- Email: [email protected]
The impact of left ventricular ejection fraction on fractional flow reserve: Insights from the FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) trial
International Journal of Cardiology 204 (2016) 206–210
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
The impact of left ventricular ejection fraction on fractional flow reserve: Insights from the FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) trial Yuhei Kobayashi a,b, Pim A.L. Tonino c, Bernard De Bruyne d, Hyoung-Mo Yang a,e, Hong-Seok Lim a,e, Nico H.J. Pijls c, William F. Fearon a,b,⁎, on Behalf of the FAME Study Investigators: a
Stanford University Medical Center, Stanford, CA, USA Stanford Cardiovascular Institute, Stanford, CA, USA c Catharina Hospital, Eindhoven, the Netherlands d Cardiovascular Center Aalst, Aalst, Belgium e Ajou University School of Medicine, Suwon, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 5 June 2015 Received in revised form 22 November 2015 Accepted 23 November 2015 Available online 27 November 2015 Keywords: Fractional flow reserve Multivessel revascularization Ejection fraction
a b s t r a c t Background: Fractional flow reserve (FFR)-guided percutaneous coronary intervention (PCI) significantly improves outcomes compared with angio-guided PCI in patients with multivessel coronary artery disease. However, there is a theoretical concern that in patients with reduced left ventricular ejection fraction (EF) FFR may be less accurate and FFR-guided PCI less beneficial. Methods: From the FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) trial database, we compared FFR values between patients with reduced EF (both ≤40%, n = 90 and ≤50%, n = 252) and preserved EF (N40%, n = 825 and N50%, n = 663) according to the angiographic stenosis severity. We also compared differences in 1 year outcomes between FFR- vs. angio-guided PCI in patients with reduced and preserved EF. Results: Both groups had similar FFR values in lesions with 50–70% stenosis (p = 0.49) and with 71–90% stenosis (p = 0.89). The reduced EF group had a higher mean FFR compared to the preserved EF group across lesions with 91–99% stenosis (0.55 vs. 0.50, p = 0.02), although the vast majority of FFR values remained ≤0.80. There was a similar reduction in the composite end point of death, nonfatal myocardial infarction, and repeat revascularization with FFR-guided compared to angio-guided PCI for both the reduced (14.5% vs. 19.0%, relative risk = 0.76, p = 0.34) and the preserved EF group (13.8 vs. 17.0%, relative risk = 0.81, p = 0.25). The results were similar with an EF cutoff of 40%. Conclusion: Reduced EF has no influence on the FFR value unless the stenosis is very tight, in which case a theoretically explainable, but clinically irrelevant overestimation might occur. As a result, FFR-guided PCI remains beneficial regardless of EF. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Left ventricular dysfunction is present in as many as 10–30% of patients undergoing percutaneous coronary intervention (PCI) and has been shown to be a predictor of increased mortality [1–4] and medical costs [2] in these patients. Fractional flow reserve (FFR) is an invasive Abbreviations: CAD, coronary artery disease; FAME, Fractional flow reserve versus Angiography for Multivessel Evaluation; FFR, fractional flow reserve; MACE, major adverse cardiac event(s); MI, myocardial infarction; Pa, mean proximal coronary pressure; Pd, mean distal coronary pressure; Pv, mean central venous pressure; PCI, percutaneous coronary intervention; QCA, quantitative coronary angiography. ⁎ Corresponding author at: Division of Cardiovascular Medicine, Stanford University Medical Center, 300 Pasteur Drive, H2103, Stanford, CA 94305-5218, USA. E-mail address: [email protected] (W.F. Fearon).
http://dx.doi.org/10.1016/j.ijcard.2015.11.169 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
index to assess the functional severity of epicardial coronary artery disease (CAD), originally validated in patients with preserved left ventricular ejection fraction (EF) [5]. The FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) trial demonstrated the superiority of FFR-guided PCI over angiography-guided PCI in patients with multivessel CAD [6–8]. However, data remain limited on the use of FFR in patients with reduced left ventricular EF. There is a theoretical concern that FFR may be less accurate and FFR-guided PCI less effective in patients with reduced EF because of the effect of elevated left ventricular end diastolic pressure, venous pressure and nonviable left ventricular myocardium on maximal flow down a vessel and the resultant FFR. Accordingly, the primary goal of the present study is to investigate the impact of reduced EF on FFR and FFR-guided PCI in patients enrolled in the FAME trial.
Y. Kobayashi et al. / International Journal of Cardiology 204 (2016) 206–210
207
2. Methods
2.4. Statistical analysis
2.1. Study design and patient population
All patients were included in the reanalysis of the present study according to the intention-to-treat principle, as long as the EF value was available. Categorical variables, including the primary endpoint and its individual components, are presented as counts and percentages. Pearson's χ [2] test or Fisher's exact test was used for comparisons of categorical variables, as appropriate. The interaction between EF and treatment strategy was analyzed with a Breslow–Day test [11]. Continuous variables are presented as mean and standard deviation. Normality of the continuous variables was confirmed with a Shapiro–Wilk test. Depending on the result of a Levene test for homoscedasticity, variables with normal distribution were compared with a Student t-test or a Welch t-test, as appropriate. If the normality test failed, variables were compared with a Mann–Whitney U test. The Spearman's correlation coefficient (ρ) between FFR values and EF was obtained. Kaplan–Meier curves are shown for the time-to-event distributions of MACE in all enrolled patients stratified by EF group and treatment strategy. Patients were censored at 1 year (365 days) or when events occurred. Lesions with FFR measurements were divided into the 4 quadrants using cutoff values of FFR = 0.80 and %diameter stenosis = 50%, and were demonstrated on the scatter-plot graphic. An overall difference of EF among subgroups was determined by one-way ANOVA test, and differences between individual subgroups were estimated using a Games–Howell test. A two-sided p value of b 0.05 was considered statistically significant. All analyses were performed using SPSS 21 software® (SPSS Inc., Chicago, Illinois).
The detailed study protocol has been published previously [6,8,9]. In brief, the FAME trial is a prospective, randomized, controlled, multicenter trial investigating the superiority of FFR-guided PCI over angioguided PCI in patients with multivessel CAD (NCT00267774). In patients with multivessel CAD amenable to PCI, the investigators indicated which lesions had at least 50% diameter stenosis and were thought to require PCI. Thereafter, patients were randomly assigned to either FFR-guided or angiography-guided PCI. In patients assigned to FFRguided PCI, only functionally significant lesions with FFR ≤ 0.80 were treated with PCI, whereas in patients assigned to angiographyguided PCI, all indicated lesions were treated without the measurement of FFR. Patients with ST-segment elevation myocardial infarction (MI) could be enrolled if the infarction had occurred at least 5 days before PCI. On the other hand, patients with unstable angina or non-STsegment elevation MI were allowed to be enrolled earlier than 5 days if the peak creatinine kinase was b 1000 IU. Patients were excluded if they had significant left main coronary artery disease, previous coronary artery bypass surgery, cardiogenic shock, or extremely tortuous or calcified coronary arteries. Patients were further excluded from the present substudy if the EF value was not available. This study was approved by an institutional review committee of each participating site and informed consent was obtained from all patients.
3. Results 2.2. FFR measurement and treatment PCI was performed according to standard coronary interventional techniques primarily with drug-eluting stents. FFR was measured with a 0.014 in. pressure sensor guidewire (St. Jude Medical, Uppsala, Sweden). After equalization to the guide catheter pressure with the sensor positioned at the ostium of the coronary artery, the pressure guidewire was advanced down the target coronary artery. To induce maximal hyperemia, intravenous adenosine was administered at 140 μg/kg/min through a central vein. Simultaneous measurement of the mean proximal coronary pressure with the guide catheter and the mean distal coronary pressure with the pressure guidewire was performed. FFR was calculated as the ratio of the mean distal to proximal coronary pressure at hyperemia. All patients received dual antiplatelet therapy with aspirin and clopidogrel for at least 1 year after PCI [6,8,9].
2.3. End points An independent clinical events committee whose members were blinded to treatment strategy adjudicated all events. The primary endpoint of this reanalysis was same as that of the original FAME trial: major adverse cardiac events (MACE, defined as a composite of allcause death, MI, or any repeat revascularization), and its components (all-cause death, MI, repeat revascularization, and death or MI) at 1 year after the index procedure in the preserved and reduced EF groups with the cutoff value of 50%. To date, preserved EF is variably defined as an LVEF N40%, N45%, or 50% in the context of patients with heart failure [10]. In this substudy, we found that the 25th percentile of the EF value was 50% and therefore we chose this cutoff value to investigate the effect of any degree of left ventricular dysfunction. Secondary goals of this reanalysis were to assess the impact of EF on FFR values and compare FFR values between patients with preserved and reduced EF according to the angiographic stenosis severity. The above mentioned analyses were repeated with an EF cutoff of 40% to test whether the results of this substudy can be applied to the patients with more severe LV dysfunction.
An EF value was available in 915 out of 1005 patients from the FAME trial database. Overall the mean EF was 57.3 ± 10.9%, ranging from 20 to 80%. The assessment method was unknown in 14 patients and included EF calculated from left ventriculography (n = 461), echocardiography (n = 289), or scintigraphy (n = 79), and visually estimated from any of these modalities (n = 72) in the other 901 patients. Accordingly, we performed reanalysis of the 915 patients, consisting of 458 patients in the FFR-guided PCI arm and 457 patients in the angiography-guided PCI arm. 3.1. Comparisons of baseline data between the preserved vs. the reduced EF group Comparisons of clinical, angiographic, and procedural characteristics between the preserved (EF N 50%) and reduced (EF ≤ 50%) EF groups are summarized in Table 1. Mean EF values of the preserved and reduced EF groups were 62.5 ± 7.0 and 43.0 ± 8.8%, respectively (p b 0.001). Baseline patient clinical characteristics including age, sex, and risk factors were similar between two groups, except for the higher incidence of hypertension and current cigarette smoking in the reduced EF group (69.0 vs. 61.7%, p = 0.04 and 33.7 vs. 27.0%, p = 0.045, respectively). A history of MI was significantly more frequent in the reduced EF group than the preserved EF group (55.2 vs. 30.2%, p b 0.001). The number of lesions intended to treat and the total number/ length of stents used were similar between the two groups. The complexity of CAD as assessed by the SYNTAX score tended to be higher in the reduced EF group than the preserved EF group (15.7 ± 9.8 vs. 14.2 ± 8.3, p = 0.07). In the reduced EF group, procedure time was significantly longer (72.5 ± 38.0 vs. 69.6 ± 45.3 min, p = 0.03), and the volume of contrast agent used during the procedure tended to be higher (297.6 ± 127.5 vs. 284.1 ± 135.0 ml, p = 0.08). 3.2. Clinical outcomes Comparisons of outcomes at 1 year between the preserved and reduced EF patients are summarized in Table 2. The primary endpoint of
208
Y. Kobayashi et al. / International Journal of Cardiology 204 (2016) 206–210
Table 1 Clinical, angiographic and procedural characteristics.
Left ventricular EF, % Clinical characteristics Age, years Male, n (%) Diabetes, n (%) Hypertension, n (%) Hypercholesterolemia, n (%) Family history, n (%) Current smoker, n (%) Previous MI, n (%) Previous PCI, n (%) Unstable angina, n (%) EQ-5D score Angiographic and procedural characteristics Diseased vessels, n (%) 2 vessel disease 3 vessel disease Lesions intended to treat, n SYNTAX score Total implanted stents per patient, n Total stented length per patient, mm Procedure time, min Volume of contrast agent used, ml
Preserved EF (n = 663)
Reduced EF (n = 252)
p value
62.5 ± 7.0
43.0 ± 8.8
b0.001
64.4 ± 10.1 480 (72.4) 156 (23.5) 409 (61.7) 477 (71.9) 270 (40.7) 179 (27.0) 200 (30.2) 172 (25.9) 218 (32.9) 69.5 ± 16.3
64.5 ± 10.5 191 (75.8) 68 (27.0) 174 (69.0) 192 (76.2) 99 (39.3) 85 (33.7) 139 (55.2) 73 (29.0) 76 (30.2) 64.0 ± 19.4
0.94 0.30 0.28 0.04 0.20 0.69 0.045 b0.001 0.36 0.43 0.001
0.26 493 (74.4) 170 (25.6) 2.8 ± 1.0 14.2 ± 8.3 2.4 ± 1.3 44.6 ± 26.7 69.6 ± 45.3 284.1 ± 135.0
178 (70.6) 74 (29.4) 2.8 ± 0.9 15.7 ± 9.8 2.5 ± 1.3 46.2 ± 29.6 72.5 ± 38.0 297.6 ± 127.5
0.88 0.07 0.43 0.71 0.03 0.08
Values are mean ± SD or n (%). EF = ejection fraction; EQ-5D = European Quality of Life5 Dimensions; MI = myocardial infarction; PCI = percutaneous coronary intervention; SYNTAX = synergy between percutaneous coronary intervention with taxus and cardiac surgery.
MACE was similarly reduced in the FFR-guided arm both in the preserved EF group (13.8 vs. 17.0%, relative risk = 0.81, p = 0.25) and the reduced EF group (14.5 vs. 19.0%, relative risk = 0.76, p = 0.34). Death, MI, repeat revascularization, and death or MI showed similar favorable results in the FFR-guided arm (Table 2). No significant interaction between treatment strategy and EF stratification was found by the Breslow–Day test (p N 0.05 for all). Kaplan–Meier curves stratified by EF groups and treatment strategy are presented in Fig. 1. MACE at 2 years was also similarly reduced in the FFR-guided arm both in the preserved EF group [19.0 vs. 20.8%, relative risk = 0.91 (0.67–1.24), p = 0.55] and the reduced EF group [16.8 vs. 22.3%, relative risk = 0.75 (0.45–1.25), p = 0.27], without significant interaction between treatment strategy and EF stratification (Breslow–Day test, p = 0.52).
3.3. The impact of EF on FFR values In the FFR-guided PCI arm, 1199 lesions from 458 patients were interrogated with FFR measurement. A weak correlation was found between FFR values and EF (Spearman's ρ = 0.07, p = 0.01). The preserved and the reduced EF groups had similar mean FFR values across lesions with 50–70% stenosis (p = 0.49) and lesions with 71–90% stenosis (p = 0.89). The reduced EF group had a higher mean
Fig. 1. Kaplan–Meier curves stratified by EF groups (cutoff EF of 50%) and treatment strategy. Angio = angiography; EF = ejection fraction; FFR = fractional flow reserve. Blue and red solid lines represent the FFR-guided and angiography-guided arm with preserved EF (N50%). Blue and red dotted lines represent the FFR-guided and angiography-guided arm with reduced EF (≤50%). FFR-guided arm (blue solid and dotted lines) had better long-term outcomes than angiography-guided arm (red solid and dotted lines) irrespective of EF.
FFR value compared to the preserved EF group across lesions with 91–99% stenosis (FFR 0.55 ± 0.14 vs. 0.50 ± 0.14, p = 0.02), with an equal percentage of FFR values remaining at ≤ 0.80 in both groups (95.5% vs. 97.2%, p = 0.67) (Fig. 2). Among 1199 lesions, quantitative coronary angiography (QCA) data were available in 975 lesions. With cutoff values of 0.80 for FFR and 50% diameter stenosis for QCA, FFR and QCA of 975 lesions were compared and divided into 4 quadrants based on the cutoff values. Mean EF values were significantly different among the 4 quadrants (ANOVA, p = 0.002). However, when comparing one quadrant to the other, the only significant difference in mean EF value occurred between lesions with low FFR/high grade stenosis and high FFR/low grade stenosis (56.3 ± 10.6 vs. 59.5 ± 8.2%, p = 0.04). Mean EF values were similar between lesions with low FFR/high grade stenosis and lesions with high FFR/high grade stenosis (56.3 ± 10.6% vs. 57.0 ± 11.0, p = 0.82).
3.4. The results of reanalysis with EF cutoff of 40% To test the potential effect of more severe LV dysfunction, the same analyses were repeated with an EF cutoff of 40%. The mean EF values of the preserved and reduced EF groups were 59.7 ± 8.1 and 34.8 ± 6.2%, respectively (p b 0.001). Comparison of outcomes at 1 year between the preserved and reduced EF patients are summarized in Supplementary Table 1. The primary endpoint of MACE was similarly reduced in the FFR-guided arm both in the preserved EF group (EF N 40%, 13.2 vs. 16.1%, relative risk = 0.82, p = 0.24) and the reduced EF group (EF ≤ 40%, 21.4 vs.
Table 2 Comparisons of outcomes at 1 year with EF cutoff of 50%. Preserved EF (EF N 50%, n = 663)
MACE Death MI Repeat revascularization Death or MI
Reduced EF (EF ≤ 50%, n = 252)
FFR-guided (n = 327)
Angiography-guided (n = 336)
p value
Relative risk with FFR guidance
FFR-guided (n = 131)
Angiography-guided (n = 121)
p value
Relative risk with FFR guidance
Breslow–Day p value
45 (13.8) 5 (1.5) 19 (5.8) 24 (7.3) 23 (7.0)
57 (17.0) 6 (1.8) 29 (8.6) 31 (9.2) 32 (9.5)
0.25 0.80 0.16 0.38 0.25
0.81 (0.57 to 1.16) 0.86 (0.26 to 2.78) 0.67 (0.39 to 1.18) 0.80 (0.48 to 1.33) 0.74 (0.44 to 1.23)
19 (14.5) 4 (3.1) 8 (6.1) 7 (5.3) 12 (9.2)
23 (19.0) 8 (6.6) 11 (9.1) 7 (5.8) 19 (15.7)
0.34 0.19 0.37 0.88 0.11
0.76 (0.44 to 1.33) 0.46 (0.14 to 1.50) 0.67 (0.28 to 1.61) 0.92 (0.33 to 2.56) 0.58 (0.30 to 1.15)
0.85 0.45 0.99 0.79 0.56
Values are n (%). FFR = fractional flow reserve; MACE = major adverse cardiac events (defined as a composite of all-cause death, MI, or any repeat revascularization); MI = myocardial infarction; PCI = percutaneous coronary intervention. Breslow–Day indicates Breslow–Day test for heterogeneity of odds ratio of FFR-guided PCI versus the odds ratio of angiographyguided PCI within each EF group.
Y. Kobayashi et al. / International Journal of Cardiology 204 (2016) 206–210
Fig. 2. Comparisons of FFR values between preserved (EF N 50%) vs. reduced EF (EF ≤ 50%) according to the angiographic stenosis severity. EF = ejection fraction; FFR = fractional flow reserve. The preserved and reduced EF group had similar mean FFR values across the lesions with 50–70% stenosis and the lesions with 71–90% stenosis. The reduced EF group had a higher mean FFR value compared to the preserved EF group across lesions with 91–99% stenosis, although the vast majority of FFR values remained ≤0.80 in both groups.
29.2%, relative risk = 0.74, p = 0.40), without significant interaction between treatment strategy and EF stratification (Breslow–Day test, p = 0.74). Death, MI, repeat revascularization, and death or MI showed similar favorable results in the FFR-guided arm. No significant interaction between treatment strategy and EF stratification was found by the Breslow–Day test (p N 0.05 for all). Kaplan–Meier curves stratified by EF groups and treatment strategy are presented in Supplementary Fig. 1. MACE at 2 years was also similarly reduced in the FFR-guided arm both in the preserved EF group (EF N 40%, 17.5 vs. 19.8%, relative risk = 0.89, p = 0.41) and the reduced EF group (EF ≤ 40%, 26.2 vs. 33.3%, relative risk = 0.79, p = 0.46), without significant interaction between treatment strategy and EF stratification (Breslow–Day test, p = 0.70). The preserved and the reduced EF groups had similar mean FFR values across lesions with 50–70% stenosis (p = 0.74) and lesions with 71–90% stenosis (p = 0.35). The reduced EF group had a tendency to have a higher mean FFR value compared to the preserved EF group across lesions with 91–99% stenosis (FFR 0.56 ± 0.12 vs. 0.52 ± 0.15, p = 0.10) with an equal percentage of FFR values remaining at ≤ 0.80 in both groups (95.5% vs. 96.7%, p = 0.56) (Supplementary Fig. 2).
4. Discussion The principal finding of the present study is that FFR-guided PCI as compared to angiography-guided PCI appears to be equally beneficial in patients both with preserved EF and with reduced EF. In addition, across various degrees of stenosis, FFR was not different between patients with reduced and normal EF except very severe stenosis (91– 99% stenosis), in which case the mean FFR value was slightly higher in patients with reduced EF. However, because these lesions were so severe (overall the mean FFR value was 0.52 ± 0.14), this difference did not affect the proportion of lesions classified as significant based on an FFR value ≤0.80 in this study. FFR-guided PCI has been shown to be superior to angiographyguided PCI [6,8] and medical therapy [12,13] in patients with CAD. A recently published meta-analysis demonstrated that FFR-guided PCI led to revascularization roughly half as often as anatomy-based PCI, but with 20% fewer adverse events and 10% better angina relief [14]. Most patients included in these studies had preserved EF, and no study to date has specifically evaluated the impact of reduced EF on the accuracy of FFR and benefit of FFR-guided PCI.
209
The finding that FFR values differed between patients with low EF and preserved EF only across very severe stenoses demonstrates the theoretical concern regarding measuring FFR in patients with low EF and is related to the derivation of FFR itself. The actual definition of FFR is (Pd − Pv) / (Pa − Pv) at hyperemia; however, we typically ignore Pv as it is negligible in most stable patients with preserved EF, and simplify the definition to Pd / Pa (where Pa = mean proximal coronary pressure, Pd = mean distal coronary pressure, and Pv = mean central venous pressure) [15]. Patients with a low EF are more likely to have an elevated Pv, which if not accounted for in the calculation of FFR will have a greater impact on the FFR value across severe stenoses as compared to mild ones. For example, in a vessel with a mild lesion the Pa may be 100 mm Hg, the Pd may be 80 mm Hg, the Pv may be 10 mm Hg, and FFR taking into account Pv would be calculated as (80 − 10) / (100 − 10) = 0.78, whereas without Pv, it would be 80 / 100 = 0.80, resulting in a negligible 0.02 overestimation of the FFR value. However, in a vessel with a very tight stenosis, the Pa may be 100 mm Hg, the Pd may be 50 mm Hg, the Pv may be 10 mm Hg, and FFR taking into account Pv would be (50 − 10) / (100 − 10) = 0.44, whereas without Pv, it would be 50 / 100 = 0.50, resulting in 0.06 overestimation of the FFR value. Because Pv was not routinely measured in this study and FFR was calculated as Pd / Pa, one would expect the patients with low EF and very tight lesions to have slightly higher FFR values, as was seen. Since the Pd must be very low for the elevated Pv to have an effect, this overestimation of FFR only occurs across lesions with very low FFR, where it will not change the decision regarding the need for PCI, as was also seen in this study; there was an equal percentage of very tight lesions with FFR ≤ 0.80 in both patients with and without preserved EF (about 96%). Evidence supporting this finding was also found in the previous report by Ntalianis and colleagues, where patients with elevated left ventricular end diastolic pressure (a surrogate for elevated Pv) had an overestimation of FFR in non-culprit vessels at the time of an ST segment elevation MI [16]. Another explanation for why patients with reduced EF might have higher FFR values for a vessel with a given stenosis is because the vessel may supply less viable myocardium resulting in lower peak flow across the stenosis and a lower pressure gradient. For example, in a previous report by De Bruyne and colleagues, patients with previous MI and reduced EF had higher FFR values across the residual culprit stenosis, compared to patients with a similar stenosis and preserved EF [17]. In the present study, we found a similar relationship; however, it only occurred in the setting of a very severe stenosis (91–99%). This may be because vessels with stenoses this tight are more likely to have been transiently occluded and recanalized resulting in a greater chance that these vessels subtend previously infarcted myocardium, particularly in patients with reduced EF. The higher rate of previous MI in the reduced EF group supports this theory. On the other hand, less severe stenoses in patients with reduced EF may be just as likely to occur in vessels subtending viable myocardium as in patients with preserved EF. Despite this small, but significant difference in FFR values across very tight narrowings in patients with and without preserved EF, the guidance of PCI by FFR rather than angiography alone was associated with similar relative risk reduction in MACE, as well as in the individual endpoints death, MI, and repeat revascularization, and in the composite of death or MI in patients with both preserved and reduced EF. These reductions did not achieve statistical significant because this is a substudy and lacked the necessary statistical power; however, importantly the trends were similar regardless of EF and there was no significant interaction between treatment strategy and EF. Moreover, there was no significant difference in mean EF values between lesions with low FFR/high grade stenosis and high FFR/high grade stenosis. These data may have important implications because a significant proportion of multivessel CAD patients have reduced EF, and further demonstrating the role of FFR guidance may have a significant effect on treatment approach in this group.
210
Y. Kobayashi et al. / International Journal of Cardiology 204 (2016) 206–210
There are some limitations to our study. First, because several different modalities for assessing EF were used, there might be some variability in the reported EF. Second, the cutoff value for EF of 50% or 40% was somewhat arbitrary; however these were used in other studies [1,4,18]. Third, the etiology of the low EF was not well-characterized in this study; the accuracy of FFR guidance may vary in patients with ischemic cardiomyopathy as compared to non-ischemic cardiomyopathy. Fourth, because Pv and left ventricular end diastolic pressure (a possible surrogate for elevated Pv) were not routinely measured, we cannot prove that the slightly higher FFR values seen across very tight lesions in the low EF group was a result of not accounting for the elevated Pv when calculating FFR in these patients. Finally, because the FAME trial was designed to demonstrate the superiority of FFR-guided PCI over angiographyguided PCI, comparisons among subgroups stratified by EF and treatment strategy were under-powered to show statistical significance. The results of the present study should be interpreted as hypothesisgenerating. 5. Conclusion Reduced EF has no influence on the FFR value across a stenosis unless the stenosis is very tight in which case a small, theoretically explainable, but clinically irrelevant overestimation might occur. As a result, FFR-guided PCI remains beneficial compared to angiography guided PCI in patients with multivessel CAD regardless of EF. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijcard.2015.11.169. Funding sources The FAME study was sponsored by St. Jude Medical. Conflict of interest Bernard De Bruyne and Nico H.J. Pijls are consultants for St. Jude Medical. William F. Fearon receives an institutional research support from St. Jude Medical. The other authors have nothing to disclose relevant to the contents of this paper. References [1] P.C. Keelan, J.M. Johnston, T. Koru-Sengul, K.M. Detre, D.O. Williams, J. Slater, et al., Comparison of in-hospital and one-year outcomes in patients with left ventricular ejection fractions b or =40%, 41% to 49%, and N or =50% having percutaneous coronary revascularization, Am. J. Cardiol. 91 (2003) 1168–1172.
[2] M. Pohlen, H. Bunzemeier, W. Husemann, N. Roeder, G. Breithardt, H. Reinecke, Risk predictors for adverse outcomes after percutaneous coronary interventions and their related costs, Clin. Res. Cardiol. 97 (2008) 441–448. [3] T.W. Wallace, J.S. Berger, A. Wang, E.J. Velazquez, D.L. Brown, Impact of left ventricular dysfunction on hospital mortality among patients undergoing elective percutaneous coronary intervention, Am. J. Cardiol. 103 (2009) 355–360. [4] M.A. Mamas, S.G. Anderson, P.D. O'Kane, B. Keavney, J. Nolan, K.G. Oldroyd, et al., Impact of left ventricular function in relation to procedural outcomes following percutaneous coronary intervention: insights from the British Cardiovascular Intervention Society, Eur. Heart J. 35 (2014) 3004–3012a. [5] N.H. Pijls, B. De Bruyne, K. Peels, P.H. Van Der Voort, H.J. Bonnier, J.K.J.J. Bartunek, et al., Measurement of fractional flow reserve to assess the functional severity of coronary-artery stenoses, N. Engl. J. Med. 334 (1996) 1703–1708. [6] P.A. Tonino, B. De Bruyne, N.H. Pijls, U. Siebert, F. Ikeno, M. van’ t Veer, et al., Fractional flow reserve versus angiography for guiding percutaneous coronary intervention, N. Engl. J. Med. 360 (2009) 213–224. [7] W.F. Fearon, B. Bornschein, P.A. Tonino, R.M. Gothe, B.D. Bruyne, N.H. Pijls, et al., Economic evaluation of fractional flow reserve-guided percutaneous coronary intervention in patients with multivessel disease, Circulation 122 (2010) 2545–2550. [8] N.H. Pijls, W.F. Fearon, P.A. Tonino, U. Siebert, F. Ikeno, B. Bornschein, 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 (2010) 177–184. [9] W.F. Fearon, P.A. Tonino, B. De Bruyne, U. Siebert, N.H. Pijls, F.S. Investigators, Rationale and design of the fractional flow reserve versus angiography for multivessel evaluation (fame) study, Am. Heart J. 154 (2007) 632–636. [10] Heart Failure Society of America, J. Lindenfeld, N.M. Albert, J.P. Boehmer, S.P. Collins, J.A. Ezekowitz, et al., Hfsa 2010 comprehensive heart failure practice guideline, J. Card. Fail. 16 (e1-194) (2010). [11] H.-S. Lim, P.A.L. Tonino, B. De Bruyne, A.S.C. Yong, B.-K. Lee, N.H.J. Pijls, et al., The impact of age on fractional flow reserve-guided percutaneous coronary intervention: a FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) trial substudy, Int. J. Cardiol. 177 (2014) 66–70. [12] B. De Bruyne, N.H. Pijls, B. Kalesan, E. Barbato, P.A. Tonino, Z. Piroth, et al., Fractional flow reserve-guided pci versus medical therapy in stable coronary disease, N. Engl. J. Med. 367 (2012) 991–1001. [13] B. De Bruyne, W.F. Fearon, N.H. Pijls, E. Barbato, P. Tonino, Z. Piroth, et al., Fractional flow reserve-guided PCI for stable coronary artery disease, N. Engl. J. Med. (2014). [14] N.P. Johnson, G.G. Toth, D. Lai, H. Zhu, G. Acar, P. Agostoni, et al., Prognostic value of fractional flow reserve: linking physiologic severity to clinical outcomes, J. Am. Coll. Cardiol. 64 (2014) 1641–1654. [15] 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 (1993) 1354–1367. [16] A. Ntalianis, J.W. Sels, G. Davidavicius, N. Tanaka, O. Muller, C. Trana, et al., Fractional flow reserve for the assessment of nonculprit coronary artery stenoses in patients with acute myocardial infarction, J. Am. Coll. Cardiol. Intv. 3 (2010) 1274–1281. [17] B. De Bruyne, N.H. Pijls, J. Bartunek, K. Kulecki, J.W. Bech, H. De Winter, et al., Fractional flow reserve in patients with prior myocardial infarction, Circulation 104 (2001) 157–162. [18] E.M. Holper, J. Blair, F. Selzer, K.M. Detre, A.K. Jacobs, D.O. Williams, et al., The impact of ejection fraction on outcomes after percutaneous coronary intervention in patients with congestive heart failure: an analysis of the National Heart, Lung, and Blood Institute Percutaneous Transluminal Coronary Angioplasty Registry and Dynamic Registry, Am. Heart J. 151 (2006) 69–75.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
The impact of left ventricular ejection fraction on fractional flow reserve: Insights from the FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) trial Yuhei Kobayashi a,b, Pim A.L. Tonino c, Bernard De Bruyne d, Hyoung-Mo Yang a,e, Hong-Seok Lim a,e, Nico H.J. Pijls c, William F. Fearon a,b,⁎, on Behalf of the FAME Study Investigators: a
Stanford University Medical Center, Stanford, CA, USA Stanford Cardiovascular Institute, Stanford, CA, USA c Catharina Hospital, Eindhoven, the Netherlands d Cardiovascular Center Aalst, Aalst, Belgium e Ajou University School of Medicine, Suwon, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 5 June 2015 Received in revised form 22 November 2015 Accepted 23 November 2015 Available online 27 November 2015 Keywords: Fractional flow reserve Multivessel revascularization Ejection fraction
a b s t r a c t Background: Fractional flow reserve (FFR)-guided percutaneous coronary intervention (PCI) significantly improves outcomes compared with angio-guided PCI in patients with multivessel coronary artery disease. However, there is a theoretical concern that in patients with reduced left ventricular ejection fraction (EF) FFR may be less accurate and FFR-guided PCI less beneficial. Methods: From the FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) trial database, we compared FFR values between patients with reduced EF (both ≤40%, n = 90 and ≤50%, n = 252) and preserved EF (N40%, n = 825 and N50%, n = 663) according to the angiographic stenosis severity. We also compared differences in 1 year outcomes between FFR- vs. angio-guided PCI in patients with reduced and preserved EF. Results: Both groups had similar FFR values in lesions with 50–70% stenosis (p = 0.49) and with 71–90% stenosis (p = 0.89). The reduced EF group had a higher mean FFR compared to the preserved EF group across lesions with 91–99% stenosis (0.55 vs. 0.50, p = 0.02), although the vast majority of FFR values remained ≤0.80. There was a similar reduction in the composite end point of death, nonfatal myocardial infarction, and repeat revascularization with FFR-guided compared to angio-guided PCI for both the reduced (14.5% vs. 19.0%, relative risk = 0.76, p = 0.34) and the preserved EF group (13.8 vs. 17.0%, relative risk = 0.81, p = 0.25). The results were similar with an EF cutoff of 40%. Conclusion: Reduced EF has no influence on the FFR value unless the stenosis is very tight, in which case a theoretically explainable, but clinically irrelevant overestimation might occur. As a result, FFR-guided PCI remains beneficial regardless of EF. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Left ventricular dysfunction is present in as many as 10–30% of patients undergoing percutaneous coronary intervention (PCI) and has been shown to be a predictor of increased mortality [1–4] and medical costs [2] in these patients. Fractional flow reserve (FFR) is an invasive Abbreviations: CAD, coronary artery disease; FAME, Fractional flow reserve versus Angiography for Multivessel Evaluation; FFR, fractional flow reserve; MACE, major adverse cardiac event(s); MI, myocardial infarction; Pa, mean proximal coronary pressure; Pd, mean distal coronary pressure; Pv, mean central venous pressure; PCI, percutaneous coronary intervention; QCA, quantitative coronary angiography. ⁎ Corresponding author at: Division of Cardiovascular Medicine, Stanford University Medical Center, 300 Pasteur Drive, H2103, Stanford, CA 94305-5218, USA. E-mail address: [email protected] (W.F. Fearon).
http://dx.doi.org/10.1016/j.ijcard.2015.11.169 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
index to assess the functional severity of epicardial coronary artery disease (CAD), originally validated in patients with preserved left ventricular ejection fraction (EF) [5]. The FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) trial demonstrated the superiority of FFR-guided PCI over angiography-guided PCI in patients with multivessel CAD [6–8]. However, data remain limited on the use of FFR in patients with reduced left ventricular EF. There is a theoretical concern that FFR may be less accurate and FFR-guided PCI less effective in patients with reduced EF because of the effect of elevated left ventricular end diastolic pressure, venous pressure and nonviable left ventricular myocardium on maximal flow down a vessel and the resultant FFR. Accordingly, the primary goal of the present study is to investigate the impact of reduced EF on FFR and FFR-guided PCI in patients enrolled in the FAME trial.
Y. Kobayashi et al. / International Journal of Cardiology 204 (2016) 206–210
207
2. Methods
2.4. Statistical analysis
2.1. Study design and patient population
All patients were included in the reanalysis of the present study according to the intention-to-treat principle, as long as the EF value was available. Categorical variables, including the primary endpoint and its individual components, are presented as counts and percentages. Pearson's χ [2] test or Fisher's exact test was used for comparisons of categorical variables, as appropriate. The interaction between EF and treatment strategy was analyzed with a Breslow–Day test [11]. Continuous variables are presented as mean and standard deviation. Normality of the continuous variables was confirmed with a Shapiro–Wilk test. Depending on the result of a Levene test for homoscedasticity, variables with normal distribution were compared with a Student t-test or a Welch t-test, as appropriate. If the normality test failed, variables were compared with a Mann–Whitney U test. The Spearman's correlation coefficient (ρ) between FFR values and EF was obtained. Kaplan–Meier curves are shown for the time-to-event distributions of MACE in all enrolled patients stratified by EF group and treatment strategy. Patients were censored at 1 year (365 days) or when events occurred. Lesions with FFR measurements were divided into the 4 quadrants using cutoff values of FFR = 0.80 and %diameter stenosis = 50%, and were demonstrated on the scatter-plot graphic. An overall difference of EF among subgroups was determined by one-way ANOVA test, and differences between individual subgroups were estimated using a Games–Howell test. A two-sided p value of b 0.05 was considered statistically significant. All analyses were performed using SPSS 21 software® (SPSS Inc., Chicago, Illinois).
The detailed study protocol has been published previously [6,8,9]. In brief, the FAME trial is a prospective, randomized, controlled, multicenter trial investigating the superiority of FFR-guided PCI over angioguided PCI in patients with multivessel CAD (NCT00267774). In patients with multivessel CAD amenable to PCI, the investigators indicated which lesions had at least 50% diameter stenosis and were thought to require PCI. Thereafter, patients were randomly assigned to either FFR-guided or angiography-guided PCI. In patients assigned to FFRguided PCI, only functionally significant lesions with FFR ≤ 0.80 were treated with PCI, whereas in patients assigned to angiographyguided PCI, all indicated lesions were treated without the measurement of FFR. Patients with ST-segment elevation myocardial infarction (MI) could be enrolled if the infarction had occurred at least 5 days before PCI. On the other hand, patients with unstable angina or non-STsegment elevation MI were allowed to be enrolled earlier than 5 days if the peak creatinine kinase was b 1000 IU. Patients were excluded if they had significant left main coronary artery disease, previous coronary artery bypass surgery, cardiogenic shock, or extremely tortuous or calcified coronary arteries. Patients were further excluded from the present substudy if the EF value was not available. This study was approved by an institutional review committee of each participating site and informed consent was obtained from all patients.
3. Results 2.2. FFR measurement and treatment PCI was performed according to standard coronary interventional techniques primarily with drug-eluting stents. FFR was measured with a 0.014 in. pressure sensor guidewire (St. Jude Medical, Uppsala, Sweden). After equalization to the guide catheter pressure with the sensor positioned at the ostium of the coronary artery, the pressure guidewire was advanced down the target coronary artery. To induce maximal hyperemia, intravenous adenosine was administered at 140 μg/kg/min through a central vein. Simultaneous measurement of the mean proximal coronary pressure with the guide catheter and the mean distal coronary pressure with the pressure guidewire was performed. FFR was calculated as the ratio of the mean distal to proximal coronary pressure at hyperemia. All patients received dual antiplatelet therapy with aspirin and clopidogrel for at least 1 year after PCI [6,8,9].
2.3. End points An independent clinical events committee whose members were blinded to treatment strategy adjudicated all events. The primary endpoint of this reanalysis was same as that of the original FAME trial: major adverse cardiac events (MACE, defined as a composite of allcause death, MI, or any repeat revascularization), and its components (all-cause death, MI, repeat revascularization, and death or MI) at 1 year after the index procedure in the preserved and reduced EF groups with the cutoff value of 50%. To date, preserved EF is variably defined as an LVEF N40%, N45%, or 50% in the context of patients with heart failure [10]. In this substudy, we found that the 25th percentile of the EF value was 50% and therefore we chose this cutoff value to investigate the effect of any degree of left ventricular dysfunction. Secondary goals of this reanalysis were to assess the impact of EF on FFR values and compare FFR values between patients with preserved and reduced EF according to the angiographic stenosis severity. The above mentioned analyses were repeated with an EF cutoff of 40% to test whether the results of this substudy can be applied to the patients with more severe LV dysfunction.
An EF value was available in 915 out of 1005 patients from the FAME trial database. Overall the mean EF was 57.3 ± 10.9%, ranging from 20 to 80%. The assessment method was unknown in 14 patients and included EF calculated from left ventriculography (n = 461), echocardiography (n = 289), or scintigraphy (n = 79), and visually estimated from any of these modalities (n = 72) in the other 901 patients. Accordingly, we performed reanalysis of the 915 patients, consisting of 458 patients in the FFR-guided PCI arm and 457 patients in the angiography-guided PCI arm. 3.1. Comparisons of baseline data between the preserved vs. the reduced EF group Comparisons of clinical, angiographic, and procedural characteristics between the preserved (EF N 50%) and reduced (EF ≤ 50%) EF groups are summarized in Table 1. Mean EF values of the preserved and reduced EF groups were 62.5 ± 7.0 and 43.0 ± 8.8%, respectively (p b 0.001). Baseline patient clinical characteristics including age, sex, and risk factors were similar between two groups, except for the higher incidence of hypertension and current cigarette smoking in the reduced EF group (69.0 vs. 61.7%, p = 0.04 and 33.7 vs. 27.0%, p = 0.045, respectively). A history of MI was significantly more frequent in the reduced EF group than the preserved EF group (55.2 vs. 30.2%, p b 0.001). The number of lesions intended to treat and the total number/ length of stents used were similar between the two groups. The complexity of CAD as assessed by the SYNTAX score tended to be higher in the reduced EF group than the preserved EF group (15.7 ± 9.8 vs. 14.2 ± 8.3, p = 0.07). In the reduced EF group, procedure time was significantly longer (72.5 ± 38.0 vs. 69.6 ± 45.3 min, p = 0.03), and the volume of contrast agent used during the procedure tended to be higher (297.6 ± 127.5 vs. 284.1 ± 135.0 ml, p = 0.08). 3.2. Clinical outcomes Comparisons of outcomes at 1 year between the preserved and reduced EF patients are summarized in Table 2. The primary endpoint of
208
Y. Kobayashi et al. / International Journal of Cardiology 204 (2016) 206–210
Table 1 Clinical, angiographic and procedural characteristics.
Left ventricular EF, % Clinical characteristics Age, years Male, n (%) Diabetes, n (%) Hypertension, n (%) Hypercholesterolemia, n (%) Family history, n (%) Current smoker, n (%) Previous MI, n (%) Previous PCI, n (%) Unstable angina, n (%) EQ-5D score Angiographic and procedural characteristics Diseased vessels, n (%) 2 vessel disease 3 vessel disease Lesions intended to treat, n SYNTAX score Total implanted stents per patient, n Total stented length per patient, mm Procedure time, min Volume of contrast agent used, ml
Preserved EF (n = 663)
Reduced EF (n = 252)
p value
62.5 ± 7.0
43.0 ± 8.8
b0.001
64.4 ± 10.1 480 (72.4) 156 (23.5) 409 (61.7) 477 (71.9) 270 (40.7) 179 (27.0) 200 (30.2) 172 (25.9) 218 (32.9) 69.5 ± 16.3
64.5 ± 10.5 191 (75.8) 68 (27.0) 174 (69.0) 192 (76.2) 99 (39.3) 85 (33.7) 139 (55.2) 73 (29.0) 76 (30.2) 64.0 ± 19.4
0.94 0.30 0.28 0.04 0.20 0.69 0.045 b0.001 0.36 0.43 0.001
0.26 493 (74.4) 170 (25.6) 2.8 ± 1.0 14.2 ± 8.3 2.4 ± 1.3 44.6 ± 26.7 69.6 ± 45.3 284.1 ± 135.0
178 (70.6) 74 (29.4) 2.8 ± 0.9 15.7 ± 9.8 2.5 ± 1.3 46.2 ± 29.6 72.5 ± 38.0 297.6 ± 127.5
0.88 0.07 0.43 0.71 0.03 0.08
Values are mean ± SD or n (%). EF = ejection fraction; EQ-5D = European Quality of Life5 Dimensions; MI = myocardial infarction; PCI = percutaneous coronary intervention; SYNTAX = synergy between percutaneous coronary intervention with taxus and cardiac surgery.
MACE was similarly reduced in the FFR-guided arm both in the preserved EF group (13.8 vs. 17.0%, relative risk = 0.81, p = 0.25) and the reduced EF group (14.5 vs. 19.0%, relative risk = 0.76, p = 0.34). Death, MI, repeat revascularization, and death or MI showed similar favorable results in the FFR-guided arm (Table 2). No significant interaction between treatment strategy and EF stratification was found by the Breslow–Day test (p N 0.05 for all). Kaplan–Meier curves stratified by EF groups and treatment strategy are presented in Fig. 1. MACE at 2 years was also similarly reduced in the FFR-guided arm both in the preserved EF group [19.0 vs. 20.8%, relative risk = 0.91 (0.67–1.24), p = 0.55] and the reduced EF group [16.8 vs. 22.3%, relative risk = 0.75 (0.45–1.25), p = 0.27], without significant interaction between treatment strategy and EF stratification (Breslow–Day test, p = 0.52).
3.3. The impact of EF on FFR values In the FFR-guided PCI arm, 1199 lesions from 458 patients were interrogated with FFR measurement. A weak correlation was found between FFR values and EF (Spearman's ρ = 0.07, p = 0.01). The preserved and the reduced EF groups had similar mean FFR values across lesions with 50–70% stenosis (p = 0.49) and lesions with 71–90% stenosis (p = 0.89). The reduced EF group had a higher mean
Fig. 1. Kaplan–Meier curves stratified by EF groups (cutoff EF of 50%) and treatment strategy. Angio = angiography; EF = ejection fraction; FFR = fractional flow reserve. Blue and red solid lines represent the FFR-guided and angiography-guided arm with preserved EF (N50%). Blue and red dotted lines represent the FFR-guided and angiography-guided arm with reduced EF (≤50%). FFR-guided arm (blue solid and dotted lines) had better long-term outcomes than angiography-guided arm (red solid and dotted lines) irrespective of EF.
FFR value compared to the preserved EF group across lesions with 91–99% stenosis (FFR 0.55 ± 0.14 vs. 0.50 ± 0.14, p = 0.02), with an equal percentage of FFR values remaining at ≤ 0.80 in both groups (95.5% vs. 97.2%, p = 0.67) (Fig. 2). Among 1199 lesions, quantitative coronary angiography (QCA) data were available in 975 lesions. With cutoff values of 0.80 for FFR and 50% diameter stenosis for QCA, FFR and QCA of 975 lesions were compared and divided into 4 quadrants based on the cutoff values. Mean EF values were significantly different among the 4 quadrants (ANOVA, p = 0.002). However, when comparing one quadrant to the other, the only significant difference in mean EF value occurred between lesions with low FFR/high grade stenosis and high FFR/low grade stenosis (56.3 ± 10.6 vs. 59.5 ± 8.2%, p = 0.04). Mean EF values were similar between lesions with low FFR/high grade stenosis and lesions with high FFR/high grade stenosis (56.3 ± 10.6% vs. 57.0 ± 11.0, p = 0.82).
3.4. The results of reanalysis with EF cutoff of 40% To test the potential effect of more severe LV dysfunction, the same analyses were repeated with an EF cutoff of 40%. The mean EF values of the preserved and reduced EF groups were 59.7 ± 8.1 and 34.8 ± 6.2%, respectively (p b 0.001). Comparison of outcomes at 1 year between the preserved and reduced EF patients are summarized in Supplementary Table 1. The primary endpoint of MACE was similarly reduced in the FFR-guided arm both in the preserved EF group (EF N 40%, 13.2 vs. 16.1%, relative risk = 0.82, p = 0.24) and the reduced EF group (EF ≤ 40%, 21.4 vs.
Table 2 Comparisons of outcomes at 1 year with EF cutoff of 50%. Preserved EF (EF N 50%, n = 663)
MACE Death MI Repeat revascularization Death or MI
Reduced EF (EF ≤ 50%, n = 252)
FFR-guided (n = 327)
Angiography-guided (n = 336)
p value
Relative risk with FFR guidance
FFR-guided (n = 131)
Angiography-guided (n = 121)
p value
Relative risk with FFR guidance
Breslow–Day p value
45 (13.8) 5 (1.5) 19 (5.8) 24 (7.3) 23 (7.0)
57 (17.0) 6 (1.8) 29 (8.6) 31 (9.2) 32 (9.5)
0.25 0.80 0.16 0.38 0.25
0.81 (0.57 to 1.16) 0.86 (0.26 to 2.78) 0.67 (0.39 to 1.18) 0.80 (0.48 to 1.33) 0.74 (0.44 to 1.23)
19 (14.5) 4 (3.1) 8 (6.1) 7 (5.3) 12 (9.2)
23 (19.0) 8 (6.6) 11 (9.1) 7 (5.8) 19 (15.7)
0.34 0.19 0.37 0.88 0.11
0.76 (0.44 to 1.33) 0.46 (0.14 to 1.50) 0.67 (0.28 to 1.61) 0.92 (0.33 to 2.56) 0.58 (0.30 to 1.15)
0.85 0.45 0.99 0.79 0.56
Values are n (%). FFR = fractional flow reserve; MACE = major adverse cardiac events (defined as a composite of all-cause death, MI, or any repeat revascularization); MI = myocardial infarction; PCI = percutaneous coronary intervention. Breslow–Day indicates Breslow–Day test for heterogeneity of odds ratio of FFR-guided PCI versus the odds ratio of angiographyguided PCI within each EF group.
Y. Kobayashi et al. / International Journal of Cardiology 204 (2016) 206–210
Fig. 2. Comparisons of FFR values between preserved (EF N 50%) vs. reduced EF (EF ≤ 50%) according to the angiographic stenosis severity. EF = ejection fraction; FFR = fractional flow reserve. The preserved and reduced EF group had similar mean FFR values across the lesions with 50–70% stenosis and the lesions with 71–90% stenosis. The reduced EF group had a higher mean FFR value compared to the preserved EF group across lesions with 91–99% stenosis, although the vast majority of FFR values remained ≤0.80 in both groups.
29.2%, relative risk = 0.74, p = 0.40), without significant interaction between treatment strategy and EF stratification (Breslow–Day test, p = 0.74). Death, MI, repeat revascularization, and death or MI showed similar favorable results in the FFR-guided arm. No significant interaction between treatment strategy and EF stratification was found by the Breslow–Day test (p N 0.05 for all). Kaplan–Meier curves stratified by EF groups and treatment strategy are presented in Supplementary Fig. 1. MACE at 2 years was also similarly reduced in the FFR-guided arm both in the preserved EF group (EF N 40%, 17.5 vs. 19.8%, relative risk = 0.89, p = 0.41) and the reduced EF group (EF ≤ 40%, 26.2 vs. 33.3%, relative risk = 0.79, p = 0.46), without significant interaction between treatment strategy and EF stratification (Breslow–Day test, p = 0.70). The preserved and the reduced EF groups had similar mean FFR values across lesions with 50–70% stenosis (p = 0.74) and lesions with 71–90% stenosis (p = 0.35). The reduced EF group had a tendency to have a higher mean FFR value compared to the preserved EF group across lesions with 91–99% stenosis (FFR 0.56 ± 0.12 vs. 0.52 ± 0.15, p = 0.10) with an equal percentage of FFR values remaining at ≤ 0.80 in both groups (95.5% vs. 96.7%, p = 0.56) (Supplementary Fig. 2).
4. Discussion The principal finding of the present study is that FFR-guided PCI as compared to angiography-guided PCI appears to be equally beneficial in patients both with preserved EF and with reduced EF. In addition, across various degrees of stenosis, FFR was not different between patients with reduced and normal EF except very severe stenosis (91– 99% stenosis), in which case the mean FFR value was slightly higher in patients with reduced EF. However, because these lesions were so severe (overall the mean FFR value was 0.52 ± 0.14), this difference did not affect the proportion of lesions classified as significant based on an FFR value ≤0.80 in this study. FFR-guided PCI has been shown to be superior to angiographyguided PCI [6,8] and medical therapy [12,13] in patients with CAD. A recently published meta-analysis demonstrated that FFR-guided PCI led to revascularization roughly half as often as anatomy-based PCI, but with 20% fewer adverse events and 10% better angina relief [14]. Most patients included in these studies had preserved EF, and no study to date has specifically evaluated the impact of reduced EF on the accuracy of FFR and benefit of FFR-guided PCI.
209
The finding that FFR values differed between patients with low EF and preserved EF only across very severe stenoses demonstrates the theoretical concern regarding measuring FFR in patients with low EF and is related to the derivation of FFR itself. The actual definition of FFR is (Pd − Pv) / (Pa − Pv) at hyperemia; however, we typically ignore Pv as it is negligible in most stable patients with preserved EF, and simplify the definition to Pd / Pa (where Pa = mean proximal coronary pressure, Pd = mean distal coronary pressure, and Pv = mean central venous pressure) [15]. Patients with a low EF are more likely to have an elevated Pv, which if not accounted for in the calculation of FFR will have a greater impact on the FFR value across severe stenoses as compared to mild ones. For example, in a vessel with a mild lesion the Pa may be 100 mm Hg, the Pd may be 80 mm Hg, the Pv may be 10 mm Hg, and FFR taking into account Pv would be calculated as (80 − 10) / (100 − 10) = 0.78, whereas without Pv, it would be 80 / 100 = 0.80, resulting in a negligible 0.02 overestimation of the FFR value. However, in a vessel with a very tight stenosis, the Pa may be 100 mm Hg, the Pd may be 50 mm Hg, the Pv may be 10 mm Hg, and FFR taking into account Pv would be (50 − 10) / (100 − 10) = 0.44, whereas without Pv, it would be 50 / 100 = 0.50, resulting in 0.06 overestimation of the FFR value. Because Pv was not routinely measured in this study and FFR was calculated as Pd / Pa, one would expect the patients with low EF and very tight lesions to have slightly higher FFR values, as was seen. Since the Pd must be very low for the elevated Pv to have an effect, this overestimation of FFR only occurs across lesions with very low FFR, where it will not change the decision regarding the need for PCI, as was also seen in this study; there was an equal percentage of very tight lesions with FFR ≤ 0.80 in both patients with and without preserved EF (about 96%). Evidence supporting this finding was also found in the previous report by Ntalianis and colleagues, where patients with elevated left ventricular end diastolic pressure (a surrogate for elevated Pv) had an overestimation of FFR in non-culprit vessels at the time of an ST segment elevation MI [16]. Another explanation for why patients with reduced EF might have higher FFR values for a vessel with a given stenosis is because the vessel may supply less viable myocardium resulting in lower peak flow across the stenosis and a lower pressure gradient. For example, in a previous report by De Bruyne and colleagues, patients with previous MI and reduced EF had higher FFR values across the residual culprit stenosis, compared to patients with a similar stenosis and preserved EF [17]. In the present study, we found a similar relationship; however, it only occurred in the setting of a very severe stenosis (91–99%). This may be because vessels with stenoses this tight are more likely to have been transiently occluded and recanalized resulting in a greater chance that these vessels subtend previously infarcted myocardium, particularly in patients with reduced EF. The higher rate of previous MI in the reduced EF group supports this theory. On the other hand, less severe stenoses in patients with reduced EF may be just as likely to occur in vessels subtending viable myocardium as in patients with preserved EF. Despite this small, but significant difference in FFR values across very tight narrowings in patients with and without preserved EF, the guidance of PCI by FFR rather than angiography alone was associated with similar relative risk reduction in MACE, as well as in the individual endpoints death, MI, and repeat revascularization, and in the composite of death or MI in patients with both preserved and reduced EF. These reductions did not achieve statistical significant because this is a substudy and lacked the necessary statistical power; however, importantly the trends were similar regardless of EF and there was no significant interaction between treatment strategy and EF. Moreover, there was no significant difference in mean EF values between lesions with low FFR/high grade stenosis and high FFR/high grade stenosis. These data may have important implications because a significant proportion of multivessel CAD patients have reduced EF, and further demonstrating the role of FFR guidance may have a significant effect on treatment approach in this group.
210
Y. Kobayashi et al. / International Journal of Cardiology 204 (2016) 206–210
There are some limitations to our study. First, because several different modalities for assessing EF were used, there might be some variability in the reported EF. Second, the cutoff value for EF of 50% or 40% was somewhat arbitrary; however these were used in other studies [1,4,18]. Third, the etiology of the low EF was not well-characterized in this study; the accuracy of FFR guidance may vary in patients with ischemic cardiomyopathy as compared to non-ischemic cardiomyopathy. Fourth, because Pv and left ventricular end diastolic pressure (a possible surrogate for elevated Pv) were not routinely measured, we cannot prove that the slightly higher FFR values seen across very tight lesions in the low EF group was a result of not accounting for the elevated Pv when calculating FFR in these patients. Finally, because the FAME trial was designed to demonstrate the superiority of FFR-guided PCI over angiographyguided PCI, comparisons among subgroups stratified by EF and treatment strategy were under-powered to show statistical significance. The results of the present study should be interpreted as hypothesisgenerating. 5. Conclusion Reduced EF has no influence on the FFR value across a stenosis unless the stenosis is very tight in which case a small, theoretically explainable, but clinically irrelevant overestimation might occur. As a result, FFR-guided PCI remains beneficial compared to angiography guided PCI in patients with multivessel CAD regardless of EF. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijcard.2015.11.169. Funding sources The FAME study was sponsored by St. Jude Medical. Conflict of interest Bernard De Bruyne and Nico H.J. Pijls are consultants for St. Jude Medical. William F. Fearon receives an institutional research support from St. Jude Medical. The other authors have nothing to disclose relevant to the contents of this paper. References [1] P.C. Keelan, J.M. Johnston, T. Koru-Sengul, K.M. Detre, D.O. Williams, J. Slater, et al., Comparison of in-hospital and one-year outcomes in patients with left ventricular ejection fractions b or =40%, 41% to 49%, and N or =50% having percutaneous coronary revascularization, Am. J. Cardiol. 91 (2003) 1168–1172.
[2] M. Pohlen, H. Bunzemeier, W. Husemann, N. Roeder, G. Breithardt, H. Reinecke, Risk predictors for adverse outcomes after percutaneous coronary interventions and their related costs, Clin. Res. Cardiol. 97 (2008) 441–448. [3] T.W. Wallace, J.S. Berger, A. Wang, E.J. Velazquez, D.L. Brown, Impact of left ventricular dysfunction on hospital mortality among patients undergoing elective percutaneous coronary intervention, Am. J. Cardiol. 103 (2009) 355–360. [4] M.A. Mamas, S.G. Anderson, P.D. O'Kane, B. Keavney, J. Nolan, K.G. Oldroyd, et al., Impact of left ventricular function in relation to procedural outcomes following percutaneous coronary intervention: insights from the British Cardiovascular Intervention Society, Eur. Heart J. 35 (2014) 3004–3012a. [5] N.H. Pijls, B. De Bruyne, K. Peels, P.H. Van Der Voort, H.J. Bonnier, J.K.J.J. Bartunek, et al., Measurement of fractional flow reserve to assess the functional severity of coronary-artery stenoses, N. Engl. J. Med. 334 (1996) 1703–1708. [6] P.A. Tonino, B. De Bruyne, N.H. Pijls, U. Siebert, F. Ikeno, M. van’ t Veer, et al., Fractional flow reserve versus angiography for guiding percutaneous coronary intervention, N. Engl. J. Med. 360 (2009) 213–224. [7] W.F. Fearon, B. Bornschein, P.A. Tonino, R.M. Gothe, B.D. Bruyne, N.H. Pijls, et al., Economic evaluation of fractional flow reserve-guided percutaneous coronary intervention in patients with multivessel disease, Circulation 122 (2010) 2545–2550. [8] N.H. Pijls, W.F. Fearon, P.A. Tonino, U. Siebert, F. Ikeno, B. Bornschein, 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 (2010) 177–184. [9] W.F. Fearon, P.A. Tonino, B. De Bruyne, U. Siebert, N.H. Pijls, F.S. Investigators, Rationale and design of the fractional flow reserve versus angiography for multivessel evaluation (fame) study, Am. Heart J. 154 (2007) 632–636. [10] Heart Failure Society of America, J. Lindenfeld, N.M. Albert, J.P. Boehmer, S.P. Collins, J.A. Ezekowitz, et al., Hfsa 2010 comprehensive heart failure practice guideline, J. Card. Fail. 16 (e1-194) (2010). [11] H.-S. Lim, P.A.L. Tonino, B. De Bruyne, A.S.C. Yong, B.-K. Lee, N.H.J. Pijls, et al., The impact of age on fractional flow reserve-guided percutaneous coronary intervention: a FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) trial substudy, Int. J. Cardiol. 177 (2014) 66–70. [12] B. De Bruyne, N.H. Pijls, B. Kalesan, E. Barbato, P.A. Tonino, Z. Piroth, et al., Fractional flow reserve-guided pci versus medical therapy in stable coronary disease, N. Engl. J. Med. 367 (2012) 991–1001. [13] B. De Bruyne, W.F. Fearon, N.H. Pijls, E. Barbato, P. Tonino, Z. Piroth, et al., Fractional flow reserve-guided PCI for stable coronary artery disease, N. Engl. J. Med. (2014). [14] N.P. Johnson, G.G. Toth, D. Lai, H. Zhu, G. Acar, P. Agostoni, et al., Prognostic value of fractional flow reserve: linking physiologic severity to clinical outcomes, J. Am. Coll. Cardiol. 64 (2014) 1641–1654. [15] 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 (1993) 1354–1367. [16] A. Ntalianis, J.W. Sels, G. Davidavicius, N. Tanaka, O. Muller, C. Trana, et al., Fractional flow reserve for the assessment of nonculprit coronary artery stenoses in patients with acute myocardial infarction, J. Am. Coll. Cardiol. Intv. 3 (2010) 1274–1281. [17] B. De Bruyne, N.H. Pijls, J. Bartunek, K. Kulecki, J.W. Bech, H. De Winter, et al., Fractional flow reserve in patients with prior myocardial infarction, Circulation 104 (2001) 157–162. [18] E.M. Holper, J. Blair, F. Selzer, K.M. Detre, A.K. Jacobs, D.O. Williams, et al., The impact of ejection fraction on outcomes after percutaneous coronary intervention in patients with congestive heart failure: an analysis of the National Heart, Lung, and Blood Institute Percutaneous Transluminal Coronary Angioplasty Registry and Dynamic Registry, Am. Heart J. 151 (2006) 69–75.