Advertisement

Trans-lesional fractional flow reserve gradient as derived from coronary CT improves patient management: ADVANCE registry

Open AccessPublished:September 01, 2021DOI:https://doi.org/10.1016/j.jcct.2021.08.003

      Abstract

      Background

      The role of change in fractional flow reserve derived from CT (FFRCT) across coronary stenoses (ΔFFRCT) in guiding downstream testing in patients with stable coronary artery disease (CAD) is unknown.

      Objectives

      To investigate the incremental value of ΔFFRCT in predicting early revascularization and improving efficiency of catheter laboratory utilization.

      Materials

      Patients with CAD on coronary CT angiography (CCTA) were enrolled in an international multicenter registry. Stenosis severity was assessed as per CAD-Reporting and Data System (CAD-RADS), and lesion-specific FFRCT was measured 2 ​cm distal to stenosis. ΔFFRCT was manually measured as the difference of FFRCT across visible stenosis.

      Results

      Of 4730 patients (66 ​± ​10 years; 34% female), 42.7% underwent ICA and 24.7% underwent early revascularization. ΔFFRCT remained an independent predictor for early revascularization (odds ratio per 0.05 increase [95% confidence interval], 1.31 [1.26–1.35]; p ​< ​0.001) after adjusting for risk factors, stenosis features, and lesion-specific FFRCT. Among the 3 models (model 1: risk factors ​+ ​stenosis type and location ​+ ​CAD-RADS; model 2: model 1 ​+ ​FFRCT; model 3: model 2 ​+ ​ΔFFRCT), model 3 improved discrimination compared to model 2 (area under the curve, 0.87 [0.86–0.88] vs 0.85 [0.84–0.86]; p ​< ​0.001), with the greatest incremental value for FFRCT 0.71–0.80. ΔFFRCT of 0.13 was the optimal cut-off as determined by the Youden index. In patients with CAD-RADS ≥3 and lesion-specific FFRCT ≤0.8, a diagnostic strategy incorporating ΔFFRCT >0.13, would potentially reduce ICA by 32.2% (1638–1110, p ​< ​0.001) and improve the revascularization to ICA ratio from 65.2% to 73.1%.

      Conclusions

      ΔFFRCT improves the discrimination of patients who underwent early revascularization compared to a standard diagnostic strategy of CCTA with FFRCT, particularly for those with FFRCT 0.71–0.80. ΔFFRCT has the potential to aid decision-making for ICA referral and improve efficiency of catheter laboratory utilization.

      Keywords

      Abbreviations

      ADVANCE
      assessing diagnostic value of non-invasive FFRCT in coronary care
      AUC
      area under the receiver operating characteristic curve
      CABG
      coronary artery bypass grafting
      CAD
      coronary artery disease
      CAD-RADS
      Coronary Artery Disease - Reporting and Data System
      CCTA
      coronary computed tomography angiographZ
      FFR
      fractional flow reserve
      FFRCT
      fractional flow reserve derived from computed tomography
      ICA
      Invasive coronary angiography; PCI ​= ​percutaneous coronary intervention

      1. Introduction

      Physiological assessment with fractional flow reserve (FFR) guides the revascularization in patients with stable coronary artery disease (CAD).
      • Fihn S.D.
      • Gardin J.M.
      • Abrams J.
      • et al.
      ACCF/AHA/ACP/AATS/PCNA/SCAI/STS guideline for the diagnosis and management of patients with stable ischemic heart disease: a report of the American college of cardiology foundation/American heart association task force on practice guidelines, and the American college of physicians, American association for thoracic surgery, preventive cardiovascular nurses association, society for cardiovascular angiography and interventions, and society of thoracic surgeons.
      ,
      • Neumann F.-J.
      • Sousa-Uva M.
      • Ahlsson A.
      • et al.
      ESC/EACTS Guidelines on myocardial revascularization.
      The application of computational fluid dynamics to a standard coronary computed tomography angiography (CCTA) enables non-invasive FFR measurement (FFRCT) without additional imaging, medications, radiation exposure, or hospital visits.
      • Taylor C.A.
      • Fonte T.A.
      • Min J.K.
      Computational fluid dynamics applied to cardiac computed tomography for noninvasive quantification of fractional flow reserve.
      Numerous studies have demonstrated the diagnostic performance,
      • Nørgaard B.L.
      • Leipsic J.
      • Gaur S.
      • et al.
      Diagnostic performance of noninvasive fractional flow reserve derived from coronary computed tomography angiography in suspected coronary artery disease.
      ,
      • Driessen R.S.
      • Danad I.
      • Stuijfzand W.J.
      • et al.
      Comparison of coronary computed tomography angiography, fractional flow reserve, and perfusion imaging for ischemia diagnosis.
      prognostic value,
      • Ihdayhid A.R.
      • Nørgaard B.L.
      • Leipsic J.
      • et al.
      Prognostic value and risk continuum of noninvasive fractional flow reserve derived from coronary CT angiography.
      ,
      • Patel M.R.
      • Nørgaard B.L.
      • Fairbairn T.A.
      • et al.
      1-Year impact on medical practice and clinical outcomes of FFRCT.
      and clinical utility of FFRCT in real-world practice.
      • Patel M.R.
      • Nørgaard B.L.
      • Fairbairn T.A.
      • et al.
      1-Year impact on medical practice and clinical outcomes of FFRCT.
      • Fairbairn T.A.
      • Nieman K.
      • Akasaka T.
      • et al.
      Real-world clinical utility and impact on clinical decision-making of coronary computed tomography angiography-derived fractional flow reserve: lessons from the ADVANCE Registry.

      Nørgaard BL, Mortensen MB, Parner E, et al. Clinical outcomes following real-world computed tomography angiography-derived fractional flow reserve testing in chronic coronary syndrome patients with calcification. Eur Heart J - cardiovasc Imaging. Published online August 13, 2020:jeaa173. doi:10.1093/ehjci/jeaa173.

      FFRCT is derived along the epicardial coronary tree. This allows for a flexible and lesion-specific approach that goes beyond the standard assessment which focuses on whether vessel specific FFRCT falls below a specific cut-off. The change in FFRCT values across a stenosis (ΔFFRCT) represents an estimate of lesion-specific pressure loss and have been shown to discriminate a more focal phenotype of physiology and identify high risk plaques.
      • Lee J.M.
      • Choi G.
      • Koo B.-K.
      • et al.
      Identification of high-risk plaques destined to cause Acute coronary syndrome using coronary computed tomographic angiography and computational fluid dynamics.
      ,
      • Collet C.
      • Sonck Jeroen
      • Vandeloo Bert
      • et al.
      Measurement of hyperemic pullback pressure gradients to characterize patterns of coronary atherosclerosis.
      The ADVANCE (Assessing Diagnostic Value of Non-invasive FFRCT in Coronary Care) registry (NCT02499679) is an international multicenter prospective registry that enrolled stable patients with CAD who were investigated with CCTA and FFRCT.
      • Patel M.R.
      • Nørgaard B.L.
      • Fairbairn T.A.
      • et al.
      1-Year impact on medical practice and clinical outcomes of FFRCT.
      ,
      • Fairbairn T.A.
      • Nieman K.
      • Akasaka T.
      • et al.
      Real-world clinical utility and impact on clinical decision-making of coronary computed tomography angiography-derived fractional flow reserve: lessons from the ADVANCE Registry.
      More than half of patients who underwent ICA in gray-zone FFRCT value did not receive subsequent revascularization in the ADVANCE registry. There could be space to utilize FFRCT beyond the standard measurement of FFRCT in terms of catheter laboratory utilization. In this analysis, we hypothesized that ΔFFRCT would improve the identification of those who required early revascularization and investigated the incremental value of ΔFFRCT at improving the efficiency of downstream invasive testing as assessed by the revascularization to invasive coronary angiography (ICA) ratio.

      2. Materials and methods

      2.1 Study design and population

      The design and outcomes of the ADVANCE registry have been described previously.
      • Patel M.R.
      • Nørgaard B.L.
      • Fairbairn T.A.
      • et al.
      1-Year impact on medical practice and clinical outcomes of FFRCT.
      ,
      • Fairbairn T.A.
      • Nieman K.
      • Akasaka T.
      • et al.
      Real-world clinical utility and impact on clinical decision-making of coronary computed tomography angiography-derived fractional flow reserve: lessons from the ADVANCE Registry.
      Patients being investigated for clinically suspected CAD with documented >30% stenosis on CCTA were prospectively enrolled at 38 sites in Europe, Japan, and North America from July 2015, to October 2017. Exclusion criteria were poor CCTA image quality, life expectancy <1-year, or an inability to comply with follow-up requirements. The decision to request an FFRCT analysis was independently determined by the clinician reporting the CCTA. All patients provided written informed consent following institutional review board review and approval. In this secondary analysis, patients not referred for FFRCT analysis or in whom FFRCT was unanalyzable or unavailable were excluded (Supplemental Figure 1).

      2.2 CCTA acquisition and interpretation

      CCTA was performed as per local practice and international guidelines.
      • Leipsic J.
      • Abbara S.
      • Achenbach S.
      • et al.
      SCCT guidelines for the interpretation and reporting of coronary CT angiography: a report of the Society of Cardiovascular Computed Tomography Guidelines Committee.
      ,
      • Abbara S.
      • Blanke P.
      • Maroules C.D.
      • et al.
      SCCT guidelines for the performance and acquisition of coronary computed tomographic angiography: a report of the society of Cardiovascular Computed Tomography Guidelines Committee.
      The sites investigators graded coronary stenosis severity as normal, 0%–29%, 30%–49%, 50%–69%, 70%–90%, >90%, occluded (100%). For this sub-analysis, the per-patient anatomical severity was classified according to the Coronary Artery Disease – Reporting and Data System (CAD-RADS™) (Supplemental Table 1).
      • Cury R.C.
      • Abbara S.
      • Achenbach S.
      • et al.
      CAD-RADSTM coronary artery disease – reporting and data system. An expert consensus document of the society of cardiovascular computed tomography (SCCT), the American college of Radiology (ACR) and the North American society for cardiovascular imaging (NASCI). Endorsed by the American college of cardiology.
      This evaluation did not include high-risk plaque findings in this study.

      2.3 FFRCT analysis and measurements

      The analysis was blindly performed at HeartFlow (Redwood, CA, United States). For all patients, 3-dimensional anatomic models of epicardial coronary arteries and aortic root were generated from CCTA images.
      In accordance with the expert consensus for interpretation of FFRCT,
      • Nørgaard B.L.
      • Fairbairn T.A.
      • Safian R.D.
      • et al.
      Coronary CT angiography-derived fractional flow reserve testing in patients with stable coronary artery disease: recommendations on interpretation and reporting.
      we obtained both lesion-specific FFRCT and ΔFFRCT for each coronary vessel using the patient-specific 3-dimensional FFRCT model. A central core laboratory (Duke Clinical Research Institute, Durham, NC, United States) blinded to clinical information reviewed all FFRCT. Lesion-specific FFRCT was measured at 2 ​cm distal to stenosis for each coronary artery.
      • Nørgaard B.L.
      • Fairbairn T.A.
      • Safian R.D.
      • et al.
      Coronary CT angiography-derived fractional flow reserve testing in patients with stable coronary artery disease: recommendations on interpretation and reporting.
      An FFRCT of ≤0.8 was defined as a positive value. Additional analyses, blinded to clinical information, were performed in our core laboratory (St. Paul's Hospital, Vancouver, BC, Canada), where we reviewed all FFRCT models and measured ΔFFRCT. The ΔFFRCT represents the change in FFRCT across a stenosis and was measured as the difference in FFRCT values proximal and distal to a stenosis. The proximal and distal reference points were both manually identified at the most adjacent points to visible stenosis on the 3-dimensional FFRCT model (Fig. 1).
      • Takagi H.
      • Ishikawa Y.
      • Orii M.
      • et al.
      Optimized interpretation of fractional flow reserve derived from computed tomography: comparison of three interpretation methods.
      The distance between the proximal and distal ΔFFRCT reference points was visually assessed to characterize the stenosis type (Fig. 1D–F):
      • Focal – length <1 coronary segment, assuming <39 ​mm
      • Diffuse – length >1 segment, assuming ≥40 ​mm
      Fig. 1
      Fig. 1Methodology for determining ​ΔFFRCT ​and stenosis type. First the presence of and extent of stenosis is visually determined by analysing the 3-dimensional (3D) model in multiple projections. Proximal and distal reference points are then markedon the 3D model at regions immediately adjacent to the stenosis at regions which appear free of luminal stenosis (B and C). The ​ΔFFRCT ​was defined as the difference of FFRCT ​values between these two points.The stenosis type for each ​ΔFFRCT ​measurement was visually categorized as ​focal ​or ​diffuse, ​based on lesion length visually assessed on the 3D coronary model as follows (D and E).
      In a case with diffuse stenosis, after placing proximal reference at the most adjacent to the visible stenosis, carefully looked along downstream coronary and placed the distal reference at the point with visually normal diameter being the most adjacent to stenosis. The reproducibility of ΔFFRCT was excellent (Supplemental document). The lesion location was determined according to the Society of Cardiovascular Computed Tomography guidelines.
      • Leipsic J.
      • Abbara S.
      • Achenbach S.
      • et al.
      SCCT guidelines for the interpretation and reporting of coronary CT angiography: a report of the Society of Cardiovascular Computed Tomography Guidelines Committee.
      Per-patient lesion-specific FFRCT was recorded as the lowest lesion-specific FFRCT in major epicardial coronary arteries, and ΔFFRCT associated with the minimum lesion-specific FFRCT was deemed per-patient ΔFFRCT.

      2.4 Patient management and clinical end points

      The site investigator and the institution's heart team reviewed clinical data and interpreted all available diagnostic tests, including CCTA and FFRCT. The clinical management decisions including revascularization or medical therapy, entirely rested with the site physician and heart team.
      • Chinnaiyan K.M.
      • Akasaka T.
      • Amano T.
      • et al.
      Rationale, design and goals of the HeartFlow assessing diagnostic value of non-invasive FFRCT in Coronary Care (ADVANCE) registry.
      Early revascularization was defined as percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG) performed within 90 days after enrollment.
      • Fairbairn T.A.
      • Nieman K.
      • Akasaka T.
      • et al.
      Real-world clinical utility and impact on clinical decision-making of coronary computed tomography angiography-derived fractional flow reserve: lessons from the ADVANCE Registry.
      ,
      • Chinnaiyan K.M.
      • Akasaka T.
      • Amano T.
      • et al.
      Rationale, design and goals of the HeartFlow assessing diagnostic value of non-invasive FFRCT in Coronary Care (ADVANCE) registry.
      Patients who did not undergo early revascularization were deemed to have undergone medical therapy alone.
      The primary endpoint of this study was the early revascularization. The secondary endpoints were the number of ICA and the ratio of early revascularization.

      2.5 Statistical analysis

      Descriptive statistics were presented as mean ​± ​standard deviation for continuous variables and raw number (percentages) for categorical variables. Independent variables were compared using unpaired t or Fisher's exact test as appropriate. Multivariable logistic regression analysis was conducted to assess the association between ΔFFRCT and early revascularization. The multivariable adjustment was performed for clinical risk factors (age, sex, symptom status, hypertension, diabetes, hyperlipidemia, and current smoking), CAD-RADS, lesion-specific FFRCT, lesion location, and stenosis type, and the interaction between lesion-specific FFRCT and ΔFFRCT. Heterogenicity of the relationship between ΔFFRCT and early revascularization was assessed according to subgroups including symptom, CAD-RADS, lesion-specific FFRCT, stenosis location, and stenosis type. Three models were created to assess the incremental value of ΔFFRCT to a standard CCTA with FFRCT strategy: model 1, risk factors ​+ ​CAD-RADS ​+ ​stenosis type and location; model 2, model 1 ​+ ​lesion-specific FFRCT; and model 3, model 2 ​+ ​ΔFFRCT. The area under the curve (AUC) was compared using DeLong's test.
      • DeLong E.R.
      • DeLong D.M.
      • Clarke-Pearson D.L.
      Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach.
      Heterogenicity of the incremental value was assessed according to CAD-RADS and lesion-specific FFRCT severity. A 2-sided p-value of <0.05 was considered statistically significant in all tests. Computation was performed using JMP PRO version 14 (SAS Institute Inc., Cary, NC, USA) or R version 4.1 (R Foundation, Vienna, Austria).

      2.6 Simulation of efficacy of ICA referral

      We conducted an ICA referral simulation to assess the impact of ΔFFRCT on the efficiency of catheter laboratory utilization. We randomly selected 2839 (60.0%) patients for determining the cut-off value of ΔFFRCT according to the Youden index and validated the cut-off value with the remaining patients. This analysis allows for the greatest extent of confirmation possible without a separate cohort. Subsequently, we simulated referral for ICA according to three potential strategies: Anatomical, ICA referral for patients with CAD-RADS ≥3; Lesion-specific FFRCT, CAD-RADS ≥3 and lesion-specific FFRCT ≤0.80; and ΔFFRCT, CAD-RADS ≥3, lesion-specific FFRCT ≤0.80, and ΔFFRCT ​> ​cut-off value. To account for other clinical factors related to a decision for early revascularization, we applied this simulation to patients who underwent ICA, meaning that more likely to undergoing ICA. For each of these strategies, the potential impact of ΔFFRCT at reducing the number of ICA and improving the ratio of subsequent revascularization was assessed.

      3. Results

      3.1 Patient characteristics

      Of the 5083 patients enrolled in the registry, FFRCT analysis was requested in 4893 (96.2%). FFRCT analysis was feasible in 4737 (93.2%) and accessible for this sub-analysis in 4730 (93.1%) (Supplemental Figure 1). A total of 2092 (42.7%) patients underwent ICA within 90 days, with 1168 (24.7%) patients requiring early revascularization (PCI: 1017 [87.1%]; CABG 151 [22.9%]). Patients who underwent revascularization were more likely to be male and to have typical angina, hypertension, diabetes mellites, hyperlipidemia, and active smoking (Table 1).
      Table 1Patient characteristics.
      VariablesTotal (n ​= ​4730)Medication (n ​= ​3562)Revascularization (n ​= ​1168)P-value
      Demographics
      Age, yr66 ​± ​1066 ​± ​1066 ​± ​100.678
      Female sex, n (%)1602 (34%)1285 (36%)317 (27%)<0.001
      Body mass index, kg/m226 ​± ​526 ​± ​526 ​± ​40.446
      Previous coronary stenting, n (%)159 (4%)126 (4%)33 (3%)0.389
      Angina status, n (%)<0.001
      Typical1024 (22%)586 (17%)438 (38%)
      Atypical1724 (36%)1381 (39%)343 (29%)
      Dyspnea472 (10%)375 (11%)97 (8%)
      Non-cardia Pain296 (6%)245 (7%)51 (4%)
      None1162 (25%)928 (26%)234 (20%)
      Risk factors
      Hypertension, n (%)2831 (60%)2091 (59%)740 (63%)0.017
      Diabetes mellites, n (%)1034 (22%)719 (20%)315 (27%)<0.001
      Hyperlipidemia, n (%)2749 (58%)1999 (56%)750 (64%)<0.001
      Current smoker, n (%)797 (17%)560 (16%)237 (20%)<0.001
      Note. — data are presented as mean ​± ​standard deviation or percentages with raw data in parenthesis. Contentious and categorical variables were compared among groups using the unpaired t-test and Fisher's exact test.

      3.2 Relationship of CAD severity with actual treatment

      Table 2 summarizes anatomical and physiological CAD characteristics. Patients with early revascularization showed higher CAD-RADS grading, as well lower lesion-specific FFRCT and larger ΔFFRCT. A larger ΔFFRCT was observed with increasing stenosis severity (Supplemental Figure 2A); further, a larger ΔFFRCT was observed in patients with early revascularization across each anatomical severity (Fig. 2A). A larger ΔFFRCT was associated with lower lesion-specific FFRCT (Supplemental Figure 2B); furthermore, a larger ΔFFRCT was observed in patients requiring early revascularization across each group stratified by 0.05 increments in lesion-specific FFRCT (Fig. 2B).
      Table 2Coronary artery disease extent.
      Total (n ​= ​4730)Medications (n ​= ​3562)Revascularization (n ​= ​1168)P-value
      Anatomical severity
      CAD-RADS, n (%)<0.001
      ≤21261 (27%)1227 (34%)34 (3%)
      31773 (38%)1497 (42%)276 (24%)
      ≥41696 (36%)838 (24%)858 (73%)
      3-vessel >70% disease136 (3%)44 (1%)92 (8%)<0.001
      Left main ≥50% disease163 (3%)85 (2%)78 (7%)<0.001
      FFRCT findings
      Minimum lesion-specific FFRCT
      Data are mean ​± ​standard deviation.
      0.74 ​± ​0.120.77 ​± ​0.100.63 ​± ​0.11<0.001
      Minimum lesion-specific FFRCT, n (%)<0.001
      >0.801588 (34%)1518 (43%)70 (6%)
      0.71–0.801615 (34%)1340 (38%)275 (24%)
      ≤0.701527 (32%)704 (19%)823 (70%)
      ΔFFRCT
      Data are mean ​± ​standard deviation.
      0.13 ​± ​0.120.10 ​± ​0.090.24 ​± ​0.15<0.001
      Lesion location<0.001
      Left main798 (17%)669 (19%)129 (11%)
      Proximal1618 (34%)1166 (32%)452 (39%)
      Mid1430 (30%)1033 (29%)397 (34%)
      Distal587 (12%)460 (13%)127 (11%)
      Branch297 (6%)234 (7%)63 (5%)
      Stenosis type<0.001
      Focal4260 (90%)3271 (92%)989 (85%)
      Diffuse470 (10%)291 (8%)179 (15%)
      Note. — data are percentages, with raw data in parenthesis, otherwise noted. Contentious and categorical variables were compared among groups using the unpaired t-test and Fisher's exact test. CAD-RADS ​= ​coronary artery disease reporting ad data system.
      a Data are mean ​± ​standard deviation.
      Fig. 2
      Fig. 2Relationship of ΔFFRCT with CAD-RADS (A) and lesion-specific FFRCT (B). ΔFFRCT was compared between patients with vs. without early revascularization in CAD-RADS (A) and FFRCT category (B).
      Early revascularization was associated with a larger ΔFFRCT as compared to patients treated medically (0.24 ​± ​0.15 vs. 0.10 ​± ​0.09; p ​< ​0.001). With increasing ΔFFRCT, patients were more likely to undergo ICA and revascularization and were associated with an increase in the revascularization to ICA ratio (Fig. 3). The revascularization rate in patients with CAD-RADS 3 and ​≥ ​4 was 15.6% (276/1773) and 50.6% (858/1696), respectively. Patients with a lesion-specific FFRCT of >0.80, 0.71–0.80, and ≤0.70 underwent revascularization at a rate of 4.4% (70/1588), 17.0% (275/1615), and 53.9% (823/1527), respectively.
      Fig. 3
      Fig. 3Relationship of ΔFFRCT with actual treatment at 90 days (A) and ICA results (B). Panel A shows actual treatment, including medications alone, percutaneous coronary intervention (PCI), or coronary artery bypass grafting (CABG) stratified by 0.05 ΔFFRCT increments. Panel B shows the ratio of ICA with or without revascularization and the ratio of revascularization to ICA stratified by 0.05 ΔFFRCT increments.

      3.3 ΔFFRCT as an independent predictor for early revascularization

      ΔFFRCT remained an independent predictor for early revascularization after adjusting for age, sex, hypertension, hyperlipidemia, diabetes mellites, angina status, CAD-RADS, stenosis type and location, and FFRCT (Table 3). The adjusted odds ratio for early revascularization per 0.05-unit increase in ΔFFRCT is illustrated in Fig. 4A. After adjusting for confounders, each 0.05 increase in ΔFFRCT was independently associated with a greater incidence of early revascularization. Although the predictive value of ΔFFRCT was demonstrated across various subgroups, there was heterogeneity: ΔFFRCT was more predictive for early revascularization in patients with CAD-RADS ≤3, FFRCT 0.71–0.80, or focal and tubular lesions as compared to those with CAD-RADS 4, FFRCT <0.7, or diffuse disease, respectively (Supplemental Figure 3).
      Table 3Multivariable logistic regression analysis for revascularization at 90-day follow-up.
      PredictorsOR (95% CIs)P-value
      Patient characteristics
      Age ≥65 ​yr.0.83 (0.70–1.00)0.044
      Female sex0.79 (0.65–0.95)0.015
      Hyperlipidemia1.40 (1.17–1.68)<0.001
      Diabetes0.98 (0.80–1.19)0.811
      Hypertension1.00 (0.83–1.20)0.999
      Current smoker1.07 (0.86–1.34)0.526
      Angina status (vs. asymptomatic)
      Typical2.32 (1.83–2.94)<0.001
      Atypical1.33 (1.06–1.67)0.014
      Non-cardiac1.19 (0.78–1.80)0.417
      Dyspnea1.36 (0.98–1.89)0.067
      Imaging findings
      CAD-RADS (ref. ≤2)
      34.28 (2.89–6.341)<0.001
      ≥414.79 (10.04–21.79)<0.001
      FFRCT ≤0.80 (vs. distal)2.40 (1.49–3.85)<0.001
      ΔFFRCT (per 0.05 increase)1.47 (1.27–1.70)<0.001
      Lesion location (vs. distal)
      LM2.09 (1.46–2.99)<0.001
      Proximal1.81 (1.35–2.42)<0.001
      Mid1.63 (1.21–2.18)0.003
      Branch1.48 (0.96–2.27)0.075
      Stenosis type (vs. diffuse)
      Focal0.82 (0.63–1.07)0.1433
      Note. — OR ​= ​odds ratio; CAD-RADS ​= ​Coronary Artery Disease - Reporting and Data System.
      Fig. 4
      Fig. 4Multivariable logistic regression analysis for predicting early revascularization. Panel A shows odds ratio (solid line) with 95% confidence interval (dotted line) of ΔFFRCT compared with ΔFFRCT of 0.00–0.04 after adjusting risk factors, CAD-RADS, stenosis type and location, and FFRCT.Panel B shows receiver operating characteristic curves for three logistic models for early revascularization: model 1 ​= ​risk factors, CAD-RADS, stenosis type and location; model 2 ​= ​model 1 ​+ ​FFRCT; and model 3 ​= ​model 2 ​+ ​ΔFFRCT. Model 2 demonstrated higher AUC as compared to model 1 (AUC difference with 95% CI, 0.02 [0.02–0.03], p ​< ​0.001). Model 3 demonstrated the highest AUC and was superior to model 1 (0.05 [0.04–0.05], p ​< ​0.001) and model 2 (0.02 [0.02–0.03], p ​< ​0.001).AUC ​= ​area under the curve; CI ​= ​confidence interval.

      3.4 Incremental value of ΔFFRCT

      Receiver operating characteristic curves and the AUC of 3 logistic models for early revascularization are given in Fig. 4B. Model 2 showed higher AUC compared to model 1 (0.82 [0.81–0.83] vs. 0.85 [0.84–0.86], p ​< ​0.001). Model 3 showed a higher AUC compared to model 2 (0.85 [0.84–0.86] vs. 0.87 [0.86–0.88]), p ​< ​0.001), indicating that ΔFFRCT had incremental value to model 2 for predicting early revascularization. The incremental value of ΔFFRCT was observed across CAD-RADS severities (Supplemental Figure 4A). Heterogenicity of the incremental value was observed according to lesion-specific FFRCT. AUC improvement was observed in patients with gray-zone lesion-specific FFRCT of 0.71–0.80, with no difference in those with FFRCT ≤0.70 or FFRCT >0.80 (Supplemental Figure 4B).

      3.5 ΔFFRCT impact on catheter laboratory utilization

      A ΔFFRCT of 0.13 was the optimal cut-off for predicting revascularization (Supplemental Figure 5), and we applied this cut-off value to the ICA referral simulation. Actual ICA results and simulated number of ICA and the ratio of subsequent revascularization for each strategy are given in Fig. 5. Although the number of ICA was decreased and ratio of revascularization was increased as compared to actual results, the anatomical strategy demonstrated the highest referral for ICA and lowest revascularization ratio among 3 strategies. The Lesion-specific FFRCT strategy demonstrated a lower number of ICA and higher revascularization ratio as compared to the anatomical strategy. The ΔFFRCT demonstrated the lowest referrals for ICA and the highest revascularization ratio; potentially reducing ICA by 32.2% (1638–1110, p ​< ​0.001), and improving the revascularization to ICA ratio from 65.2% [1068/1638] to 73.1% [811/1110] as compared to the lesion-specific FFRCT strategy (Fig. 5). Applying a ΔFFRCT strategy, the largest improvement in revascularization to ICA ratio was observed in patients with lesion-specific FFRCT between 0.71 and 0.80 (from 43.7% [275/629] to 60.3% [143/237]) as compared to a small improvement in those with an FFRCT of ≤0.70 (from 72.6% [823/1134] to 76.5% [668/873]) (Supplemental Table 2).
      Fig. 5
      Fig. 5Efficiency of catheter laboratory based on a referral strategy of the 2092 patients who underwent ICA in the ADVANCE registry. Modeling was performed to simulate the efficiency of catheter laboratory utilization according to three ICA referral strategies: anatomical strategy based solely on anatomical findings (CAD-RADS ≥3), lesion-specific FFRCT strategy that also included lesion-specific FFRCT ≤0.80 and a ΔFFRCT strategy that incorporated ΔFFRCT >0.13 in addition to CAD-RADS ≥3 and lesion-specific FFRCT ≤0.80. Shown are the number of ICA referrals with (green) or without actual revascularization (red) and the ratio of revascularization to ICA according to ICA referral strategy. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

      4. Discussion

      This analysis of the ADVANCE registry investigated the utility of ΔFFRCT at predicting early revascularization and discriminating patients with higher revascularization to ICA ratio; both of which may improve efficiency of care of patients with CAD. The main findings of this investigation are as follows: 1) ΔFFRCT values represent a continuum with larger values independently associated with early revascularization, 2) ΔFFRCT demonstrated incremental value at predicting early revascularization compared to a standard strategy of CCTA with lesion-specific FFRCT, with the greatest benefit in patients with gray-zone FFRCT values between 0.71 and 0.80, and 3) incorporating ΔFFRCT in addition to standard CCTA and lesion-specific FFRCT diagnostic strategy may reduce the number of ICA and improve the ratio of subsequent revascularization.
      While there is increasing evidence supporting the use of FFRCT to improve the efficiency of catheter laboratory utilization,
      • Lu M.T.
      • Ferencik M.
      • Roberts R.S.
      • et al.
      Noninvasive FFR derived from coronary CT angiography.
      • Nørgaard B.L.
      • Terkelsen C.J.
      • Mathiassen O.N.
      • et al.
      Clinical outcomes using coronary CT angiography and FFRCT-guided management of stable chest pain patients.
      • Douglas P.S.
      • De Bruyne B.
      • Pontone G.
      • et al.
      1-Year outcomes of FFRCT-guided care in patients with suspected coronary disease.
      the results of the ADVANCE registry highlight some of the real-world clinical challenges of interpreting FFRCT and guiding downstream decision making. In the ADVANCE registry, 72.3% of patients undergoing ICA with lesion-specific FFRCT of ≤0.80 underwent revascularization.
      • Fairbairn T.A.
      • Nieman K.
      • Akasaka T.
      • et al.
      Real-world clinical utility and impact on clinical decision-making of coronary computed tomography angiography-derived fractional flow reserve: lessons from the ADVANCE Registry.
      However, several patients were recommended for medications alone even with positive lesion-specific FFRCT results (<0.80), and some underwent ICA even with negative lesion-specific FFRCT results (0.80), highlighting that there is a space for interpreting the FFRCT results beyond the lesion-specific FFRCT. In particular, among patients with lesion-specific FFRCT between 0.71 and 0.80 and who underwent ICA, 56.3% did not subsequently undergo early revascularization. The results of this sub-analysis highlight that ΔFFRCT may improve physician decision-making in identifying patients who require revascularization, particularly those with gray-zone lesion-specific FFRCT values between 0.71 and 0.80. The results of the ISCHEMIA (International Study of Comparative Health Effectiveness with Medical and Invasive Approaches) trial demonstrated that a routine invasive approach does not provide prognostic benefit compared to medical therapy alone.

      Maron DJ, Hochman JS, Reynolds HR, et al. Initial invasive or conservative strategy for stable coronary disease. N Engl J Med. Published online March 30, 2020:NEJMoa1915922. doi:10.1056/NEJMoa1915922.

      Accordingly, there is a renewed imperative to consider the risk and benefits of the different treatment options and improve the identification of lesoins that would benefit from revascularization. Given the concerns that a first-line CCTA strategy may result in over referral for ICA,
      • Foy A.J.
      • Dhruva S.S.
      • Peterson B.
      • Mandrola J.M.
      • Morgan D.J.
      • Redberg R.F.
      Coronary computed tomography angiography vs functional stress testing for patients with suspected coronary artery disease: a systematic review and meta-analysis.
      the ability of ΔFFRCT to identify lesions that require revascularization and potentially improve resource utilization is highly relevant and warrants further investigation with prospective studies.
      Our findings also highlight the potential value of non-invasively characterizing the physiological pattern of CAD. Standard lesion-specific FFRCT is affected by coronary atherosclerosis upstream of a measurement point, and the presence of coronary plaque causes FFRCT to decrease even without the obstructive disease.
      • Doris M.K.
      • Otaki Y.
      • Arnson Y.
      • et al.
      Non-invasive fractional flow reserve in vessels without severe obstructive stenosis is associated with coronary plaque burden.
      In contrast to lesion-specific FFRCT, ΔFFRCT represents a more stenosis-specific physiological severity and is not affected by coronary plaque beyond the stenosis. Our results suggest that adding ΔFFRCT to lesion-specific FFRCT may inform on disease severity and physiological phenotype. Recent invasive studies have provided similar results with an invasive FFR pullback able to characterize several physiological patterns of CAD.
      • Collet C.
      • Sonck Jeroen
      • Vandeloo Bert
      • et al.
      Measurement of hyperemic pullback pressure gradients to characterize patterns of coronary atherosclerosis.
      ,
      • De Bruyne B.
      • Hersbach F.
      • Pijls N.H.J.
      • et al.
      Abnormal epicardial coronary resistance in patients with diffuse atherosclerosis but “normal” coronary angiography.
      ,

      Lee SH, Shin D, Lee JM, et al. Automated algorithm using pre-intervention fractional flow reserve pullback curve to predict post-intervention physiological results. JACC Cardiovasc Interv. Published online October 2020:S1936879820314795. doi:10.1016/j.jcin.2020.06.062.

      A high ΔFFRCT provides an opportunity to identify subjects with a “focal phenotype” of physiology as described by Collet et al.
      • Collet C.
      • Sonck Jeroen
      • Vandeloo Bert
      • et al.
      Measurement of hyperemic pullback pressure gradients to characterize patterns of coronary atherosclerosis.
      Despite the clinical benefit observed with FFR-guided PCI,
      • Zhang D.
      • Lv S.
      • Song X.
      • et al.
      Fractional flow reserve versus angiography for guiding percutaneous coronary intervention: a meta-analysis.
      one-third of patients experience suboptimal post-PCI results, associated with major adverse cardiac events.
      • Choi K.H.
      • Lee J.M.
      • Koo B.-K.
      • et al.
      Prognostic implication of functional incomplete revascularization and residual functional SYNTAX score in patients with coronary artery disease.
      ,
      • Piroth Z.
      • Toth G.G.
      • Tonino P.A.L.
      • et al.
      Prognostic value of fractional flow reserve measured immediately after drug-eluting stent implantation.
      Therefore, there is an increasing emphasis on achieving a physiologically optimal result post-PCI. In cases with a large focal pressure gradient, PCI is more likely to achieve an ideal functional result and symptomatic benefit for the patient

      Lee SH, Shin D, Lee JM, et al. Automated algorithm using pre-intervention fractional flow reserve pullback curve to predict post-intervention physiological results. JACC Cardiovasc Interv. Published online October 2020:S1936879820314795. doi:10.1016/j.jcin.2020.06.062.

      ; on the contrary, revascularization in vessels with diffuse pressure loss is associated with limited FFR or symptomatic improvement and even potential harm.
      • Baranauskas A.
      • Peace A.
      • Kibarskis A.
      • et al.
      FFR result post PCI is suboptimal in long diffuse coronary artery disease.
      The current approach for reading FFRCT involves interpreting an FFRCT value at one point on the coronary tree, typically 20–30 ​mm distal to a stenosis (i.e. lesion-specific FFRCT).
      • Nørgaard B.L.
      • Fairbairn T.A.
      • Safian R.D.
      • et al.
      Coronary CT angiography-derived fractional flow reserve testing in patients with stable coronary artery disease: recommendations on interpretation and reporting.
      Although this provides insight into total pressure loss upstream of the coronary artery measurement point, this method is limited in its capacity to characterize the lesion specific physiological phenotype requiring revascularization.
      • Rønnow Sand N.P.
      • Nissen L.
      • Winther S.
      • et al.
      Prediction of coronary revascularization in stable Angina.
      Our results highlight that ΔFFRCT can provide clinically relevant insight into the physiological pattern of disease requiring revascularization. With the emerging use of CCTA and FFRCT to guide PCI,
      • Nagumo S.
      • Collet C.
      • Norgaard B.L.
      • et al.
      Rationale and design of the precise percutaneous coronary intervention plan ( P3 ) study: prospective evaluation of a virtual computed tomography-based percutaneous intervention planner.
      ΔFFRCT may provide further non-invasive guidance for optimizing revascularization strategies and outcomes.
      This study has several limitations. First, the findings related to ΔFFRCT are observational in nature with inherent physician bias for both ICA referral and decisions on revascularization. Second, the endpoints in this study were driven by revascularization. The optimal cut-off value for ΔFFRCT was not powered to evaluate cardiac death and myocardial infarction. However, this does not undermine the opportunity for ΔFFRCT to improve the efficiency of catheter laboratory utilization and adds to recent data from the EMERALD (Exploring the Mechanism of Plaque Rupture in Acute Coronary Syndrome Using Coronary CT Angiography and Computational Fluid Dynamics), demonstrating the prognostic utility of ΔFFRCT at identifying lesions potentially at risk of future myocardial infarction.
      • Lee J.M.
      • Choi G.
      • Koo B.-K.
      • et al.
      Identification of high-risk plaques destined to cause Acute coronary syndrome using coronary computed tomographic angiography and computational fluid dynamics.
      Third, revascularization was not guided by invasive FFR. However, the cut-off value derived from the early revascularization as a clinical endpoint was similar to one which compared invasive FFR ≤0.80 or not in previous study,
      • Takagi H.
      • Ishikawa Y.
      • Orii M.
      • et al.
      Optimized interpretation of fractional flow reserve derived from computed tomography: comparison of three interpretation methods.
      which may support that the decision for early revascularization was based on coronary physiology in the ADVANCE registry. Finally, the ICA referral simulation did not take into account symptoms and risk factors. The decision to revascularize is multi-factorial, and stenosis and FFRCT are just one part of many factors taken into consideration in the decision making process. Also, the estimates were theoretical, not observed ones. Therefore, further study is warranted. We applied the simulation to patients who underwent ICA to affect clinical factors other than stenosis and FFRCT severity. However, these confounding might not be fully adjusted.

      5. Conclusions

      In this analysis of the ADVANCE registry, ΔFFRCT improves the identification of patients who required early revascularization compared to a standard diagnostic strategy with CCTA and lesion-specific FFRCT, with the greatest incremental benefit in patients with gray-zone FFRCT values between 0.71 and 0.80. Applying a criterion of ΔFFRCT to ICA referral, efficiency of resource utilization may be improved. Prospective validation of these findings will be important to translate these findings into broader practice. Appendix A. Supplementary data.

      Declaration of competing interest

      This study was supported by HeartFlow, Inc., Redwood City, California, via individual Clinical Study Agreements with each enrolling institution and with the Duke Clinical Research Institute ( DCRI ) for Core Laboratory activities and Clinical Event Committee adjudication of adverse events. Dr. Leipsic receives institutional grants to provide core lab services to Edwards Life Sciences, Medtronic and is a consultant to Circle CVI and HeartFlow. Dr. Fairbairn is on the Speakers Bureau for HeartFlow. Dr. Nørgaard has received unrestricted institutional research grants from Siemens and HeartFlow. Dr. Berman has received unrestricted research support from HeartFlow. Dr. Chinnaiyan has received institutional grants from HeartFlow. Dr. Hurwitz-Koweek is on the Speakers Bureau for HeartFlow; and has unrestricted grant funding from Siemens and HeartFlow. Dr. Pontone is a consultant for GE Healthcare; and has research grants from GE Healthcare and HeartFlow. Dr. Rabbat has received institutional grants from HeartFlow. Dr. Mullen is an employee of HeartFlow. Dr. Rogers is an employee of and has equity in HeartFlow. Dr. Bax has received unrestricted research grants from Edwards Lifescience, Medtronic , Boston Scientific , Biotronik , and GE Healthcare; and is on the Speakers Bureau with Abbott. Dr. Douglas receives an institutional research grant from HeartFlow. Dr. Patel has received grants from HeartFlow, Jansen, Bayer, AstraZeneca, and NHLBI; and has served as a consultant for Jansen, Bayer, AstraZeneca, Genzyme, and Merck. Dr. Nieman reports institutional research support from Siemens Healthineers , Bayer , HeartFlow Inc. and is a consultant to Siemens Medical Solutions USA . Dr. Ihdayhid is supported by the National Health and Medical Research Council of Australia and National Heart Foundation Scholarships; and has received honoraria from Canon Medical and Boston Scientific . All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.

      Acknowledgments

      The authors would like to thank Ms. Whitney Huey and Ms. Amy Flynt for their contribution towards data collection and analysis.

      Appendix A. Supplementary data

      The following is the supplementary data to this article:

      Funding

      This study was supported by HeartFlow, Inc., Redwood City, CA, USA, via individual Clinical Study Agreements with each enrolling institution and with the Duke Clinical Research Institute (DCRI) for Core Laboratory activities and Clinical Event Committee adjudication of adverse events.

      References

        • Fihn S.D.
        • Gardin J.M.
        • Abrams J.
        • et al.
        ACCF/AHA/ACP/AATS/PCNA/SCAI/STS guideline for the diagnosis and management of patients with stable ischemic heart disease: a report of the American college of cardiology foundation/American heart association task force on practice guidelines, and the American college of physicians, American association for thoracic surgery, preventive cardiovascular nurses association, society for cardiovascular angiography and interventions, and society of thoracic surgeons.
        Circulation. 2012; 126 (2012): e354-e471https://doi.org/10.1161/CIR.0b013e318277d6a0
        • Neumann F.-J.
        • Sousa-Uva M.
        • Ahlsson A.
        • et al.
        ESC/EACTS Guidelines on myocardial revascularization.
        Eur Heart J. 2018; 40 (2019): 87-165https://doi.org/10.1093/eurheartj/ehy394
        • Taylor C.A.
        • Fonte T.A.
        • Min J.K.
        Computational fluid dynamics applied to cardiac computed tomography for noninvasive quantification of fractional flow reserve.
        J Am Coll Cardiol. 2013; 61: 2233-2241https://doi.org/10.1016/j.jacc.2012.11.083
        • Nørgaard B.L.
        • Leipsic J.
        • Gaur S.
        • et al.
        Diagnostic performance of noninvasive fractional flow reserve derived from coronary computed tomography angiography in suspected coronary artery disease.
        J Am Coll Cardiol. 2014; 63: 1145-1155https://doi.org/10.1016/j.jacc.2013.11.043
        • Driessen R.S.
        • Danad I.
        • Stuijfzand W.J.
        • et al.
        Comparison of coronary computed tomography angiography, fractional flow reserve, and perfusion imaging for ischemia diagnosis.
        J Am Coll Cardiol. 2019; 73: 161-173https://doi.org/10.1016/j.jacc.2018.10.056
        • Ihdayhid A.R.
        • Nørgaard B.L.
        • Leipsic J.
        • et al.
        Prognostic value and risk continuum of noninvasive fractional flow reserve derived from coronary CT angiography.
        Radiology. 2019; (Published online June 11)https://doi.org/10.1148/radiol.2019182264
        • Patel M.R.
        • Nørgaard B.L.
        • Fairbairn T.A.
        • et al.
        1-Year impact on medical practice and clinical outcomes of FFRCT.
        JACC Cardiovasc Imaging. 2020; 13: 97-105https://doi.org/10.1016/j.jcmg.2019.03.003
        • Fairbairn T.A.
        • Nieman K.
        • Akasaka T.
        • et al.
        Real-world clinical utility and impact on clinical decision-making of coronary computed tomography angiography-derived fractional flow reserve: lessons from the ADVANCE Registry.
        Eur Heart J. 2018; 39: 3701-3711https://doi.org/10.1093/eurheartj/ehy530
      1. Nørgaard BL, Mortensen MB, Parner E, et al. Clinical outcomes following real-world computed tomography angiography-derived fractional flow reserve testing in chronic coronary syndrome patients with calcification. Eur Heart J - cardiovasc Imaging. Published online August 13, 2020:jeaa173. doi:10.1093/ehjci/jeaa173.

        • Lee J.M.
        • Choi G.
        • Koo B.-K.
        • et al.
        Identification of high-risk plaques destined to cause Acute coronary syndrome using coronary computed tomographic angiography and computational fluid dynamics.
        JACC Cardiovasc Imaging. 2019; 12: 1032-1043https://doi.org/10.1016/j.jcmg.2018.01.023
        • Collet C.
        • Sonck Jeroen
        • Vandeloo Bert
        • et al.
        Measurement of hyperemic pullback pressure gradients to characterize patterns of coronary atherosclerosis.
        J Am Coll Cardiol. 2019; 74: 13
        • Leipsic J.
        • Abbara S.
        • Achenbach S.
        • et al.
        SCCT guidelines for the interpretation and reporting of coronary CT angiography: a report of the Society of Cardiovascular Computed Tomography Guidelines Committee.
        J Cardiovasc Comput Tomogr. 2014; 8: 342-358https://doi.org/10.1016/j.jcct.2014.07.003
        • Abbara S.
        • Blanke P.
        • Maroules C.D.
        • et al.
        SCCT guidelines for the performance and acquisition of coronary computed tomographic angiography: a report of the society of Cardiovascular Computed Tomography Guidelines Committee.
        J Cardiovasc Comput Tomogr. 2016; 10: 435-449https://doi.org/10.1016/j.jcct.2016.10.002
        • Cury R.C.
        • Abbara S.
        • Achenbach S.
        • et al.
        CAD-RADSTM coronary artery disease – reporting and data system. An expert consensus document of the society of cardiovascular computed tomography (SCCT), the American college of Radiology (ACR) and the North American society for cardiovascular imaging (NASCI). Endorsed by the American college of cardiology.
        J Cardiovasc Comput Tomogr. 2016; 10: 269-281https://doi.org/10.1016/j.jcct.2016.04.005
        • Nørgaard B.L.
        • Fairbairn T.A.
        • Safian R.D.
        • et al.
        Coronary CT angiography-derived fractional flow reserve testing in patients with stable coronary artery disease: recommendations on interpretation and reporting.
        Radiol Cardiothorac Imaging. 2019; 1e190050https://doi.org/10.1148/ryct.2019190050
        • Takagi H.
        • Ishikawa Y.
        • Orii M.
        • et al.
        Optimized interpretation of fractional flow reserve derived from computed tomography: comparison of three interpretation methods.
        J Cardiovasc Comput Tomogr. 2019; 2: 134-141https://doi.org/10.1016/j.jcct.2018.10.027
        • Chinnaiyan K.M.
        • Akasaka T.
        • Amano T.
        • et al.
        Rationale, design and goals of the HeartFlow assessing diagnostic value of non-invasive FFRCT in Coronary Care (ADVANCE) registry.
        J Cardiovasc Comput Tomogr. 2016; 11: 62-67https://doi.org/10.1016/j.jcct.2016.12.002
        • DeLong E.R.
        • DeLong D.M.
        • Clarke-Pearson D.L.
        Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach.
        Biometrics. 1988; 44: 837-845
        • Lu M.T.
        • Ferencik M.
        • Roberts R.S.
        • et al.
        Noninvasive FFR derived from coronary CT angiography.
        JACC Cardiovasc Imaging. 2017; 10: 1350-1358https://doi.org/10.1016/j.jcmg.2016.11.024
        • Nørgaard B.L.
        • Terkelsen C.J.
        • Mathiassen O.N.
        • et al.
        Clinical outcomes using coronary CT angiography and FFRCT-guided management of stable chest pain patients.
        J Am Coll Cardiol. Published online. August 2018; https://doi.org/10.1016/j.jacc.2018.07.043
        • Douglas P.S.
        • De Bruyne B.
        • Pontone G.
        • et al.
        1-Year outcomes of FFRCT-guided care in patients with suspected coronary disease.
        J Am Coll Cardiol. 2016; 68: 435-445https://doi.org/10.1016/j.jacc.2016.05.057
      2. Maron DJ, Hochman JS, Reynolds HR, et al. Initial invasive or conservative strategy for stable coronary disease. N Engl J Med. Published online March 30, 2020:NEJMoa1915922. doi:10.1056/NEJMoa1915922.

        • Foy A.J.
        • Dhruva S.S.
        • Peterson B.
        • Mandrola J.M.
        • Morgan D.J.
        • Redberg R.F.
        Coronary computed tomography angiography vs functional stress testing for patients with suspected coronary artery disease: a systematic review and meta-analysis.
        JAMA Intern Med. 2017; 177: 1623https://doi.org/10.1001/jamainternmed.2017.4772
        • Doris M.K.
        • Otaki Y.
        • Arnson Y.
        • et al.
        Non-invasive fractional flow reserve in vessels without severe obstructive stenosis is associated with coronary plaque burden.
        J Cardiovasc Comput Tomogr. 2018; 12: 379-384https://doi.org/10.1016/j.jcct.2018.05.003
        • De Bruyne B.
        • Hersbach F.
        • Pijls N.H.J.
        • et al.
        Abnormal epicardial coronary resistance in patients with diffuse atherosclerosis but “normal” coronary angiography.
        Circulation. 2001; 104: 2401-2406https://doi.org/10.1161/hc4501.099316
      3. Lee SH, Shin D, Lee JM, et al. Automated algorithm using pre-intervention fractional flow reserve pullback curve to predict post-intervention physiological results. JACC Cardiovasc Interv. Published online October 2020:S1936879820314795. doi:10.1016/j.jcin.2020.06.062.

        • Zhang D.
        • Lv S.
        • Song X.
        • et al.
        Fractional flow reserve versus angiography for guiding percutaneous coronary intervention: a meta-analysis.
        Heart. 2015; 101: 455-462https://doi.org/10.1136/heartjnl-2014-306578
        • Choi K.H.
        • Lee J.M.
        • Koo B.-K.
        • et al.
        Prognostic implication of functional incomplete revascularization and residual functional SYNTAX score in patients with coronary artery disease.
        JACC Cardiovasc Interv. 2018; 11: 237-245https://doi.org/10.1016/j.jcin.2017.09.009
        • Piroth Z.
        • Toth G.G.
        • Tonino P.A.L.
        • et al.
        Prognostic value of fractional flow reserve measured immediately after drug-eluting stent implantation.
        Circ Cardiovasc Interv. 2017; 10https://doi.org/10.1161/CIRCINTERVENTIONS.116.005233
        • Baranauskas A.
        • Peace A.
        • Kibarskis A.
        • et al.
        FFR result post PCI is suboptimal in long diffuse coronary artery disease.
        EuroIntervention. 2016; 12: 1473-1480https://doi.org/10.4244/EIJ-D-15-00514
        • Rønnow Sand N.P.
        • Nissen L.
        • Winther S.
        • et al.
        Prediction of coronary revascularization in stable Angina.
        JACC Cardiovasc Imaging. 2020; 13: 994-1004https://doi.org/10.1016/j.jcmg.2019.06.028
        • Nagumo S.
        • Collet C.
        • Norgaard B.L.
        • et al.
        Rationale and design of the precise percutaneous coronary intervention plan ( P3 ) study: prospective evaluation of a virtual computed tomography-based percutaneous intervention planner.
        Clin Cardiol. Published online March. 2021; 3 (clc): 23551https://doi.org/10.1002/clc.23551