Author + information
- Received September 11, 2015
- Revision received November 4, 2015
- Accepted November 10, 2015
- Published online April 13, 2016.
- Bjarne L. Nørgaard, MD, PhDa,∗ (, )
- Jakob Hjort, MPHa,
- Sara Gaur, MDa,
- Nicolaj Hansson, MDa,
- Hans Erik Bøtker, MD, DMScia,
- Jonathon Leipsic, MD, PhDb,
- Ole N. Mathiassen, MD, PhDa,
- Erik L. Grove, MD, PhDa,c,
- Kamilla Pedersen, BSca,
- Evald H. Christiansen, MD, PhDa,
- Anne Kaltoft, MD, PhDa,
- Lars C. Gormsen, MD, PhDd,
- Michael Mæng, MD, PhDa,
- Christian J. Terkelsen, MD, DmScia,
- Steen D. Kristensen, MD, DmScia,
- Lars R. Krusell, MDa and
- Jesper M. Jensen, MD, PhDa
- aDepartment of Cardiology, Aarhus University Hospital Skejby, Aarhus, Denmark
- bDepartment of Radiology, St. Paul’s Hospital, University of British Columbia, British Columbia, Canada
- cFaculty of Health, Institute of Clinical Medicine, Aarhus University, Aarhus, Denmark
- dDepartment of Nuclear Medicine, Aarhus University Hospital Skejby, Aarhus, Denmark
- ↵∗Reprint requests and correspondence:
Dr. Bjarne L. Nørgaard, Department of Cardiology, Aarhus University Hospital Skejby, Skejby DK-8200 Aarhus N, Denmark.
Objectives The goal of this study was to assess the real-world clinical utility of fractional flow reserve (FFR) derived from coronary computed tomography angiography (FFRCT) for decision-making in patients with stable coronary artery disease (CAD).
Background FFRCT has shown promising results in identifying lesion-specific ischemia. The real-world feasibility and influence on the diagnostic work-up of FFRCT testing in patients suspected of having CAD are unknown.
Methods We reviewed the complete diagnostic work-up of nonemergent patients referred for coronary computed tomography angiography over a 12-month period at Aarhus University Hospital, Denmark, including all patients with new-onset chest pain with no known CAD and with intermediate-range coronary lesions (lumen reduction, 30% to 70%) referred for FFRCT. The study evaluated the consequences on downstream diagnostic testing, the agreement between FFRCT and invasively measured FFR or instantaneous wave-free ratio (iFR), and the short-term clinical outcome after FFRCT testing.
Results Among 1,248 patients referred for computed tomography angiography, 189 patients (mean age 59 years; 59% male) were referred for FFRCT, with a conclusive FFRCT result obtained in 185 (98%). FFRCT was ≤0.80 in 31% of patients and 10% of vessels. After FFRCT testing, invasive angiography was performed in 29%, with FFR measured in 19% and iFR in 1% of patients (with a tendency toward declining FFR-iFR guidance during the study period). FFRCT ≤0.80 correctly classified 73% (27 of 37) of patients and 70% (37 of 53) of vessels using FFR ≤0.80 or iFR ≤0.90 as the reference standard. In patients with FFRCT >0.80 being deferred from invasive coronary angiography, no adverse cardiac events occurred during a median follow-up period of 12 (range 6 to 18 months) months.
Conclusions FFRCT testing is feasible in real-world symptomatic patients with intermediate-range stenosis determined by coronary computed tomography angiography. Implementation of FFRCT for clinical decision-making may influence the downstream diagnostic workflow of patients. Patients with an FFRCT value >0.80 being deferred from invasive coronary angiography have a favorable short-term prognosis.
Current guidelines recommend noninvasive functional imaging testing as the first-line strategy in patients with suspected stable coronary artery disease (CAD) (1). However, shortcomings of current noninvasive diagnostic strategies are apparent from the frequent inaccurate selection of patients for invasive coronary angiography (ICA) (2). Randomized trials have shown that a fractional flow reserve (FFR) threshold of 0.80 distinguishes patients and lesions that will (3–5) and will not (4–6) benefit from coronary revascularization. FFR has therefore emerged as the reference standard for guiding coronary revascularization in intermediate-range lesions (7,8). Noninvasive anatomic assessment using coronary computed tomography angiography (CTA) reveals high diagnostic performance for detection or exclusion of CAD. However, coronary CTA tends to overestimate obstructive CAD, and the correlation of stenoses to downstream myocardial ischemia is poor (9). Consequently, guidelines recommend additional stress testing in ambiguously symptomatic patients with suspected obstructive CAD determined by coronary CTA (1).
Recent advances in computational fluid dynamics and individual image-based modeling permit calculation of coronary blood flow and pressure from standard acquired coronary CTA datasets (10–12). This technique allows for noninvasive calculation of FFR (coronary computed tomography angiography–derived fractional flow reserve [FFRCT]). In 3 prospective multicenter trials that included 609 patients (1,050 vessels) with known or suspected CAD and blinded comparison to FFR, FFRCT exhibited a high and superior diagnostic performance compared with stenosis assessment according to coronary CTA (10–12). However, because of the potential selection bias in these studies, concerns have been raised regarding to what extent FFRCT in stable CAD can be translated into clinical practice (13,14). We report the first real-world experience of coronary CTA with FFRCT as gatekeeper to the catheterization laboratory in nonemergent symptomatic patients with intermediate-range coronary lesions.
This single-center, observational all-comers study was conducted in symptomatic patients with suspected CAD referred to coronary CTA at Aarhus University Hospital between May 1, 2014, and April 30, 2015. In this institution, coronary CTA testing is the preferred initial diagnostic modality in patients with new-onset chest pain with no known CAD and low-intermediate pre-test probability of disease (1). Patients were referred to nonemergent ambulatory coronary CTA from our outpatient clinic, 2 community hospitals, and 2 private cardiologist practices. Before performance of the coronary CTA, all patients underwent 12-lead electrocardiogram and routine biochemistry, echocardiography, and clinical evaluation by a cardiovascular physician; noninvasive ischemia testing (using exercise electrocardiography) was performed only in patients being referred from private cardiologist practices. Demographic and clinical characteristics, data on downstream diagnostic testing, and clinical follow-up were obtained from patient files and registries (15). The study was approved by the Danish Data Protection Agency (1-16-02-414-15) with a waiver for individual informed consent by the regional ethical committee.
Coronary CTA Acquisition, Analysis, and Diagnostic Consequence
Coronary CTA was performed using a dual-source CT scanner (SOMATOM Definition Flash; Siemens, Forchheim, Germany). Contraindications were renal insufficiency, pregnancy, or allergy to contrast. Image acquisition was performed in accordance with society guidelines (16). Oral and/or intravenous beta-blockers or oral ivabradine were administered if necessary, targeting a heart rate <60 beats/min, and all patients received sublingual nitrates. An initial nonenhanced scan for calcium scoring was performed using high-pitch spiral acquisition mode and 120 kV. Coronary CTA was performed using prospective electrocardiographic triggering in all patients. In the event of a heart rate ≤65 beats/min or >65 beats/min, the recommended RR scan intervals were 65% to 75% and 40% to 70%, respectively. Data acquisition was performed with 100- or 120-kV tube voltage in patients weighing ≤70 or >70 kg. Filtered back-projection was used for image reconstruction. Scans were assessed by using axial images and multiplanar reconstructions by experienced cardiologists. Vessel segments ≥2 mm were evaluated for lumen narrowing, and the per-vessel maximum stenosis category (30% to 50%, 51% to 70%, 71% to 90%, and >90%) was reported. The strategy of stenosis quantification was at the discretion of the CT scan interpreter. The cardiologists reading the CT scans grouped patients into 1 of 3 risk categories on the basis of the coronary anatomy (1); recommendations regarding downstream patient management were based on these categories (Table 1). Subsequent myocardial perfusion imaging (MPI) with rest-stress 82-rubidium positron emission tomography was recommended in patients in whom the coronary CTA interpreter deemed the result as inconclusive.
A standard coronary CTA dataset was transmitted for analysis (HeartFlow, Inc., Redwood City, California) (12). Clinical information related to FFRCT analysis was restricted to patient weight, height, heart rate, and blood pressure. The principles behind FFRCT computation have been described previously (10–12). FFRCT was displayed for each point in the coronary tree. FFRCT values distally in the major epicardial (left main, left anterior descending [LAD], left circumflex [LCx], and right coronary [RCA]) arteries (including side branches) >2 mm in diameter were registered. FFRCT ≤0.80 was considered diagnostic of lesion-specific ischemia (10–12). Recommendations during the study period on downstream patient management according to the FFRCT result are shown in Table 1. Figure 1 displays a representative case.
The proportion of patients having ICA performed as a consequence of the FFRCT result was registered. The proportion of patients having subsequent coronary revascularization (percutaneous coronary intervention [PCI] or coronary artery bypass grafting [CABG]) was registered. Patients were followed up for a median of 12 (range 6 to 18 months) months during which the occurrence of cardiac events (death, acute myocardial infarction, and angina leading to hospital admission or visit in the outpatient clinic) and diagnostic procedures (ICA, coronary CTA, and MPI) were recorded. No patients were lost to follow-up.
ICA and FFR
ICA and FFR were performed according to standard practice by experienced invasive cardiologists. All information relevant for patient management, including the FFRCT result, was available for decision-making. Decisions on measuring FFR and coronary revascularization were at the discretion of the interventionalist. Measurement of FFR was performed using the PressureWire (St. Jude Medical, St. Paul, Minnesota) or the ComboWire XT guidewire (Volcano Therapeutics, Cordova, California). The pressure sensor was advanced past the most distal stenosis, and FFR was measured during intravenous infusion of adenosine (140 to 180 μg/kg/min) via the antecubital vein. Intracoronary nitroglycerin (100 to 250 μg) was given before introduction of the FFR guidewire. Instantaneous wave-free ratio (iFR) was measured in few patients (17). FFR ≤0.80 or iFR ≤0.90 was indicative of ischemia.
Cumulative radiation exposure including all diagnostic tests are reported in millisieverts using the formula mSv = (dose length product) × 0.014 for coronary CTA, a conversion factor of 0.18 mSv/(Gy cm2) for ICA (18), and 0.00126 MBq/mSv for MPI (19).
Categorical data are presented as numbers or proportions. Continuous data are presented as mean ± SD, median (interquartile range), or median (range) as appropriate. Fisher’s exact test, chi-squared test, Mann-Whitney U test, or the Student t test were used for comparisons as appropriate. Correlation of per-vessel FFRCT to FFR was assessed using the Pearson correlation coefficient and Bland-Altman analysis. Statistical analyses were performed using Stata version 14.0 (Stata Corp, College Station, Texas).
Figure 2 displays the diagnostic flow of patients referred for coronary CTA during the study period. Contrast was not administered in 75 patients because of irregular heart rhythm and/or severe coronary calcification and/or inadequate patient cooperation. In this patient group, 77% were in sinus rhythm; the mean heart rate was 72 (range 53 to 131 beats/min) beats/min; and the median Agatston score was 617 (interquartile range: 56 to 2,024; range 0 to 4,723). Because of inconclusive coronary CTA results (low contrast, severe motion, misalignment, and/or blooming artifacts), 33 (3%) patients were referred for subsequent MPI. In 11 (1%) patients with a conclusive coronary CTA result and ≥1 intermediate lesion, MPI was the preferred downstream test strategy. Baseline characteristics of patients in whom coronary CTA was performed are shown in Table 2. Before coronary CTA testing, exercise electrocardiography was performed in 12% of the patients without temporal changes in proportions during the study. Baseline characteristics of patients referred directly to ICA based on the coronary CTA result or to MPI are shown in Table 3. No proportional changes during the study period in direct referrals to ICA or MPI could be detected.
Table 3 presents the baseline characteristics of patients scheduled for FFRCT testing. A conclusive FFRCT result was available within 24 h in 185 (98%) patients. In 4 patients, FFRCT computation was not possible due to calcium blooming, motion, and/or low contrast (n = 3) or missing diastole reconstruction for myocardial segmentation (n = 1). The per-patient and per-vessel distribution of FFRCT values are shown in Figure 3. In 31% (57 of 185) of patients and 10% (72 of 740) of vessels, FFRCT was ≤0.80. Among these 185 patients, intermediate stenoses were present in 303 vessels (left main, n = 9; LAD, n = 156; LCx, n = 62; RCA, n = 76). FFRCT in these vessels was 0.85 ± 0.08 (range 0.43 to 0.98) versus 0.92 ± 0.05 (range 0.78 to 1.00) in vessels without stenosis or with <30% stenosis (p < 0.0001). In the latter group, FFRCT was ≤0.80 (range 0.78 to 0.80) in 5 (1%) vessels (LCx, n = 4; RCA, n = 1).
The diagnostic work-up of patients after FFRCT testing is shown in Figure 4. Median time between CT scan and catheterization was 30 (range 1 to 121) days. In 37 patients, FFR (n = 35) or iFR interrogation was performed. The proportion of patients having FFR or iFR measured in the first two-thirds versus the last one-third of the study period was 77% (27 of 35) and 53% (10 of 19), respectively (p = 0.08). FFR was measured in 51 vessels and iFR in 2 vessels (LAD, n = 34; LCx, n = 8; RCA, n = 11).
There was good direct correlation of per-vessel FFRCT to FFR, with a slight systematic underestimation of FFRCT compared with FFR (Figure 5). Overall, FFRCT ≤0.80 correctly classified 73% (27 of 37) of patients and 70% (37 of 53) of vessels by using FFR ≤0.80 or iFR ≤0.90 as the reference standard. There was an inverse continuous relationship between the FFRCT numerical value and the probability of detecting ischemia (Table 4). The diagnostic agreement of FFRCT ≤0.80 according to coronary stenosis severity using FFR or iFR as the reference standard, as well as median (interquartile range; range) FFRCT in vessels with “true or false-positive” or “true or false-negative” FFRCT test results are shown in Table 5. FFRCT values in 3 “false-negative” vessels ranged between 0.81 and 0.83, all with a corresponding FFR value of 0.79. On a patient level, no FFRCT “false-negative” outcomes were observed.
Coronary Revascularization and Follow-up
In patients with FFRCT ≤0.80 being referred to ICA, 45% (22 of 49) underwent coronary revascularization (PCI, n = 12; CABG, n = 10). In 23 patients with FFRCT ≤0.75 being referred to ICA (FFR or iFR measured, n = 13), revascularization was performed in 70% (PCI, n = 10; CABG, n = 6). In patients referred from coronary CTA directly to ICA, FFR or iFR was measured in 20% without temporal proportional changes during the study, and 65% (53 of 82) underwent revascularization (PCI, n = 42; CABG, n = 11). No patients having FFRCT, ICA, or MPI performed experienced a serious adverse cardiac event during follow-up, including the 123 (66%) patients with FFRCT >0.80 in whom ICA was deferred. In 1 patient with the lowest FFRCT value of 0.88 (LAD) and 1 patient with FFRCT = 0.78 (LAD) in whom ICA originally was deferred, subsequent ICAs were performed after 4 and 6 months of follow-up, respectively, because of continuing chest pain. ICA revealed minor vessel irregularities (no FFR interrogation) in the former patient, and diffuse nonobstructive disease (measured FFR = 0.84) in the other patient. In the MPI group, ICA was performed in 2 patients during follow-up (normal test results).
This report demonstrates for the first time the real-world feasibility of FFRCT testing in consecutive patients with intermediate-range coronary stenosis. We made several important observations: 1) conclusive FFRCT results were obtained in 98% of patients who had CT scans referred for FFRCT testing; 2) implementation of FFRCT for clinical decision-making may influence the downstream diagnostic workflow; 3) in the event of FFRCT ≤0.75, the risk of a false-positive result using FFR as the reference standard is very low (<10%), whereas in patients with FFRCT ranging between 0.76 and 0.80, a non-negligible number of false-positive results may be expected; and 4) patients with FFRCT >0.80 being deferred from ICA have a favorable short-term prognosis.
In the most recent trial (NXT [Analysis of Coronary Blood Flow Using CT Angiography: Next Steps]) investigating the diagnostic performance of FFRCT in stable CAD, 13% of the patients had nonevaluable coronary CTA images (12). In this real-world report, we found that only 2% of patients failed to meet the image quality requirements for FFRCT analysis. Taking into account those patients in whom contrast was not administered because of an anticipated nondiagnostic CT image quality and those with an inconclusive test outcome, a conclusive CT scan–based anatomic or anatomic-physiological result was available in >90% of all patients referred to coronary CTA testing. These findings most likely reflect that CT image acquisition in this report was performed in accordance with guidelines, with careful attention to heart rate control and use of nitroglycerin (16,20). Of note, it was recently shown that FFRCT also provides high and superior diagnostic performance compared with coronary CTA alone in patients with high calcium scores (21). This outcome, together with future improvements in CT spatial resolution, may expand the eligibility of coronary CTA to patients in whom a nondiagnostic CT image quality is anticipated and to those with an inconclusive test result due to severe coronary calcification. Accordingly, MPI may potentially be avoided in a substantial proportion of patients in the diagnostic pathway of coronary CTA.
The recent multicenter PLATFORM (Prospective Longitudinal Trial of FFRCT: Outcome and Resource Impacts) trial demonstrated, in patients referred to ICA, that a diagnostic strategy of coronary CTA and FFRCT resulted in safe cancellation of planned ICAs in 61% of the patients (22). Similarly, in the present study that includes a contemporary coronary CTA cohort, a “normal” FFRCT result was present in 69% of the patients, among whom ICA was successfully deferred. On the other hand, the coronary revascularization rate in patients with FFRCT ≤0.80 in this study was surprisingly low (45%). This finding is partially explained by the high proportion of false-positive results in patents with FFRCT between 0.75 and 0.80. Accordingly, the revascularization rate increased to 70% in patients with FFRCT ≤0.75. Moreover, decisions on revascularization, especially in the FFR range between 0.75 and 0.80, are often influenced by factors other than FFR (e.g., severity of angina) (4,5). Finally, it should be acknowledged that “thresholds” by individual CT operators in prescribing subsequent FFRCT testing govern the proportion of normal and abnormal results. In the present study, this decision was recommended primarily on the basis of coronary anatomy; however, integration to a varying degree of other factors (including preference of downstream testing modality and information on pre-test risk of CAD and symptoms) naturally may have influenced decision-making regarding downstream patient management.
The high negative predictive value of FFRCT for the prediction of ischemia (10–12), as well as the findings in the PLATFORM trial (22) and in this report of a favorable short-term clinical outcome in patients with FFRCT >0.80, may encourage confidence for clinicians that such patients can safely be deferred for subsequent ICA. However, in the NXT trial, using an FFRCT dichotomous interpretation approach, despite diagnostic superiority compared with coronary CTA stenosis assessment alone, the FFRCT per-patient specificity and positive predictive value in predicting ischemia were modest (79% and 65%, respectively) (12). Thus, a substantial rate of false-positive results remained. After our initial clinical experiences with FFRCT testing as described in this report, our institution recently implemented an FFRCT testing decision-rule algorithm (Figure 6). We recommend in patients with FFRCT ranging between 0.75 and 0.80 that decisions on referral to ICA should be based on all available information, in particular regarding the severity of angina (i.e., the main target of PCI), whereas in patients with FFRCT >0.80 or ≤0.75, we propose a dichotomous interpretation strategy. Several factors support the relevance of this approach. First, in accordance with this report, the overall prognosis in patients with stable chest pain in contemporary practice is favorable (22,23), and the annual risk of cardiac death or myocardial infarction after deferral of coronary revascularization in intermediate-range lesions with FFR ≥0.75 is very low (<1%) (6). Moreover, it should be acknowledged that FFR exhibits a continuous relationship between its numerical value and the clinical outcome, with the largest benefit of revascularization being obtained at lower FFR values (5). In this study, the lower limit of FFRCT values among patients with a false-positive result was modestly low (0.74), and, in the event of FFRCT ≤0.75, the probability of having ischemia was high (92%), whereas if FFRCT ranged between 0.76 and 0.80, ischemia was present in only 55% of the patients. Accordingly, it has been shown that the variation in repeated FFR (24) measurements and FFRCT calculations (25) has the largest impact on interpretation of results when approaching the established cutpoint of 0.80. The finding in this study and in previous studies of a slight systematic underestimation of FFRCT compared with measured FFR (10–12) support the proposed FFRCT guidance algorithm. Safety and cost-efficiency of this FFRCT guidance diagnostic workflow in symptomatic patients with intermediate-range coronary lesions need delineation in future studies.
We observed a higher proportion of FFR guidance (69% vs. 20%) and lower revascularization rate (45% vs. 65%) in patients being referred to ICA based on FFRCT assessment compared with patients being referred directly from coronary CTA. These findings may reflect the higher severity in risk, symptoms, and CAD in the latter group, thus encouraging direct PCI in the majority of patients. In addition, the extensive use of FFR testing in patients being referred to ICA based on the FFRCT result may reflect an FFRCT diagnostic uncertainty among the interventionalists, thus prompting “reassurance” FFR measurements. This scenario is supported by the observation that FFR guidance in this group was reduced by 30% during the study period. Finally, in accordance with previous trials and real-world observations (3–5,26), the lower revascularization rate in this group may at least partially reflect the more extensive use of FFR testing per se. It can be speculated that the implementation of FFRCT and the extensive use of FFR in patients referred to ICA after FFRCT testing may stimulate the overall institutional adoption of FFR. However, in this study, FFR guidance in patients being referred directly from coronary CTA to ICA was relatively infrequent throughout the study period. We cannot exclude that the use of uncaptured PCI guidance procedures such as intravascular ultrasound and optical coherence tomography may have influenced the PCI operators’ overall threshold for FFR testing (26).
This study was a single-center, observational trial with inherent limitations. Data were collected in a nonselected cohort of patients and involved multiple CT and PCI operators unblinded to the FFRCT results. However, it included all patients encountered in a defined time period and represents consecutive data from a relevant study cohort in a real-world setting. FFRCT–FFR agreement estimates were based on few observations and most likely were influenced by selection bias. However, measurement of FFR in more low- or high-grade lesions may have resulted in higher agreement estimates (27). No current guidelines provide recommendations about the clinical use of FFRCT testing and its interpretation. Thus, in this study, FFRCT testing was recommended in patients with intermediate-range lesions in whom coronary CTA interpretation is most challenging and for whom guidelines recommend performance of additional ischemia testing (1). We did not consider the reduction in angina. However, during 12 months of follow-up, the rate of symptom-driven ICA was low. No information on the downstream primary physician contact was obtainable, leaving questions about changes in medication unanswered. The implementation of a “standard diagnostic testing” comparison group and a longer follow-up period would have added valuable information. Patients with emergent chest pain were not included in this study. FFRCT testing requires offsite computer processing, with a current 3- to 6-h turnaround time. A reduced order model for on-site (<1 h) CT scan–based derivation of FFR, which may be relevant in the acute setting, was recently introduced (28). However, further investigations in prospective multicenter trials are needed to determine the diagnostic performance of this technique.
FFRCT testing is feasible in real-world patients with intermediate-range coronary stenosis determined by coronary CTA. Implementation of FFRCT for clinical decision-making may influence the downstream diagnostic workflow of patients. Patients with FFRCT >0.80 being deferred from ICA have a favorable short-term prognosis. We propose an “FFRCT guidance algorithm” to be evaluated in future studies.
COMPETENCY IN MEDICAL KNOWLEDGE: FFRCT is feasible in real-world symptomatic patients with intermediate-range coronary stenosis. Real-world implementation of FFRCT for clinical decision-making may influence the downstream diagnostic workflow of patients. Patients being deferred from ICA based on an FFRCT value >0.80 have a favorable short-term prognosis.
TRANSLATIONAL OUTLOOK: FFRCT is a promising gatekeeper to the catheterization laboratory in patients with intermediate-range coronary stenosis. In this scenario, future studies are needed to evaluate the safety and cost-efficiency of FFRCT relative to conventional noninvasive ischemia-testing modalities. We propose an “FFRCT guidance algorithm” to be evaluated in future studies.
Drs. Nørgaard and Hansson have received unrestricted research grants from Edwards Lifesciences. Dr. Leipsic has received speakers honorarium from GE Healthcare and is a consultant for Edwards Lifesciences and HeartFlow. Dr. Grove has no conflicts related to this study but reports the following general conflicts: he has received speaker honoraria from AstraZeneca, Baxter, Bayer, Boehringer Ingelheim, and Pfizer; and has participated in advisory board meetings for AstraZeneca, Bayer, Boehringer Ingelheim, and Bristol-Myers Squibb. Dr. Christiansen has received unrestricted research grants from St. Jude Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- coronary artery bypass grafting
- coronary artery disease
- computed tomography angiography
- fractional flow reserve
- coronary computed tomography angiography–derived fractional flow reserve
- invasive coronary angiography
- instantaneous wave-free ratio
- left anterior descending artery
- left circumflex artery
- myocardial perfusion imaging
- right coronary artery
- Received September 11, 2015.
- Revision received November 4, 2015.
- Accepted November 10, 2015.
- American College of Cardiology Foundation
- Montalescot G.,
- Sechtem U.,
- Achenbach S.,
- et al.
- Johnson N.P.,
- Tóth G.G.,
- Lai D.,
- et al.
- Pijls N.H.,
- van Schaardenburgh P.,
- Manoharan G.,
- et al.
- Wijns W.,
- Kolh P.,
- Danchin N.,
- et al.
- Levine G.N.,
- Bates E.R.,
- Blankenship J.C.,
- et al.
- Meijboom W.B.,
- Van Mieghem C.A.,
- van Pelt N.,
- et al.
- Koo B.K.,
- Erglis A.,
- Doh J.H.,
- et al.
- Nørgaard B.L.,
- Leipsic J.,
- Gaur S.,
- et al.
- Hecht H.S.
- Schoenhagen P.,
- Desai M.Y.
- Nielsen L.H.,
- Nørgaard B.L.,
- Tilsted H.H.,
- et al.
- Sen S.,
- Escaned J.,
- Malik I.S.,
- et al.
- Einstein A.J.,
- Moser K.W.,
- Thompson R.C.,
- Cerqueira M.D.,
- Henzlova M.J.
- Senthamizhchelvan S.,
- Bravo P.E.,
- Esaias C.,
- et al.
- Nørgaard B.L.,
- Gaur S.,
- Leipsic J.,
- et al.
- Douglas P.S.,
- Pontone G.,
- Hlatky M.A.,
- et al.,
- PLATFORM Investigators
- Petraco R.,
- Sen S.,
- Nijjer S.,
- et al.
- Park S.J.,
- Ahn J.M.,
- Park G.M.,
- et al.
- Martus P.,
- Schueler S.,
- Dewey M.