The importance of integrating coronary physiology into

Transcription

1 Controversies in Cardiovascular Medicine Fractional Flow Reserve-Guided Percutaneous Coronary Intervention: Is it a Valid Concept? Percutaneous Coronary Intervention Should Be Guided by Fractional Flow Reserve Measurement William F. Fearon, MD The importance of integrating coronary physiology into percutaneous coronary intervention (PCI) has been recognized since its inception. When Andreus Gruentzig first described percutaneous transluminal coronary angioplasty, he stressed the need to identify the translesional pressure gradient before and after intervention as a means of guiding a successful procedure. 1 Subsequently, other investigators examined the use of the Doppler wire to measure coronary flow velocity reserve before and after PCI. 2 Unfortunately, the lack of a low-profile pressure-monitoring device and the lack of appreciation of the importance of measuring hyperemic pressure gradients inhibited the full integration of coronary physiology into the PCI procedure. Additionally, with the advent of stents, the ease and safety of PCI improved dramatically. In conjunction with financial reimbursement for PCI and patient demand, these factors resulted in an inclination toward performing PCI, regardless of the functional significance of the stenosis. Response by Arbab-Zadeh p 1870 While this enthusiasm over PCI was occurring, Pijls and coworkers 3 and De Bruyne and colleagues 4 first described fractional flow reserve (FFR). Like any new technique, FFR has required time to mature. Over the ensuing 20 years since its introduction, improvements in the pressure wire handling characteristics, the integration of data acquisition into the catheterization laboratory workflow, the completion of a number of multicenter, randomized studies, and the increased emphasis on appropriate use of PCI have contributed to FFR achieving its current status as an indispensable and critical component of PCI. Concept of FFR FFR is defined as the ratio of myocardial blood flow down a coronary artery in the presence of an epicardial stenosis compared with the flow down the same vessel in the theoretical absence of any stenosis. 3 The derivation of FFR has been described in detail and is shown in brief in Figure 1. 3,4 FFR can be calculated easily by measuring the mean distal coronary pressure with a coronary pressure wire and dividing it by the mean proximal coronary pressure, measured with a guiding catheter, during maximal hyperemia (Figure 2A and 2B). FFR was first validated in a landmark study by Pijls, De Bruyne, and colleagues 5 in which they measured FFR in 45 patients with single-vessel intermediate coronary disease. Because there is no true noninvasive reference standard for diagnosing myocardial ischemia and to improve the accuracy of existing tests, every patient underwent exercise testing, dobutamine stress echocardiography, and exercise thallium scintigraphy. According to the Bayes theorem and on the basis of the diagnostic accuracy of each individual stress test, the composite accuracy of these 3 consecutive tests for myocardial ischemia approaches 100%. In every patient with an FFR <0.75, myocardial ischemia was present on at least 1 test, and in 21 of 24 patients with FFR values of 0.75, myocardial ischemia was absent on all 3 tests, giving a sensitivity, specificity, and diagnostic accuracy of FFR for identifying lesions capable of inducing myocardial ischemia of 88%, 100%, and 93%, respectively. In the 24 patients with an FFR of 0.75, there were no ischemic events during 14 months of follow-up. This small study formed the cornerstone for a number of larger studies that validated The opinions expressed in this article are not necessarily those of the editors or of the American Heart Association. From the Division of Cardiovascular Medicine, Stanford University Medical Center, Stanford, CA. This article is Part I of a 2-part article. Part II appears on p Correspondence to William F. Fearon, MD, Director, Interventional Cardiology, Associate Professor of Medicine, Stanford University Medical Center, 300 Pasteur Dr, H2103, Stanford, CA wfearon@stanford.edu (Circulation. 2014;129: ) 2014 American Heart Association, Inc. Circulation is available at DOI: /CIRCULATIONAHA

2 Fearon FFR-Guided PCI 1861 Derivation of Fractional Flow Reserve FFR = Maximum Myocardial Flow (Stenosis) Maximum Myocardial Flow (Normal) Δ Pressure Myocardial Flow = Resistance FFR = (P d-p v )/ Resistance at maximal hyperemia (P a -P v ) / Resistance P d -P v FFR = P a -P v P d FFR = P a Figure 1. Derivation of fractional flow reserve (FFR). During maximal hyperemia, resistance is minimized and assumed to be unchanged in the presence and absence of an epicardial stenosis; therefore, it is removed from the equation. Venous pressure is generally negligible in relation to coronary pressure and assumed to equal zero. P a indicates proximal coronary or aortic pressure; P d, distal coronary pressure; and P v, venous pressure. FFR in a variety of patient populations and clinical scenarios, and it demonstrated the safety of deferring revascularization of lesions that are not functionally significant (nonischemic FFR). Practical Aspects of FFR FFR has a number of unique aspects that distinguish it from other indexes of coronary physiology such as coronary flow velocity reserve (Table). FFR has an upper normal value of 1.0 in every patient and every vessel. In a study of 37 angiographically normal-appearing coronary arteries in patients without signs of myocardial ischemia, the average FFR was FFR is extremely reproducible and independent of hemodynamic perturbations. In another study, the coefficient of variation of 2 repeated FFR measurements was 1.6%, and there was no significant difference in FFR after changing the heart rate with pacing, the blood pressure with sodium nitroprusside administration, or the left ventricular contractility with dobutamine administration. 7 FFR has a narrow cutoff range, distinguishing lesions capable of producing ischemia from lesions that are not hemodynamically significant. Multiple studies performed by numerous different groups in a variety of patient populations have shown that FFR values <0.75 to 0.80 have a very high specificity for identifying ischemia based on a variety of noninvasive imaging studies. 8 If the FFR value is >0.80, it is very unlikely that the vessel or lesion interrogated is responsible for significant ischemia. FFR has superb spatial resolution. By placing the pressure wire in the distal portion of an epicardial vessel, one can measure the FFR and determine whether the disease in that vessel is responsible for ischemia. If the FFR is abnormal, and particularly if the vessel has tandem areas of narrowing, one can slowly pull the pressure wire back to the proximal vessel during maximal hyperemia and identify which lesion is responsible for the greatest portion of the gradient. It is important to note that with the pressure sensor positioned between the 2 narrowings, distal to the proximal one, it is not possible to determine the FFR of the proximal narrowing. The reason is that each narrowing affects flow across the other and the removal of 1 narrowing, for example with PCI, can lead to an increase in flow down the vessel and a larger pressure gradient across the remaining lesion. In the setting of tandem stenoses, equations have been described that allow the operator to determine the FFR of each lesion in the theoretical absence of the other lesion; however, the complexity of these equations and the requirement Figure 2. A, Photograph of an angiographic image of a right coronary artery (RCA) lesion and the associated fractional flow reserve (FFR) tracing of a 67-year-old man presenting with chest pain. The red tracing represents the mean and phasic proximal coronary pressure; green line, the mean and phasic distal coronary pressure; horizontal yellow line, the beat-to-beat ratio of distal pressure to proximal pressure; and vertical yellow line, where the FFR measurement is made. On the left side of the tracing is the resting pressure gradient. On the right side of the tracing is the hyperemic pressure gradient after administration of intravenous adenosine. B, Photograph of an angiographic image of a left anterior descending (LAD) coronary artery lesion in the same patient with the associated FFR measurement. The right coronary lesion is not significant, whereas the LAD lesion is significant. Pa indicates proximal coronary or aortic pressure; and Pd, distal coronary pressure.

3 1862 Circulation May 6, 2014 Table. Unique Aspects of Fractional Flow Reserve Normal value of 1.0 in every patient/vessel Narrow ischemic threshold ( ) Highly reproducible Independent of hemodynamic changes Specific for epicardial artery stenosis Independent of the microvasculature Accounts for collateral flow for measuring the coronary wedge pressure have limited their clinical utility. 9 FFR is independent of the microvasculature, is specific for epicardial coronary artery disease, and accounts for both coronary and collateral flow. In patients with remote myocardial infarction (MI), for example, FFR still correlates with the presence or absence of ischemia on noninvasive stress imaging. 10 In this setting, the flow across a given stenosis subtending infarcted myocardium will be less and the pressure gradient will be less, leading to a higher FFR value. The higher FFR is not falsely elevated but reflects the smaller amount of viable myocardium supplied by the vessel and still provides information on the expected gain in flow after PCI. On the other hand, if there is an intermediate lesion in a vessel supplying collaterals to viable myocardium subtended by a chronically occluded vessel, the flow down the collateral-supplying vessel will be greater than in the normal setting, the pressure gradient across the intermediate stenosis will be higher than expected, and the FFR will be lower. These scenarios help to explain why FFR is more accurate than angiography or intravascular anatomic methods such as intravascular ultrasound for identifying lesions capable of producing myocardial ischemia. One of the limitations of FFR is that it is not reliable in the culprit vessel of a patient with an ST-segment elevation MI. The reason is that FFR assumes that microvascular resistance will be minimized and fixed. In the culprit vessel of an ST-segment elevation MI, there is a variable degree of reversible microvascular stunning, resulting in a lower maximum achievable flow, a lower pressure gradient, and a higher FFR. With time, the microvasculature may recover, maximum achievable flow may increase, and a larger gradient with a lower FFR may be measured across a given stenosis. For this reason, the FFR measured in the culprit vessel in the acute setting may be higher than the FFR measured in the same vessel 1 month later. However, in the setting of non ST-segment elevation acute coronary syndromes, in which the degree of reversible microvascular dysfunction is likely to be low, measuring FFR in the culprit and nonculprit vessels has been shown to be safe and effective. 11 Moreover, in nonculprit vessels of patients with ST-segment elevation MI, FFR remains accurate. 12 Achieving maximal hyperemia is a prerequisite to measuring FFR accurately. A potential pitfall to FFR measurement is the lack of maximal vasodilation, which will result in lower maximal flow across a stenosis, a lower pressure gradient, and a falsely high FFR. In the validation studies and major clinical trials, FFR has been measured after administration of intravenous adenosine delivered via a central vein at 140 μg kg 1 min 1. This method has been shown to be extremely safe and reliable. 8 Most patients have symptoms of either chest pressure or dyspnea related to the infusion, but they are well tolerated in the vast majority of cases and resolve as soon as the infusion is discontinued. In one respect, these symptoms are reassuring because they inform the operator that hyperemia has occurred. Why Do We Need FFR? Myocardial ischemia is associated with symptoms and adverse cardiac outcomes. Noninvasive stress imaging identifies the presence and degree of myocardial ischemia and can predict death and MI. 13 However, the spatial resolution (ie, the ability to identify the culprit lesion or even vessel) of noninvasive stress imaging is lacking, particularly in patients with multivessel coronary disease. 14 The noninvasive test is a useful gatekeeper to decide which patients warrant invasive coronary angiography, but once the patient is in the cardiac catheterization laboratory, FFR is necessary to identify which lesion is responsible for the patient s symptoms. This was recently highlighted in 2 studies comparing FFR and nuclear perfusion scanning in patients with multivessel disease, both of which demonstrated significant discordance between the 2 modalities and prompted the editorialists to conclude that FFR is the reference standard for identifying lesions capable of producing ischemia Not only does the inaccuracy of noninvasive imaging make FFR necessary, but the fact that noninvasive stress imaging has not been performed in the majority of patients undergoing PCI further emphasizes the need for a cardiac catheterization laboratory based technique for identifying lesions capable of producing ischemia. 18 One might argue that the coronary angiogram is accurate enough to guide the decision about PCI once a patient makes it to the catheterization suite. However, in the Fractional Flow Reserve Versus Angiography for Multi-vessel Evaluation (FAME) trial, FFR was measured in >500 patients with multivessel coronary disease and compared with the visual interpretation of the angiogram. 19 Approximately 35% of the lesions graded as being between 50% and 70% narrowed had an ischemic FFR, whereas roughly 20% of the lesions between 71% and 90% had nonischemic FFR values (Figure 3). Multiple smaller studies have asked experienced interventional cardiologists to predict which lesions will have an ischemic FFR value and which will not; in all cases, the diagnostic accuracy of the visual interpretation of the angiogram has been unacceptable for guiding decisions about the need for PCI. 20,21 Quantitative coronary angiography has not improved the lack of concordance between the angiographic severity of a stenosis and its ischemic potential. 22 Others might argue that FFR is not necessary because intravascular imaging techniques such as intravascular ultrasound or optical coherence tomography are available and can accurately identify which lesions warrant PCI. These techniques provide detailed visualization of coronary plaque and allow quantification of lumen area and other anatomic parameters such as

4 Fearon FFR-Guided PCI 1863 Figure 3. Comparison of fractional flow reserve (FFR) values and visual interpretation of the coronary angiogram in 509 patients randomized to FFR-guided percutaneous coronary intervention in the Fractional Flow Reserve Versus Angiography for Multi-vessel Evaluation (FAME) trial. Of the lesions interpreted to be narrowed between 50% and 70%, 35% had an ischemic FFR. Of the lesions narrowed between 71% and 90%, 20% had a nonischemic FFR. Reprinted with permission from Tonino et al 19 with permission from the publisher. Copyright 2010 Elsevier B.V. lesion length that contribute to the ischemic potential of a stenosis. Unfortunately, they currently do not provide information on the viscous and separation forces or flow across a stenosis, all of which are accounted for by measuring the pressure gradient and all of which determine the functional significance of a stenosis. 8 Most studies evaluating the accuracy of either intravascular ultrasound or optical coherence tomography for identifying lesions responsible for ischemia have compared the minimum lumen area assessed by either technique with FFR and have found poor correlations The reason is that the cutoff for a significant minimum lumen area will vary, depending on where the measurement is made along the vessel, the size of the vessel, and the amount of viable myocardium subtended by the vessel. One of the advantages of FFR is that the same cutoff is applied in every patient and every vessel. Data Supporting FFR-Guided PCI Initially, FFR was promoted as a method for identifying which intermediate coronary lesions could be treated safely medically without PCI. At the 18-month follow-up of a retrospective cohort of 100 patients with chest pain, intermediate coronary stenoses, and nonischemic FFR values in whom PCI was deferred in favor of medical therapy, there were no cardiac deaths, 1 target vessel MI, and 3 target vessel revascularizations. 26 These data were reassuring, but it was not until a multicenter, prospective, randomized trial was published by Bech and colleagues 27 that the safety of this approach began to gain acceptance. The DEFER trial included 325 patients with chest pain and intermediate coronary lesions who underwent FFR assessment. The 181 patients with nonischemic FFR values were randomized to performance of PCI or to deferral of PCI. At the 2-year followup, the event-free survival rate was 89% in the deferral group and 83.3% in the perform group (P=0.27). The cardiac death rate was 2.2% and the study lesion revascularization rate was 5.6%, without any MI in the deferral group. Angina relief also was similar between the 2 groups. Subsequently, the 5-year death and MI rates of the 2 groups have been reported. 28 Although not statistically significant, the 3.3% rate in the deferral group was less than half the 7.9% rate in the perform group (P=0.21). The findings from the DEFER trial established the safety of treating functionally nonsignificant intermediate epicardial coronary narrowings with medicine as opposed to PCI, but they did not address an important and often vexing subgroup of intermediate coronary lesions, those involving the left main coronary artery. The inadequacy of the angiogram alone for assessing the severity of moderate left main coronary disease has been appreciated for years. However, relying on a novel technique such as FFR to make the critical decision about revascularization in this setting has gained acceptance more slowly. Initially, a number of small, single-center studies demonstrated excellent short-term outcomes in patients with moderate left main disease in whom revascularization was deferred because the FFR was not in the ischemic range. 29 More recently, Hamilos and colleagues 21 reported 213 patients with moderate left main disease in whom FFR was measured. FFR was nonischemic in 138 of these patients, and revascularization was deferred. The 5-year survival rate was 90% in these patients compared with 85% in those with ischemic FFR values who underwent revascularization (P=0.48). The event-free survival also was similar between the 2 groups. In this study, as in previous ones, experienced interventional cardiologists were unable to predict accurately which lesions would have a significant FFR on the basis of their interpretation of the angiogram, further emphasizing the importance of an FFR-guided approach to PCI even when involving the left main coronary artery (Figure 4). The DEFER trial and others like it established FFR as a method for deferring PCI, not necessarily as a technique to be embraced by the interventional cardiology community. Two studies evaluating FFR after stenting showed that it could be used to determine optimal stent deployment and to predict outcomes after PCI. 30,31 However, it was not until the FAME trial was performed and published by Tonino and colleagues 32 that FFR gained traction as a method for guiding PCI. In the FAME trial, 1005 patients with coronary lesions of >50% diameter stenosis in 2 or 3 major epicardial vessels that were amenable to PCI with drug-eluting stents were randomized either to the standard of care, angiography-guided PCI, in which case PCI was performed on the basis of the noninvasive clinical data and the angiographic appearance of the lesions, or to FFR-guided PCI, in which case FFR was measured across every stenosis and only if the FFR was 0.80 was PCI performed on the particular stenosis. There were 3 lesions identified per patient in both groups, but the angiography-guided PCI group received significantly more stents per patient (2.7±1.2 versus 1.9±1.3; P<0.001). In addition, significantly more contrast media was administered to the angiography-guided PCI patients (302±127 versus

5 1864 Circulation May 6, 2014 Figure 4. Comparison of 2 expert interventional cardiologists predictions of whether a left main coronary lesion is functionally significant or nonsignificant or the cardiologist is unsure of the significance and the actual fractional flow reserve (FFR) value. Neither reviewer was able to accurately predict which lesions would have an ischemic or nonischemic FFR. Reprinted from Hamilos et al ±133 ml; P<0.001). Importantly, the time of the procedure was identical between the 2 groups; although measuring FFR adds time to the procedure, avoiding unnecessary PCI saves time. Despite the fewer stents, the percentage of patients free of angina at 1 year was numerically higher in patients randomized to FFR guidance (81.3% versus 77.9%; P=0.20). The primary end point of the FAME trial was the 1-year rate of death, MI, or repeat revascularization, which occurred in 18.3% of the angiography-guided PCI patients compared with 13.2% of the FFR-guided PCI patients (P<0.02; Figure 5). This significant reduction resulted from 30% to 40% relative risk reductions in each component of the composite end point. In addition, the rate of death and MI was reduced significantly in the FFR-guided PCI group (7.3% versus 11.1%; P=0.04). Subsequently, the 2-year results of FAME were reported and revealed a persistent reduction in the composite of death, MI, and repeat revascularization (17.9% versus 22.4%; P=0.08). 33 Importantly, the hard end points of death and MI remained significantly lower in the FFR-guided PCI patients (8.4% versus 12.9%; P=0.02). The Kaplan-Meier curves at 1 and 2 years showed that the benefit attributed to FFR guidance occurred early, as a result of fewer procedure-related events, and continued during follow-up, as a result of fewer late events. The FAME trial provided further evidence for the safety of deferring PCI of lesions that are not functionally significant. In the FFR-guided patients, there were 513 angiographically significant-appearing lesions (>50% diameter stenosis) that had Figure 5. One-year adverse events in the Fractional Flow Reserve Versus Angiography for Multi-vessel Evaluation (FAME) trial. MACE indicates major adverse cardiac events (death, myocardial infarction [MI], or repeat revascularization). FFR values >0.80 and therefore were treated medically without PCI. The 2-year rate of MI resulting from 1 of these lesions was 0.2% and the revascularization rate was 3.2%. 33 This is particularly noteworthy given that approximately one fourth of these deferred lesions appeared >70% narrowed on visual interpretation of the angiogram. Data like these led the European Society of Cardiology to give FFR it highest recommendation, Class IA, in favor of FFR-guided PCI when objective evidence of ischemia is lacking. 34 Older guidelines from the American cardiology societies give FFR a Class IIA recommendation for assessing the need for PCI of an intermediate lesion. 35 By demonstrating improved outcomes despite more judicious use of stents, the FAME trial validated the concept of functional angioplasty, stenting lesions responsible for ischemia and treating medically those lesions that are not hemodynamically significant. In this manner, the benefit of the stent is optimized and the risks are minimized. Although this approach results in a functionally complete revascularization, it does not result in an anatomically complete one. In patients with multivessel coronary disease, this is particularly relevant because many of our treatment decisions about revascularization are based on angiographic (or anatomic) assessment of the severity of coronary disease. For example, the Synergy between Percutaneous Coronary Intervention with Taxus can Cardiac Surgery (SYNTAX) score is an angiography-based scoring system designed to quantify the complexity of coronary disease. Patients with high SYNTAX scores tend to have better outcomes with coronary artery bypass graft surgery, whereas those patients with low scores appear to do equally well with PCI. 36 Because the SYNTAX score is based on visual interpretation of the angiogram, it is inherently limited by the inaccuracy of the angiogram. A substudy from the FAME trial asked whether incorporating FFR into the SYNTAX score and calculating a functional SYNTAX score (FSS) might convert patients with a higher-risk SYNTAX score to a lower risk and whether it might improve the risk stratification of patients with multivessel coronary disease undergoing PCI. 37 In the 497 patients in the FFR-guided arm of FAME, the SYNTAX score was calculated in the usual fashion, and the patients were divided into tertiles based on the SYNTAX score. The FSS was then determined by recalculating the SYNTAX score and taking into account only those lesions

6 Fearon FFR-Guided PCI 1865 with an FFR More than one third of patients moved from a higher-risk group to a lower-risk group after calculation of the FSS. In addition, the FSS was a significant predictor of death or MI, whereas the classic SYNTAX score was not (Figure 6). In this manner, the FSS simplifies the approach to PCI in many of these challenging patients with multivessel disease and may make PCI a more appropriate treatment than coronary artery bypass graft surgery. This hypothesis is now being tested prospectively in the FAME 3 trial, comparing FFR-guided PCI with coronary artery bypass graft surgery in patients with 3-vessel coronary disease. The FAME trial also further established the safety and benefit of measuring FFR in patients with non ST-segment elevation acute coronary syndromes. A small study of 70 patients with unstable angina and non ST-segment elevation MI suggested that measuring FFR, even in the culprit vessel, was a safe and reliable method for triaging treatment decisions. 38 The FAME trial, however, included a much larger cohort of 328 patients with unstable angina or non ST-segment elevation MI who underwent urgent angiography and were randomized to FFR-guided or to angiography-guided PCI, including FFR assessment of both the culprit vessel and nonculprit vessel. A substudy of this cohort revealed a relative risk reduction from FFR-guided compared with angiography-guided PCI in the patients with non ST-segment elevation acute coronary syndrome similar to that seen in the stable patients enrolled in the FAME trial. 39 Although the FAME trial successfully demonstrated that decisions about PCI in patients with multivessel coronary disease should be based on FFR guidance and not the angiogram alone, one of the criticisms of the FAME trial was the use of the zotarolimus-eluting Endeavor stent in >40% of patients because of concern that it might be inferior to the 2 other drug-eluting Figure 6. Risk of death or myocardial infarction (MI) based on low-, medium-, or high-risk SYNTAX score tertile in the 509 patients randomized to fractional flow reserve (FFR) guided percutaneous coronary intervention in the Fractional Flow Reserve Versus Angiography for Multi-vessel Evaluation (FAME) trial (left) and the same risk depicted for the patients once they are reclassified on the basis of the functional SYNTAX score (right). Of note, there is a shift of higher-risk patients to a lower-risk score after taking into account the FFR value of each lesion being scored. The functional SYNTAX score was a better predictor of death or MI. Adapted from Nam et al 37 with permission from the publisher. Copyright 2011 Elsevier B.V. stents available at the time of the study, the paclitaxel-eluting Taxus stent and the sirolimus-eluting Cypher stent. However, a subsequent study has demonstrated lower rates of periprocedural MI and similar rates of target lesion revascularization and target vessel failure with the Endeavor stent compared with the Taxus stent. 40 Moreover, a longer-term comparison of the Endeavor and Cypher stents revealed significantly lower rates of mortality, MI, and major adverse cardiac events. 41 Another criticism that has emerged is that the results of FAME would be different if the second-generation everolimuseluting stent had been used because studies have shown that this stent is superior to first-generation drug-eluting stents. 42 However, any improved outcome seen with the everolimuseluting stent would occur in both the angiography- and FFRguided patients, with the FFR-guided patients still having improved outcomes compared with the angiography-guided patients. For the angiography-guided patients to achieve results similar to or better than those achieved by the FFRguided patients, the event rate after everolimus-eluting stenting would need to be similar to or lower than the event rate after medical treatment of nonischemic lesions. At the 2-year follow-up in FAME, the deferred lesions had a 0.2% MI rate and a 3.2% revascularization rate; at the 2-year follow-up after everolimus-eluting stenting, the MI rate was 2.5% and the target vessel revascularization rate was 6.8%. 33,43 A final criticism of FAME was the lack of a medical therapy arm in the randomization scheme, particularly in the patients with stable coronary artery disease. The Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial randomized >2000 patients with stable single-vessel or multivessel coronary disease to best medical therapy alone or to PCI in addition to best medical therapy and found no difference in the rate of death or MI at a median follow-up of 4.6 years. 44 On the basis of these results, some wondered what the event rate in FAME would have been if a medical therapy arm had been included in the two thirds of stable patients in FAME not presenting with an acute coronary syndrome. One of the goals of the FAME 2 trial was to address this concern by comparing PCI of lesions with an abnormal FFR with best medical therapy in patients with stable single-vessel or multivessel coronary disease. 45 One of the criticisms of the COURAGE trial is that low-risk patients without significant myocardial ischemia were enrolled. The FAME 2 trial was designed specifically to ensure the inclusion of patients with a large burden of ischemia. This was achieved by randomizing only patients with at least 1 lesion in 1 of the 3 major epicardial vessels with >50% narrowing and an FFR Those patients with narrowings of >50% but with an FFR >0.80 were not included in the randomized portion of the study but were followed up in a registry and treated with best medical therapy. The goal was to randomize roughly 1600 patients and to follow them for 2 years with a primary composite end point of death, MI, or unplanned hospitalization requiring urgent revascularization. After 888 patients had been randomized, recruitment into FAME 2 was stopped prematurely by the data and safety

7 1866 Circulation May 6, 2014 monitoring board because of a highly significant difference in the primary end point between the 2 groups. After a mean follow-up of only 7 months, the primary end point occurred in 12.7% of patients in the best medical therapy arm compared with 4.3% in the PCI plus best medical therapy arm. There was no significant difference in death (0.7% versus 0.2%; P=0.31) or MI (3.2% versus 3.4%; P=0.89), but there was a highly significant difference in the rate of unplanned hospitalization requiring urgent revascularization (11.1% versus 1.6%; P<0.001) between the best medical therapy arm and PCI arm, respectively. The rate of any revascularization was also significantly higher in the best medical therapy arm (19.5% versus 3.1%; P<0.001). Both groups in FAME 2 received excellent medical therapy, with the vast majority prescribed aspirin, a statin, a β-blocker, and an angiotensin-converting enzyme inhibitor at rates similar to those seen in the COURAGE trial. The majority of patients in the best medical therapy arm also received a calcium channel blocker or nitrate for refractory angina. Despite this, the hospitalization rate for an acute coronary syndrome requiring urgent revascularization in the best medical therapy arm was 11.1% at just 7 months. In comparison, in the COURAGE trial, the rate of an acute coronary syndrome at a median follow-up of 4.6 years was only 11.8%. 44 This difference likely resulted because, in the FAME 2 trial, in which FFR was measured first, patients with a larger burden of ischemia were randomized. The concept of more severe ischemia resulting in worse outcomes is supported further by the fact that stratifying outcomes in FAME 2 on the basis of an FFR value of 0.65 demonstrated a significant interaction, meaning that those patients with more severe ischemia at baseline had even greater benefit with PCI compared with medical therapy alone. 45 Importantly, there were 332 patients in the FAME 2 trial with angiographic coronary disease that was not significant on the basis of the FFR assessment, and 166 of these patients were randomly assigned follow-up in a registry receiving best medical therapy alone. The primary end point occurred in only 3.0% of these registry patients with a 2.4% rate of urgent revascularization. This finding further emphasizes the importance of not only identifying angiographic evidence of coronary disease but also characterizing whether the disease is responsible for ischemia. Patients with these types of lesions benefit from PCI, whereas patients with angiographic disease that is not functionally significant as assessed by FFR respond well to medical therapy alone. The FAME 2 trial has been criticized because it was stopped early, because there was no significant difference in the rate of death or MI, and because the primary end point included unplanned hospitalization requiring urgent revascularization, which some consider a subjective or soft end point. The data and safety monitoring board concluded that, from a scientific standpoint, the difference in the rate of the composite end point between the 2 groups was so statistically significant that it was extremely unlikely that completing enrollment would alter the results. Even though there was no difference in death and MI, the board considered unplanned hospitalization requiring urgent revascularization as an important safety outcome. Although this is a subjective end point compared with death and MI, it was adjudicated by an independent clinical events committee. Approximately one half of the unplanned hospitalizations requiring urgent revascularization were associated with biomarker elevation or ECG changes consistent with ischemia. With the use of this stricter and more objective definition of acute coronary syndrome, there was still a highly significant difference in the rate of unplanned hospitalization requiring urgent revascularization between the 2 groups (0.9% versus 5.2%; P<0.001; unpublished data). The FAME 2 trial was not powered to detect a significant difference between the 2 randomized groups with respect to death and MI; however, a landmark analysis showed that, during the first 7 days after enrollment, there was a higher rate in the PCI arm (1.8% versus 0.2%; P=0.04) resulting primarily from periprocedural MI in the PCI patients, whereas after 7 days, there was a higher rate of death and MI in the best medical therapy arm (1.6% versus 3.7%; P=0.05) resulting primarily from spontaneous MI. 45 Other studies have found that late spontaneous MI is a much stronger predictor of long-term mortality than peri-pci MI. 46 Longer-term follow-up of the FAME 2 trial will be important to evaluate this further. Many accepted therapies that we provide our patients, for example, orthopedic procedures, do not reduce death or MI, but they improve quality of life and decrease the need for hospitalization. The fact that PCI guided by FFR has not yet been shown to decrease death or MI rates compared with medical therapy should not diminish its value. Another important consideration in assessing the value of a new approach is the economic and quality-of-life implications of the approach compared with the standard of care. Both an early analysis using computer modeling and a randomized, single-center study suggested that measuring FFR was cost-effective in patients with intermediate single-vessel coronary disease compared with a noninvasive stress imaging strategy. 38,47 In the FAME trial comparing FFR-guided PCI and angiography-guided PCI in patients with multivessel coronary disease, the FFR-guided approach significantly reduced costs both at the time of the procedure and during follow-up. 48 Although the coronary pressure wire and adenosine cost approximately $800, by decreasing the number of drugeluting stents necessary for revascularization, the FFR-guided approach resulted in roughly $700 savings in procedural costs. Because of the lower rate of adverse events in the FFR-guide group, the 1-year costs were reduced significantly by $2400. Quality of life in the FAME trial was assessed with the EuroQual 5 dimension health survey at baseline, 1 month, and 1 year. The quality-adjusted life-years (QALYs) were numerically higher in the FFR-guided patients, although not statistically different at 1 year (0.853 versus 0.838; P=0.20). Bootstrap simulation revealed that the FFR-guided strategy was cost-saving in >90% of cases and cost-effective (at a

8 Fearon FFR-Guided PCI 1867 Figure 7. A, Bootstrap simulation of the incremental cost-effectiveness of fractional flow reserve (FFR) guided percutaneous coronary intervention (PCI) compared with angiography-guided PCI. Data are from 5000 bootstrap replications. A willingness-to-pay threshold of $ per quality-adjusted life-year (QALY) is added (solid line). Reprinted with permission from Boden et al 44 with permission from the publisher. Copyright 2007 Massachusetts Medical Society. B, Bootstrap replications of the incremental cost-effectiveness of the PCI in the setting of an abnormal FFR strategy compared with best medical therapy. Each of the points represents the results of 1 bootstrap replication. The difference in cumulative costs is displayed in the y axis, and the difference in qualityadjusted life-years (QALY) is displayed on x axis. Willingness-to-pay thresholds of $ per QALY added (solid line), $ per QALY added (dashed line), and $ per QALY added (dotted line) are indicated in the plane. The fractions of replications in each sector of the plane are indicated (eg, of the replications had a cost difference <0 and QALY difference >0). Reprinted with permission from De Bruyne et al 45 with permission from the publisher. Copyright 2012 Massachusetts Medical Society. threshold of $ per QALY gained) in 99.96% of cases (Figure 7A). Thus, compared with an angiography-guided approach, an FFR-guided strategy is a dominant one. More recently, the cost-effectiveness of an FFR-guided strategy for PCI has been compared with medical therapy in patients with stable coronary disease. In the FAME 2 trial,

9 1868 Circulation May 6, 2014 performance of PCI on lesions with an abnormal FFR resulted in a roughly $6000 initial cost increase compared with best medical therapy. 49 However, during the course of 1 year, this difference narrowed to just under $3000 because of increased revascularization costs in the best medical therapy arm. Quality of life improved significantly in the PCI group of patients, whereas there was no change in the best medical therapy arm. It was assumed, from previous data from the COURAGE trial and other studies, that this difference in quality of life would gradually dissipate over 3 years but that the cost difference between the 2 strategies would not narrow any further. From these assumptions, the cost-effectiveness of performing PCI on lesions with an abnormal FFR was $ per QALY gained compared with best medical therapy. Bootstrap simulation revealed that for 80% of cases the cost-effectiveness ratio was <$ per QALY gained and for 99.5% of cases it was <$ per QALY gained (Figure 7B). Conclusions Should FFR be measured in all cases of PCI? If a patient has typical angina and single-vessel coronary disease with a significant-appearing lesion in a vessel subtending myocardium that is ischemic on a noninvasive stress imaging study, then FFR measurement is not necessary. If a patient is having an acute MI with significant single-vessel coronary disease, FFR is not necessary. However, these 2 scenarios represent the minority of cases seen in the cardiac catheterization laboratory. The vast majority of cases involve patients with multivessel disease, lesions of indeterminate severity, or discordant or absent stress imaging data. In these scenarios, there is now a wealth of data supporting the routine measurement of FFR to guide PCI. Our primary goal in the catheterization laboratory should be to relieve the symptoms and to improve the prognosis of our patients. This is best accomplished by identifying and relieving stenoses responsible for myocardial ischemia. Measuring FFR is currently the most cost-effective and accurate method for achieving this goal. Acknowledgments Some of the ideas presented in this article and rebuttal came from conversations and discussions over the years with Nico H.J. Pijls, MD, PhD; Bernard De Bruyne, MD, PhD; and Alan C. Yeung, MD. Source of Funding This work was supported in part by grant R01 HL A1 from the National Institutes of Health, Heart, Lung, and Blood Institute, Bethesda, MD. Disclosures Dr Fearon receives research support from St. Jude Medical and consulting fees from Medtronic Corp. References 1. Grüntzig AR, Senning A, Siegenthaler WE. Nonoperative dilatation of coronary-artery stenosis: percutaneous transluminal coronary angioplasty. N Engl J Med. 1979;301: Wilson RF, Johnson MR, Marcus ML, Aylward PE, Skorton DJ, Collins S, White CW. The effect of coronary angioplasty on coronary flow reserve. Circulation. 1988;77: Pijls NHJ, van Son JAM, Kirkeeide RL, De Bruyne B, Gould KL. 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. 1993;86: De Bruyne B, Baudhuin T, Melin JA, Pijls NH, Sys SU, Bol A, Paulus WJ, Heyndrickx GR, Wijns W. Coronary flow reserve calculated from pressure measurements in humans: validation with positron emission tomography. Circulation. 1994;89: Pijls NH, De Bruyne B, Peels K, Van Der Voort PH, Bonnier HJ, Bartunek J Koolen JJ, Koolen JJ. Measurement of fractional flow reserve to assess the functional severity of coronary-artery stenoses. N Engl J Med. 1996;334: De Bruyne B, Hersbach F, Pijls NH, Bartunek J, Bech JW, Heyndrickx GR, Gould KL, Wijns W. Abnormal epicardial coronary resistance in patients with diffuse atherosclerosis but normal coronary angiography. Circulation. 2001;104: Ng MK, Yeung AC, Fearon WF. Invasive assessment of the coronary microcirculation: superior reproducibility and less hemodynamic dependence of index of microcirculatory resistance compared with coronary flow reserve. Circulation. 2006;113: Kern MJ, Samady H. Current concepts of integrated coronary physiology in the catheterization laboratory. J Am Coll Cardiol. 2010;55: Pijls NH, De Bruyne B, Bech GJ, Liistro F, Heyndrickx GR, Bonnier HJ, Koolen JJ. Coronary pressure measurement to assess the hemodynamic significance of serial stenoses within one coronary artery: validation in humans. Circulation. 2000;102: De Bruyne B, Pijls NH, Bartunek J, Kulecki K, Bech JW, De Winter H, Van Crombrugge P, Heyndrickx GR, Wijns W. Fractional flow reserve in patients with prior myocardial infarction. Circulation. 2001;104: Sels JW, Tonino PA, Siebert U, Fearon WF, Van t Veer M, De Bruyne B, Pijls NH. Fractional flow reserve in unstable angina and non-st-segment elevation myocardial infarction experience from the FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) study. JACC Cardiovasc Interv. 2011;4: Ntalianis A, Sels JW, Davidavicius G, Tanaka N, Muller O, Trana C, Barbato E, Hamilos M, Mangiacapra F, Heyndrickx GR, Wijns W, Pijls NH, De Bruyne B. Fractional flow reserve for the assessment of nonculprit coronary artery stenoses in patients with acute myocardial infarction. JACC Cardiovasc Interv. 2010;3: Hachamovitch R, Berman DS, Shaw LJ, Kiat H, Cohen I, Cabico JA, Friedman J, Diamond GA. Incremental prognostic value of myocardial perfusion single photon emission computed tomography for the prediction of cardiac death: differential stratification for risk of cardiac death and myocardial infarction. Circulation. 1998;97: Lima RS, Watson DD, Goode AR, Siadaty MS, Ragosta M, Beller GA, Samady H. Incremental value of combined perfusion and function over perfusion alone by gated SPECT myocardial perfusion imaging for detection of severe three-vessel coronary artery disease. J Am Coll Cardiol. 2003;42: Ragosta M, Bishop AH, Lipson LC, Watson DD, Gimple LW, Sarembock IJ, Powers ER. Comparison between angiography and fractional flow reserve versus single-photon emission computed tomographic myocardial perfusion imaging for determining lesion significance in patients with multivessel coronary disease. Am J Cardiol. 2007;99: Melikian N, De Bondt P, Tonino P, De Winter O, Wyffels E, Bartunek J, Heyndrickx GR, Fearon WF, Pijls NH, Wijns W, De Bruyne B. Fractional flow reserve and myocardial perfusion imaging in patients with angiographic multivessel coronary artery disease. JACC Cardiovasc Interv. 2010;3: Beller GA, Ragosta M. Decision making in multivessel coronary disease: the need for physiological lesion assessment. JACC Cardiovasc Interv. 2010;3: Lin GA, Dudley RA, Lucas FL, Malenka DJ, Vittinghoff E, Redberg RF. Frequency of stress testing to document ischemia prior to elective percutaneous coronary intervention. JAMA. 2008;300:

10 Fearon FFR-Guided PCI Tonino PA, Fearon WF, De Bruyne B, Oldroyd KG, Leesar MA, Ver Lee PN, Maccarthy PA, Van t Veer M, Pijls NH. Angiographic versus functional severity of coronary artery stenoses in the FAME study: Fractional Flow Reserve Versus Angiography in Multivessel Evaluation. J Am Coll Cardiol. 2010;55: Fischer JJ, Samady H, McPherson JA, Sarembock IJ, Powers ER, Gimple LW, Ragosta M. Comparison between visual assessment and quantitative angiography versus fractional flow reserve for native coronary narrowings of moderate severity. Am J Cardiol. 2002;90: Hamilos M, Muller O, Cuisset T, Ntalianis A, Chlouverakis G, Sarno G, Nelis O, Bartunek J, Vanderheyden M, Wyffels E, Barbato E, Heyndrickx GR, Wijns W, De Bruyne B. Long-term clinical outcome after fractional flow reserve-guided treatment in patients with angiographically equivocal left main coronary artery stenosis. Circulation. 2009;120: Christou MA, Siontis GC, Katritsis DG, Ioannidis JP. Meta-analysis of fractional flow reserve versus quantitative coronary angiography and noninvasive imaging for evaluation of myocardial ischemia. Am J Cardiol. 2007;99: Koo BK, Yang HM, Doh JH, Choe H, Lee SY, Yoon CH, Cho YK, Nam CW, Hur SH, Lim HS, Yoon MH, Park KW, Na SH, Youn TJ, Chung WY, Ma S, Park SK, Kim HS, Tahk SJ. Optimal intravascular ultrasound criteria and their accuracy for defining the functional significance of intermediate coronary stenoses of different locations. JACC Cardiovasc Interv. 2011;4: Waksman R, Legutko J, Singh J, Orlando Q, Marso S, Schloss T, Tugaoen J, DeVries J, Palmer N, Haude M, Swymelar S, Torguson R. FIRST: Fractional Flow Reserve and Intravascular Ultrasound Relationship Study. J Am Coll Cardiol. 2013;61: Gonzalo N, Gonzalo N, Escaned J, Alfonso F, Nolte C, Rodriguez V, Jimenez-Quevedo P, Bañuelos C, Fernández-Ortiz A, Fernández-Ortiz A, Garcia E, Hernandez-Antolin R, Macaya C. Morphometric assessment of coronary stenosis relevance with optical coherence tomography: a comparison with fractional flow reserve and intravascular ultrasound. J Am Coll Cardiol. 2012;59: Bech GJ, De Bruyne B, Bonnier HJ, Bartunek J, Wijns W, Peels K, Heyndrickx GR, Koolen JJ, Pijls NH. Long-term follow-up after deferral of percutaneous transluminal coronary angioplasty of intermediate stenosis on the basis of coronary pressure measurement. J Am Coll Cardiol. 1998;31: Bech GJ, De Bruyne B, Pijls NH, de Muinck ED, Hoorntje JC, Escaned J, Stella PR, Boersma E, Bartunek J, Koolen JJ, Wijns W. Fractional flow reserve to determine the appropriateness of angioplasty in moderate coronary stenosis: a randomized trial. Circulation. 2001;103: Pijls NH, van Schaardenburgh P, Manoharan G, Boersma E, Bech JW, van t Veer M, Bär F, Hoorntje J, Koolen J, Wijns W, de Bruyne B. Percutaneous coronary intervention of functionally nonsignificant stenosis: 5-year follow-up of the DEFER Study. J Am Coll Cardiol. 2007;49: Lindstaedt M. Patient stratification in left main coronary artery disease rationale from a contemporary perspective. Int J Cardiol. 2008;130: Fearon WF, Luna J, Samady H, Powers ER, Feldman T, Dib N, Tuzcu EM, Cleman MW, Chou TM, Cohen DJ, Ragosta M, Takagi A, Jeremias A, Fitzgerald PJ, Yeung AC, Kern MJ, Yock PG. Fractional flow reserve compared with intravascular ultrasound guidance for optimizing stent deployment. Circulation. 2001;104: Pijls NH, Klauss V, Siebert U, Powers E, Takazawa K, Fearon WF, Escaned J, Tsurumi Y, Akasaka T, Samady H, De Bruyne B; Fractional Flow Reserve (FFR) Post-Stent Registry Investigators. Coronary pressure measurement after stenting predicts adverse events at follow-up: a multicenter registry. Circulation. 2002;105: Tonino PAL, De Bruyne B, Pijls NHJ, Siebert U, Ikeno F, van t Veer M, Klauss V, Manoharan G, Engstrom T, Oldroyd KG, Ver Lee PN, MacCarthy PA, Fearon WF. Fractional flow reserve versus angiography for guiding percutaneous coronary intervention in patients with multivessel coronary artery disease. New Engl J Med. 2009;360: Pijls NH, Fearon WF, Tonino PA, Siebert U, Ikeno F, Bornschein B, van t Veer M, Klauss V, Manoharan G, Engstrøm T, Oldroyd KG, Ver Lee PN, MacCarthy PA, De Bruyne B; FAME Study Investigators. 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. 2010;56: Task Force on Myocardial Revascularization of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS); European Association for Percutaneous Cardiovascular Interventions (EAPCI), Wijns W, Kolh P, Danchin N, Di Mario C, Falk V, Folliguet T, Garg S, Huber K, James S, Knuuti J, Lopez-Sendon J, Marco J, Menicanti L, Ostojic M, Piepoli MF, Pirlet C, Pomar JL, Reifart N, Ribichini FL, Schalij MJ, Sergeant P, Serruys PW, Silber S, Sousa Uva M, Taggart D. Guidelines on myocardial revascularization. Eur Heart J. 2010;31: Kushner FG, Hand M, Smith SC Jr, King SB 3rd, Anderson JL, Antman EM, Bailey SR, Bates ER, Blankenship JC, Casey DE Jr, Green LA, Hochman JS, Jacobs AK, Krumholz HM, Morrison DA, Ornato JP, Pearle DL, Peterson ED, Sloan MA, Whitlow PL, Williams DO; American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines Focused updates: ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction (updating the 2004 guideline and 2007 focused update) and ACC/ AHA/SCAI guidelines on percutaneous coronary intervention (updating the 2005 guideline and 2007 focused update): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2009;120: Serruys PW, Morice MC, Kappetein AP, Colombo A, Holmes DR, Mack MJ, Ståhle E, Feldman TE, van den Brand M, Bass EJ, Van Dyck N, Leadley K, Dawkins KD, Mohr FW; SYNTAX Investigators. Percutaneous coronary intervention versus coronary-artery bypass grafting for severe coronary artery disease. N Engl J Med. 2009;360: Nam CW, Mangiacapra F, Entjes R, Chung IS, Sels JW, Tonino PA, De Bruyne B, Pijls NH, Fearon WF; FAME Study Investigators. Functional SYNTAX score for risk assessment in multivessel coronary artery disease. J Am Coll Cardiol. 2011;58: Leesar MA, Abdul-Baki T, Akkus NI, Sharma A, Kannan T, Bolli R. Use of fractional flow reserve versus stress perfusion scintigraphy after unstable angina: effect on duration of hospitalization, cost, procedural characteristics, and clinical outcome. J Am Coll Cardiol. 2003;41: Sels JW, Tonino PA, Siebert U, Fearon WF, Van t Veer M, De Bruyne B, Pijls NH. Fractional flow reserve in unstable angina and non-st-segment elevation myocardial infarction experience from the FAME (Fractional flow reserve versus Angiography for Multivessel Evaluation) study. JACC Cardiovasc Interv. 2011;4: Leon MB, Mauri L, Popma JJ, Cutlip DE, Nikolsky E, O Shaughnessy C, Overlie PA, McLaurin BT, Solomon SL, Douglas JS Jr, Ball MW, Caputo RP, Jain A, Tolleson TR, Reen BM 3rd, Kirtane AJ, Fitzgerald PJ, Thompson K, Kandzari DE; ENDEAVOR IV Investigators. A randomized comparison of the Endeavor zotarolimus-eluting stent versus the TAXUS paclitaxeleluting stent in de novo native coronary lesions 12-month outcomes from the ENDEAVOR IV trial. J Am Coll Cardiol. 2010;55: Kandzari DE, Mauri L, Popma JJ, Turco MA, Gurbel PA, Fitzgerald PJ, Leon MB. Late-term clinical outcomes with zotarolimus- and sirolimuseluting stents. 5-year follow-up of the ENDEAVOR III (A Randomized Controlled Trial of the Medtronic Endeavor Drug [ABT-578] Eluting Coronary Stent System Versus the Cypher Sirolimus-Eluting Coronary Stent System in De Novo Native Coronary Artery Lesions). JACC Cardiovasc Interv. 2011;4: Stone GW, Rizvi A, Newman W, Mastali K, Wang JC, Caputo R, Doostzadeh J, Cao S, Simonton CA, Sudhir K, Lansky AJ, Cutlip DE, Kereiakes DJ; SPIRIT IV Investigators. Everolimus-eluting versus paclitaxel-eluting stents in coronary artery disease. N Engl J Med. 2010;362: Stone GW, Rizvi A, Sudhir K, Newman W, Applegate RJ, Cannon LA, Maddux JT, Cutlip DE, Simonton CA, Sood P, Kereiakes DJ; SPIRIT IV Investigators. Randomized comparison of everolimus- and paclitaxeleluting stents. 2-year follow-up from the SPIRIT (Clinical Evaluation of the XIENCE V Everolimus Eluting Coronary Stent System) IV trial. J Am Coll Cardiol. 2011;58: Boden WE, O Rourke RA, Teo KK, Hartigan PM, Maron DJ, Kostuk WJ, Knudtson M, Dada M, Casperson P, Harris CL, Chaitman BR, Shaw L, Gosselin G, Nawaz S, Title LM, Gau G, Blaustein AS, Booth DC, Bates ER, Spertus JA, Berman DS, Mancini GB, Weintraub WS; COURAGE Trial Research Group. Optimal medical therapy with or without PCI for stable coronary disease. N Engl J Med. 2007;356: