Data Anal. 2003;43:121–137. doi: 10.1016/S0167-9473(02)00225-6 [DOI] [Google Scholar]
- 30. Habib PJ, Green J, Butterfield RC, Kuntz GM, Murthy R, Kraemer DF, Percy RF, Miller AB, Strom JA. Association of cardiac events with coronary artery disease detected by 64‐slice or greater coronary CT angiography: a systematic review and meta‐analysis. Int J Cardiol. 2013;169:112–120. doi: 10.1016/j.ijcard.2013.08.096 [DOI] [PubMed] [Google Scholar]
- 31. Arbab‐Zadeh A, Fuster V. The risk continuum of atherosclerosis and its implications for defining CHD by coronary angiography. J Am Coll Cardiol. 2016;68:2467–2478. doi: 10.1016/j.jacc.2016.08.069 [DOI] [PubMed] [Google Scholar]
- 32. Schuijf JD, Matheson MB, Ostovaneh MR, Arbab‐Zadeh A, Kofoed KF, Scholte A, Dewey M, Steveson C, Rochitte CE, Yoshioka K, et al. Ischemia and no obstructive stenosis (INOCA) at CT angiography, CT myocardial perfusion, invasive coronary angiography, and SPECT: the CORE320 study. Radiology. 2020;294:61–73. doi: 10.1148/radiol.2019190978 [DOI] [PubMed] [Google Scholar]
- 33. Choi BJ, Prasad A, Gulati R, Best PJ, Lennon RJ, Barsness GW, Lerman LO, Lerman A. Coronary endothelial dysfunction in patients with early coronary artery disease is associated with the increase in intravascular lipid core plaque. Eur Heart J. 2013;34:2047–2054. doi: 10.1093/eurheartj/eht132 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Rahman H, Corcoran D, Aetesam‐Ur‐Rahman M, Hoole SP, Berry C, Perera D. Diagnosis of patients with angina and non‐obstructive coronary disease in the catheter laboratory. Heart. 2019;105:1536–1542. doi: 10.1136/heartjnl-2019-315042 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Eskerud I, Gerdts E, Larsen TH, Simon J, Maurovich‐Horvat P, Lønnebakken MT. Total coronary atherosclerotic plaque burden is associated with myocardial ischemia in non‐obstructive coronary artery disease. Int J Cardiol Heart Vasc. 2021;35:100831. doi: 10.1016/j.ijcha.2021.100831 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Barbato E, Toth GG, Johnson NP, Pijls NH, Fearon WF, Tonino PA, Curzen N, Piroth Z, Rioufol G, Jüni P, et al. A prospective natural history study of coronary atherosclerosis using fractional flow reserve. J Am Coll Cardiol. 2016;68:2247–2255. doi: 10.1016/j.jacc.2016.08.055 [DOI] [PubMed] [Google Scholar]
- 37. Park J, Lee JM, Koo BK, Shin ES, Nam CW, Doh JH, Hwang D, Zhang J, Hu X, Wang J, et al. Clinical relevance of functionally insignificant moderate coronary artery stenosis assessed by 3‐vessel fractional flow reserve measurement. J Am Heart Assoc. 2018;7:e008055. doi: 10.1161/JAHA.117.008055 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Jeremias A, Kirtane AJ, Stone GW. A test in context: fractional flow reserve: accuracy, prognostic implications, and limitations. J Am Coll Cardiol. 2017;69:2748–2758. doi: 10.1016/j.jacc.2017.04.019 [DOI] [PubMed] [Google Scholar]
- 39. Parikh RV, Liu G, Plomondon ME, Sehested TSG, Hlatky MA, Waldo SW, Fearon WF. Utilization and outcomes of measuring fractional flow reserve in patients with stable ischemic heart disease. J Am Coll Cardiol. 2020;75:409–419. doi: 10.1016/j.jacc.2019.10.060 [DOI] [PubMed] [Google Scholar]
- 40. Tu S, Barbato E, Köszegi Z, Yang J, Sun Z, Holm NR, Tar B, Li Y, Rusinaru D, Wijns W, et al. Fractional flow reserve calculation from 3‐dimensional quantitative coronary angiography and TIMI frame count: a fast computer model to quantify the functional significance of moderately obstructed coronary arteries. JACC Cardiovasc Interv. 2014;7:768–777. doi: 10.1016/j.jcin.2014.03.004 [DOI] [PubMed] [Google Scholar]
- 41. Papafaklis MI, Muramatsu T, Ishibashi Y, Lakkas LS, Nakatani S, Bourantas CV, Ligthart J, Onuma Y, Echavarria‐Pinto M, Tsirka G, et al. Fast virtual functional assessment of intermediate coronary lesions using routine angiographic data and blood flow simulation in humans: comparison with pressure wire—fractional flow reserve. EuroIntervention. 2014;10:574–583. doi: 10.4244/EIJY14M07_01 [DOI] [PubMed] [Google Scholar]
- 42. Tu S, Westra J, Yang J, von Birgelen C, Ferrara A, Pellicano M, Nef H, Tebaldi M, Murasato Y, Lansky A, et al. Diagnostic accuracy of fast computational approaches to derive fractional flow reserve from diagnostic coronary angiography: the international multicenter FAVOR pilot study. JACC Cardiovasc Interv. 2016;9:2024–2035. doi: 10.1016/j.jcin.2016.07.013 [DOI] [PubMed] [Google Scholar]
- 43. Tsugu T, Tanaka K, Belsack D, Nagatomo Y, Tsugu M, Argacha J‐F, Cosyns B, Buls N, De Maeseneer M, De Mey J. Impact of vessel morphology on CT‐derived fractional‐flow‐reserve in non‐obstructive coronary artery disease in right coronary artery. Eur Radiol. 2024;34:1836–1845. doi: 10.1007/s00330-023-09972-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Tsugu T, Tanaka K, Nagatomo Y, Belsack D, De Maeseneer M, De Mey J. Paradoxical changes of coronary computed tomography derived fractional flow reserve. Echocardiography. 2022;39:398–403. doi: 10.1111/echo.15307 [DOI] [PubMed] [Google Scholar]
- 45. Tsugu T, Tanaka K, Nagatomo Y, Belsack D, Argacha JF, Cosyns B, De Maeseneer M, De Mey J. Impact of ramus coronary artery on computed tomography derived fractional flow reserve (FFR(CT)) in no apparent coronary artery disease. Echocardiography. 2023;40:103–112. doi: 10.1111/echo.15526 [DOI] [PubMed] [Google Scholar]
- 46. Mushtaq S, De Araujo GP, Garcia‐Garcia HM, Pontone G, Bartorelli AL, Bertella E, Campos CM, Pepi M, Serruys PW, Andreini D. Long‐term prognostic effect of coronary atherosclerotic burden: validation of the computed tomography‐Leaman score. Circ Cardiovasc Imaging. 2015;8:e002332. doi: 10.1161/CIRCIMAGING.114.002332 [DOI] [PubMed] [Google Scholar]
- 47. Chow BJ, Small G, Yam Y, Chen L, McPherson R, Achenbach S, Al‐Mallah M, Berman DS, Budoff MJ, Cademartiri F, et al. Prognostic and therapeutic implications of statin and aspirin therapy in individuals with nonobstructive coronary artery disease: results from the CONFIRM (COronary CT angiography EvaluatioN for clinical outcomes: an InteRnational multicenter registry) registry. Arterioscler Thromb Vasc Biol. 2015;35:981–989. doi: 10.1161/ATVBAHA.114.304351 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Hadamitzky M, Achenbach S, Al‐Mallah M, Berman D, Budoff M, Cademartiri F, Callister T, Chang HJ, Cheng V, Chinnaiyan K, et al. Optimized prognostic score for coronary computed tomographic angiography: results from the CONFIRM registry (COronary CT angiography EvaluatioN for clinical outcomes: an InteRnational multicenter registry). J Am Coll Cardiol. 2013;62:468–476. doi: 10.1016/j.jacc.2013.04.064 [DOI] [PubMed] [Google Scholar]
- 49. Rasmussen LD, Fordyce CB, Nissen L, Hill CL Jr, Alhanti B, Hoffmann U, Udelson J, Bøttcher M, Douglas PS, Winther S. The PROMISE minimal risk score improves risk classification of symptomatic patients with suspected CAD. JACC Cardiovasc Imaging. 2022;15:1442–1454. doi: 10.1016/j.jcmg.2022.03.009 [DOI] [PubMed] [Google Scholar]
- 50. Udelson JE, Kelsey MD, Nanna MG, Fordyce CB, Yow E, Clare RM, Mark DB, Patel MR, Rogers C, Curzen N, et al. Deferred testing in stable outpatients with suspected coronary artery disease: a prespecified secondary analysis of the PRECISE randomized clinical trial. JAMA Cardiol. 2023;8:915–924. doi: 10.1001/jamacardio.2023.2614 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Mathew RC, Gottbrecht M, Salerno M. Computed tomography fractional flow reserve to guide coronary angiography and intervention. Interv Cardiol Clin. 2018;7:345–354. doi: 10.1016/j.iccl.2018.03.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. Douglas PS, Hoffmann U, Patel MR, Mark DB, Al‐Khalidi HR, Cavanaugh B, Cole J, Dolor RJ, Fordyce CB, Huang M, et al. Outcomes of anatomical versus functional testing for coronary artery disease. N Engl J Med. 2015;372:1291–1300. doi: 10.1056/NEJMoa1415516 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Tables S1–S7
Figure S1