When should fractional flow reserve be performed to assess the ...

International Journal of Cardiology 222 (2016) 606–610

Contents lists available at ScienceDirect

International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard

When should fractional flow reserve be performed to assess the significance of borderline coronary artery lesions: Derivation of a simplified scoring system Fadi A. Matar ⁎, Shayan Falasiri, Charles B. Glover, Asma Khaliq, Calvin C. Leung, Jad Mroue, George Ebra Department of Cardiovascular Sciences, The University of South Florida, Morsani College of Medicine, Tampa, FL, USA

a r t i c l e

i n f o

Article history: Received 4 June 2016 Accepted 27 July 2016 Available online 2 August 2016 Keywords: Coronary artery disease Coronary angiography Fractional flow reserve Percutaneous coronary intervention

a b s t r a c t Objectives: To derive a simplified scoring system (SSS) that can assist in selecting patients who would benefit from the application of fractional flow reserve (FFR). Background: Angiographers base decisions to perform FFR on their interpretation of % diameter stenosis (DS), which is subject to variability. Recent studies have shown that the amount of myocardium at jeopardy is an important factor in determining the degree of hemodynamic compromise. Methods: We conducted a retrospective multivariable analysis to identify independent predictors of hemodynamic compromise in 289 patients with 317 coronary vessels undergoing FFR. A SSS was derived using the odds ratios as a weighted factor. The receiver operator characteristics curve was used to identify the optimal cutoff (≥3) to discern a functionally significant lesion (FFR ≤ 0.8). Results: Male gender, left anterior descending artery apical wrap, disease proximal to lesion, minimal lumen diameter and % DS predicted abnormal FFR (≤0.8) and lesion location in the left circumflex predicted a normal FFR. Using a cutoff score of ≥3 on the SSS, a specificity of 90.4% (95% CI: 83.0–95.3) and a sensitivity of 38.0% (95% CI: 31.5–44.9) was generated with a positive predictive value of 89.0% (95% CI: 80.7%–94.6%) and negative predictive value of 41.6% (95% CI: 35.1%–48.3%). Conclusions: The decision to use FFR should be based not only on the % DS but also the size of the myocardial mass jeopardized. A score of ≥3 on the SSS should prompt further investigation with a pressure wire. © 2016 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Fractional flow reserve (FFR) is increasingly being used to assess the hemodynamic significance of borderline coronary artery stenosis at the time of angiography. FFR-guided percutaneous coronary intervention (PCI) has been associated with reduced major adverse cardiac events and to yield enhanced clinical outcomes in patients presenting with multi-vessel coronary artery disease (CAD) [1]. Since the publication of the Fractional Flow Reserve versus Angiography for Multivessel Evaluation (FAME) trial [2], there has been an increased use of FFR. However, data from U.S. Medicare beneficiaries in 2012 reported that the percent of diagnostic catheterizations undergoing FFR remained low (4%) [3].

Abbreviations: AUC, area under the curve; CABG, coronary artery bypass grafting; CAD, coronary artery disease; CI, confidence interval; DS, diameter stenosis; FFR, fractional flow reserve; LAD, left anterior descending; LCX, left circumflex; MLD, minimum lumen diameter; NPV, negative predictive value; PCI, percutaneous coronary intervention; PPV, positive predictive value; RCA, right coronary artery; SSS, simplified scoring system. ⁎ Corresponding author at: University of South Florida, Morsani College of Medicine, 509 South Armenia Avenue Suite 200, Tampa, FL 33609, USA. E-mail address: [email protected] (F.A. Matar).

http://dx.doi.org/10.1016/j.ijcard.2016.07.171 0167-5273/© 2016 Elsevier Ireland Ltd. All rights reserved.

A strategy of routine FFR measurement of all vessels with any degree of luminal narrowing in patients with stable angina at the time of diagnostic coronary angiography can change the number deemed significant and required revascularization in 26% of patients [4]. Operators may not deem that FFR of all diseased vessels to be feasible in routine daily practice environment given its impact on cost, procedure time, excess contrast load and reduced but definite procedural complication risk [4]. During coronary angiography, the operator is often faced with deciding if a lesion is ischemic and whether hemodynamic severity assessment with FFR is warranted. These decisions are usually based on the operator's subjective interpretation of angiographic percent diameter stenosis (% DS) which is subject to wide inter and intra-observer variability [5,6]. Recent studies have shown that angiographic factors relating to the amount of myocardium at jeopardy, such as left anterior descending (LAD) artery apical wrap and proximal lesion location, may also be important factors affecting the hemodynamic impact of a stenosis [7–10]. The purpose of this study is to identify demographic/clinical and angiographic predictors of hemodynamically significant lesions which can assist in targeted selection of patients who are more likely to benefit from the application of FFR assessment during cardiac catheterization.

F.A. Matar et al. / International Journal of Cardiology 222 (2016) 606–610

Moreover, to develop a simplified scoring system (SSS) for the identification of vessels warranting FFR assessment. 2. Methods 2.1. Study population From 2007 to 2012, 14,021 patients underwent diagnostic coronary angiography at a single center, of which 9728 were performed by faculty members of the University of South Florida, Morsani College of Medicine. From this cohort, 645 patients underwent clinically driven, FFR assessment. Excluded from this group were patients (n = 356) who had one or more of the following: left main lesion (≥50%), chronic total coronary artery occlusion anywhere in the coronary circulation, sequential lesions (two or more discrete and separate lesions 30% in diameter stenosis in the same vessel by visual assessment), a history of coronary artery bypass grafting, the presence of hemodynamically significant valvular stenosis or regurgitation, history of acute or prior myocardial infarction or abnormal ventricular ejection fraction (b50%). The final study population consisted of 289 patients with 317 single and multiple vessels undergoing FFR assessment. The study was approved by the Institutional Review Board and a written waiver of informed consent was granted. 2.2. Study procedures Following informed consent, coronary angiography was performed through radial or femoral access based on patient suitability and operator discretion. A 6 Fr. Guiding catheter was used for the initial angiography prior to FFR assessment. If the systolic blood pressure was over 100 mm Hg, 100 to 200 microgram intracoronary nitroglycerine was generally administered prior to FFR measurement. Intravenous anticoagulants with either heparin or bivalirudin were also administered prior to the introduction of the pressure wire. Maximum hyperemia was achieved by intravenous adenosine infusion at 140 μg/kg/min. The St. Jude Radi wire (St. Jude Medical Inc., Saint Paul, MN, USA) was used for FFR measurement in all patients. FFR was defined as the ratio between the mean coronary pressure distal to an observed coronary artery stenosis and the mean aortic pressure at maximum hyperemia. An FFR cutoff of ≤0.8 was used to determine myocardial ischemia and generally in need of revascularization. 2.3. Angiographic analysis Coronary angiograms were retrospectively subjected to qualitative and quantitative analysis using computerized automated edge detection (QAngio XA, Leiden, Netherlands) with the contrast-filled guiding catheter as a calibration reference. A lesion was identified to be in a main branch if it was located in a major epicardial vessel such as the LAD, right coronary artery (RCA) or left circumflex artery (LCX); or in a side branch if located in a diagonal, obtuse marginal, posterior descending or posterolateral branch. Lesions were considered proximal if present in the main vessel proximal to the first major diagonal in the LAD, prior to the first major acute marginal in the RCA or prior to the first obtuse marginal in the LCX. LAD apical wrap was defined as one that terminated more than one-third of the way on the diaphragmatic surface. Disease proximal to the lesion included moderate, non-discrete luminal irregularities (30–50% DS) proximal to the lesion being investigated. 2.4. Simplified scoring system for predicting FFR (≤0.8) A simplified scoring system (FFR-SSS) was devised to assist the operator in determining the need for FFR, based on selected demographic/clinical variables. The scoring scheme was based on the results of a multivariable model designed to predict the probability of a functionally significant lesion (FFR ≤ 0.8). Each significant demographic/clinical factor

607

generated from the multivariable analysis was assigned a value of “2”. An LAD apical wrap and a lesion in a non-LCX vessel were assigned a value of “1”. The final score for a given lesion was based on the sum of the values assigned to each of the significant variables (odds ratio N 1) and ranged from “0” to “10”. The higher the final score, the greater the likelihood of a functionally significant lesion. 2.5. Statistical analysis Data analyses were performed on a per-patient basis for demographic/clinical variables and on a per-lesion characteristic for the remainder of the calculations. Demographic, clinical and lesion-specific data are presented as frequency distributions and simple percentages. Values of continuous variables are expressed as mean ± the standard deviation. Two-sample unpaired Student's t-tests were used to assess the equality of means in continuous variables. To identify independent determinants of FFR ≤ 0.8, univariable analyses were performed using 18 demographic/clinical variables. Those with a probability value ≤0.10 were included in a multivariable logistic regression model, with flow limitation (FFR ≤ 0.8) as the dependent variable. The final multivariable logistic regression model included six covariates. The area under the curve (AUC) was determined by a receiver operator characteristic curve with 95% confidence interval (CI) to detect FFR ≤ 0.8 from the value generated by the SSS. The results of this analysis was used to identify the optimal cut-off point of the scoring scheme to determine sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) with 95% CI. All probability values reported are two-sided and not adjusted for multiple testing. A probability value of 0.05 or less indicated a significant difference between measures. All analyses were performed using the Number Cruncher Statistical Systems software (Version 9; NCSS, Kaysville, UT, USA). 3. Results There were 289 patients with 317 single and multiple vessels assessed with FFR of which 81 (28.0%) underwent revascularization. Among those 289 patients, 32 (11.1%) underwent PCI of additional vessels (n = 35) without FFR assessment. In the study group (N = 289), 175 (60.6%) were male and the mean age was 62.6 ± 10.9 years (range, 34 to 87). There were 97 (33.6%) patients who had diabetes mellitus, 226 (78.2%) hypertension and 147 (50.9%) a history of known coronary artery disease. Pre-procedure stress testing was performed in 144 (49.8%) patients. The patient baseline demographic/clinical characteristics are summarized in Table 1 and were similar to previously reported FFR studies [1–11]. The information in Table 2 presents the angiographic lesion characteristics of the study group. Of the 317 lesions that underwent FFR assessment, 91 (28.7%) were abnormal (≤0.8) and 226 (71.3%) normal. Fig. 1 shows the distribution of % DS in the normal and abnormal FFR groups. Although the mean % DS in the abnormal FFR group was higher than in the normal FFR group (53.3 ± 11.5 versus 47.5 ± 9.6; p b 0.001), there was significant overlap in the two cohorts. There were 36 of 173 (20.7%) lesions with b50% DS that had an abnormal FFR and 88 of 143 (61.5%) with ≥50% DS with a normal FFR. The information in Table 3 displays the univariable and multivariable predictors of abnormal FFR in the 317 lesions subjected to pressure wire assessment as deemed by the operator to be angiographically indeterminate. Of the demographic/clinical, lesion/vessel and luminal dimension variables included in the multivariable model, six were found to be significant: male gender (p = 0.002), disease proximal to the lesion (p = 0.014), LAD apical wrap (p b 0.001), MLD (p = 0.015) and % DS (p b 0.001). Lesion location in the LCX was independently predictive of a normal FFR (p = 0.024).

608


Table 1 Baseline demographic and patient clinical characteristics. Study population Variable

Number

Percent

No. of patients Sex Male Female Age (years) Preprocedure coronary risk factors Current smoker Hypertension Diabetes mellitus Dyslipidemia Preprocedure risk factors Chronic obstructive pulmonary disease Renal dysfunction Prior known coronary artery disease Stable angina Acute coronary syndrome Left ventricular ejection fraction (%) Stress test performed Abnormal stress test Non diagnostic stress test

289

100.0

175 114 62.6 ± 10.9

60.6 39.4

46 226 97 202

15.9 78.2 33.6 69.9

39 26 147 91 82 59.3 ± 5.5 144 113 4

13.5 9.0 50.9 32.9 29.6 49.8 78.5 2.8

The FFR-SSS generated a total score for each vessel based on the values assigned to the independent predictors (Table 4). The mean score for the study cohort was 4.57 ± 2.06. An ROC analysis demonstrated an AUC of 73.5% (95% CI: 67.3–78.8). The results of this analysis were used to assess the best cutoff value to identify functionally significant lesions using FFR. A score of ≥3 on the FFR-SSS tool was selected as an appropriate cut-off value to screen for abnormal FFR. Using this value, a specificity of 90.4% (95% CI: 83.0–95.3) and a sensitivity of 38.0% (95% CI: 31.5–44.9) was generated with a PPV of 89.0% (95% CI: 80.7%–94.6%) and NPV of 41.6% (95% CI: 35.1%–48.3%). 4. Discussion Historically, cardiologists who visually evaluate the angiographic appearance of the coronary artery tree during cardiac catheterization determine the treatment alternative of choice. The operator's ability to discriminate between lesions that are hemodynamically significant on the basis of angiography only has been determined to be inaccurate. [6]. FFR has become a valuable tool to provide a more accurate assessment of the functional severity of lesions identified through coronary

Table 2 Angiographic lesion characteristics of vessels undergoing fractional flow reserve. Variable

Number

Percent

Total FFR vessels LAD LCX RCA Lesion length (mm) Vessel diameter (mm) MLD (mm) Diameter stenosis (%) 30–39 40–49 50–69 70–89 90 or more In-stent restenosis Type B2/Ca Severe calcification

317 165 58 94 10.4 ± 5.4 2.8 ± 0.7 1.4 ± 0.5

100.0 52.1 18.3 29.7

61 113 133 10 0 48 191 26

19.2 35.7 42.0 3.2 0.0 15.1 60.3 8.2

FFR = fractional flow reserve; LAD = left anterior descending, LCX = left circumflex; RCA = right coronary artery; MLD = minimum lumen diameter. a ACC/AHA lesion classification.

Fig. 1. Scatter plot of % diameter stenosis (% DS) vs. abnormal FFR (≤0.80) in 317 vessels subjected to FFR assessment (normal FFR = 226, abnormal FFR = 91).

angiography. It is computed as the ratio of the coronary pressure distal and proximal to a stenosis. This is a simplification of the true definition, which is the maximum myocardial blood flow in the presence of a stenosis divided by the theoretical maximum flow in the absence of stenosis [12]. Blood flow in a coronary segment is proportional to the amount of myocardial mass it supplies [11]. The larger the mass, the smaller the degree of luminal narrowing required to compromise flow. Conversely, the smaller the mass, the larger the degree of luminal narrowing required to cause hemodynamic compromise [13]. Therefore, it is not surprising that measures of luminal narrowing, such as % DS, poorly correlate with FFR values as demonstrated in the present study and others [2,14–16]. This study demonstrated that the hemodynamic severity of intermediate coronary stenosis is not only determined by the degree of luminal narrowing such as MLD, % DS and proximal luminal disease, but also by other variables such as male gender, LAD apical wrap and LCX. This study has confirmed that male gender is an independent predictor of abnormal FFR [8,11]. This may be attributed to an increased prevalence of microvascular disease in women which can result in higher FFR values or to males having a larger myocardial mass [17–19]. The latter explanation is more likely since Leone et al. noted that male gender was no longer an independent predictor of abnormal FFR, once indices of myocardial jeopardy were included in the prediction model [8]. Lesion location in an LAD with apical wrap has also been found to be an independent predictor of an FFR ≤0.8 [7,15]. This is consistent with the fact that such vessels supply large myocardial distributions [20]. A large myocardial mass with a plaque of even lesser degree of luminal narrowing can also cause compromise. This study also shows that a lesion located in a LCX was a negative predictor of abnormal FFR. This may be attributed to the fact that LCX coronary arteries supply a relatively smaller myocardial distribution. Therefore, a lesion in such a vessel is less likely to be hemodynamically significant. To our knowledge, no other studies using FFR have shown such a relationship. The decision to perform ad hoc diagnostic FFR during coronary angiography integrates clinical presentation, stress test results if available and the operator's uncertainty regarding angiographic lesion severity. In the present study, less than one-half (49.8%) of the patients had a

609

F.A. Matar et al. / International Journal of Cardiology 222 (2016) 606–610 Table 3 Univariable and multivariable logistic regression analysis of predictors for flow limitation (FFR ≤ 0.80) in 317 vessels subjected to fractional flow reserve assessment. Univariable Variable Demographic Age Gender—male Clinical Diabetes mellitus Dyslipidemia Hypertension Current smoker Lesion/vessel characteristic Lesion in main vessel Disease proximal to lesion Proximal LAD LAD apical wrap LCX RCA Severe calcification Type B2/C lesion In-stent re-stenosis Luminal dimensions Lesion length ≥ 20 mm MLD b 1.4 mm Diameter stenosis ≥ 50%

Multivariable

β estimate

SE

Odds ratio (95% CL)

p value

β estimate

SE

Odds ratio (95% CL)

p value

−0.0074 0.6852

0.0115 1.9841

0.99 (0.97–1.02) 1.98 (1.17–3.35)

0.521 0.011

0.9426

0.2979

2.57 (1.43–4.60)

0.002

−0.1310 0.0200 −0.3358 −0.1283

0.2672 0.2715 0.2940 0.3504

0.88 (0.51–1.46) 1.02 (0.60–1.74) 0.71 (0.40–1.27) 0.88 (0.44–1.75)

0.624 0.941 0.253 0.714

0.9476 0.5204 0.3943 1.0396 −1.0809 −0.5471 0.4827 0.3390 0.1435

2.5794 0.3049 0.2834 0.2582 0.4034 0.2901 0.4238 0.2590 0.3396

2.58 (1.16–5.72) 1.68 (0.93–3.06) 1.48 (0.85–2.58) 2.83 (1.71–4.69) 0.34 (0.15–0.75) 0.58 (0.33–1.02) 1.62 (0.71–3.72) 1.40 (0.84–2.33) 1.15 (0.59–2.25)

0.020 0.088 0.164 b0.001 0.007 0.059 0.255 0.191 0.673

0.8424

0.3439

2.32 (1.18–4.56)

0.014

1.2541 −1.0211

0.3088 0.4518

3.50 (1.91–6.42) 0.36 (0.15–0.87)

b0.001 0.024

0.3219 0.8744 0.8737

0.5234 0.2554 0.2541

1.38 (0.49–3.85) 2.40 (1.45–3.96) 2.39 (1.46–3.94)

0.539 b0.001 b0.001

0.7127 1.0378

0.2931 0.3026

2.04 (1.15–3.62) 2.82 (1.56–5.11)

0.015 b0.001

SE = standard error, CL = confidence limits, LAD = left anterior descending, LCX = left circumflex; RCA = right coronary artery; MLD = minimum lumen diameter.

stress test prior to coronary angiography. Similar findings have been reported by others [10,11]. Increased reliance on the degree of angiographic luminal obstruction for deciding whether FFR assessment should be performed is likely to lead to over-treatment with PCI in smaller supply areas or under-treatment of larger supply areas such as an LAD with apical wrap. Natsumeda et al. used QCA to derive a scoring system (STABLED Score-Stenosis 2 points, TAndem lesion 1 point, Bifurcation 1 point, LEsion length 1 point, Distance from ostium 1 point) to predict FFR abnormalities. No other surrogates of myocardial supply mass were included in this prediction model [9]. Biasco et al. identified five angiographic parameters to independently predict flow limitation, viz., 30–50% lesion proximal to the lesion of interest, lesion length N20 mm, distal take-off of all diagonal branches, N 2 mm diameter, apical wrap of LAD and collaterals to an occluded LCX/RCA. Based on these results, risk scores for prediction of flow limitation in intermediate LAD lesions were determined. Their study was limited to LAD lesions and no other surrogates of myocardial supply mass such as gender were included in their predictive model [7]. Wong et al. created the DILEMMA score to predict an FFR ≤ 0.8. It combined MLD, lesion length and BARI MJI (Bypass Angioplasty Revascularization Investigation [BARI] Myocardial Jeopardy Index [MJI]) as a surrogate of myocardial supply mass subtended by the stenotic coronary artery. The MJI was computed by assigning a value (0–3) to all terminal arteries based on their length and caliber. A value of “0” represented an almost insignificant vessel size, whereas a value of “3” defined a large Table 4 Simplified scoring system for identifying abnormal FFR (N0.80) based on independent predictors. Predictor

Assigned Value

Male gender MLD b 1.4 mm DS ≥ 50% Disease proximal to lesion Non LCX vessel LAD apical wrap Total Score

2 2 2 2 1 1 10

FFR = fractional flow reserve; MLD = minimum lumen diameter; DS = diameter stenosis; LCX – left circumflex; LAD = left anterior descending.

artery with a length greater than two-thirds the distance between the cardiac base and the apex. However, computing such a complex index is not practical at the point of care by an operator in the catheterization laboratory [10]. The proposed FFR-SSS is an easily administered scale that allows operators to make an informed decision concerning the need to perform FFR in a diseased coronary artery. Based on the results of multivariable and ROC analyses, a score of ≥3 on this tool represents a useful cut-off to screen for the need for FFR assessment. This tool is of value for lesions that do not appear narrowed enough by visual inspection to prompt the use of a pressure wire for myocardial ischemia assessment. The scale takes into account luminal compromise, e.g., MLD, % DS and disease proximal to the lesion, as well as indicators of jeopardized myocardial mass, e.g., male gender, LAD with apical wrap as surrogates of larger and LCX as an indicator of smaller territories. In the present study, a PPV of 89.0% (95% CI: 80.7%–94.6%) of vessels were correctly classified as abnormal and confirmed on FFR. Moreover, a NPV 41.6% (95% CI: 35.1%–48.3%) of vessels were correctly classified as normal on FFR. 4.1. Limitations of the study Despite the valuable information generated from this study concerning the value of FFR, several important limitations must be acknowledged. The study was conceived as an observational, nonrandomized investigation, conducted by a retrospective review of prospectively collected data. Non-randomization may have introduced confounding clinical factors not accounted for in the study that may have contributed to the use of FFR. These results cannot be generalized to all lesions since patients with left ventricular systolic dysfunction, prior myocardial infarction, the presence of a chronic total occlusion and a history of CABG were excluded from the study cohort. All of these factors may likely influence the need for FFR. Although data for the current study were derived from a single-center database, several operators participated over time and these may have adversely affected the use of FFR. Single source data also may have influenced the analytical biases of FFR use. Moreover, the exclusion criteria may have limited the applicability of the results to a more general population. For example, the effect of impaired left ventricular hypertrophy on FFR was not assessed. In the present study, there is a small disparity in the

610


number of patients and the lesions assessed by FFR. Further analysis with an increased sample size to assess the relationship between patient and lesion-specific characteristics may more clearly define the role of FFR in cardiac catheterization. Recognizing its inherent limitations, this study does have the advantage of representing a daily practice in an environment that serves a diverse patient population. Moreover, data collected on pre- and intra-procedure demographic/clinical variables were gathered in a standardized manner using specified definitions. This FFR-SSS warrants further investigation with an increased sample size to validate its appropriate use and its application to other clinical subgroups. 4.2. Clinical implications The decision to use FFR should be based on the severity of vessel stenosis and the size of the myocardial mass jeopardized. The FFR-SSS can serve as a valuable decision-making tool to discern further investigation with a pressure wire. It provides an easily administered quantitative technique for evaluating the physiologic significance of a vessel stenosis. It can assist in cost-containment efforts, reduces potential risk exposure and enhance patient outcomes. Author disclosures This study was supported, in part, by an unrestricted grant from St. Jude Medical Inc., Saint Paul, Minnesota. Conflict of interest The authors report no relationships that could be construed as a conflict of interest. Acknowledgments The authors thank Dr. Debra D. Guest for her assistance in statistical analysis and manuscript preparation. References [1] N.H. Pijls, P. van Schaardenburgh, G. Manoharan, et al., Percutaneous coronary intervention of functionally nonsignificant stenosis: 5-year follow-up of the DEFER study, J. Am. Coll. Cardiol. 49 (21) (2007) 2105–2111.

[2] P.A. Tonino, B. De Bruyne, N.H. Pijls, et al., Fractional flow reserve versus angiography for guiding percutaneous coronary intervention, N. Engl. J. Med. 360 (3) (2009) 213–224. [3] N.V. Pothineni, N.S. Shah, Y. Rochlani, et al., U.S. trends in inpatient utilization of fractional flow reserve and percutaneous coronary intervention, J. Am. Coll. Cardiol. 67 (6) (2016) 732–733. [4] N. Curzen, O. Rana, Z. Nicholas, et al., Does routine pressure wire assessment influence management strategy at coronary angiography for diagnosis of chest pain?: the RIPCORD study, Circ. Cardiovasc. Interv. 7 (2) (2014) 248–255. [5] L.L. Leape, R.E. Park, T.M. Bashore, J.K. Harrison, C.J. Davidson, R.H. Brook, Effect of variability in the interpretation of coronary angiograms on the appropriateness of use of coronary revascularization procedures, Am. Heart J. 139 (1 Pt 1) (2000) 106–113. [6] L.M. Zir, S.W. Miller, R.E. Dinsmore, J.P. Gilbert, J.W. Harthorne, Interobserver variability in coronary angiography, Circulation 53 (4) (1976) 627–632. [7] L. Biasco, F. Pedersen, J. Lønborg, et al., Angiographic characteristics of intermediate stenosis of the left anterior descending artery for determination of lesion significance as identified by fractional flow reserve, Am. J. Cardiol. 115 (11) (2015) 1475–1480. [8] A.M. Leone, A.R. De Caterina, E. Basile, et al., Influence of the amount of myocardium subtended by a stenosis on fractional flow reserve, Circ. Cardiovasc. Interv. 6 (1) (2013) 29–36. [9] M. Natsumeda, G. Nakazawa, T. Murakami, et al., Coronary angiographic characteristics that influence fractional flow reserve, Circ. J. 79 (4) (2015) 802–807. [10] D.T. Wong, O. Narayan, B.S. Ko, et al., A novel coronary angiography index (DILEMMA score) for prediction of functionally significant coronary artery stenoses assessed by fractional flow reserve: a novel coronary angiography index, Am. Heart J. 169 (4) (2015) 564–571 e4. [11] B. De Bruyne, N.H. Pijls, B. Kalesan, et al., Fractional flow reserve-guided PCI versus medical therapy in stable coronary disease, N. Engl. J. Med. 367 (11) (2012) 991–1001. [12] N.H. Pijls, B. De Bruyne, Coronary pressure measurement and fractional flow reserve, Heart 80 (6) (1998) 539–542. [13] B. De Bruyne, N.H. Pijls, J. Bartúnek, et al., Fractional flow reserve in patients with prior myocardial infarction, Circulation 104 (2) (2001) 157–162. [14] J. Bartúnek, S.U. Sys, G.R. Heyndrickx, N.H. Pijls, B. De Bruyne, Quantitative coronary angiography in predicting functional significance of stenoses in an unselected patient cohort, J. Am. Coll. Cardiol. 26 (2) (1995) 328–334. [15] S.J. Park, S.J. Kang, J.M. Ahn, et al., Visual-functional mismatch between coronary angiography and fractional flow reserve, JACC Cardiovasc. Interv. 5 (10) (2012) 1029–1036. [16] P.A. Tonino, W.F. Fearon, B. De Bruyne, et al., Angiographic versus functional severity of coronary artery stenoses in the FAME study fractional flow reserve versus angiography in multivessel evaluation, J. Am. Coll. Cardiol. 55 (25) (2010) 2816–2821. [17] M. Fineschi, G. Guerrieri, D. Orphal, et al., The impact of gender on fractional flow reserve measurements, EuroIntervention 9 (3) (2013) 360–366. [18] S.J. Kang, J.M. Ahn, S. Han, et al., Sex differences in the visual-functional mismatch between coronary angiography or intravascular ultrasound versus fractional flow reserve, JACC Cardiovasc. Interv. 6 (6) (2013) 562–568. [19] H.S. Kim, P.A. Tonino, B. De Bruyne, et al., The impact of sex differences on fractional flow reserve-guided percutaneous coronary intervention: a FAME (fractional flow reserve versus angiography for multivessel evaluation) substudy, JACC Cardiovasc. Interv. 5 (10) (2012) 1037–1042. [20] N. Kobayashi, A. Maehara, S.J. Brener, et al., Usefulness of the left anterior descending coronary artery wrapping around the left ventricular apex to predict adverse clinical outcomes in patients with anterior wall ST-segment elevation myocardial infarction (from the harmonizing outcomes with revascularization and stents in acute myocardial infarction trial), Am. J. Cardiol. 116 (11) (2015) 1658–1665.