- Email: [email protected]
TCT-529 The correlation between Quantitative Flow Ratio (QFR) and Fractional Flow Reserve (FFR)
B213
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY, VOL. 68, NO. 18, SUPPL B, 2016
BACKGROUND In this study we evaluate the performance of a machine-learning based algorithm, which combines anatomical and functional information acquired during rest-state, against invasive FFR. METHODS Invasive aortic (proximal) and coronary (distal) pressures were measured at rest and at adenosine-induced hyperemia using a pressure wire in 64 patients (125 lesions) with suspected coronary artery disease (CAD). The measured pressure traces were processed offline to determine Pd/Parest and FFR. Routine coronary angiograms were acquired under resting conditions and two angiographic views were selected to reconstruct a three-dimensional anatomical model of each diseased vessel. The 125 lesions were randomly divided into a training (62 lesions) and a validation set (63 lesions). A machine learning algorithm was trained on the training set with invasive FFR as the ground truth. Geometric characteristics extracted automatically from the reconstructed 3D anatomical model and the Pd/Parest value were used as features. The trained model was then applied on the unseen validation set. RESULTS From the 63 lesions of the validation set, 17 were hemodynamically significant (FFR 0.8). When compared against invasive FFR, the proposed algorithm demonstrated a diagnostic accuracy of 90.5% (sensitivity: 82.4%, specificity: 93.5%, positive predictive value: 82.4%, negative predictive value: 93.5%), a correlation of r ¼0.92, mean difference of -0.02 with a standard deviation of 0.05. The diagnostic accuracy of Pd/Parest (cut-off 0.92) was 74.6% (sensitivity: 82.4%, specificity: 71.8%, PPV: 51.8%, NPV: 91.6). The area under the curve (AUC) of the receiver-operator characteristic curves was 0.95 for the proposed algorithm and 0.78 for Pd/Parest. CONCLUSION The proposed machine learning algorithm, which augments the resting state measurements with patient-specific anatomical features extracted from routine angiograms, demonstrated a high diagnostic accuracy and correlation against invasive FFR, thus potentially obviating the need for hyperemia. The feature (mentioned herein) is not commercially available. Due to regulatory reasons its future availability cannot be guaranteed. CATEGORIES IMAGING: FFR and Physiologic Lesion Assessment TCT-528 GLAsgow MRI ROtational AtheRectOmy Study: Physiology Reflects Procedural Success. The GLAMOROS Study. Barry Hennigan,1 Colin Berry,2 David Corcoran,3 David Carrick,4 Mitchell Lindsay,5 Hany Eteiba,6 Stuart Watkins,7 Keith Oldroyd,8 Margaret McEntegart9 1 University of Glasgow, Glasgow, United Kingdom; 2Golden Jubilee National Hospital, Glasgow, United Kingdom; 3The Prince Charles Hospital, Glasgow, United Kingdom; 4Golden Jubilee National Hospital; 5 Golden Jubilee National Hospital - Scotland; 6Golden Jubilee National Hospital, Clydebank, United Kingdom; 7Golden Jubilee National Hospital, Glasgow, United Kingdom; 8West of Scotland Regional Heart and Lung Centre, Glasgow, United Kingdom; 9Golden Jubilee National Hospital, Glasgow, United Kingdom BACKGROUND PCI with high speed rotational atherectomy (HSRA) is used to treat highly complex and calcified coronary artery stenoses. Theoretically, HSRA may have deleterious effects on the coronary microcirculation though it’s true impact has not been systematically studied. METHODS We enrolled near consecutive elective patients undergoing HSRA in a prospective study and assessed their coronary physiology using thermodilution derived coronary flow reserve (CFR), the Index of Microvascular Resistance (IMR) and fractional flow reserve (FFR) at three time points; before HSRA (T1), after HSRA (T2) and after implantation and post dilation of a stent(s)(T3). We used Certus wires and a RadiAnalyzer (St. Jude Medical, U.S.A.). To avoid peri-procedural hypotension we developed a strategy without verapamil or nitrate infusate during HSRA. Results are expressed as means and standard deviations and were analysed using Student’s t test for paired data with Bonferroni correction. A p value <¼ 0.0167 is significant. RESULTS 58 patients were recruited of whom 56 had physiological data (see Table for results). 46 (79%) patients were male and 24 (41%) of vessels treated were the LAD. Uncorrected IMR demonstrated a non-significant reduction post PCI which disappeared after correction for coronary wedge pressure. CFR fell after HSRA but recovered above baseline levels after stenting. FFR improved significantly after HSRA and showed further improvement after stenting. 84% of patients achieved a final FFR value > 0.80. Table 1. Results expressed as means (standard deviations) and n¼total number per group. Log transformation and paired T tests to compare data. Significant p values are in bold.
Pre HSRA
Post HSRA
(T1)
(T2)
(T3)
T1 vs T2
T1 vs T3
IMR
23.8 (14.4)
24.4 (22.3)
20.5 (24.3)
p¼0.414
p[0.017
IMRc
18.1 (13.3)
p¼0.125
p¼0.466
p¼0.030
p¼0.031
p<0.001
p<0.001
N/A
N/A
N/A
N/A
N/A
N/A
n¼37 n¼27 CFR
1.58 (0.66) n¼44
FFR
0.56 (0.13)
n¼48 24.6 (24) n¼32 1.29 (0.4) n¼52 0.71 (0.15)
n¼36
n¼39
Number of
4 (9.1%)
2 (3.8%)
patients
n¼44
n¼52
Post Stenting
n¼46 19.3 (19.6) n¼26 2.17 (1.92) n¼53 0.87 (0.08) n¼37 13 (24.5%) n¼53
with CFR2.5 % of total Number of patients
15 (40.5%) n¼37
14 (29.2%) n¼48
7 (15.2%) n¼46
with IMR25 % of total Number of patients
2 (5.6%) n¼36
9 (23.1%) n¼39
31 (83.8%) n¼37
with FFR0.80 % of total
CONCLUSION This data confirms that PCI with HSRA and no vasodilator cocktail can be performed without compromising coronary microvascular function. Consequently, post procedural FFR values can be used to assess the physiological success of the procedure. CATEGORIES IMAGING: FFR and Physiologic Lesion Assessment TCT-529 The correlation between Quantitative Flow Ratio (QFR) and Fractional Flow Reserve (FFR) Hiroki Emori,1 Takashi Kubo,2 Takeyoshi Kameyama,3 Yasushi Ino,4 Yoshiki Matsuo,5 Atsushi Tanaka,6 Takashi Akasaka7 1 Dept. of Cardiology-Fondazione IRCCS Policlinico San Matteo, Pavia/ Italy, Wakayama, Japan; 2Wakayama Medical University, Wakayama, Japan; 3Cedars-Sinai Medical Center; 4Wakayama Medical University, Wakayama, Japan; 5Unknown, Wakayama, Japan; 6Massachusetts General Hospital, Boston, Massachusetts, United States; 7Wakayama Medical University, Wakayama, Japan BACKGROUND Fractional Flow Reserve (FFR) is an indispensable device to identify coronary stenoses causing myocardial ischemia. But FFR is more invasive than coronary angiography and require additional cost, time and efforts. Quantitative Flow Ratio (QFR) is a novel method for rapid computation of FFR with three-dimensional quantitative coronary angiography (QCA). The aim of this study was to investigate the correlation between QFR and the pressure wire-based FFR in patients with intermediate coronary stenosis. METHODS We enrolled 73 vessels (35 left anterior descending (LAD) arteries, 17 left circumflex arteries, 21 right coronary arteries) in 49 patients with stable coronary artery disease who had intermediate coronary stenosis on coronary angiography and underwent pressure wire-based FFR measurements. QFR was computed based on the vessel anatomy and TIMI (Thrombolysis In Myocardial Infarction) frame counting. The reconstructed 3D vessel was divided into subsegments from proximal to distal, and the pressure drop for the different segments of the vessel was calculated from the lumen sizing with respect to the reference sizing. Pearson correlation was used to quantify the correlation between FFR and QFR.
B214
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY, VOL. 68, NO. 18, SUPPL B, 2016
RESULTS QFR correlated well with FFR (r ¼ 0.69, p < 0.005). In LAD lesion, better correlation between QFR and FFR was observed (r ¼ 0.76, p < 0.001) than that in non-LAD lesion (r ¼ 0.61, p < 0.005). Applying the FFR cut-off value of < 0.80 to QFR resulted in 32 true positives, 27 true negatives, 7 false positives, and 7 false negatives. QFR < 0.80 predicted an FFR < 0.80 with sensitivity of 82%, specificity of 79%, positive predictive value of 82%, and negative predictive value of 79%. CONCLUSION QFR can reliably predict coronary lesions causing myocardial ischemia. It may emerge as a safe, efficient, and costreducing device for evaluation of coronary stenoses severity during diagnostic coronary angiography. CATEGORIES IMAGING: FFR and Physiologic Lesion Assessment TCT-530 Femoral vs antecubital vein adenosine infusion for fractional flow reserve assessment Jacek Legutko,1 Pawel Kleczynski,2 Maciej Bagienski,3 Lukasz Rzeszutko,4 Stanislaw Bartus,5 Dariusz Dudek6 1 University Hospital in Krakow, Kraków, Poland; 2Krakowski Szpital Specjalistyczny, Krakow, Poland; 3Institue of Cardiology; 4University Hospital, Krakow, Poland, Krakow, Poland; 5Institue of Cardiology; 6 Jagiellonian University, Krakow, Poland BACKGROUND Adenosine is often used to induce maximal hyperemia when measuring fractional flow reserve (FFR). The gold standard is continuous infusion via a large central vein. Peripheral venous access is frequently obtained in the antecubital vein, but concern exists as to whether adenosine delivery from this site can achieve adequate vasodilation for accurate FFR assessment. Therefore, we sought to assess adenosine infusion via antecubital vein compared with femoral venous access to achieve peak hyperemia during FFR. METHODS Consecutive patients with borderline coronary lesions, who were scheduled for FFR, were enrolled. Subjects received intravenous adenosine infusion via 2 routes, via a 20-gauge cannula in the antecubital vein, and, after a washout period, via a 6-F femoral venous sheath. Adenosine was administered at 140 mg/kg/min from each site. Additionally adenosine infusion at 280 mg/kg/min was performed. Minimal FFR achieved with intravenous adenosine from each infusion site was recorded as was the time to reaction and time to peak hyperemia. RESULTS Antecubital and femoral vein adenosine at 140 and 280 mg/ kg/min recordings from 81 vessels in 32 patients were suitable for blinded analysis. The median FFR measured using adenosine administered via antecubital vein and femoral routes at 140 mg/kg/min was 0.82 [IQR 0.73-0.9] and 0.81 [IQR 0.71-0.89], respectively, p¼0.72. Mean time to peak hyperemia was longer on average with antecubitaladministered adenosine compared with femoral adenosine administration (58 s vs. 42 s; mean difference 15 s with a 95% confidence interval: 14 s to 25 s; p<0.001). We also showed no significant evidence of a difference in FFR values after 140 and 280 mg/kg/min infusion (p¼0.63). CONCLUSION Antecubital vein adenosine infusion achieved FFR values are very similar to those obtained using femoral vein adenosine administration. However, time to maximal hyperemia is longer with the infusion via antecubital vein. Moreover, the FFR values achieved by adenosine infusion at 140 and 280 mg/kg/min are also similar. CATEGORIES IMAGING: FFR and Physiologic Lesion Assessment TCT-531 INTRACORONARY VS. INTRAVENOUS ADENOSINE FOR FRACTIONAL FLOW RESERVE MEASUREMENT: A META-ANALYSIS Stefano Rigattieri,1 Giuseppe Biondi-Zoccai,2 Alessandro Sciahbasi,3 Cristian Di RUsso,4 Maria Cera,5 Jonathan Beaudoin,6 Silvio Fedele,7 Francesco Pugliese8 1 Interventional Cardiology, Sandro Pertini Hospital, Rome, Italy, Rome, Italy; 2Sapienza University of Rome, Ospedaletti, Italy; 3Complejo Hospitalario de Jaén; 4Hospital de Puerto Real, Cádiz; 5Università degli studi di Milano, Milano, Italy; 6Québec Heart & Lung Insitute, Québec, Quebec, Canada; 7Interventional Cardiology, Sandro Pertini Hospital, Rome; 8Centre hospitalier sud francilien BACKGROUND Intravenous (IV) infusion of adenosine represents the gold standard for measuring Fractional Flow Reserve (FFR) in order to assess physiological significance of coronary lesions. However, IV adenosine preferably requires a central venous catheter and is more expensive and time-consuming as compared to
intracoronary (IC) boluses of adenosine. On the other hand, IC adenosine may be limited in inducing maximal hyperaemia and the optimal bolus dosage is still debated. We aimed to systematically review published studies comparing head-to-head IC to IV adenosine for FFR assessment in the same coronary lesions of patients undergoing coronary angiography. METHODS We performed electronic search of major databases (PubMed, Scopus, Web of Science) and retrieved published studies comparing IC to IV adenosine and reporting at least one of the following: Pd/Pa ratio values at peak hyperaemia; rate of abnormal FFR results (0.80 or <0.75 according to study definition); incidence of adverse events, such as angina-like symptoms, dyspnea, flushing and transient atrio-ventricular (AV) block . Weighted mean differences (WMD) were calculated with the inverse variance method for continuous variables, whereas relative risks (RR) were calculated with the Mantel-Haenszel method for dichotomous variables. We performed pre-specified subgroup analysis in order to appraise studies using low dose (<100 mcg) or high dose IC adenosine (100 mcg). RESULTS We retrieved 8 studies amounting to 456 patients and 489 lesions. In most studies a 140 mcg/Kg/min IV adenosine infusion was used; regarding IC adenosine, 3 studies evaluated lowdose boluses (range 20 to 80 mcg), 2 studies high-dose boluses (range 120 to 150 mcg) and 3 studies both low and high-dose boluses (range 60 to 600 mcg). In the latter studies, we extracted data relative to the lowest and the highest IC adenosine dose given. IC adenosine (both at low and high doses) was as effective as IV adenosine in inducing hyperaemia; indeed, WMD in Pd/Pa ratio at peak hyperaemia was 0.01 (95% CI -0.01, 0.02; p¼0.40). Accordingly, the rate of abnormal FFR results was similar (28.2% for IC adenosine and 31.1% for IV adenosine, RR 0.90, 95% CI 0.76, 1.08; p¼0.27). Adverse events were less frequent with IC adenosine as compared to IV adenosine (RR 0.19, 95% CI 0.07, 0.47; p<0.001). The incidence of transient AV block was 8.3% with IC adenosine (mostly related to high doses) and 4.0% with IV adenosine. CONCLUSION IC boluses of adenosine are equally effective as and might be better tolerated than IV infusion of adenosine for FFR measurement. As compared to low-dose, high-dose intracoronary adenosine has similar diagnostic accuracy but is associated with more AV blocks. CATEGORIES IMAGING: FFR and Physiologic Lesion Assessment TCT-532 Morphometric assessment of hemodynamically severe coronary stenosis relevance with optical frequency domain imaging according to vessel diameter Norihiro Kogame,1 Raisuke Iijima,2 Hidehiko Hara,3 Hitoshi Anzai,4 Masato Nakamura5 1 Toho University Ohashi Medical Center, Tokyo, Japan; 2Munich, Japan; 3Toho University Ohashi Medical Center, Tokyo, Japan; 4Ota Memorila Hospital, Yokoyama, Japan; 5Toho University Ogashi Medical Center, Tokyo, Japan BACKGROUND Fractional flow reserve (FFR) is considered the gold standard for assessing intermediate coronary lesions. Many studies have investigated the correlation between FFR value and minimum lumen area (MLA) measured by intracoronary imaging. However, it’s still unclear whether MLA of optical frequency domain imaging (OFDI) had diagnostic efficiency of coronary stenosis based on FFR value. The aim of this study was to evaluate the diagnostic efficiency of MLA measured by OFDI (OFDI-MLA) to predict FFR 0.80. METHODS We investigated all patients with suspected ischemic heart disease admitted to our hospitals between August 2013 and June 2016. If coronary angiogram showed the culprit lesion as intermediate angiographic severity, we performed OFDI and FFR examinations, and a correlation between FFR value and OFDI-MLA were evaluated. In addition, to study efficiency for different ranges of vessel diameter. RESULTS Both OFDI and FFR were done in 134 patients who had 157 coronary lesions. Angiographic stenosis severity was 51.2 13 % in diameter stenosis. Positive correlation was observed between FFR and OFDI-MLA (r ¼ 0.589, p <0.05). Significant stenoses with FFR 0.80 were observed in 114 lesions (72.6%). The diagnostic efficiency of OFDI-MLA in identifying significant stenosis was moderate with an area under the curve (AUC) 0.76, 95%CI 0.670.84. The best cut-off value of OFDI-MLA to predict FFR 0.80 was 1.98 mm2 [sensitivity 54%, specificity 85%]. When looked at
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY, VOL. 68, NO. 18, SUPPL B, 2016
BACKGROUND In this study we evaluate the performance of a machine-learning based algorithm, which combines anatomical and functional information acquired during rest-state, against invasive FFR. METHODS Invasive aortic (proximal) and coronary (distal) pressures were measured at rest and at adenosine-induced hyperemia using a pressure wire in 64 patients (125 lesions) with suspected coronary artery disease (CAD). The measured pressure traces were processed offline to determine Pd/Parest and FFR. Routine coronary angiograms were acquired under resting conditions and two angiographic views were selected to reconstruct a three-dimensional anatomical model of each diseased vessel. The 125 lesions were randomly divided into a training (62 lesions) and a validation set (63 lesions). A machine learning algorithm was trained on the training set with invasive FFR as the ground truth. Geometric characteristics extracted automatically from the reconstructed 3D anatomical model and the Pd/Parest value were used as features. The trained model was then applied on the unseen validation set. RESULTS From the 63 lesions of the validation set, 17 were hemodynamically significant (FFR 0.8). When compared against invasive FFR, the proposed algorithm demonstrated a diagnostic accuracy of 90.5% (sensitivity: 82.4%, specificity: 93.5%, positive predictive value: 82.4%, negative predictive value: 93.5%), a correlation of r ¼0.92, mean difference of -0.02 with a standard deviation of 0.05. The diagnostic accuracy of Pd/Parest (cut-off 0.92) was 74.6% (sensitivity: 82.4%, specificity: 71.8%, PPV: 51.8%, NPV: 91.6). The area under the curve (AUC) of the receiver-operator characteristic curves was 0.95 for the proposed algorithm and 0.78 for Pd/Parest. CONCLUSION The proposed machine learning algorithm, which augments the resting state measurements with patient-specific anatomical features extracted from routine angiograms, demonstrated a high diagnostic accuracy and correlation against invasive FFR, thus potentially obviating the need for hyperemia. The feature (mentioned herein) is not commercially available. Due to regulatory reasons its future availability cannot be guaranteed. CATEGORIES IMAGING: FFR and Physiologic Lesion Assessment TCT-528 GLAsgow MRI ROtational AtheRectOmy Study: Physiology Reflects Procedural Success. The GLAMOROS Study. Barry Hennigan,1 Colin Berry,2 David Corcoran,3 David Carrick,4 Mitchell Lindsay,5 Hany Eteiba,6 Stuart Watkins,7 Keith Oldroyd,8 Margaret McEntegart9 1 University of Glasgow, Glasgow, United Kingdom; 2Golden Jubilee National Hospital, Glasgow, United Kingdom; 3The Prince Charles Hospital, Glasgow, United Kingdom; 4Golden Jubilee National Hospital; 5 Golden Jubilee National Hospital - Scotland; 6Golden Jubilee National Hospital, Clydebank, United Kingdom; 7Golden Jubilee National Hospital, Glasgow, United Kingdom; 8West of Scotland Regional Heart and Lung Centre, Glasgow, United Kingdom; 9Golden Jubilee National Hospital, Glasgow, United Kingdom BACKGROUND PCI with high speed rotational atherectomy (HSRA) is used to treat highly complex and calcified coronary artery stenoses. Theoretically, HSRA may have deleterious effects on the coronary microcirculation though it’s true impact has not been systematically studied. METHODS We enrolled near consecutive elective patients undergoing HSRA in a prospective study and assessed their coronary physiology using thermodilution derived coronary flow reserve (CFR), the Index of Microvascular Resistance (IMR) and fractional flow reserve (FFR) at three time points; before HSRA (T1), after HSRA (T2) and after implantation and post dilation of a stent(s)(T3). We used Certus wires and a RadiAnalyzer (St. Jude Medical, U.S.A.). To avoid peri-procedural hypotension we developed a strategy without verapamil or nitrate infusate during HSRA. Results are expressed as means and standard deviations and were analysed using Student’s t test for paired data with Bonferroni correction. A p value <¼ 0.0167 is significant. RESULTS 58 patients were recruited of whom 56 had physiological data (see Table for results). 46 (79%) patients were male and 24 (41%) of vessels treated were the LAD. Uncorrected IMR demonstrated a non-significant reduction post PCI which disappeared after correction for coronary wedge pressure. CFR fell after HSRA but recovered above baseline levels after stenting. FFR improved significantly after HSRA and showed further improvement after stenting. 84% of patients achieved a final FFR value > 0.80. Table 1. Results expressed as means (standard deviations) and n¼total number per group. Log transformation and paired T tests to compare data. Significant p values are in bold.
Pre HSRA
Post HSRA
(T1)
(T2)
(T3)
T1 vs T2
T1 vs T3
IMR
23.8 (14.4)
24.4 (22.3)
20.5 (24.3)
p¼0.414
p[0.017
IMRc
18.1 (13.3)
p¼0.125
p¼0.466
p¼0.030
p¼0.031
p<0.001
p<0.001
N/A
N/A
N/A
N/A
N/A
N/A
n¼37 n¼27 CFR
1.58 (0.66) n¼44
FFR
0.56 (0.13)
n¼48 24.6 (24) n¼32 1.29 (0.4) n¼52 0.71 (0.15)
n¼36
n¼39
Number of
4 (9.1%)
2 (3.8%)
patients
n¼44
n¼52
Post Stenting
n¼46 19.3 (19.6) n¼26 2.17 (1.92) n¼53 0.87 (0.08) n¼37 13 (24.5%) n¼53
with CFR2.5 % of total Number of patients
15 (40.5%) n¼37
14 (29.2%) n¼48
7 (15.2%) n¼46
with IMR25 % of total Number of patients
2 (5.6%) n¼36
9 (23.1%) n¼39
31 (83.8%) n¼37
with FFR0.80 % of total
CONCLUSION This data confirms that PCI with HSRA and no vasodilator cocktail can be performed without compromising coronary microvascular function. Consequently, post procedural FFR values can be used to assess the physiological success of the procedure. CATEGORIES IMAGING: FFR and Physiologic Lesion Assessment TCT-529 The correlation between Quantitative Flow Ratio (QFR) and Fractional Flow Reserve (FFR) Hiroki Emori,1 Takashi Kubo,2 Takeyoshi Kameyama,3 Yasushi Ino,4 Yoshiki Matsuo,5 Atsushi Tanaka,6 Takashi Akasaka7 1 Dept. of Cardiology-Fondazione IRCCS Policlinico San Matteo, Pavia/ Italy, Wakayama, Japan; 2Wakayama Medical University, Wakayama, Japan; 3Cedars-Sinai Medical Center; 4Wakayama Medical University, Wakayama, Japan; 5Unknown, Wakayama, Japan; 6Massachusetts General Hospital, Boston, Massachusetts, United States; 7Wakayama Medical University, Wakayama, Japan BACKGROUND Fractional Flow Reserve (FFR) is an indispensable device to identify coronary stenoses causing myocardial ischemia. But FFR is more invasive than coronary angiography and require additional cost, time and efforts. Quantitative Flow Ratio (QFR) is a novel method for rapid computation of FFR with three-dimensional quantitative coronary angiography (QCA). The aim of this study was to investigate the correlation between QFR and the pressure wire-based FFR in patients with intermediate coronary stenosis. METHODS We enrolled 73 vessels (35 left anterior descending (LAD) arteries, 17 left circumflex arteries, 21 right coronary arteries) in 49 patients with stable coronary artery disease who had intermediate coronary stenosis on coronary angiography and underwent pressure wire-based FFR measurements. QFR was computed based on the vessel anatomy and TIMI (Thrombolysis In Myocardial Infarction) frame counting. The reconstructed 3D vessel was divided into subsegments from proximal to distal, and the pressure drop for the different segments of the vessel was calculated from the lumen sizing with respect to the reference sizing. Pearson correlation was used to quantify the correlation between FFR and QFR.
B214
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY, VOL. 68, NO. 18, SUPPL B, 2016
RESULTS QFR correlated well with FFR (r ¼ 0.69, p < 0.005). In LAD lesion, better correlation between QFR and FFR was observed (r ¼ 0.76, p < 0.001) than that in non-LAD lesion (r ¼ 0.61, p < 0.005). Applying the FFR cut-off value of < 0.80 to QFR resulted in 32 true positives, 27 true negatives, 7 false positives, and 7 false negatives. QFR < 0.80 predicted an FFR < 0.80 with sensitivity of 82%, specificity of 79%, positive predictive value of 82%, and negative predictive value of 79%. CONCLUSION QFR can reliably predict coronary lesions causing myocardial ischemia. It may emerge as a safe, efficient, and costreducing device for evaluation of coronary stenoses severity during diagnostic coronary angiography. CATEGORIES IMAGING: FFR and Physiologic Lesion Assessment TCT-530 Femoral vs antecubital vein adenosine infusion for fractional flow reserve assessment Jacek Legutko,1 Pawel Kleczynski,2 Maciej Bagienski,3 Lukasz Rzeszutko,4 Stanislaw Bartus,5 Dariusz Dudek6 1 University Hospital in Krakow, Kraków, Poland; 2Krakowski Szpital Specjalistyczny, Krakow, Poland; 3Institue of Cardiology; 4University Hospital, Krakow, Poland, Krakow, Poland; 5Institue of Cardiology; 6 Jagiellonian University, Krakow, Poland BACKGROUND Adenosine is often used to induce maximal hyperemia when measuring fractional flow reserve (FFR). The gold standard is continuous infusion via a large central vein. Peripheral venous access is frequently obtained in the antecubital vein, but concern exists as to whether adenosine delivery from this site can achieve adequate vasodilation for accurate FFR assessment. Therefore, we sought to assess adenosine infusion via antecubital vein compared with femoral venous access to achieve peak hyperemia during FFR. METHODS Consecutive patients with borderline coronary lesions, who were scheduled for FFR, were enrolled. Subjects received intravenous adenosine infusion via 2 routes, via a 20-gauge cannula in the antecubital vein, and, after a washout period, via a 6-F femoral venous sheath. Adenosine was administered at 140 mg/kg/min from each site. Additionally adenosine infusion at 280 mg/kg/min was performed. Minimal FFR achieved with intravenous adenosine from each infusion site was recorded as was the time to reaction and time to peak hyperemia. RESULTS Antecubital and femoral vein adenosine at 140 and 280 mg/ kg/min recordings from 81 vessels in 32 patients were suitable for blinded analysis. The median FFR measured using adenosine administered via antecubital vein and femoral routes at 140 mg/kg/min was 0.82 [IQR 0.73-0.9] and 0.81 [IQR 0.71-0.89], respectively, p¼0.72. Mean time to peak hyperemia was longer on average with antecubitaladministered adenosine compared with femoral adenosine administration (58 s vs. 42 s; mean difference 15 s with a 95% confidence interval: 14 s to 25 s; p<0.001). We also showed no significant evidence of a difference in FFR values after 140 and 280 mg/kg/min infusion (p¼0.63). CONCLUSION Antecubital vein adenosine infusion achieved FFR values are very similar to those obtained using femoral vein adenosine administration. However, time to maximal hyperemia is longer with the infusion via antecubital vein. Moreover, the FFR values achieved by adenosine infusion at 140 and 280 mg/kg/min are also similar. CATEGORIES IMAGING: FFR and Physiologic Lesion Assessment TCT-531 INTRACORONARY VS. INTRAVENOUS ADENOSINE FOR FRACTIONAL FLOW RESERVE MEASUREMENT: A META-ANALYSIS Stefano Rigattieri,1 Giuseppe Biondi-Zoccai,2 Alessandro Sciahbasi,3 Cristian Di RUsso,4 Maria Cera,5 Jonathan Beaudoin,6 Silvio Fedele,7 Francesco Pugliese8 1 Interventional Cardiology, Sandro Pertini Hospital, Rome, Italy, Rome, Italy; 2Sapienza University of Rome, Ospedaletti, Italy; 3Complejo Hospitalario de Jaén; 4Hospital de Puerto Real, Cádiz; 5Università degli studi di Milano, Milano, Italy; 6Québec Heart & Lung Insitute, Québec, Quebec, Canada; 7Interventional Cardiology, Sandro Pertini Hospital, Rome; 8Centre hospitalier sud francilien BACKGROUND Intravenous (IV) infusion of adenosine represents the gold standard for measuring Fractional Flow Reserve (FFR) in order to assess physiological significance of coronary lesions. However, IV adenosine preferably requires a central venous catheter and is more expensive and time-consuming as compared to
intracoronary (IC) boluses of adenosine. On the other hand, IC adenosine may be limited in inducing maximal hyperaemia and the optimal bolus dosage is still debated. We aimed to systematically review published studies comparing head-to-head IC to IV adenosine for FFR assessment in the same coronary lesions of patients undergoing coronary angiography. METHODS We performed electronic search of major databases (PubMed, Scopus, Web of Science) and retrieved published studies comparing IC to IV adenosine and reporting at least one of the following: Pd/Pa ratio values at peak hyperaemia; rate of abnormal FFR results (0.80 or <0.75 according to study definition); incidence of adverse events, such as angina-like symptoms, dyspnea, flushing and transient atrio-ventricular (AV) block . Weighted mean differences (WMD) were calculated with the inverse variance method for continuous variables, whereas relative risks (RR) were calculated with the Mantel-Haenszel method for dichotomous variables. We performed pre-specified subgroup analysis in order to appraise studies using low dose (<100 mcg) or high dose IC adenosine (100 mcg). RESULTS We retrieved 8 studies amounting to 456 patients and 489 lesions. In most studies a 140 mcg/Kg/min IV adenosine infusion was used; regarding IC adenosine, 3 studies evaluated lowdose boluses (range 20 to 80 mcg), 2 studies high-dose boluses (range 120 to 150 mcg) and 3 studies both low and high-dose boluses (range 60 to 600 mcg). In the latter studies, we extracted data relative to the lowest and the highest IC adenosine dose given. IC adenosine (both at low and high doses) was as effective as IV adenosine in inducing hyperaemia; indeed, WMD in Pd/Pa ratio at peak hyperaemia was 0.01 (95% CI -0.01, 0.02; p¼0.40). Accordingly, the rate of abnormal FFR results was similar (28.2% for IC adenosine and 31.1% for IV adenosine, RR 0.90, 95% CI 0.76, 1.08; p¼0.27). Adverse events were less frequent with IC adenosine as compared to IV adenosine (RR 0.19, 95% CI 0.07, 0.47; p<0.001). The incidence of transient AV block was 8.3% with IC adenosine (mostly related to high doses) and 4.0% with IV adenosine. CONCLUSION IC boluses of adenosine are equally effective as and might be better tolerated than IV infusion of adenosine for FFR measurement. As compared to low-dose, high-dose intracoronary adenosine has similar diagnostic accuracy but is associated with more AV blocks. CATEGORIES IMAGING: FFR and Physiologic Lesion Assessment TCT-532 Morphometric assessment of hemodynamically severe coronary stenosis relevance with optical frequency domain imaging according to vessel diameter Norihiro Kogame,1 Raisuke Iijima,2 Hidehiko Hara,3 Hitoshi Anzai,4 Masato Nakamura5 1 Toho University Ohashi Medical Center, Tokyo, Japan; 2Munich, Japan; 3Toho University Ohashi Medical Center, Tokyo, Japan; 4Ota Memorila Hospital, Yokoyama, Japan; 5Toho University Ogashi Medical Center, Tokyo, Japan BACKGROUND Fractional flow reserve (FFR) is considered the gold standard for assessing intermediate coronary lesions. Many studies have investigated the correlation between FFR value and minimum lumen area (MLA) measured by intracoronary imaging. However, it’s still unclear whether MLA of optical frequency domain imaging (OFDI) had diagnostic efficiency of coronary stenosis based on FFR value. The aim of this study was to evaluate the diagnostic efficiency of MLA measured by OFDI (OFDI-MLA) to predict FFR 0.80. METHODS We investigated all patients with suspected ischemic heart disease admitted to our hospitals between August 2013 and June 2016. If coronary angiogram showed the culprit lesion as intermediate angiographic severity, we performed OFDI and FFR examinations, and a correlation between FFR value and OFDI-MLA were evaluated. In addition, to study efficiency for different ranges of vessel diameter. RESULTS Both OFDI and FFR were done in 134 patients who had 157 coronary lesions. Angiographic stenosis severity was 51.2 13 % in diameter stenosis. Positive correlation was observed between FFR and OFDI-MLA (r ¼ 0.589, p <0.05). Significant stenoses with FFR 0.80 were observed in 114 lesions (72.6%). The diagnostic efficiency of OFDI-MLA in identifying significant stenosis was moderate with an area under the curve (AUC) 0.76, 95%CI 0.670.84. The best cut-off value of OFDI-MLA to predict FFR 0.80 was 1.98 mm2 [sensitivity 54%, specificity 85%]. When looked at