Author + information
- Received September 21, 2017
- Revision received November 7, 2017
- Accepted November 8, 2017
- Published online April 2, 2018.
- Amir Ahmadi, MDa,b,
- Jonathon Leipsic, MDb,
- Kristian A. Øvrehus, MDc,
- Sara Gaur, MDc,
- Emilia Bagiella, PhDa,
- Brian Ko, MDd,
- Damini Dey, PhDe,
- Gina LaRocca, MDa,
- Jesper M. Jensen, MDc,
- Hans Erik Bøtker, MDc,
- Stephan Achenbach, MDf,
- Bernard De Bruyne, MD, PhDg,
- Bjarne L. Nørgaard, MD, PhDc and
- Jagat Narula, MD, PhDa,∗ ()
- aDivision of Cardiology, Icahn School of Medicine at Mount Sinai, New York, New York
- bDivision of Cardiology, Department of Radiology, University of British Columbia, Vancouver, British Columbia, Canada
- cDivision of Cardiology, Aarhus University Hospital, Aarhus, Denmark
- dDivision of Cardiology, Monash University, Melbourne, Australia
- eDivision of Cardiology, Cedars-Sinai Medical Center, Los Angeles, California
- fDivision of Cardiology, University of Erlangen-Nuremberg, Erlangen, Germany
- gDivision of Cardiology, Cardiovascular Center Aalst, Aalst, Belgium
- ↵∗Address for correspondence:
Dr. Jagat Narula, Icahn School of Medicine at Mount Sinai, 1190 Fifth Avenue, GP-1 West, N-125, New York, New York 10029.
Objectives The aims of the present study were: 1) to investigate the contribution of the extent of luminal stenosis and other lesion composition-related factors in predicting invasive fractional flow reserve (FFR); and 2) to explore the distribution of various combinations of morphological characteristics and the severity of stenosis among lesions demonstrating normal and abnormal FFR.
Background In patients with stable ischemic heart disease, FFR-guided revascularization, as compared with medical therapy alone, is reported to improve outcomes. Because morphological characteristics are the basis of plaque rupture and acute coronary events, a relationship between FFR and lesion characteristics may exist.
Methods This is a subanalysis of NXT (HeartFlowNXT: HeartFlow Analysis of Coronary Blood Flow Using Coronary CT Angiography), a prospective, multicenter study of 254 patients (age 64 ± 10 years, 64% male) with suspected stable ischemic heart disease; coronary computed tomography angiography including plaque morphology assessment, invasive angiography, and FFR were obtained for 383 lesions. Ischemia was defined by invasive FFR ≤0.80. Computed tomography angiography–defined morphological characteristics of plaques and their vascular location were used in univariate and multivariate analyses to examine their predictive value for invasive FFR. The distribution of various combinations of plaque morphological characteristics and the severity of stenosis among lesions demonstrating normal and abnormal FFR were examined.
Results The percentage of luminal stenosis, low-attenuation plaque (LAP) or necrotic core volume, left anterior descending coronary artery territory, and the presence of multiple lesions per vessel were the predictors of FFR. When grouped on the basis of degree of luminal stenosis, FFR-negative lesions had consistently smaller LAP volumes compared with FFR-positive lesions. The distribution of plaque characteristics in lesions with normal and abnormal FFR demonstrated that whereas FFR-negative lesions excluded likelihood of stenotic plaques with moderate to high LAP volumes, only one-third of FFR-positive lesions demonstrated obstructive plaques with moderate to high LAP volumes.
Conclusions In addition to the severity of luminal stenosis, necrotic core volume is an independent predictor of FFR. The distribution of plaque characteristics among lesions with varying luminal stenosis and normal and abnormal FFR may explain the outcomes associated with FFR-guided therapy.
- coronary artery stenosis
- myocardial ischemia
- percutaneous coronary intervention
- stable ischemic heart disease
- vulnerable plaque
On the basis of the FAME (Fractional Flow Reserve Versus Angiography for Multivessel Evaluation) family of studies (1,2), fractional flow reserve (FFR)–guided therapy has become the current standard for treatment of patients with stable ischemic heart disease. However, what continues to be intriguing is how a marker describing the functional significance of an anatomic lesion would identify plaque behavior and clinical outcomes. The prognostic outcome has traditionally been linked to the morphological features of high-risk plaques (HRPs) vulnerable to rupture. Therefore, it is important to address whether plaque morphology could influence the FFR abnormality, and if so, whether a normal FFR would predict relatively stable plaque morphology and hence a lower rate of future major adverse cardiovascular events. It may also be interesting to inquire whether an abnormal FFR would predict a higher likelihood of major adverse cardiovascular events.
Luminal stenosis is an established predictor of FFR (3). However, the relationship between the 2 is far from perfect (4–6), and there are a considerable number of stenotic lesions without ischemia and ischemic lesions without stenosis (7,8). The imperfect relationship between luminal stenosis and FFR has been attributed to the limitations of 2-dimensional interpretation of luminal stenosis on invasive coronary angiography (ICA), even though more accurate measures of luminal narrowing (e.g., minimal luminal diameter and minimal luminal area with intravascular ultrasound) have not helped improve the relationship between the degree of stenosis and FFR (9). Other anatomic characteristics such as the lesion length, entrance and exit angles, and size of the reference vessel have been proposed to explain the discrepancy, but without convincing evidence (10).
Coronary computed tomography angiography (CTA) is well placed to evaluate HRP characteristics and correlate them with the FFR-based hemodynamic relevance of the coronary artery lesions. On the basis of a multivariable analysis of various CTA-defined vessel factors, including quantitative percentage of luminal stenosis, minimal luminal area, minimal luminal diameter, and total lesion length in vessel, as well as the total plaque volume within the vessel, it was reported that total low-attenuation plaque (LAP) volume >30 mm3 in the culprit vessel was an independent predictor of FFR and additive to the CTA-verified degree of luminal stenosis of the most stenotic lesion (11). Because the lesion-specific large LAP volume is known to be associated with short- and intermediate-term outcomes (12,13) and because vessel-specific LAP correlates with FFR (11), it is logical that FFR may be a marker of severely stenotic lesions, large LAP volume, or a combination thereof (8). Conversely, a normal FFR should define a plaque with a lack of these features and hence with a benign prognostic outcome.
In the present study, we focused on the most stenotic lesion of the vessel and explored: 1) the contribution of various lesion-related and vessel-related factors to the invasive FFR value; and 2) the distribution of various combinations of plaque characteristics (morphology) and the severity of stenosis (anatomy) among lesions demonstrating normal and abnormal FFR.
Study population and protocol
This is a subanalysis of NXT (HeartFlowNXT: HeartFlow Analysis of Coronary Blood Flow Using Coronary CT Angiography; NCT01757678) comprising patients suspected of stable coronary artery disease (CAD) (14). Coronary CTA was performed within 60 days before clinically indicated nonemergency ICA. Excluded were patients with prior stent implantation or coronary bypass surgery; contraindications to beta-blockers, nitrates, or adenosine; suspicion of acute coronary syndrome; significant arrhythmia; and body mass index ≥35 kg/m2 (15,16).
In 254 patients, 484 vessels were evaluated by CTA, ICA, and FFR. Vessels without stenosis noted on coronary CTA (n = 81) and vessels with complete occlusion on ICA (n = 20) were excluded from the analyses (Figure 1). The study complied with the Declaration of Helsinki, and the local ethics committees approved the study protocol. All patients provided written informed consent.
Coronary computed tomography angiography acquisition
Coronary CTA was performed using computed tomography (CT) scanners with 64 or more detector rows, as previously described (15,16). Beta-blockers were administered if necessary, targeting a heart rate of 60 beats/min. Sublingual nitrates were administered before scanning in all patients (14).
Coronary plaque analysis
The coronary plaque analysis strategy has previously been described (11). In short, coronary segments ≥2 mm with plaque were analyzed using semiautomated software (AutoPlaq version 9.7, Los Angeles, California). Two readers (S.G. and K.A.Ø.) who were blinded to ICA and invasive FFR results performed the analyses using multiplanar coronary CTA images. Scan-specific thresholds for noncalcified plaque and calcified plaque were automatically generated (17). Plaque components were quantified within the manually designated area by using adaptive algorithms (17); adjustments were made if necessary. LAP volume was defined by the volume of part of the plaque with attenuation <30 Hounsfield units. The remodeling index was obtained by dividing the maximum vessel area within the lesion by the area of a proximal normal reference point. Positive remodeling was defined by the presence of a remodeling index ≥1.1. The most stenotic lesion in each vessel was identified, and various plaque components specific to that lesion were used in the analysis (14).
Invasive coronary angiography and fractional flow reserve measurements
ICA and invasive FFR were performed according to standard practice (15,16). The FFR pressure wire was positioned a minimum of 20 mm distal to the stenosis in vessel segments ≥2 mm. Hyperemia was induced by intravenous adenosine (140 to 180 mg/kg/min). FFR ≤0.80 was defined as the presence of lesion-specific ischemia.
The following statistical analyses were performed. 1) Various vessel- and lesion-related factors were investigated as independent predictors of FFR by using univariate and multivariate analyses in 2 models: in model 1, vessels with only 1 lesion per vessel were identified, and the analysis was performed on this subset; in model 2, the analysis was expanded to all vessels in the study. 2) The relationship between LAP volume and the degree of luminal stenosis for the most obstructive lesions with normal and abnormal FFR was explored for all vessels (n = 383). 3) The distribution of various combinations of the degree of luminal stenosis and the extent of LAP volume was investigated for all lesions with normal and abnormal FFR (Figure 1).
The CTA-verified most stenotic lesions of each vessel were identified. Descriptive analyses of lesion characteristics of the most stenotic lesion in the vessel together with various vessel-related variables for prediction of invasive FFR were conducted using means and SDs, or medians and interquartile ranges for continuous data and frequency and proportion for binary and categorical data. The lesion-related variables used in this analysis were total plaque volume, noncalcified plaque volume, calcified plaque volume, LAP volume, lesion length, CT-based diameter stenosis (quantitative coronary angiography), and remodeling index. The vessel-related variables were epicardial vessel territory (left anterior descending coronary artery [LAD], right coronary artery [RCA], left circumflex coronary artery [LCx]), location of the lesion within the vessel (proximal, middle, distal), and number of lesions per vessel.
Univariate analyses were conducted to determine the relationship between variables of interest and FFR. On the basis of these analyses, multivariate models were built to identify independent contributors of FFR. The analysis was conducted on vessels with only 1 lesion (model 1) and all vessels (model 2) separately. The analysis of the vessels with only 1 lesion (model 1) was conducted using linear regression models. For the analysis of all vessels (model 2), linear models with generalized estimating equations were used to correct for the correlation among vessels on the same patient. All analyses were conducted using SAS software version 9.4 (SAS Institute, Inc., Cary, North Carolina). All tests were 2-sided at the 0.05 significance level.
From the NXT study database of 254 patients and 484 vessels, after the exclusion of 81 CTA-verified normal vessels and 20 ICA-verified totally occluded vessels, analysis was performed on 383 vessels for exploring relationship between FFR and morphological composition of the lesion (Figure 1). Of these vessels, 128 showed only 1 lesion, and these vessels were also analyzed separately. Baseline characteristics of the NXT study population have been reported earlier (14). Briefly, the mean age of NXT patients was 64 ± 10 years, 162 (64%) of the patients were male, and 220 (87%) of the patients demonstrated an intermediate pre-test risk of significant CAD at 20% to 80% on the updated Diamond-Forrester risk scale.
Independent predictors of FFR
Model 1: vessels with 1 lesion only
Of the 383 vessels, 128 demonstrated a single lesion (32% LAD, 41% LCx, 27% RCA). Of these, 15 vessels demonstrated FFR ≤0.8 (12%); 23% of LAD lesions were FFR-positive (FFR[+]) as compared with 8% LCx and 6% RCA lesions. The determinants of invasive FFR were explored in a multivariable model including CTA-verified lesion-specific variables (degree of stenosis, lesion length, and LAP volume) and vessel-related variables (vessel territory, location of the lesion on the vessel, and number of involved segments per vessel) (Table 1). Only LAP volume and degree of luminal stenosis by CTA emerged as the independent lesion-related predictors of invasive FFR (Table 1). In this model, LAP volume (estimate −0.002; p < 0.0001) was a superior predictor as compared with the degree of luminal stenosis (estimate −0.001; p < 0.04). The lesion length was not a predictor of FFR (estimate −0.001; p = 0.3). In addition, location of the lesion in the LAD was the only independent vessel-related predictor of FFR (estimate −0.08; p < 0.0001).
Model 2: all vessels
The analyses of all 383 vessels (49% LAD, 28% RCA, and 23% LCx) demonstrated invasive FFR of ≤0.8 in 83 (22%) vessels; 33% of LADs were FFR(+) compared with 10% of RCAs and 13% of LCxs (p < 0.001). As described earlier, multivariable analysis of lesion- and vessel-related variables revealed the degree of stenosis and LAP volume to be the independent predictors of invasive FFR. Notably, the lesion length, similar to model 1, was not a predictor of FFR. With regard to vessel-related factors, localization of the lesion in the LAD (as opposed to RCA or LCx), the number of vessel segments involved, and the proximal location of the lesion within the vessel emerged as the independent predictors of FFR (Table 1).
Relationship between the severity of stenosis and LAP volume in FFR-negative and FFR-positive lesions
All lesions were divided into 3 groups on the basis of the CTA-verified degree of luminal stenosis: <50%, 50% to 70%, and >70%. The percentage of FFR(+) lesions increased with the increase in degree of luminal stenosis (Figure 2). Similarly, on classifying the lesions into 3 categories on the basis of the LAP volume, the proportion of FFR(+) lesions increased significantly with the increase in LAP volume (Figure 2), independent of the degree of luminal stenosis.
When lesions were grouped together on the basis of the degree of luminal stenosis, there were FFR(+) and FFR-negative (FFR[−]) lesions within each group. There were no significant differences between the degree of stenosis of FFR(+) and FFR(−) lesions within each subgroup. However, statistically significant differences were noted between the LAP volumes of FFR(+) and FFR(−) lesions within each subgroup (Figures 3A and 3B).
Distribution of the severity of stenosis and LAP volumes in lesions with normal and abnormal FFR
The LAP volume of each lesion and the degree of luminal stenosis were compared between lesions with normal and abnormal FFR. For this purpose, 50% ICA-verified luminal stenosis was used as the cutoff value to differentiate obstructive from nonobstructive lesions, and small, moderate, and large LAP volumes were defined by cutoff values of <10 mm3, 10 to 25 mm3, and >25 mm3, respectively. The cutoff values for LAP were determined on the basis of LAP volume tertiles of the FFR(+) lesions.
Of 300 lesions with FFR >0.80, 53% were nonobstructive with small LAP volumes, whereas 4% of lesions were obstructive with moderate LAP, and 2% were obstructive with large LAP volumes. Interestingly, among lesions with FFR ≤0.80, 54% were nonobstructive; 30% of lesions were obstructive with moderate or large LAP volumes, and 28% had small LAP volumes (Table 2).
It has been previously proposed that the presence of a large necrotic core may somehow contribute to lesion-specific ischemia expressed by abnormal FFR, independent of the degree of stenosis (7,8). It has also been reported that the total LAP volume within the entire vessel and the degree of luminal stenosis were the independent predictors of FFR (11). Unlike earlier studies (11) wherein vessel-level characteristics were observed to be predictors of normal or abnormal FFR, the present study focused on the most stenotic lesion, which is clinically more relevant. We used CTA as a tool that allowed identification and quantification of morphological characteristics of the most stenotic lesion, as well as the degree of luminal stenosis, to explore their relationship with lesion-specific ischemia by FFR. The degree of luminal stenosis and the LAP volume of the most stenotic lesion were the independent predictors of FFR. In addition to the lesion-specific characteristics, the presence of the lesion in the LAD artery, the location of the lesion in the proximal part of the vessel, and multiple lesions within the same vessel were also the independent predictors of invasive FFR. Importantly, lesion length, usually considered an important determinant of FFR, was found not to be related to ischemia. Moreover, we observed that among the lesions with a similar degree of luminal stenosis, the lesions with FFR ≤0.80 had significantly larger LAP volumes compared with the lesions with FFR >0.80. Therefore, it is conceivable that the difference in LAP volume could explain the stenosis-FFR mismatch in some cases, whereas in others, vascular territory and the additive effect of multiple lesions in the same vessel may contribute to the mismatch. We observed that most FFR(−) lesions were either nonobstructive or had small LAP volumes, findings that most likely explain the favorable prognosis associated with FFR(−) lesions reported in the DEFER and FAME studies (1). Conversely, the abnormal FFR may result from variable combinations of plaque stenosis and morphological traits of the lesion. In the current study, more than one-half of lesions with an FFR value ≤0.80 had <50% luminal stenosis, and only one-third of FFR(+) lesions demonstrated obstructive plaques with moderate to large LAP volumes. This finding could explain why adverse cardiovascular events occur in a relatively small portion of lesions with abnormal FFR when these patients are treated with medical therapy only (8.6%) compared with revascularization (7.2%) (2). Therefore, FFR could be accepted as a sensitive but not a specific marker of obstructive lesions associated with moderate to large LAP volumes or HRPs.
LAP volume and severity of luminal stenosis are independent predictors of FFR
The imperfect relationship between stenosis and FFR is well documented (4,5), and large numbers of lesions with stenosis-ischemia mismatch have been described. Importantly, among lesions with intermediate degrees of luminal stenosis, the ratio of FFR(+) to FFR(−) lesions is almost 50:50 (8), and it has been attributed to possible limitations of 2-dimensional angiography, lesion length, and other factors (9,10,18). The present study, in a multivariable model, did not find lesion length to be important and instead suggested that in addition to luminal stenosis LAP volume is an independent lesion-related predictor of FFR. There were FFR(+) and FFR(−) lesions in all lumen stenosis categories wherein the FFR(+) lesions demonstrated significantly larger LAP volumes. This relationship suggests that the addition of the lesion-specific LAP volume may enhance the prognostic importance of luminal obstruction and stenosis-related ischemia in stable CAD. The importance of the LAP volume may be more pronounced in mildly to moderately stenotic lesions because lesion-specific ischemia in such patients is not likely to result from anatomic stenosis but rather from development of functionally significant stenosis at the time of maximum hyperemia, as described later (8).
During FFR measurement, the infusion of adenosine dilates the distal arteriolar bed with a drop in the resistance in the distal coronary bed that leads to development of an increased gradient between the aorta and distal coronary bed to drive increased coronary blood flow through the stenotic lesion or a maximally hyperemic state. Epicardial coronary artery autoregulatory mechanisms also respond to the state of maximal hyperemia by further dilatation, which, in addition, is exacerbated by nitroglycerin administration (a standard component of FFR protocol). In the case of severe and fixed luminal stenosis, because the vessel at the level of stenosis cannot dilate any further, a post-stenotic pressure drop occurs. Conversely, in the case of mild to moderate luminal stenosis with a large necrotic core, it is proposed (8) that the epicardial vessel likely develops dynamic stenosis with an inability to dilate at the lesion site. With the rest of the vessel dilating during maximal hyperemia, the once mild to moderate stenosis may become functionally significant. The reason for the local inability of the vessel to dilate at the site of a large necrotic core is not clear, but this inability could be caused by significant positive remodeling at that level. The maximally stretched smooth muscle layer (or maximal positive remodeling) would restrict further dilatation analogous to the limits of Glagov phenomenon, after which the lipid-rich lesions encroach on the luminal diameter (19). Furthermore, the large, lipid-rich necrotic core should add inflammatory insult, oxidative stress, and hence local endothelial dysfunction (7,20–22). FFR is not a direct measure of ischemia, but rather is a surrogate that measures a ratio in pressure drop across the lesion. The transstenotic pressure decay is inversely proportional to the fourth power of the lumen radius, and it contributes significantly to the specific curvilinear pressure-flow relationship and defines the physiological importance of a fixed or dynamic narrowing (23). As a consequence, a change in luminal diameter relative to other segments of the same vessel, caused by local vasodilatory impairment at the time of maximal hyperemia, produces a marked hemodynamic effect, leading to abnormal FFR measurement. It is also reasonable to presume that the impairment in vasodilatation may vary depending on the size of the necrotic core, the volume of the reference vessel, and the vessel’s overall ability to dilate.
Combination plaque morphology and the degree of luminal stenosis may explain outcomes in lesions with normal and abnormal FFR
Analyses of the DEFER and FAME trial results (1,2,8) suggest that: 1) the benefit of FFR-guided therapy predominantly results from the safe deferral of FFR(−) lesions to medical therapy only, thereby avoiding the complications of unnecessary revascularization; and 2) only a small portion of FFR(+) lesions (8.6%) will result in future myocardial infarction if they are not revascularized (8,24,25). It is not clear why relatively normal post-stenotic pressure (normal FFR) would strongly predict freedom from subsequent events known to be caused by HRP morphology and plaque rupture.
From numerous intravascular ultrasound and CTA studies (12,13,26,27), it is known that obstructive lesions with high-risk features (large necrotic core and thin fibrous cap) portend a maximum likelihood of future events, ranging up to 22%, and nonobstructive lesions with high-risk features often undergo rapid plaque progression to become increasingly prone to rupture (28). More importantly, the absence of a large necrotic core or a thin fibrous cap, even in obstructive plaques, renders these lesions at low risk for future events; the negative predictive value for future events ranges from 96% to 100% for non-HRPs (12,13,26,27,29–32). It could thus be surmised that: 1) obstructive lesions with a large necrotic core should be at high risk of plaque rupture and an acute event; 2) nonobstructive lesions with a moderate to large necrotic core may result in unstable symptoms or subsequent events on plaque progression; and 3) lesions with a small or absent necrotic core, regardless of the degree of stenosis, should remain at a relatively low risk for future events.
Exploring the distribution of various lesion types (the plaque characteristics and severity of luminal stenosis) in FFR(−) and FFR(+) cohorts may explain the event rates in both groups. In the FFR(−) cohort, only 2% of lesions were obstructive by ICA and had a large LAP volume; the rest of the lesions were either nonobstructive or had smaller LAP volumes. This could explain why lesions with normal FFR could safely be treated with medical therapy alone and only a minority may result in future events including revascularization. Conversely, among FFR(+) lesions, only 30% were obstructive, with moderate to large LAP volume, and hence at high risk of future events; 29% had a small LAP volume and were therefore at low risk for future events. These observations can perhaps explain why only a relatively small portion of FFR(+) plaques would result in events if these lesions were left to medical therapy alone (24). These observations also beg to question the validity of revascularization recommendations for all FFR(+) lesions when only 8.6% of FFR(+) lesions in patients randomized to medical therapy alone in the FAME 2 trial resulted in death or myocardial infarction (2). These findings question the necessity of revascularization in all abnormal FFR lesions and bring up the possibility of the use of CTA plaque characterization as a means to stratify the lesions with normal and abnormal FFR further.
This was a post hoc analysis of data from the NXT trial. The pre-specified selection criteria for inclusion in this study resulted in a higher proportion of patients with obstructive CAD than in a nonselected coronary CTA population. Patients with acute coronary syndromes or previous revascularization were excluded in this study. Thus, applicability of results to those patient categories needs further study.
The current study demonstrated that the CT-defined LAP volume (a surrogate for necrotic core) of the most stenotic lesion in the vessel and the degree of luminal stenosis were the independent predictors of FFR. We also observed that in the plaques with a similar degree of stenosis, the presence of greater LAP volume was associated with lesion-specific ischemia or abnormal FFR. Further, most lesions with normal FFR did not demonstrate HRP characteristics, and only a small subset of FFR(+) lesions had these characteristics, thereby making FFR a more sensitive but less specific predictor of high-risk coronary lesions. Additional information about plaque morphological features, such as LAP volume, may offer a logical next step beyond FFR measurement in therapeutic decision making. This proposal would need to be tested in a randomized prospective fashion in patients presenting with stable CAD.
COMPETENCY IN MEDICAL KNOWLEDGE: The ischemia-stenosis mismatch could probably be explained in many cases by the lesion’s morphological composition. The relationship of lesion morphology and FFR may also help explain superior outcomes associated with FFR-guided therapy.
TRANSLATIONAL OUTLOOK: The links among luminal obstruction, plaque morphology, and functional characteristics of the plaque can open a new door for future studies to investigate the value of combining these features in lesion stratification and guiding revascularization.
Dr. Leipsic is a consultant for and has received stock options from Circle CVI and HeartFlow; and has received fellowship support from GE Healthcare. Dr. Ko has received speaker fees from St. Jude Medical, Merck Sharpe & Dohme, Novartis, Bristol-Myers Squibb, and Specialised Therapeutics. Dr. Dey may receive royalties for software licenses from Cedars-Sinai Medical Center. Dr. Jensen has received a speaker honorarium from Bracco Imaging. Dr. De Bruyne has served as an institutional consultant for Abbott, Boston Scientific Corporation, and Opsens. Dr. Nørgaard’s institution has received unrestricted research grants from HeartFlow and Siemens. Dr. Botker’s institution has received an unrestricted research grant from HeartFlow.
- Abbreviations and Acronyms
- coronary artery disease
- computed tomography
- computed tomography angiography
- invasive coronary angiography
- fractional flow reserve
- high-risk plaque
- left anterior descending coronary artery
- low-attenuation plaque
- left circumflex coronary artery
- right coronary artery
- Received September 21, 2017.
- Revision received November 7, 2017.
- Accepted November 8, 2017.
- Gould K.L.,
- Lipscomb K.,
- Calvert C.
- Park S.J.,
- Kang S.J.,
- Ahn J.M.,
- et al.
- Tonino P.A.,
- Fearon W.F.,
- De Bruyne B.,
- et al.
- Ahmadi A.,
- Kini A.,
- Narula J.
- Ahmadi A.,
- Stone G.W.,
- Leipsic J.,
- et al.
- Johnson N.P.,
- Kirkeeide R.L.,
- Gould K.L.
- Kern M.J.,
- Samady H.
- Gaur S.,
- Ovrehus K.A.,
- Dey D.,
- et al.
- Motoyama S.,
- Sarai M.,
- Harigaya H.,
- et al.
- Motoyama S.,
- Ito H.,
- Sarai M.,
- et al.
- Nørgaard B.L.,
- Leipsic J.,
- Gaur S.,
- et al.
- Nakazato R.,
- Shalev A.,
- Doh J.H.,
- et al.
- Lavi S.,
- Bae J.H.,
- Rihal C.S.,
- et al.
- Lavi S.,
- McConnell J.P.,
- Rihal C.S.,
- et al.
- Spaan J.A.,
- Piek J.J.,
- Hoffman J.I.,
- Siebes M.
- Pijls N.H.,
- Fearon W.F.,
- Tonino P.A.,
- et al.
- Motoyama S.,
- Kondo T.,
- Sarai M.,
- et al.
- Ahmadi A.,
- Leipsic J.,
- Blankstein R.,
- et al.
- Kaul S.,
- Narula J.
- Narula J.,
- Nakano M.,
- Virmani R.,
- et al.
- Oemrawsingh R.M.,
- Cheng J.M.,
- Garcia-Garcia H.M.,
- et al.