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

Study design

The design of this study has been described previously (8). In eligible patients with multivessel CAD, the investigator indicated which lesions had stenosis of at least 50% and were thought to require PCI on the basis of angiographic appearance and clinical data. Patients were then randomly assigned to either angiography-guided PCI or FFR-guided PCI. (Fig. 1).Patients assigned to angiographic guidance underwent stenting of all indicated lesions with DES. In patients assigned to FFR guidance, first FFR was measured in each diseased coronary artery and DES were placed in indicated lesions only if the FFR was 0.80 or less. The study protocol was approved by the internal review board of each participating center. An independent clinical event committee blinded to treatment assignment adjudicated all events. Data management and statistical analysis were performed by an independent data coordinating center. The study sponsors had no role in the methods, data acquisition, data analysis, reporting, or publication of this study.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The FAME Study Flowchart

Flowchart of the FAME (Fractional Flow Reserve Versus Angiography for Multivessel Evaluation) study. In all, 1,005 patients with multivessel coronary artery disease were randomly assigned to angiography-guided percutaneous coronary intervention (PCI) (n = 496) or fractional flow reserve (FFR)-guided PCI (n = 509). The key secondary end points at 2 years were individual rates of death, myocardial infarction, revascularization, major adverse cardiac events (MACE), and functional status. DES = drug-eluting stent(s).

Study population

Patients were included in the study if they had multivessel CAD, defined as coronary artery stenoses ≥50% diameter stenosis in at least 2 of the 3 major epicardial coronary arteries, which were thought to require PCI. Patients with ST-segment elevation myocardial infarction could be included if the infarction had occurred at least 5 days before PCI. Enrollment and PCI of patients with non–ST-segment elevation myocardial infarction could occur earlier than 5 days if the peak creatinine kinase was <1,000 IU. Patients who had undergone previous PCI could be included in the study. Patients with angiographically significant left main coronary artery disease, previous coronary artery bypass surgery, cardiogenic shock, extremely tortuous or calcified coronary arteries, a life expectancy of <2 years, pregnancy, or a contraindication to DES placement were excluded.

Treatment

The PCI was performed using standard techniques. The FFR was measured with a coronary pressure guidewire (Radi, St. Jude Medical, Uppsala, Sweden) at maximal hyperemia induced by intravenous adenosine, administered at 140 μg/kg/min through a central vein. Hyperemic pressure pull-back recordings were performed as described previously (8,9). All included patients were treated with aspirin and clopidogrel for at least 1 year. Quantitative coronary angiography was performed off line and is extensively described elsewhere (10,11).

End points and 2-year follow-up

MACE were defined as a composite of death, myocardial infarction, or any repeat revascularization. The primary end point was the rate of MACE by 1 year. Secondary end points included death and myocardial infarction rates and MACE and its individual components at 2 years, procedure time, amount of contrast agent used, functional class at 1 and 2 years, and number of antianginal medications. Cost effectiveness was a secondary end point as well. Death was defined as all-cause mortality. Myocardial infarction was defined as a threefold or greater elevation of creatine kinase-myocardial band (CK-MB) level or new Q-waves in ≥2 contiguous leads of the electrocardiogram (ECG) (12). Total CK and CK-MB levels were measured in all patients between 12 and 24 h after PCI. After discharge, follow-up was performed at 1 month, 6 months, 1 year, and 2 years (13).

Statistical analysis

All enrolled patients were included in the analysis of primary and secondary end points according to the intention-to-treat principle. Categorical variables including the primary end point, and its components are described as proportions and were compared by chi-square test. Continuous variables are described as mean and standard deviation and were compared by unpaired ttest or Mann-Whitney Utest. A 2-sided p value of <0.05 was considered to indicate statistical significance. Kaplan-Meier curves are presented to describe the time-to-event distributions for the different end points. All statistical analyses were performed with SAS software version 9.1 (SAS Institute, Cary, North Carolina).

Baseline characteristics and angiographic data

Between January 2006 and September 2007, 1,005 patients were included in 20 centers in the U.S. and Europe. Of these, 496 were randomly assigned to angiography-guided PCI and 509 to FFR-guided PCI. Baseline characteristics of the 2 groups were similar, as were the number of indicated lesions and angiographic extent and severity of CAD (Table 1).

Procedural results

In all, 2,415 stents were placed, of which 2,339 (96.9%) were DES. Significantly more stents per patient were placed in the angiography-guided group compared with the FFR-guided group (2.7 ± 1.2 vs. 1.9 ± 1.3; p < 0.001) (Table 2).In the FFR-guided group, FFR was successfully measured in 94% of all lesions. In 874 (63%) lesions, FFR was 0.80 or less, and these lesions were stented as per protocol. In 513 (37%) lesions, FFR was >0.80, and these lesions were not stented. The procedure time was similar in the 2 groups (70 ± 44 min in the angiography-guided group and 71 ± 43 min in the FFR-guided group; p = 0.51). Significantly more contrast agent was used in the angiography-guided group than in the FFR-guided group (302 ± 127 ml vs. 272 ± 133 ml; p < 0.001). Length of hospital stay was 3.7 ± 3.5 days in the angiography-guided group versus 3.4 ± 3.3 days in the FFR-guided group (p = 0.05).

Adverse events and freedom from angina at 2 years

Complete 2-year follow-up was obtained in 93.6% of patients (36 were lost to follow-up in the angiography-guided group, and 29 were lost to follow-up in the FFR-guided group [p = 0.31]). All-cause mortality at 2 years was 3.8% (n = 19) in the angiography-guided group and 2.6% (n = 13) in the FFR-guided group (p = 0.25) (Table 3).Myocardial infarction occurred in 9.9% (n = 49) in the angiography-guided group and 6.1% (n = 31) in the FFR-guided group (p = 0.03). Of these patients, 13 (2.6%) fulfilled the Academic Research Consortium definition of definite or probable stent thrombosis in the angiography-guided group versus 8 patients (1.6%) in the FFR-guided group.

The 2-year rate of death or myocardial infarction, which was not a pre-specified secondary end point but is an important clinical variable, was 12.9% (n = 64) in the angiography-guided group and 8.4% (n = 43) in the FFR-guided group (p = 0.02). Revascularization rate was 12.7% (n = 63) in the angiography-guided group and 10.6% (n = 54) in the FFR-guided group (p = 0.30). After 2 years, MACE had occurred in 111 patients (22.4%) in the angiography-guided group and in 91 patients (17.9%) in the FFR group (p = 0.08). Event-free survival is shown by Kaplan-Meier curves (Fig. 2).At 2 years, 76% of the patients were free from angina in the angiography-guided group compared with 80% in the FFR-guided group (p = 0.14) (Fig. 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Kaplan-Meier Survival Curves

Kaplan-Meier survival curves according to study group for survival free from (A)major adverse cardiac event (MACE), (B)death, (C)death or myocardial infarction (MI), and (D)revascularization. The red curveindicates the angiography-guided percutaneous coronary intervention (PCI) group; the blue curveindicates the fractional flow reserve (FFR)-guided PCI group. CABG = coronary artery bypass graft surgery.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Patient Follow-Up

Percentage of patients treated by angiography-guided strategy (red bars)and fractional flow reserve (FFR)-guided strategy (blue bars)who were completely free from angina at baseline, and at 1- and 2-year follow-up.

Outcome of the deferred lesions in the FFR-guided group

In the 509 patients in the FFR-guided group, 1,329 stenoses were successfully measured by FFR, of which 816 were stented (FFR ≤0.80) and 513 were deferred (FFR >0.80). During the follow-up of 2 years, among these 509 patients, 9 (1.8%) myocardial infarctions occurred after the index hospitalization of which 8 (1.6%) were related to a new lesion or were stent related, whereas only 1 (0.2%) myocardial infarction occurred in an originally deferred lesion (Fig. 4).In 53 (10.4%) patients in the FFR-guided group, 1 or more repeat revascularizations were performed, of which 37 (7.2%) were related to a new or restenotic lesion and only 16 (3.2%) to an originally deferred lesion.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

2-Year Outcome of Stenoses in FFR Group Initially Deferred on Basis of FFR >0.80

Numbers of late myocardial infarction and repeat revascularization of the stenoses in the fractional flow reserve (FFR) group initially deferred from stenting on the basis of FFR >0.80, and in stenoses in the FFR group that were stented because of FFR ≤0.80.

Strengths and limitations of the FAME study

The FAME study has several specific strengths that should be mentioned. First, because of the liberal inclusion and few exclusion criteria, the study truly reflects everyday practice in performing PCI for multivessel disease. This position is further corroborated by the fact that 53% of all screened patients were actually included in the study (7) and that 43 patients were included per center per year, both numbers exceptionally high for this kind of study (14,15).

Second, the majority of the stenoses was located in the proximal or mid-segments of the 3 major coronary arteries, and in almost 40% of the patients, a proximal left anterior descending artery stenosis was 1 of the target lesions. Furthermore, patients who had undergone PCI in the past were not excluded, as is mostly the case in other studies comparing outcome after different treatment modalities in multivessel CAD (14–17).

Third, the reduction of event rate by 30% was consistently present for all types of events. Fourthly, the high percentage of patients free from angina after 2 years underlines that PCI—especially when applied appropriately—is a very effective way to eliminate ischemia and improve quality of life. Lastly, the very low late infarction and late revascularization rate of initially deferred lesions underscores the safety of an FFR-guided approach to PCI for multivessel disease.

Our results also suggest that outcomes of PCI as compared with those achieved by medical treatment, such as in the COURAGE (Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation) trial (12,14) or by coronary artery bypass graft surgery, such as in the SYNTAX (Synergy Between PCI With Taxus and Cardiac Surgery) trial (15), might be improved if the PCI is performed with FFR guidance, ensuring a functionally complete revascularization with more appropriate use of stents (10). A substudy of the COURAGE trial (18), which demonstrated lower rates of death or myocardial infarction in those patients with the greatest relief of ischemia, further supports the concept that PCI should be guided by physiologic considerations and not solely by anatomic ones.

It has been suggested that including stenoses of 50% to 70% in the FAME study might have biased the outcome because many interventionalists would not stent such lesions. However, just for these lesions (1,174 of a total of 2,764 stenoses), the correlation between anatomic and physiologic severity was worse. Not stenting such stenoses on the basis of the angiogram would have left untreated a hemodynamically significant stenosis in 35% of all lesions (11).

It can be questioned whether the difference in events is due to the first 48 h with catch-up of events thereafter. However, that was not the case. The number of small, periprocedural infarctions (CK-MB 3 to 5 times the upper limit of normal) was low: 3.2% and 2.4%, respectively; and absolute differences in MACE were 2.2% at hospital discharge, 5.1% at 1 year, and 4.5% at 2 years, whereas the difference in rate of myocardial infarction even further increased at the advantage of the FFR-guided group during the second year of follow-up.

Another potential limitation could be that FFR measurement requires some extra manipulations, preferably a central venous line if intravenous adenosine is used, and some basic knowledge about interpretation. But given a well-designed infrastructure in the catheterization laboratory and an experienced interventionalist, it is a simple and quick procedure to be performed without significant delay, as shown in this study.

Finally, the current data are restricted to a 2-year follow-up. Theoretically, unstented lesions in the FFR-guided group could progress and lead to future events beyond this time horizon. We intend to collect data up to a follow-up of 5 years for the present study.