- Email: [email protected]
Comparison of instantaneous wave-free ratio (iFR) and fractional flow reserve (FFR) — First real world experience
International Journal of Cardiology 199 (2015) 1–7
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Comparison of instantaneous wave-free ratio (iFR) and fractional flow reserve (FFR) — First real world experience Tobias Härle a,⁎,1, Waldemar Bojara b,1, Sven Meyer a,c,1, Albrecht Elsässer a,1 a b c
Klinikum Oldenburg, Klinik für Kardiologie, Oldenburg, Germany Gemeinschaftsklinikum Koblenz-Mayen, Medizinische Klinik II, Koblenz, Germany University of Groningen, University Medical Center Groningen, Department of Cardiology, Groningen, The Netherlands
a r t i c l e
i n f o
Article history: Received 16 October 2014 Received in revised form 10 March 2015 Accepted 1 July 2015 Available online 6 July 2015 Keywords: Instantaneous wave-free ratio Real-time Physiological assessment Coronary stenosis Intermediate stenosis Hydrostatic pressure
a b s t r a c t Background: The instantaneous wave-free ratio (iFR) is a new adenosine-independent index of coronary stenosis severity. Most published data have been based on off-line analyses of pressure recordings in a core laboratory. We prospectively compared real-time iFR and fractional flow reserve (FFR) measurements. Methods and results: iFR and FFR were measured in 151 coronary stenoses in 108 patients. Repeated iFR measurements were technically simple, showed excellent agreement [rs = 0.99; p b 0.0001], and the mean difference between consecutive iFR values was 0.0035 (limits of agreement: −0.019, 0.026). Mean iFR showed a significant correlation with FFR [rs = 0.81; p b 0.0001]. Receiver-operating characteristic analysis identified an optimal iFR cut-off value of 0.896 for categorization based on an FFR cut-off value 0.8. We compared two different iFRbased diagnostic strategies (iFR-only and hybrid iFR–FFR) with standard FFR: The iFR-only strategy showed good classification agreement (83.4%) with standard FFR. Use of the hybrid iFR–FFR strategy, assessing lesions in an iFR-gray zone of 0.86–0.93 by FFR, improved classification accuracy to 94.7%, and diagnosis would have been established in 61% of patients without adenosine-induced hyperemia. Notably, both iFR and FFR values were significantly higher in the posterior coronary vessels. Conclusions: Real-time iFR measurements are easily performed, have excellent diagnostic performance and confirm available off-line core laboratory data. The excellent agreement between repeated iFR measurements demonstrates the reliability of single measurements. Combining iFR with FFR in a hybrid strategy enhances diagnostic accuracy, exposing fewer patients to adenosine. Overall, iFR is a promising method, but still requires prospective clinical endpoint trial evaluation. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Evidence of inducible myocardial ischemia is a fundamental prerequisite for revascularization in stable coronary artery disease [1]. Unfortunately, no universal gold standard for diagnosis of myocardial ischemia has been established, and the correlation between angiographic findings and functional stenosis severity is limited [2]. Over the past two decades, functional assessment of coronary stenoses using intracoronary pressure guidewires has become a standard diagnostic tool in the catheterization laboratory. The implementation of fractional flow reserve (FFR) as the most widely used index in clinical practice was a milestone in interventional cardiology [3]. FFR has demonstrated the inaccuracy of angiography for evaluating the functional significance of coronary lesions with a 50% to 90% diameter stenosis
⁎ Corresponding author at: Klinikum Oldenburg gGmbH, Klinik für Kardiologie, Rahel-Straus-Str. 10, 26133 Oldenburg, Germany. E-mail address: [email protected] (T. Härle). 1 This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
http://dx.doi.org/10.1016/j.ijcard.2015.07.003 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
[4]. In the FAME trial, only 14% of patients with angiographic 3-vessel disease had functional 3-vessel disease, while 43% had 2-vessel disease, 34% suffered from single-vessel disease, and 9% had no functionally relevant coronary artery disease at all [4]. In another recent study, FFR guidance was associated with reclassification of revascularization strategy in about half of the patients [5]. Furthermore, FFR guided percutaneous coronary intervention has been shown to improve clinical outcome and procedural cost-effectiveness [6–9]. The working theory of FFR is based on the linear relationship between coronary pressure and flow under conditions of constant and minimal coronary microvascular resistance [10,11]. Therefore, FFR is calculated from pressure measurements during pharmacologicallyinduced (usually adenosine-induced) hyperemia in order to achieve the necessary resistance condition [12]. In 2012, the instantaneous wave-free ratio (iFR) was introduced as an adenosine-independent index of coronary stenosis severity. While FFR is time-averaged over several cardiac cycles, iFR is calculated as the ratio of the distal transstenotic pressure to the proximal coronary or aortic pressure during a specific diastolic wave-free period in a single cardiac cycle, when coronary resistance is most stable and minimized
2
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7 planned invasive functional assessment were included (Table 1). Despite contraindications for adenosine administration, there were no exclusion criteria. The study complies with the Declaration of Helsinki, and consent for anonymous analysis of their data was obtained from all patients.
Table 1 Patient characteristics. Number (%) Patients Age [yr ± SD] Female Single-vessel disease Multi-vessel disease Lesions Coronary vessel Left anterior descending Diagonal Intermediate Circumflex Obtuse marginal Right coronary Adenosine administration Central venous Intracoronary bolus
108 67 39 75 33 151
±11 (36.1) (69.4) (30.6)
66 12 2 20 11 42
(43.7) (7.9) (1.3) (13.2) (7.3) (27.8)
138 13
(91.4) (8.6)
2.2. Process of pressure measurements Cardiac catheterization was performed via a femoral approach. Unfractionated heparin (5000 IU) was given intravenously at the start of the procedure. A 0.014-inch pressure sensor-tipped wire (PrimeWire PrestigeTM, Volcano Corporation, San Diego, USA) was positioned at the tip of a guiding-catheter. After intracoronary administration of a nitrate and pressure equalization at the tip of the catheter, the wire was advanced into the target vessel as distally as reasonably possible for pressure recordings. First, iFR was automatically calculated on-line using the Volcano CORE System version 3.3.0 (Volcano Corporation). In 130 lesions (86.1%) iFR measurement was repeated in order to test reproducibility. In these cases, iFR was defined as the mean of both measured values. Subsequently, FFR was measured during adenosine-induced hyperemia (91.4% central intravenous administration [140 μg/kg/min], 8.6% intracoronary bolus [120 μg]). At the end of each measurement, the pressure sensor was retracted to the catheter tip to preclude pressure drift. In the event of pressure drift N0.01, the measurement was repeated. Clinical decisions were based exclusively on FFR measurements. 2.3. Diagnostic strategies and definitions
over the cardiac cycle [13]. A good classification agreement between iFR and FFR has been demonstrated, and the optimal calculated iFR cut-off was 0.89 for predicting an FFR value of 0.8 [14]. Furthermore, a hybrid iFR–FFR decision-making strategy with a revascularization cut-off b0.86, a defer cut-off N0.93, and FFR measurement between these values was theoretically calculated and recommended [15]. However, the reliability of iFR has been cause for debate in the literature since its introduction [16–18]. To date, most available data have been based on off-line analyses of pressure recordings performed in a core laboratory. The aim of this study was to prospectively compare iFR and FFR measurements in a real-world setting, with immediate online calculation of both values in the catheterization laboratory using an automated software algorithm. 2. Materials and methods 2.1. Study population This study was a prospectively designed all-comers registry. From June 2013 to April 2014, 108 consecutive patients with 151 intermediate coronary stenoses (i.e. 50–70% diameter stenosis classified by visual estimation by experienced interventionalists) and
The study design and diagnostic strategies are presented in Fig. 1. The currently recommended treatment cut-off value of ≤0.8 was used for FFR measurements. To compare iFR and FFR measurements, an iFR-only strategy and a hybrid iFR–FFR strategy were analyzed. For the iFR-only strategy, the previously reported treatment cut-off value of ≤0.89 was used [14]. For the hybrid iFR–FFR strategy, a revascularization cut-off value of b0.86 and a defer cut-off value of N0.93 were defined, as previously suggested in respect to a lack of clinical data for a distinct iFR cut-off value [15]. For iFR values between 0.86 and 0.93, agreement with a FFR-only strategy was postulated, as further investigation using additional FFR measurement would have been necessary in real life for decision making in these cases. 2.4. Statistical analyses Continuous variables are presented as mean with standard deviation (SD) and median with inter-quartile range. Categorical variables are presented as counts and percentages. We assessed the short-time reproducibility of repeated iFR measurements using Spearman's rank correlation and the Bland–Altman method, the latter providing the bias, i.e. the mean difference between both measurements, and the limits of agreement between both measurements. Comparison of the reference method, namely FFR, with the mean of two consecutive iFR measurements was performed using the Mann–Whitney U test. Correlation between FFR and mean iFR was assessed by Spearman's rank correlation (rs) and linear regression analysis. Diagnostic plots were assessed to evaluate transformations of FFR and iFR for the linear regression model and residuals were assessed for normality. No transformations were required. Conventional summary statistics for diagnostic tests, compared to a patient's true disease status as indicated by FFR ≤0.80, were
Fig. 1. Study design flow chart. Strategies for coronary revascularization decision process. A: The FFR-only strategy was used as gold standard reference for our analysis. B: In the iFR-only strategy, revascularization decisions would be based on iFR measurement with a cut-off ≤0.89 exclusively. C: In the hybrid iFR–FFR strategy, an iFR gray zone of 0.86–0.93 was defined in which decisions are based on additional FFR measurement exclusively.
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7
Fig. 2. Frequency histogram of FFR values in the study population. This distribution represents a typical clinical population for physiologically intermediate coronary stenosis with mean FFR 0.82 (SD 0.1), and 68% of values between 0.7 and 0.9. Each bar represents a 0.05 quantile.
calculated from a 2 × 2 contingency table, comparing either the iFR-only strategy or the hybrid iFR–FFR strategy with standard FFR. Nonparametric receiver-operating characteristic (ROC) analysis was performed to assess the area under the ROC curve and verify the optimal threshold for mean iFR by using the minimally important change (MIC) threshold as the cut-off level, corresponding to a 45 degree tangent line intersection. Reclassification was assessed via a 4 × 4 contingency table of true and false positives, and true and false negatives in both diagnostic approaches. STATA version 11 (StataCorp, College Station, TX, USA) was used for the analyses. 2.5. Limitations This study was neither randomized nor blinded. As clinical decisions were based on an FFR-only strategy exclusively, no clinical follow-up was performed, considering the existing data on FFR. No quantitative coronary analysis was performed, so we were not able to correlate iFR and FFR measurements with the angiographic degree of stenosis severity.
3. Results From June 2013 to April 2014, reliability of iFR was evaluated prospectively in 151 lesions in 108 patients with intermediate coronary stenoses or multivessel disease. Both iFR and FFR were measured in all patients, but clinical decisions were based exclusively on FFR. The characteristics of the patients are summarized in Table 1. The distribution of FFR values within the analyzed population was typical for intermediate stenosis, with a mean FFR of 0.82 (SD 0.1) [median 0.84 (0.75–0.88)] and 68% of values between 0.7 and 0.9 (Fig. 2). After positioning of the guide wire, automated online calculation of iFR took less than 5 s. iFR
3
measurement was technically feasible in all patients, and there were no procedure-related complications. Mean iFR was 0.89 (SD 0.09) [median 0.91 (0.83–0.97)], and short-term reproducibility of the values was excellent (rs = 0.99; p b 0.0001), and the mean difference between consecutive iFR values was 0.0035 (limits of agreement: − 0.019, 0.026), with 8 of 130 (6.2%) of measurements lying outside the limits of agreement (Fig. 3). The iFR correlated strongly with FFR (rs = 0.81; p b 0.0001) (Fig. 4). ROC analysis identified an area under the curve (AUC) of 0.9106, suggesting high accuracy of iFR as a diagnostic test for FFR (Fig. 5). The estimation of MIC thresholds revealed an iFR of 0.896 (95% confidence interval: 0.885–0.907) as the best cut-off for prediction of an FFR of 0.8 in our population. Analysis of an iFR-only strategy with a treatment cut-point ≤0.89 revealed a diagnostic classification agreement with the FFR-only strategy in 126 lesions (83.4%) with a sensitivity of 79.7%, a specificity of 85.9%, a positive predictive value of 78.3%, and a negative predictive value of 86.8%. PCI would have been deferred in 91 lesions (60.3%) based on iFR and in 92 lesion (60.9%) based on FFR. The hybrid iFR–FFR strategy accurately classified 143 lesions (94.7%) compared to FFR, and PCI would have been deferred in 92 lesions (60.9%). Using this strategy, only 42 patients (38.9%) would have been exposed to adenosine for evaluation of 51 lesions (33.8%). In 8 lesions (5.3%), there was disagreement in diagnostic classification between iFR and FFR (Fig. 6). These lesions were analyzed for differences with respect to anatomical lesion distribution. All four lesions in which iFR was b0.86 while FFR was N0.8 were located in the territory of the left anterior descending artery, while all four lesions in which iFR was N0.93 and FFR was ≤0.8 were located in the right or circumflex arteries (Fig. 7). Further analysis of the scatter plot and raw data revealed a general trend towards higher iFR as well as FFR values in the posterior coronary arteries (right coronary and circumflex arteries), compared with the anterior region (left anterior descending artery) (Table 2 and Table 3). These differences were highly significant (p b 0.0001) for iFR as well as FFR, but there was no statistically significant difference between the two methods (Fig. 8). 4. Discussion We aimed to clinically evaluate the practicability of immediate online calculation of iFR in a real world setting, as all data presented to date have been based on off-line analyses of pressure recordings in a core laboratory. Practicability and performance of real-time iFR measurement were excellent in our study, including patient comfort. Despite induction of hyperemia, the procedures for iFR and FFR
Fig. 3. Reproducibility of iFR. The reproducibility of the instantaneous wave-free ratio (iFR) in two repeated measurements was excellent. A: Linear regression (r = 0.992). B: Bland– Altman plot with limits of agreement (LoA).
4
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7
Fig. 4. Correlation of iFR with FFR. The instantaneous wave-free ratio (iFR) was found to correlate with fractional flow reseve (FFR) (r = 0.81).
measurement were completely identical, including the required equipment. After positioning the guide wire, automated online calculation of iFR was very fast (less than 5 s), and repeated measurements showed excellent short-term reproducibility. These findings agree with the recently published ADVISED-in-practice study [19]. The ability to perform functional assessments of coronary stenoses without adenosineinduced hyperemia could significantly improve the workflow in the catheterization laboratory due to reduced acquisition times. In addition, adenosine-dependent procedural costs would drop, because both the drug itself and the administration equipment would be unnecessary. Furthermore, considering adenosine contraindications and sideeffects, more patients could undergo functional assessment, and patient comfort would improve significantly. In the light of these advantages, iFR is a promising tool which may increase acceptance and use of invasive functional assessment of coronary stenoses. Remarkably, neither the iFR cut-off value of ≤ 0.89 nor the hybrid range of 0.86 to 0.93 were clinically determined, but deduced from the optimal match of iFR with the FFR value of 0.8, as previously reported [14]. However, there is limited evidence for the distinct FFR cut-off value of 0.8 itself. Originally, a cut-off value b 0.75 for evidence of myocardial ischemia was determined by comparison of FFR with noninvasive stress tests in 45 patients [3], and a cut-off value N 0.8 for 90% absence of myocardial ischemia and a gray zone of 0.75 to 0.8 was postulated [20]. In the DEFER trial, treatment of 325 patients with
Fig. 5. Diagnostic characteristics of the iFR. Classification accuracy of iFR for discrimination at the FFR ≤0.8 reference threshold was assessed by calculating the area under the receiver operating characteristic curve (ROC AUC). The ROC AUC was 0.9106 and the optimal iFR threshold was 0.895, indicating good discriminatory power.
Fig. 6. Agreement between iFR and FFR. Scatter plots of measured iFR and FFR values. A: iFR-only approach: agreement (green squares) and disagreement (red squares) according to a FFR cut-point ≤0.8 and an iFR cut-point ≤0.89 are illustrated. B: Hybrid iFR–FFR approach: agreement (green areas), disagreement (red areas), and need of further evaluation using FFR (gray zone) according to an FFR cut-point ≤0.8, an iFR treatment cut-point b0.86, and an iFR defer cut-point N0.93 are illustrated.
Fig. 7. Agreement between iFR and FFR in respect to anatomical lesion distribution. Scatter plot of measured iFR and FFR values classified for the analyzed coronary vessels.
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7 Table 2 Reclassification table between the iFR-only and the hybrid iFR–FFR strategy. Diagnostic classification of coronary stenosis as true positive (pos.), false positive, true negative (neg.), or false negative (based on FFR as reference standard) using an iFR-only or a hybrid iFR–FFR strategy is listed as numbers and percentage. Diagnostic classification using an iFR-only and a hybrid iFR–FFR strategy Hybrid iFR–FFR strategy
iFR-only strategy True pos.
False pos.
True neg.
False neg.
Total
True pos.
47 100.00 0 0.00 0 0.00 0 0.00 47 100.00
0 0.00 4 30.77 9 69.23 0 0.00 13 100.00
0 0.00 0 0.00 79 100.00 0 0.00 79 100.00
8 66.67 0 0.00 0 0.00 4 33.33 12 100.00
55 36.42 4 2.65 88 58.28 4 2.65 151 100.00
False pos. True neg. False neg. Total
intermediate coronary stenosis was based on this cut-point b0.75 for intervention and ≥0.75 for deferral of intervention. Clinical outcome of these patients was excellent after 5 years [7]. In the subsequent FAME trial, the FFR cut-point for revascularization was changed to ≤0.8. The study demonstrated a significant reduction of myocardial infarction for FFR-guided treatment [21]. This study is the foundation for current recommendations on the use of functional assessment for coronary stenoses. However, the FAME trial was not designed to define the appropriate FFR cut-off, as the distribution of FFR values in this study was relatively wide [mean FFR for ischemic lesions 0.6 (SD 0.14), mean FFR for non-ischemic lesions 0.88 (SD 0.05)] [9]. As a whole, recommended iFR cut-off values are based on relationships with the FFR cut-off value ≤0.8, for which evidence is limited. In our study, FFR distribution in the analyzed population was typical for intermediate stenosis, with a mean FFR 0.82 (SD 0.1) and 68% of values between 0.7 and 0.9. Classification agreement between online calculated iFR and FFR was 83.4% according to the iFR-only strategy. This agreement was consistent with previously reported off-line data from the ADVISE registry [14]. Furthermore, the optimal iFR cut-off value to predict an FFR ≤ 0.8 was 0.895 in our population, which was consistent with previously published data [14,22]. Unsurprisingly, classification agreement was limited close to the established cut-off values. Thus, application of the recently recommended hybrid iFR–FFR strategy showed very good classification agreement with the FFR-only reference Table 3 Comparison of mean iFR, median iFR and FFR between coronary territories and between coronary vessels. Mean values with 95%-confidence interval (95%-CI) of iFR and FFR as well as median with interquartile range (IQR) are compared between anterior [left anterior descending artery (LAD) and diagonal branches (D)] and posterior coronary territories [circumflex artery (CX) with obtuse marginal branches (OM) and right coronary artery (RCA)]. Two stenoses in the intermediate artery were excluded from comparison due to difficulties in attribution. iFR
Coronary territory Anterior (LAD, D) Posterior (CX, OM, RCA) p-Value Coronary vessel LAD Diagonal Intermediate Circumflex Obtuse marginal RCA
FFR
Mean (95%-CI)
Median (IQR)
Mean (95%-CI)
0.86 (0.84–0.88)
0.87 (0.80–0.91)
0.79 (0.77–0.81)
0.93 (0.91–0.95)
0.96 (0.91–0.99)
0.86 (0.84–0.88)
b0.0001
b0.0001
b0.0001
0.85 (0.83–0.87) 0.88 (0.85–0.91) 0.85 (0.77–0.93) 0.92 (0.87–0.97) 0.92 (0.86–0.98) 0.94 (0.92–0.96)
0.86 (0.80–0.91) 0.88 (0.85–0.92) 0.85 (0.80–0.89) 0.96 (0.92–1.00) 0.93 (0.89–0.97) 0.96 (0.91–0.99)
0.78 (0.76–0.80) 0.82 (0.79–0.85) 0.82 (0.78–0.86) 0.85 (0.80–0.90) 0.84 (0.80–0.88) 0.87 (0.84–0.90)
5
in our real world scenario. This strategy would have facilitated diagnosis without requirement of adenosine-induced hyperemia in 61% of the patients. However, in a few cases there were significant classification differences between iFR and FFR in both directions. For correct interpretation of coronary pressure measurements, it is of fundamental importance to realize that coronary circulation consists of at least three compartments: the epicardial coronary vessels (conductive vessels), the microvascular bed (resistance vessels), and the collateral vessels. Myocardial blood flow is the sum of antegrade coronary flow and collateral flow [8,12]. It is essential to understand that both FFR and iFR evaluate myocardial blood flow, not coronary blood flow [12]. Therefore, collateral flow is an important factor in coronary pressure measurements [8,12]. Adenosine acts systemically and increases not only coronary flow, but also collateral flow [12]. Furthermore, there are individual differences in the microvascular anatomy, function, and its vasodilatory reserve. Therefore, it is very likely that response to adenosine varies greatly, and maybe not even constant within individuals. These differential effects of adenosine-induced hyperemia might explain at least a part of the divergent results between iFR and FFR measurements. Generally, FFR and iFR are distinct parameters which presumably take collateral flow into account differentially, but which method better reflects the true real impact of collateral flow is currently unknown. In fact, when compared with FFR, iFR showed a stronger correlation with coronary flow velocity reserve in a recently published study [23]. As there is no established gold standard for myocardial ischemia detection, and moreover, both iFR and FFR are parameters of myocardial blood flow but not myocardial ischemia, assessing the superiority of either diagnostic tool remains challenging. Interestingly, all four lesions in which iFR was b0.86 and FFR was N 0.8 were in the territory of the left anterior descending artery, while all four lesions in which iFR was N0.93 and FFR was ≤0.8 were in the territory of the right coronary or circumflex artery. Further analysis revealed highly significant differences in both iFR and FFR values between measurements in anterior and posterior coronary territories. This may be explained by differences in coronary stenosis severity. However, since the majority of lesions were classified as intermediate stenoses by visual estimation by a small group of experienced interventionalists, this would suggest a systematic angiographic overestimation of lesions in the posterior coronary territories when compared with anterior lesions. While a possibility, this explanation seems unlikely. In fact, early analysis of the accuracy of coronary pressure measurements revealed an inverse correlation between reference vessel diameter and overestimation of a pressure decrease [24]. With respect to the frequently large diameters of the right coronary artery, this aspect could at least in part explain the observed differences. On the other hand, this phenomenon may reflect an effect of hydrostatic pressure on intracoronary pressure measurements. In a supine position, a higher hydrostatic pressure in the posterior coronary territories when compared with the anterior territory may be assumed, which could influence distal coronary pressures measured and indexed for calculation of iFR as well as FFR. This would explain the incidence of iFR values N1.0 also, which was seen predominantly in posterior vessels. To date, no data on effects of hydrostatic pressure on coronary pressure measurements have been published. However, this study was not designed to answer this question, so this explanation remains hypothetical. From our point of view, the hybrid iFR–FFR strategy could be a reasonable way to implement iFR in clinical practice, but randomized trials with clinical endpoints are mandatory. Regardless, further research is necessary to define appropriate cut-off values for both methods. However, although a single cut-off value is very attractive to interventional cardiologists, no parameter – either FFR or iFR – will be correct in every single patient, considering the complex biological system with multiple unknown variables being investigated. With regard to all of these aspects, we should accept, define and promote diagnostic gray zones in which we would be well-advised to make individualized
6
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7
Fig. 8. Comparison of different coronary territories for iFR and FFR. Error bars with mean value and 95% confidence interval are shown for measurements in the anterior coronary territory of the left anterior descending artery (LAD) including the diagonal branches compared with measurements in the posterior territories of the circumflex (CX) and right coronary artery (RCA) including their side branches. A: Mean of iFR measurements was highly significantly lower in the anterior coronary territory (p b 0.0001). B: Mean of FFR measurements was highly significantly lower in the anterior coronary territory (p b 0.0001). C: Mean differences of iFR mean and FFR measurements were not different between anterior and posterior coronary territories (p = 0.83).
decisions based on clinical aspects, rather than relying exclusively on measured values. 5. Conclusion Practicability and performance of online iFR measurements in a real-world setting were excellent. The excellent agreement between repeated iFR measurements in our study shows that single measurements are clinically reliable. Overall, our results confirm the available data from previous off-line core laboratory analyses. A decision strategy based on a hybrid iFR–FFR strategy showed high classification agreement with the FFR-only reference strategy, with only 38.9% of the patients requiring exposure to adenosine. However, in a few patients, iFR and FFR diverge for unknown reasons. Interestingly, both iFR and FFR values were significantly higher in the posterior coronary vessels. An influence of hydrostatic pressure on intracoronary pressure measurements may explain this phenomenon. With respect to adenosine side-effects and contraindications, economic aspects, and workflow in the catheterization laboratory, iFR is a promising diagnostic tool. However, a general recommendation for the use of iFR in clinical routine cannot be given until data from prospective trials with clinical endpoints demonstrate, at the very least, noninferiority compared with FFR. 6. Conflict of interest Dr. Härle and Dr. Bojara report non-financial support from Volcano Corp. during the conduct of the study. References [1] W. Wijns, P. Kolh, N. Danchin, et al., Guidelines on myocardial revascularization, Eur. Heart J. 31 (20) (2010) 2501–2555. [2] C.W. White, C.B. Wright, D.B. Doty, et al., Does visual interpretation of the coronary arteriogram predict the physiologic importance of a coronary stenosis? N. Engl. J. Med. 310 (13) (1984) 819–824. [3] N.H. Pijls, B. De Bruyne, K. Peels, et al., Measurement of fractional flow reserve to assess the functional severity of coronary-artery stenoses, N. Engl. J. Med. 334 (26) (1996) 1703–1708. [4] 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.
[5] E. Van Belle, G. Rioufol, C. Pouillot, et al., Outcome impact of coronary revascularization strategy reclassification with fractional flow reserve at time of diagnostic angiography: insights from a large French multicenter fractional flow reserve registry, Circulation 129 (2) (2014) 173–185. [6] W.F. Fearon, B. Bornschein, P.A. Tonino, et al., Economic evaluation of fractional flow reserve-guided percutaneous coronary intervention in patients with multivessel disease, Circulation 122 (24) (2010) 2545–2550. [7] 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. [8] N.H. Pijls, J.A. van Son, R.L. Kirkeeide, B. De Bruyne, K.L. Gould, Experimental basis of determining maximum coronary, myocardial, and collateral blood flow by pressure measurements for assessing functional stenosis severity before and after percutaneous transluminal coronary angioplasty, Circulation 87 (4) (1993) 1354–1367. [9] 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. [10] M.J. Kern, An adenosine-independent index of stenosis severity from coronary wave-intensity analysis: a new paradigm in coronary physiology for the cath lab? J. Am. Coll. Cardiol. 59 (15) (2012) 1403–1405. [11] J.A. Spaan, J.J. Piek, J.I. Hoffman, M. Siebes, Physiological basis of clinically used coronary hemodynamic indices, Circulation 113 (3) (2006) 446–455. [12] N.H. Pijls, B. Van Gelder, P. Van der Voort, et al., Fractional flow reserve. A useful index to evaluate the influence of an epicardial coronary stenosis on myocardial blood flow, Circulation 92 (11) (1995) 3183–3193. [13] S. Sen, J. Escaned, I.S. Malik, et al., Development and validation of a new adenosineindependent index of stenosis severity from coronary wave-intensity analysis: results of the ADVISE (ADenosine Vasodilator Independent Stenosis Evaluation) study, J. Am. Coll. Cardiol. 59 (15) (2012) 1392–1402. [14] R. Petraco, J. Escaned, S. Sen, et al., Classification performance of instantaneous wave-free ratio (iFR) and fractional flow reserve in a clinical population of intermediate coronary stenoses: results of the ADVISE registry, EuroIntervention 9 (1) (2013) 91–101. [15] R. Petraco, J.J. Park, S. Sen, et al., Hybrid iFR–FFR decision-making strategy: implications for enhancing universal adoption of physiology-guided coronary revascularisation, EuroIntervention 8 (10) (2013) 1157–1165. [16] G. Finet, G. Rioufol, A new adenosine-independent index of stenosis severity: why would one assess a coronary stenosis differently? J. Am. Coll. Cardiol. 59 (21) (2012) 1915 (author reply 1917–8). [17] N.H. Pijls, M. Van 't Veer, K.G. Oldroyd, et al., Instantaneous wave-free ratio or fractional flow reserve without hyperemia: novelty or nonsense? J. Am. Coll. Cardiol. 59 (21) (2012) 1916–1917 (author reply 1917–8). [18] W. Rudzinski, A.H. Waller, E. Kaluski, Instantaneous wave-free ratio and fractional flow reserve: close, but not close enough! J. Am. Coll. Cardiol. 59 (21) (2012) 1915–1916 (author reply 1917–8). [19] R. Petraco, R. Al-Lamee, M. Gotberg, et al., Real-time use of instantaneous wave-free ratio: results of the ADVISE in-practice: an international, multicenter evaluation of instantaneous wave-free ratio in clinical practice, Am. Heart J. 168 (5) (2014) 739–748. [20] N.H. Pijls, Is it time to measure fractional flow reserve in all patients? J. Am. Coll. Cardiol. 41 (7) (2003) 1122–1124.
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7 [21] N.H. Pijls, W.F. Fearon, P.A. Tonino, et al., Fractional flow reserve versus angiography for guiding percutaneous coronary intervention in patients with multivessel coronary artery disease: 2-year follow-up of the FAME (Fractional Flow Reserve Versus Angiography for Multivessel Evaluation) study, J. Am. Coll. Cardiol. 56 (3) (2010) 177–184. [22] N.P. Johnson, R.L. Kirkeeide, K.N. Asrress, et al., Does the instantaneous wave-free ratio approximate the fractional flow reserve? J. Am. Coll. Cardiol. 61 (13) (2013) 1428–1435. [23] R. Petraco, T.P. van de Hoef, S. Nijjer, et al., Baseline instantaneous wave-free ratio as a pressure-only estimation of underlying coronary flow reserve: results of the
7
JUSTIFY-CFR Study (Joined Coronary Pressure and Flow Analysis to Determine Diagnostic Characteristics of Basal and Hyperemic Indices of Functional Lesion Severity– Coronary Flow Reserve), Circ. Cardiovasc. Interv. 7 (4) (2014) 492–502. [24] B. De Bruyne, N.H. Pijls, W.J. Paulus, P.J. Vantrimpont, S.U. Sys, G.R. Heyndrickx, Transstenotic coronary pressure gradient measurement in humans: in vitro and in vivo evaluation of a new pressure monitoring angioplasty guide wire, J. Am. Coll. Cardiol. 22 (1) (1993) 119–126.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Comparison of instantaneous wave-free ratio (iFR) and fractional flow reserve (FFR) — First real world experience Tobias Härle a,⁎,1, Waldemar Bojara b,1, Sven Meyer a,c,1, Albrecht Elsässer a,1 a b c
Klinikum Oldenburg, Klinik für Kardiologie, Oldenburg, Germany Gemeinschaftsklinikum Koblenz-Mayen, Medizinische Klinik II, Koblenz, Germany University of Groningen, University Medical Center Groningen, Department of Cardiology, Groningen, The Netherlands
a r t i c l e
i n f o
Article history: Received 16 October 2014 Received in revised form 10 March 2015 Accepted 1 July 2015 Available online 6 July 2015 Keywords: Instantaneous wave-free ratio Real-time Physiological assessment Coronary stenosis Intermediate stenosis Hydrostatic pressure
a b s t r a c t Background: The instantaneous wave-free ratio (iFR) is a new adenosine-independent index of coronary stenosis severity. Most published data have been based on off-line analyses of pressure recordings in a core laboratory. We prospectively compared real-time iFR and fractional flow reserve (FFR) measurements. Methods and results: iFR and FFR were measured in 151 coronary stenoses in 108 patients. Repeated iFR measurements were technically simple, showed excellent agreement [rs = 0.99; p b 0.0001], and the mean difference between consecutive iFR values was 0.0035 (limits of agreement: −0.019, 0.026). Mean iFR showed a significant correlation with FFR [rs = 0.81; p b 0.0001]. Receiver-operating characteristic analysis identified an optimal iFR cut-off value of 0.896 for categorization based on an FFR cut-off value 0.8. We compared two different iFRbased diagnostic strategies (iFR-only and hybrid iFR–FFR) with standard FFR: The iFR-only strategy showed good classification agreement (83.4%) with standard FFR. Use of the hybrid iFR–FFR strategy, assessing lesions in an iFR-gray zone of 0.86–0.93 by FFR, improved classification accuracy to 94.7%, and diagnosis would have been established in 61% of patients without adenosine-induced hyperemia. Notably, both iFR and FFR values were significantly higher in the posterior coronary vessels. Conclusions: Real-time iFR measurements are easily performed, have excellent diagnostic performance and confirm available off-line core laboratory data. The excellent agreement between repeated iFR measurements demonstrates the reliability of single measurements. Combining iFR with FFR in a hybrid strategy enhances diagnostic accuracy, exposing fewer patients to adenosine. Overall, iFR is a promising method, but still requires prospective clinical endpoint trial evaluation. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Evidence of inducible myocardial ischemia is a fundamental prerequisite for revascularization in stable coronary artery disease [1]. Unfortunately, no universal gold standard for diagnosis of myocardial ischemia has been established, and the correlation between angiographic findings and functional stenosis severity is limited [2]. Over the past two decades, functional assessment of coronary stenoses using intracoronary pressure guidewires has become a standard diagnostic tool in the catheterization laboratory. The implementation of fractional flow reserve (FFR) as the most widely used index in clinical practice was a milestone in interventional cardiology [3]. FFR has demonstrated the inaccuracy of angiography for evaluating the functional significance of coronary lesions with a 50% to 90% diameter stenosis
⁎ Corresponding author at: Klinikum Oldenburg gGmbH, Klinik für Kardiologie, Rahel-Straus-Str. 10, 26133 Oldenburg, Germany. E-mail address: [email protected] (T. Härle). 1 This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
http://dx.doi.org/10.1016/j.ijcard.2015.07.003 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
[4]. In the FAME trial, only 14% of patients with angiographic 3-vessel disease had functional 3-vessel disease, while 43% had 2-vessel disease, 34% suffered from single-vessel disease, and 9% had no functionally relevant coronary artery disease at all [4]. In another recent study, FFR guidance was associated with reclassification of revascularization strategy in about half of the patients [5]. Furthermore, FFR guided percutaneous coronary intervention has been shown to improve clinical outcome and procedural cost-effectiveness [6–9]. The working theory of FFR is based on the linear relationship between coronary pressure and flow under conditions of constant and minimal coronary microvascular resistance [10,11]. Therefore, FFR is calculated from pressure measurements during pharmacologicallyinduced (usually adenosine-induced) hyperemia in order to achieve the necessary resistance condition [12]. In 2012, the instantaneous wave-free ratio (iFR) was introduced as an adenosine-independent index of coronary stenosis severity. While FFR is time-averaged over several cardiac cycles, iFR is calculated as the ratio of the distal transstenotic pressure to the proximal coronary or aortic pressure during a specific diastolic wave-free period in a single cardiac cycle, when coronary resistance is most stable and minimized
2
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7 planned invasive functional assessment were included (Table 1). Despite contraindications for adenosine administration, there were no exclusion criteria. The study complies with the Declaration of Helsinki, and consent for anonymous analysis of their data was obtained from all patients.
Table 1 Patient characteristics. Number (%) Patients Age [yr ± SD] Female Single-vessel disease Multi-vessel disease Lesions Coronary vessel Left anterior descending Diagonal Intermediate Circumflex Obtuse marginal Right coronary Adenosine administration Central venous Intracoronary bolus
108 67 39 75 33 151
±11 (36.1) (69.4) (30.6)
66 12 2 20 11 42
(43.7) (7.9) (1.3) (13.2) (7.3) (27.8)
138 13
(91.4) (8.6)
2.2. Process of pressure measurements Cardiac catheterization was performed via a femoral approach. Unfractionated heparin (5000 IU) was given intravenously at the start of the procedure. A 0.014-inch pressure sensor-tipped wire (PrimeWire PrestigeTM, Volcano Corporation, San Diego, USA) was positioned at the tip of a guiding-catheter. After intracoronary administration of a nitrate and pressure equalization at the tip of the catheter, the wire was advanced into the target vessel as distally as reasonably possible for pressure recordings. First, iFR was automatically calculated on-line using the Volcano CORE System version 3.3.0 (Volcano Corporation). In 130 lesions (86.1%) iFR measurement was repeated in order to test reproducibility. In these cases, iFR was defined as the mean of both measured values. Subsequently, FFR was measured during adenosine-induced hyperemia (91.4% central intravenous administration [140 μg/kg/min], 8.6% intracoronary bolus [120 μg]). At the end of each measurement, the pressure sensor was retracted to the catheter tip to preclude pressure drift. In the event of pressure drift N0.01, the measurement was repeated. Clinical decisions were based exclusively on FFR measurements. 2.3. Diagnostic strategies and definitions
over the cardiac cycle [13]. A good classification agreement between iFR and FFR has been demonstrated, and the optimal calculated iFR cut-off was 0.89 for predicting an FFR value of 0.8 [14]. Furthermore, a hybrid iFR–FFR decision-making strategy with a revascularization cut-off b0.86, a defer cut-off N0.93, and FFR measurement between these values was theoretically calculated and recommended [15]. However, the reliability of iFR has been cause for debate in the literature since its introduction [16–18]. To date, most available data have been based on off-line analyses of pressure recordings performed in a core laboratory. The aim of this study was to prospectively compare iFR and FFR measurements in a real-world setting, with immediate online calculation of both values in the catheterization laboratory using an automated software algorithm. 2. Materials and methods 2.1. Study population This study was a prospectively designed all-comers registry. From June 2013 to April 2014, 108 consecutive patients with 151 intermediate coronary stenoses (i.e. 50–70% diameter stenosis classified by visual estimation by experienced interventionalists) and
The study design and diagnostic strategies are presented in Fig. 1. The currently recommended treatment cut-off value of ≤0.8 was used for FFR measurements. To compare iFR and FFR measurements, an iFR-only strategy and a hybrid iFR–FFR strategy were analyzed. For the iFR-only strategy, the previously reported treatment cut-off value of ≤0.89 was used [14]. For the hybrid iFR–FFR strategy, a revascularization cut-off value of b0.86 and a defer cut-off value of N0.93 were defined, as previously suggested in respect to a lack of clinical data for a distinct iFR cut-off value [15]. For iFR values between 0.86 and 0.93, agreement with a FFR-only strategy was postulated, as further investigation using additional FFR measurement would have been necessary in real life for decision making in these cases. 2.4. Statistical analyses Continuous variables are presented as mean with standard deviation (SD) and median with inter-quartile range. Categorical variables are presented as counts and percentages. We assessed the short-time reproducibility of repeated iFR measurements using Spearman's rank correlation and the Bland–Altman method, the latter providing the bias, i.e. the mean difference between both measurements, and the limits of agreement between both measurements. Comparison of the reference method, namely FFR, with the mean of two consecutive iFR measurements was performed using the Mann–Whitney U test. Correlation between FFR and mean iFR was assessed by Spearman's rank correlation (rs) and linear regression analysis. Diagnostic plots were assessed to evaluate transformations of FFR and iFR for the linear regression model and residuals were assessed for normality. No transformations were required. Conventional summary statistics for diagnostic tests, compared to a patient's true disease status as indicated by FFR ≤0.80, were
Fig. 1. Study design flow chart. Strategies for coronary revascularization decision process. A: The FFR-only strategy was used as gold standard reference for our analysis. B: In the iFR-only strategy, revascularization decisions would be based on iFR measurement with a cut-off ≤0.89 exclusively. C: In the hybrid iFR–FFR strategy, an iFR gray zone of 0.86–0.93 was defined in which decisions are based on additional FFR measurement exclusively.
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7
Fig. 2. Frequency histogram of FFR values in the study population. This distribution represents a typical clinical population for physiologically intermediate coronary stenosis with mean FFR 0.82 (SD 0.1), and 68% of values between 0.7 and 0.9. Each bar represents a 0.05 quantile.
calculated from a 2 × 2 contingency table, comparing either the iFR-only strategy or the hybrid iFR–FFR strategy with standard FFR. Nonparametric receiver-operating characteristic (ROC) analysis was performed to assess the area under the ROC curve and verify the optimal threshold for mean iFR by using the minimally important change (MIC) threshold as the cut-off level, corresponding to a 45 degree tangent line intersection. Reclassification was assessed via a 4 × 4 contingency table of true and false positives, and true and false negatives in both diagnostic approaches. STATA version 11 (StataCorp, College Station, TX, USA) was used for the analyses. 2.5. Limitations This study was neither randomized nor blinded. As clinical decisions were based on an FFR-only strategy exclusively, no clinical follow-up was performed, considering the existing data on FFR. No quantitative coronary analysis was performed, so we were not able to correlate iFR and FFR measurements with the angiographic degree of stenosis severity.
3. Results From June 2013 to April 2014, reliability of iFR was evaluated prospectively in 151 lesions in 108 patients with intermediate coronary stenoses or multivessel disease. Both iFR and FFR were measured in all patients, but clinical decisions were based exclusively on FFR. The characteristics of the patients are summarized in Table 1. The distribution of FFR values within the analyzed population was typical for intermediate stenosis, with a mean FFR of 0.82 (SD 0.1) [median 0.84 (0.75–0.88)] and 68% of values between 0.7 and 0.9 (Fig. 2). After positioning of the guide wire, automated online calculation of iFR took less than 5 s. iFR
3
measurement was technically feasible in all patients, and there were no procedure-related complications. Mean iFR was 0.89 (SD 0.09) [median 0.91 (0.83–0.97)], and short-term reproducibility of the values was excellent (rs = 0.99; p b 0.0001), and the mean difference between consecutive iFR values was 0.0035 (limits of agreement: − 0.019, 0.026), with 8 of 130 (6.2%) of measurements lying outside the limits of agreement (Fig. 3). The iFR correlated strongly with FFR (rs = 0.81; p b 0.0001) (Fig. 4). ROC analysis identified an area under the curve (AUC) of 0.9106, suggesting high accuracy of iFR as a diagnostic test for FFR (Fig. 5). The estimation of MIC thresholds revealed an iFR of 0.896 (95% confidence interval: 0.885–0.907) as the best cut-off for prediction of an FFR of 0.8 in our population. Analysis of an iFR-only strategy with a treatment cut-point ≤0.89 revealed a diagnostic classification agreement with the FFR-only strategy in 126 lesions (83.4%) with a sensitivity of 79.7%, a specificity of 85.9%, a positive predictive value of 78.3%, and a negative predictive value of 86.8%. PCI would have been deferred in 91 lesions (60.3%) based on iFR and in 92 lesion (60.9%) based on FFR. The hybrid iFR–FFR strategy accurately classified 143 lesions (94.7%) compared to FFR, and PCI would have been deferred in 92 lesions (60.9%). Using this strategy, only 42 patients (38.9%) would have been exposed to adenosine for evaluation of 51 lesions (33.8%). In 8 lesions (5.3%), there was disagreement in diagnostic classification between iFR and FFR (Fig. 6). These lesions were analyzed for differences with respect to anatomical lesion distribution. All four lesions in which iFR was b0.86 while FFR was N0.8 were located in the territory of the left anterior descending artery, while all four lesions in which iFR was N0.93 and FFR was ≤0.8 were located in the right or circumflex arteries (Fig. 7). Further analysis of the scatter plot and raw data revealed a general trend towards higher iFR as well as FFR values in the posterior coronary arteries (right coronary and circumflex arteries), compared with the anterior region (left anterior descending artery) (Table 2 and Table 3). These differences were highly significant (p b 0.0001) for iFR as well as FFR, but there was no statistically significant difference between the two methods (Fig. 8). 4. Discussion We aimed to clinically evaluate the practicability of immediate online calculation of iFR in a real world setting, as all data presented to date have been based on off-line analyses of pressure recordings in a core laboratory. Practicability and performance of real-time iFR measurement were excellent in our study, including patient comfort. Despite induction of hyperemia, the procedures for iFR and FFR
Fig. 3. Reproducibility of iFR. The reproducibility of the instantaneous wave-free ratio (iFR) in two repeated measurements was excellent. A: Linear regression (r = 0.992). B: Bland– Altman plot with limits of agreement (LoA).
4
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7
Fig. 4. Correlation of iFR with FFR. The instantaneous wave-free ratio (iFR) was found to correlate with fractional flow reseve (FFR) (r = 0.81).
measurement were completely identical, including the required equipment. After positioning the guide wire, automated online calculation of iFR was very fast (less than 5 s), and repeated measurements showed excellent short-term reproducibility. These findings agree with the recently published ADVISED-in-practice study [19]. The ability to perform functional assessments of coronary stenoses without adenosineinduced hyperemia could significantly improve the workflow in the catheterization laboratory due to reduced acquisition times. In addition, adenosine-dependent procedural costs would drop, because both the drug itself and the administration equipment would be unnecessary. Furthermore, considering adenosine contraindications and sideeffects, more patients could undergo functional assessment, and patient comfort would improve significantly. In the light of these advantages, iFR is a promising tool which may increase acceptance and use of invasive functional assessment of coronary stenoses. Remarkably, neither the iFR cut-off value of ≤ 0.89 nor the hybrid range of 0.86 to 0.93 were clinically determined, but deduced from the optimal match of iFR with the FFR value of 0.8, as previously reported [14]. However, there is limited evidence for the distinct FFR cut-off value of 0.8 itself. Originally, a cut-off value b 0.75 for evidence of myocardial ischemia was determined by comparison of FFR with noninvasive stress tests in 45 patients [3], and a cut-off value N 0.8 for 90% absence of myocardial ischemia and a gray zone of 0.75 to 0.8 was postulated [20]. In the DEFER trial, treatment of 325 patients with
Fig. 5. Diagnostic characteristics of the iFR. Classification accuracy of iFR for discrimination at the FFR ≤0.8 reference threshold was assessed by calculating the area under the receiver operating characteristic curve (ROC AUC). The ROC AUC was 0.9106 and the optimal iFR threshold was 0.895, indicating good discriminatory power.
Fig. 6. Agreement between iFR and FFR. Scatter plots of measured iFR and FFR values. A: iFR-only approach: agreement (green squares) and disagreement (red squares) according to a FFR cut-point ≤0.8 and an iFR cut-point ≤0.89 are illustrated. B: Hybrid iFR–FFR approach: agreement (green areas), disagreement (red areas), and need of further evaluation using FFR (gray zone) according to an FFR cut-point ≤0.8, an iFR treatment cut-point b0.86, and an iFR defer cut-point N0.93 are illustrated.
Fig. 7. Agreement between iFR and FFR in respect to anatomical lesion distribution. Scatter plot of measured iFR and FFR values classified for the analyzed coronary vessels.
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7 Table 2 Reclassification table between the iFR-only and the hybrid iFR–FFR strategy. Diagnostic classification of coronary stenosis as true positive (pos.), false positive, true negative (neg.), or false negative (based on FFR as reference standard) using an iFR-only or a hybrid iFR–FFR strategy is listed as numbers and percentage. Diagnostic classification using an iFR-only and a hybrid iFR–FFR strategy Hybrid iFR–FFR strategy
iFR-only strategy True pos.
False pos.
True neg.
False neg.
Total
True pos.
47 100.00 0 0.00 0 0.00 0 0.00 47 100.00
0 0.00 4 30.77 9 69.23 0 0.00 13 100.00
0 0.00 0 0.00 79 100.00 0 0.00 79 100.00
8 66.67 0 0.00 0 0.00 4 33.33 12 100.00
55 36.42 4 2.65 88 58.28 4 2.65 151 100.00
False pos. True neg. False neg. Total
intermediate coronary stenosis was based on this cut-point b0.75 for intervention and ≥0.75 for deferral of intervention. Clinical outcome of these patients was excellent after 5 years [7]. In the subsequent FAME trial, the FFR cut-point for revascularization was changed to ≤0.8. The study demonstrated a significant reduction of myocardial infarction for FFR-guided treatment [21]. This study is the foundation for current recommendations on the use of functional assessment for coronary stenoses. However, the FAME trial was not designed to define the appropriate FFR cut-off, as the distribution of FFR values in this study was relatively wide [mean FFR for ischemic lesions 0.6 (SD 0.14), mean FFR for non-ischemic lesions 0.88 (SD 0.05)] [9]. As a whole, recommended iFR cut-off values are based on relationships with the FFR cut-off value ≤0.8, for which evidence is limited. In our study, FFR distribution in the analyzed population was typical for intermediate stenosis, with a mean FFR 0.82 (SD 0.1) and 68% of values between 0.7 and 0.9. Classification agreement between online calculated iFR and FFR was 83.4% according to the iFR-only strategy. This agreement was consistent with previously reported off-line data from the ADVISE registry [14]. Furthermore, the optimal iFR cut-off value to predict an FFR ≤ 0.8 was 0.895 in our population, which was consistent with previously published data [14,22]. Unsurprisingly, classification agreement was limited close to the established cut-off values. Thus, application of the recently recommended hybrid iFR–FFR strategy showed very good classification agreement with the FFR-only reference Table 3 Comparison of mean iFR, median iFR and FFR between coronary territories and between coronary vessels. Mean values with 95%-confidence interval (95%-CI) of iFR and FFR as well as median with interquartile range (IQR) are compared between anterior [left anterior descending artery (LAD) and diagonal branches (D)] and posterior coronary territories [circumflex artery (CX) with obtuse marginal branches (OM) and right coronary artery (RCA)]. Two stenoses in the intermediate artery were excluded from comparison due to difficulties in attribution. iFR
Coronary territory Anterior (LAD, D) Posterior (CX, OM, RCA) p-Value Coronary vessel LAD Diagonal Intermediate Circumflex Obtuse marginal RCA
FFR
Mean (95%-CI)
Median (IQR)
Mean (95%-CI)
0.86 (0.84–0.88)
0.87 (0.80–0.91)
0.79 (0.77–0.81)
0.93 (0.91–0.95)
0.96 (0.91–0.99)
0.86 (0.84–0.88)
b0.0001
b0.0001
b0.0001
0.85 (0.83–0.87) 0.88 (0.85–0.91) 0.85 (0.77–0.93) 0.92 (0.87–0.97) 0.92 (0.86–0.98) 0.94 (0.92–0.96)
0.86 (0.80–0.91) 0.88 (0.85–0.92) 0.85 (0.80–0.89) 0.96 (0.92–1.00) 0.93 (0.89–0.97) 0.96 (0.91–0.99)
0.78 (0.76–0.80) 0.82 (0.79–0.85) 0.82 (0.78–0.86) 0.85 (0.80–0.90) 0.84 (0.80–0.88) 0.87 (0.84–0.90)
5
in our real world scenario. This strategy would have facilitated diagnosis without requirement of adenosine-induced hyperemia in 61% of the patients. However, in a few cases there were significant classification differences between iFR and FFR in both directions. For correct interpretation of coronary pressure measurements, it is of fundamental importance to realize that coronary circulation consists of at least three compartments: the epicardial coronary vessels (conductive vessels), the microvascular bed (resistance vessels), and the collateral vessels. Myocardial blood flow is the sum of antegrade coronary flow and collateral flow [8,12]. It is essential to understand that both FFR and iFR evaluate myocardial blood flow, not coronary blood flow [12]. Therefore, collateral flow is an important factor in coronary pressure measurements [8,12]. Adenosine acts systemically and increases not only coronary flow, but also collateral flow [12]. Furthermore, there are individual differences in the microvascular anatomy, function, and its vasodilatory reserve. Therefore, it is very likely that response to adenosine varies greatly, and maybe not even constant within individuals. These differential effects of adenosine-induced hyperemia might explain at least a part of the divergent results between iFR and FFR measurements. Generally, FFR and iFR are distinct parameters which presumably take collateral flow into account differentially, but which method better reflects the true real impact of collateral flow is currently unknown. In fact, when compared with FFR, iFR showed a stronger correlation with coronary flow velocity reserve in a recently published study [23]. As there is no established gold standard for myocardial ischemia detection, and moreover, both iFR and FFR are parameters of myocardial blood flow but not myocardial ischemia, assessing the superiority of either diagnostic tool remains challenging. Interestingly, all four lesions in which iFR was b0.86 and FFR was N 0.8 were in the territory of the left anterior descending artery, while all four lesions in which iFR was N0.93 and FFR was ≤0.8 were in the territory of the right coronary or circumflex artery. Further analysis revealed highly significant differences in both iFR and FFR values between measurements in anterior and posterior coronary territories. This may be explained by differences in coronary stenosis severity. However, since the majority of lesions were classified as intermediate stenoses by visual estimation by a small group of experienced interventionalists, this would suggest a systematic angiographic overestimation of lesions in the posterior coronary territories when compared with anterior lesions. While a possibility, this explanation seems unlikely. In fact, early analysis of the accuracy of coronary pressure measurements revealed an inverse correlation between reference vessel diameter and overestimation of a pressure decrease [24]. With respect to the frequently large diameters of the right coronary artery, this aspect could at least in part explain the observed differences. On the other hand, this phenomenon may reflect an effect of hydrostatic pressure on intracoronary pressure measurements. In a supine position, a higher hydrostatic pressure in the posterior coronary territories when compared with the anterior territory may be assumed, which could influence distal coronary pressures measured and indexed for calculation of iFR as well as FFR. This would explain the incidence of iFR values N1.0 also, which was seen predominantly in posterior vessels. To date, no data on effects of hydrostatic pressure on coronary pressure measurements have been published. However, this study was not designed to answer this question, so this explanation remains hypothetical. From our point of view, the hybrid iFR–FFR strategy could be a reasonable way to implement iFR in clinical practice, but randomized trials with clinical endpoints are mandatory. Regardless, further research is necessary to define appropriate cut-off values for both methods. However, although a single cut-off value is very attractive to interventional cardiologists, no parameter – either FFR or iFR – will be correct in every single patient, considering the complex biological system with multiple unknown variables being investigated. With regard to all of these aspects, we should accept, define and promote diagnostic gray zones in which we would be well-advised to make individualized
6
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7
Fig. 8. Comparison of different coronary territories for iFR and FFR. Error bars with mean value and 95% confidence interval are shown for measurements in the anterior coronary territory of the left anterior descending artery (LAD) including the diagonal branches compared with measurements in the posterior territories of the circumflex (CX) and right coronary artery (RCA) including their side branches. A: Mean of iFR measurements was highly significantly lower in the anterior coronary territory (p b 0.0001). B: Mean of FFR measurements was highly significantly lower in the anterior coronary territory (p b 0.0001). C: Mean differences of iFR mean and FFR measurements were not different between anterior and posterior coronary territories (p = 0.83).
decisions based on clinical aspects, rather than relying exclusively on measured values. 5. Conclusion Practicability and performance of online iFR measurements in a real-world setting were excellent. The excellent agreement between repeated iFR measurements in our study shows that single measurements are clinically reliable. Overall, our results confirm the available data from previous off-line core laboratory analyses. A decision strategy based on a hybrid iFR–FFR strategy showed high classification agreement with the FFR-only reference strategy, with only 38.9% of the patients requiring exposure to adenosine. However, in a few patients, iFR and FFR diverge for unknown reasons. Interestingly, both iFR and FFR values were significantly higher in the posterior coronary vessels. An influence of hydrostatic pressure on intracoronary pressure measurements may explain this phenomenon. With respect to adenosine side-effects and contraindications, economic aspects, and workflow in the catheterization laboratory, iFR is a promising diagnostic tool. However, a general recommendation for the use of iFR in clinical routine cannot be given until data from prospective trials with clinical endpoints demonstrate, at the very least, noninferiority compared with FFR. 6. Conflict of interest Dr. Härle and Dr. Bojara report non-financial support from Volcano Corp. during the conduct of the study. References [1] W. Wijns, P. Kolh, N. Danchin, et al., Guidelines on myocardial revascularization, Eur. Heart J. 31 (20) (2010) 2501–2555. [2] C.W. White, C.B. Wright, D.B. Doty, et al., Does visual interpretation of the coronary arteriogram predict the physiologic importance of a coronary stenosis? N. Engl. J. Med. 310 (13) (1984) 819–824. [3] N.H. Pijls, B. De Bruyne, K. Peels, et al., Measurement of fractional flow reserve to assess the functional severity of coronary-artery stenoses, N. Engl. J. Med. 334 (26) (1996) 1703–1708. [4] 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.
[5] E. Van Belle, G. Rioufol, C. Pouillot, et al., Outcome impact of coronary revascularization strategy reclassification with fractional flow reserve at time of diagnostic angiography: insights from a large French multicenter fractional flow reserve registry, Circulation 129 (2) (2014) 173–185. [6] W.F. Fearon, B. Bornschein, P.A. Tonino, et al., Economic evaluation of fractional flow reserve-guided percutaneous coronary intervention in patients with multivessel disease, Circulation 122 (24) (2010) 2545–2550. [7] 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. [8] N.H. Pijls, J.A. van Son, R.L. Kirkeeide, B. De Bruyne, K.L. Gould, Experimental basis of determining maximum coronary, myocardial, and collateral blood flow by pressure measurements for assessing functional stenosis severity before and after percutaneous transluminal coronary angioplasty, Circulation 87 (4) (1993) 1354–1367. [9] 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. [10] M.J. Kern, An adenosine-independent index of stenosis severity from coronary wave-intensity analysis: a new paradigm in coronary physiology for the cath lab? J. Am. Coll. Cardiol. 59 (15) (2012) 1403–1405. [11] J.A. Spaan, J.J. Piek, J.I. Hoffman, M. Siebes, Physiological basis of clinically used coronary hemodynamic indices, Circulation 113 (3) (2006) 446–455. [12] N.H. Pijls, B. Van Gelder, P. Van der Voort, et al., Fractional flow reserve. A useful index to evaluate the influence of an epicardial coronary stenosis on myocardial blood flow, Circulation 92 (11) (1995) 3183–3193. [13] S. Sen, J. Escaned, I.S. Malik, et al., Development and validation of a new adenosineindependent index of stenosis severity from coronary wave-intensity analysis: results of the ADVISE (ADenosine Vasodilator Independent Stenosis Evaluation) study, J. Am. Coll. Cardiol. 59 (15) (2012) 1392–1402. [14] R. Petraco, J. Escaned, S. Sen, et al., Classification performance of instantaneous wave-free ratio (iFR) and fractional flow reserve in a clinical population of intermediate coronary stenoses: results of the ADVISE registry, EuroIntervention 9 (1) (2013) 91–101. [15] R. Petraco, J.J. Park, S. Sen, et al., Hybrid iFR–FFR decision-making strategy: implications for enhancing universal adoption of physiology-guided coronary revascularisation, EuroIntervention 8 (10) (2013) 1157–1165. [16] G. Finet, G. Rioufol, A new adenosine-independent index of stenosis severity: why would one assess a coronary stenosis differently? J. Am. Coll. Cardiol. 59 (21) (2012) 1915 (author reply 1917–8). [17] N.H. Pijls, M. Van 't Veer, K.G. Oldroyd, et al., Instantaneous wave-free ratio or fractional flow reserve without hyperemia: novelty or nonsense? J. Am. Coll. Cardiol. 59 (21) (2012) 1916–1917 (author reply 1917–8). [18] W. Rudzinski, A.H. Waller, E. Kaluski, Instantaneous wave-free ratio and fractional flow reserve: close, but not close enough! J. Am. Coll. Cardiol. 59 (21) (2012) 1915–1916 (author reply 1917–8). [19] R. Petraco, R. Al-Lamee, M. Gotberg, et al., Real-time use of instantaneous wave-free ratio: results of the ADVISE in-practice: an international, multicenter evaluation of instantaneous wave-free ratio in clinical practice, Am. Heart J. 168 (5) (2014) 739–748. [20] N.H. Pijls, Is it time to measure fractional flow reserve in all patients? J. Am. Coll. Cardiol. 41 (7) (2003) 1122–1124.
T. Härle et al. / International Journal of Cardiology 199 (2015) 1–7 [21] N.H. Pijls, W.F. Fearon, P.A. Tonino, et al., Fractional flow reserve versus angiography for guiding percutaneous coronary intervention in patients with multivessel coronary artery disease: 2-year follow-up of the FAME (Fractional Flow Reserve Versus Angiography for Multivessel Evaluation) study, J. Am. Coll. Cardiol. 56 (3) (2010) 177–184. [22] N.P. Johnson, R.L. Kirkeeide, K.N. Asrress, et al., Does the instantaneous wave-free ratio approximate the fractional flow reserve? J. Am. Coll. Cardiol. 61 (13) (2013) 1428–1435. [23] R. Petraco, T.P. van de Hoef, S. Nijjer, et al., Baseline instantaneous wave-free ratio as a pressure-only estimation of underlying coronary flow reserve: results of the
7
JUSTIFY-CFR Study (Joined Coronary Pressure and Flow Analysis to Determine Diagnostic Characteristics of Basal and Hyperemic Indices of Functional Lesion Severity– Coronary Flow Reserve), Circ. Cardiovasc. Interv. 7 (4) (2014) 492–502. [24] B. De Bruyne, N.H. Pijls, W.J. Paulus, P.J. Vantrimpont, S.U. Sys, G.R. Heyndrickx, Transstenotic coronary pressure gradient measurement in humans: in vitro and in vivo evaluation of a new pressure monitoring angioplasty guide wire, J. Am. Coll. Cardiol. 22 (1) (1993) 119–126.