Fractional flow reserve–guided pci versus medical therapy in stable coronary disease

Fractional Flow Reserve–Guided PCI versus Medical Therapy in Stable Coronary Disease Bernard De Bruyne, M.D., Ph.D., Nico H.J. Pijls, M.D., Ph.D., Bindu Kalesan, M.P.H., Emanuele Barbato, M.D., Ph.D., Pim A.L. Tonino, M.D., Ph.D., Zsolt Piroth, M.D., Nikola Jagic, M.D., Sven Mobius-Winckler, M.D., Gilles Rioufol, M.D., Ph.D., Nils Witt, M.D., Ph.D., Petr Kala, M.D., Philip MacCarthy, M.D., Thomas Engström, M.D., Keith G. Oldroyd, M.D., Kreton Mavromatis, M.D., Ganesh Manoharan, M.D., Peter Verlee, M.D., Ole Frobert, M.D., Nick Curzen, B.M., Ph.D., Jane B. Johnson, R.N., B.S.N., Peter Jüni, M.D., and William F. Fearon, M.D., for the FAME 2 Trial Investigators* From the Cardiovascular Center Aalst, Onze-Lieve-Vrouw Clinic, Aalst, Belgium (B.D.B., E.B.); Department of Cardiology, Catharina Hospital, and Department of Biomedical Engineering, Eindhoven Univer- The preferred initial treatment for patients with stable coronary artery disease is the sity of Technology — both in Eindhoven, best available medical therapy. We hypothesized that in patients with functionally the Netherlands (N.H.J.P., P.A.L.T.); Divi- sion of Clinical Epidemiology and Biosta- significant stenoses, as determined by measurement of fractional flow reserve (FFR), tistics, Institute of Social and Preventive percutaneous coronary intervention (PCI) plus the best available medical therapy Medicine and Clinical Trials Unit Bern, would be superior to the best available medical therapy alone.
University of Bern, Bern, Switzerland (B.K., P.J.); Hungarian Institute of Cardiology, Budapest (Z.P.); Clinical Center Kragujevac, Kragujevac, Serbia (N.J.); Heart Center In patients with stable coronary artery disease for whom PCI was being considered, Leipzig, Leipzig, Germany (S.M.-W.); Car- we assessed all stenoses by measuring FFR. Patients in whom at least one stenosis diovascular Hospital, Lyon, France (G.R.); was functionally significant (FFR, ≤0.80) were randomly assigned to FFR-guided Sodersjukhuset, Stockholm (N.W.), and Örebro University Hospital, Örebro (O.F.) PCI plus the best available medical therapy (PCI group) or the best available medical — both in Sweden; Department of Inter- therapy alone (medical-therapy group). Patients in whom all stenoses had an FFR of nal Medicine and Cardiology, University Hospital Brno, Brno, Czech Republic more than 0.80 were entered into a registry and received the best available medical (P.K.); King's College Hospital, London therapy. The primary end point was a composite of death, myocardial infarction, or (P.M.), Golden Jubilee National Hospital, Glasgow (K.G.O.), Royal Victoria Hospi-tal, Belfast (G.M.), and Department of Cardiology, Southampton University Hos-pital Trust, Southampton (N.C.) — all in Recruitment was halted prematurely after enrollment of 1220 patients (888 who the United Kingdom; Department of Car- underwent randomization and 332 enrolled in the registry) because of a significant diology, Rigshospitalet University Hospi- between-group difference in the percentage of patients who had a primary end- tal, Copenhagen (T.E.); Atlanta Veterans Affairs Medical Center, Atlanta (K.M.); point event: 4.3% in the PCI group and 12.7% in the medical-therapy group (hazard Northeast Cardiology Associates, Bangor, ratio with PCI, 0.32; 95% confidence interval [CI], 0.19 to 0.53; P<0.001). The dif- ME (P.V.); St. Jude Medical, Plymouth, MN (J.B.J.); and Stanford University Medical ference was driven by a lower rate of urgent revascularization in the PCI group than Center, Stanford, CA (W.F.F.). Address in the medical-therapy group (1.6% vs. 11.1%; hazard ratio, 0.13; 95% CI, 0.06 to 0.30; reprint requests to Dr. De Bruyne at the P<0.001); in particular, in the PCI group, fewer urgent revascularizations were trig- Cardiovascular Centre Aalst, OLV-Clinic, Moorselbaan 164, B-9300 Aalst, Belgium, gered by a myocardial infarction or evidence of ischemia on electrocardiography or at [email protected].
(hazard ratio, 0.13; 95% CI, 0.04 to 0.43; P<0.001). Among patients in the registry, *The investigators in the Fractional Flow 3.0% had a primary end-point event.
Reserve versus Angiography for Multi-vessel Evaluation 2 (FAME 2) trial are listed in the Supplementary Appendix, In patients with stable coronary artery disease and functionally significant stenoses, available at NEJM.org.
FFR-guided PCI plus the best available medical therapy, as compared with the best This article was published on August 28, available medical therapy alone, decreased the need for urgent revascularization. In 2012, at NEJM.org.
patients without ischemia, the outcome appeared to be favorable with the best N Engl J Med 2012.
available medical therapy alone. (Funded by St. Jude Medical; ClinicalTrials.gov DOI: 10.1056/NEJMoa1205361
Copyright 2012 Massachusetts Medical Society. number, NCT01132495.) n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. Percutaneous coronary interven- as necessary for the oversight of the trial. No for- tion (PCI) improves the outcome in patients mal stopping rules were specified. The academic with acute coronary syndromes.1 In contrast, members of the steering committee had full ac- for the treatment of patients with stable coronary cess to all the data in the study, vouch for the artery disease, controversy persists regarding the accuracy and completeness of the data and analy- extent of the benefit from PCI, as compared with ses and for the fidelity of the study to the proto- the best available medical therapy, as an initial col, wrote the manuscript, and had final respon- management strategy.2-5 The potential benefit of sibility for the decision to submit it for publication. revascularization depends on the presence and The research protocol is available at NEJM.org.
extent of myocardial ischemia.6-8 Performing PCI on nonischemic stenoses is not beneficial9 and is Patients
probably harmful.10 Thus, careful selection of Patients in stable condition who were appropriate ischemia-inducing stenoses is essential for deriv- candidates for PCI and who had angiographically ing the greatest benefit from revascularization in assessed one-, two-, or three-vessel coronary ar- patients with stable coronary artery disease.
tery disease suitable for PCI were included in the Fractional flow reserve (FFR) is a pressure- trial. Details of the inclusion and exclusion crite- wire–based index that is used during coronary ria are provided in the Supplementary Appendix. angiography to assess the potential of a coronary The investigator first indicated which stenoses were stenosis to induce myocardial ischemia.11-14 The thought to require stenting on the basis of the usefulness of FFR-guided PCI as compared with clinical and angiographic data. FFR was then mea- PCI guided by angiography alone is supported by sured with a coronary guidewire (PressureWire robust clinical outcome data.9,10,14-17 Certus or PressureWire Aeris, St. Jude Medical) The aim of this trial was to determine whether during adenosine-induced hyperemia to assess FFR-guided PCI with drug-eluting stents plus the the hemodynamic severity of each indicated ste- best available medical therapy is superior to the nosis. Patients who had at least one stenosis in a best available medical therapy alone in reducing major coronary artery with an FFR of 0.80 or less the rate of death, myocardial infarction, or un- were randomly assigned, by means of an interac- planned hospitalization leading to urgent revascu- tive voice-response system, to FFR-guided PCI plus larization among patients with stable coronary the best available medical therapy (hereinafter artery disease.
called the PCI group) or to the best available med- ical therapy alone (hereinafter called the medical- therapy group). The randomization schedule was computer-generated; randomization was strati- Study Design and Oversight
fied according to site and performed in blocks, The Fractional Flow Reserve versus Angiography with block sizes varied randomly. Patients with for Multivessel Evaluation 2 (FAME 2) study is a an FFR of more than 0.80 in all vessels with indi- randomized "all comers" trial (i.e., involving the cated stenoses were enrolled in a registry and consecutive enrollment of all eligible patients received the best available medical therapy. A with stable coronary artery disease). The trial random sample of 50% of the registry patients was conducted at 28 sites in Europe and North underwent the same follow-up as the patients in America and was approved by the institutional the randomized trial. The treatment assignments review board at each participating center. The were known to the patients. All patients provided members of the steering committee (see the Sup- written informed consent.
plementary Appendix, available with the full text of this article at NEJM.org) designed the study with- Treatment
out involvement of the sponsor, St. Jude Medical. All patients were prescribed aspirin at a dose of The sponsor was involved in the collection and 80 to 325 mg daily, metoprolol at a dose of 50 to source verification of the data but not in the con- 200 mg daily (or any other beta-1–selective blocker, duct of the trial or in the decision to terminate it. alone or in combination with a calcium-channel An independent data and safety monitoring blocker or a long-acting nitrate), lisinopril (≥5 mg board (see the Supplementary Appendix) oversaw daily, or another angiotensin-converting–enzyme the trial and met twice a year or more frequently, [ACE] inhibitor or an angiotensin II–receptor n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. Fractional Flow Reserve in Stable Coronary Disease blocker if the patient had unacceptable side ef- treatment assignments. For each revasculariza- fects with the ACE inhibitor), and atorvastatin tion procedure, a detailed description was in- (20 to 80 mg daily, or another statin of similar cluded. Revascularization was considered to be potency alone or in combination with ezetimibe, to urgent when a patient was admitted to the hospi- reduce the low-density-lipoprotein [LDL] level to tal with persistent or increasing chest pain (with less than 70 mg per deciliter [1.8 mmol per liter]). or without ST-segment or T-wave changes or ele- Patients who were randomly assigned to PCI vated biomarker levels) and the revascularization received a loading dose of clopidogrel (600 mg) procedure was performed during the same hos- and aspirin immediately before the procedure if pitalization.
they were not already taking these medications. All stenoses with an FFR of 0.80 or less were treated Statistical Analysis
with second-generation drug-eluting stents.18-20 The trial was powered to determine whether PCI After PCI, all patients received clopidogrel at a with the best available medical therapy was supe- dose of 75 mg per day for at least 12 months in rior to the best available medical therapy alone with addition to the best available medical therapy. All respect to the primary end point at 24 months. patients were given a medication tracking form On the basis of findings from previous studies and for recording weekly medication use and doses. using binomial proportions, we estimated that the Patients who smoked were counseled regarding cumulative incidence of the primary end point at smoking cessation. Patients with diabetes were 24 months would be 12.6% in the PCI group18-20 referred to a diabetes specialist to receive the and 18.0% in the medical-therapy group,3 corre- best available treatment for that disease.
sponding to a relative risk reduction with PCI of 30%, and that with 816 patients in each group, the study would have more than 84% power to Electrocardiography (ECG) was performed with the detect that relative risk reduction, at a two-sided patient at rest, and the creatine kinase level and type I error rate of 0.05. Continuous variables are the MB fraction of creatine kinase were measured presented as means and standard deviations, and in all patients before angiography was performed categorical data are presented as numbers and per- and between 12 and 24 hours after enrollment. centages. All patients were included in the analy sis Follow-up visits were scheduled at 1 and 6 months according to the groups to which they were orig- and at 1, 2, 3, 4, and 5 years. At baseline and all inally assigned (intention-to-treat analysis). We follow-up visits, we obtained information regard- used the Mantel–Cox method to calculate hazard ing the presence or absence (and, if present, the ratios and 95% confidence intervals for the be- severity) of angina, the patient's work status, and tween-group comparisons of clinical outcomes the number and doses of cardiac medications and and the log-rank test to calculate corresponding assessed the patient's quality of life with the use P values. We constructed Kaplan–Meier curves for of the European Quality of Life–5 Dimensions the primary end point and its components. [EQ-5D] instrument.21 In addition, we performed In an exploratory analysis, we also calculated resting ECG, measured the levels of total choles- the hazard ratio for urgent revascularization terol and cholesterol fractions, and assessed the triggered by a myocardial infarction or by un- patient's utilization of medical resources.
stable angina with evidence of ischemia on ECG. We used a chi-square test to assess the interac- End Points
tion between treatment effect and these charac- The prespecified primary end point was a compos- teristics. Landmark analyses were performed ite of death from any cause, nonfatal myocardial according to a landmark point at 7 days, with infarction, or unplanned hospitalization leading to the hazard ratio calculated separately for events urgent revascularization during the first 2 years. that occurred up to 7 days after randomization Secondary end points included individual com- and events that occurred between 8 days and the ponents of the primary end point, cardiac death, end of the follow-up period. We then performed nonurgent revascularization, and angina class. a test for the interaction between treatment and All outcomes were adjudicated by an independent time (first 7 days vs. subsequent period). In all clinical events committee (see the Supplementary time-to-event analyses (i.e., overall and land- Appendix) whose members were unaware of the mark), for each type of event, data for a patient n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. Table 1. Baseline Clinical, Angiographic, and Fractional Flow Reserve (FFR) Characteristics.*
Randomly Assigned Groups
P Value†
PCI plus Medical Medical Therapy Patient characteristics
Total no. of patients
Male sex — no. (%) Body-mass index‡ Family history of coronary artery disease — no. (%) Current smoking — no. (%) Hypertension — no. (%) Hypercholesterolemia — no. (%) Diabetes mellitus — no. (%) Renal insufficiency — no. (%)§ Peripheral vascular disease — no. (%) History of stroke or transient ischemic attack — no. (%) History of myocardial infarction — no./total no. (%) History of PCI in target vessel — no. (%) Angina — no./total no (%)¶ CCS class IV, stabilized Silent ischemia — no. (%) Left ventricular ejection fraction <50% — no./total no. (%) Angiographic findings Angiographically significant lesions — no. per patient Vessels with at least one significant lesion — no. of patients (%) At least one significant lesion in proximal or middle left anterior descending artery — no. (%) Functionally significant lesions — no. per patient Vessels with at least one significant lesion — no. of patients (%) At least one significant lesion in proximal or middle left anterior descending artery — no. (%) n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. Fractional Flow Reserve in Stable Coronary Disease Table 1. (Continued.)
Randomly Assigned Groups
P Value†
PCI plus Medical Medical Therapy Lesion characteristics
Total no. of lesions
Angiographic findings Lesions with stenosis of >50% of diameter — no. (%) Stenosis — no. (%) <50% of diameter 50–69% of diameter 70–90% of diameter >90% of diameter Lesions with FFR ≤0.80 — no. (%) Mean FFR in lesions with FFR ≤0.80 * Plus–minus values are means ±SD. There were no significant differences between the two randomly assigned groups in any of the baseline charac- teristics, with the exception of left ventricular ejection fraction of less than 50% (P = 0.04). PCI denotes percutaneous coronary intervention.
† The P values are for the combined groups that underwent randomization (the group assigned to PCI plus the best available medical therapy and the group assigned to the best available medical therapy alone) as compared with the group of patients who did not undergo random- ization (patients in whom all stenoses had an FFR of more than 0.80) and were enrolled in a registry. In patient-level analyses, the P values were calculated with the use of a chi-square test except when cell numbers were small (<15 patients), in which case Fisher's exact test was used. In lesion-level analyses, mixed maximum-likelihood logistic-regression models were used for between-group comparisons of dichoto- mous variables, and mixed maximum-likelihood linear-regression models were used for comparisons of continuous variables, to account for the correlation of multiple lesions within patients.
‡ The body-mass index is the weight in kilograms divided by the square of the height in meters.
§ Renal insufficiency was defined as a creatinine level of more than 2.0 mg per deciliter (176.8 μmol per liter).
¶ Angina was classified according to the Canadian Cardiovascular Society (CCS) functional classification, in which classes range from I to IV, with higher classes indicating greater limitations on physical activity owing to angina.
‖ Five totally occluded arteries supplied infarcted areas and were therefore not considered for revascularization by means of PCI.
were censored at the time of the first event that with an FFR of 0.80 or less in a large epicardial occurred in that patient. All analyses were per- artery: 447 patients were randomly assigned to formed by an independent statistician from an FFR-guided PCI plus the best available medical academic clinical trials unit (CTU Bern, Univer- therapy, and 441 patients to the best available sity of Bern, Switzerland) with the use of Stata medical therapy alone. In 332 patients with an- software, version 11.2.
giographically significant stenoses, none of the stenoses had an FFR of 0.80 or less; these patients were enrolled in the registry and received the best available medical therapy alone. The mean Study Termination and Patient Follow-up
(±SD) duration of follow-up was 213±128 days At the recommendation of the independent data among patients assigned to PCI plus the best and safety monitoring board, patient recruitment medical available therapy, 214±127 days among was stopped on January 15, 2012, owing to a high- patients assigned to the best available medical ly significant difference in the incidence rates of therapy alone, and 206±119 days among patients the primary end point between the PCI and med- enrolled in the registry.
ical-therapy groups. Between May 15, 2010, and January 15, 2012, a total of 1220 patients were Baseline Characteristics
enrolled (Fig. S1 in the Supplementary Appendix). Table 1 shows the baseline clinical, angiographic, A total of 888 patients had at least one stenosis and FFR characteristics of the patients who under- n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. went randomization, as compared with the pa- gent revascularization did differ significantly be- tients who were enrolled in the registry. There were tween the groups (hazard ratio with PCI, 0.13; 95% higher percentages of men, patients with periph- CI, 0.06 to 0.30; P<0.001). Among the 56 patients eral vascular disease, and patients with multivessel who underwent urgent revascularization, the pro- disease in the groups that underwent randomiza- cedure was triggered by a myocardial infarction tion than in the registry cohort. More than 25% of in 12 patients (21.4%), by unstable angina accom- the patients had diabetes, and 68% of the patients panied by evidence of ischemia on ECG in 15 pa- had angina of class II to IV on the Canadian Car- tients (26.8%), and by unstable angina diagnosed diovascular Society (CCS) scale (which ranges from on the basis of clinical features in 29 patients I to IV, with higher classes indicating greater (51.8%). In an exploratory analysis, 4 patients in limitations on physical activity owing to angina). the PCI group (0.9%) and 23 patients in the med- There were more lesions per patient and more ical-therapy group (5.2%) underwent an urgent lesions with stenosis of more than 70% of the revascularization that was triggered by a myocar- diameter of the artery among patients who under- dial infarction or by unstable angina with evi- went randomization than among patients in the dence of ischemia on ECG (hazard ratio with registry. A total of 1601 stenoses in the patients PCI, 0.13; 95% CI, 0.04 to 0.43; P<0.001). As com- who underwent randomization were considered pared with patients in the medical-therapy group, for PCI on the basis of angiographic findings, patients in the PCI group were significantly less whereas 1304 were considered for PCI on the ba- likely to undergo any revascularization (hazard sis of an FFR of 0.80 or less. Among the latter, the ratio with PCI, 0.14; 95% CI, 0.08 to 0.26) or non- FFR ranged from 0.19 to 0.80. Table S1 in the urgent revascularization (hazard ratio, 0.17; 95% Supplementary Appendix shows the medications CI, 0.08 to 0.39) (Table 2, and Fig. S2 in the Sup- the patients were taking at baseline and during plementary Appendix). Among patients in the the follow-up period.
registry, the rates of death from any cause, myo- cardial infarction, urgent revascularization, and Primary End Point
nonurgent revascularization were all low (Fig. 1, By January 15, 2012, a total of 75 patients in the and Table S2 in the Supplementary Appendix).
randomized groups had had at least one primary Figure 2 shows the results from landmark end-point event. The percentage of patients who analyses of the primary end point and its com- had a primary end-point event was lower in the PCI ponents. PCI plus the best available medical thera- group than in the medical-therapy group (4.3% vs. py was shown to be consistently more beneficial 12.7%; hazard ratio with PCI, 0.32; 95% confi- after the landmark point of 7 days after random- dence interval [CI], 0.19 to 0.53; P<0.001) (Fig. 1A ization than before; there were significant inter- and Table 2). In the registry cohort, 5 patients had actions between time and treatment with respect at least one primary end-point event (3.0%). There to the primary end point, the individual compo- was little difference in the incidence of a primary nents of death and myocardial infarction, and end-point event between patients in the PCI group the composite of death or myocardial infarction, and patients in the registry (hazard ratio for the as well as a trend toward an interaction with re- PCI group, 1.29; 95% CI, 0.49 to 3.39; P = 0.61), spect to urgent revascularization. Corresponding but there was a large difference between patients Kaplan–Meier curves are presented in Figure S3 in in the medical-therapy group and patients in the the Supplementary Appendix. Stratified analyses registry (hazard ratio for the medical-therapy according to patient characteristics are shown in group, 4.32; 95% CI, 1.75 to 10.66; P = 0.001) (Table Figure S4 in the Supplementary Appendix. Effects S2 in the Supplementary Appendix).
were similar across most subgroups; however, the benefit of PCI appeared to be more pronounced Secondary End Points
among patients who had lesions with an FFR of The Kaplan–Meier curves for the individual com- less than 0.65 than among patients who had only ponents of the primary end point are shown in lesions with larger FFR values (P = 0.01 for the in- Figures 1B, 1C, and 1D. Neither the rate of death teraction). The reduction from baseline in the per- from any cause nor the rate of myocardial infarc- centage of patients with angina of CCS grade II to tion differed significantly between the PCI group IV was greater in the PCI group than in the medi- and the medical-therapy group, but the rate of ur- cal-therapy group and the registry cohort (Fig. 3).
n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. Fractional Flow Reserve in Stable Coronary Disease A Primary End Point
B Death from Any Cause
PCI vs. medical therapy: PCI vs. medical therapy: Hazard ratio, 0.32 (95% CI, 0.19–0.53); P<0.001 Hazard ratio, 0.33 (95% CI, 0.03–3.17); P=0.31 PCI vs. registry: PCI vs. registry: Hazard ratio, 1.29 (95% CI, 0.49–3.39); P=0.61 Hazard ratio, 1.12 (95% CI, 0.05–27.33); P=0.54 Medical therapy vs. registry: Medical therapy vs. registry: Hazard ratio, 4.32 (95% CI, 1.75–10.70); P<0.001 Hazard ratio, 2.66 (95% CI, 0.14–51.18); P=0.30 Cumulative Incidence (%)
Cumulative Incidence (%)
Months since Randomization
Months since Randomization
No. at Risk
No. at Risk
441 414 370 322 283 253 220 192 162 127 100 70 37 441 423 390 350 312 281 247 219 188 154 122 90 54 447 414 388 351 308 277 243 212 175 155 117 92 53 447 423 396 359 318 288 250 220 183 163 122 95 54 166 156 145 133 117 106 93 74 64 52 41 25 13 166 156 145 134 118 107 96 76 67 55 43 27 13 C Myocardial Infarction
D Urgent Revascularization
PCI vs. medical therapy: PCI vs. medical therapy: Hazard ratio, 1.05 (95% CI, 0.51–2.19); P=0.89 Hazard ratio, 0.13 (95% CI, 0.06–0.30); P<0.001 PCI vs. registry: PCI vs. registry: Hazard ratio, 1.61 (95% CI, 0.48–5.37); P=0.41 Hazard ratio, 0.63 (95% CI, 0.19–2.03); P=0.43 Medical therapy vs. registry: Medical therapy vs. registry: Hazard ratio, 1.65 (95% CI, 0.50–5.47); P=0.41 Hazard ratio, 4.65 (95% CI, 1.72–12.62); P<0.001 Cumulative Incidence (%)
Cumulative Incidence (%)
Months since Randomization
Months since Randomization
No. at Risk
No. at Risk
441 421 386 341 304 273 239 212 182 148 117 85 48 441 414 371 325 286 256 223 195 164 129 101 71 38 447 414 388 352 309 278 244 214 177 157 119 94 54 447 421 395 356 315 285 248 217 180 160 119 93 53 166 156 145 134 118 107 95 75 65 53 42 26 13 166 156 145 133 117 106 94 75 65 53 42 26 13 Figure 1. Cumulative Incidence of the Primary End Point and Its Components.
Kaplan–Meier curves are shown for the cumulative incidence of the primary end point of death, myocardial infarction, or urgent revascu-
larization (Panel A) and the individual components of the primary end point (Panels B, C, and D) in the group that was randomly assigned
to PCI and the best available medical therapy (PCI), the group that was randomly assigned to the best available medical therapy alone
(medical therapy), and the group that did not undergo randomization and was enrolled in a registry (registry). After 12 months, a total
of two primary end-point events occurred in the PCI group, none in the medical-therapy group, and one in the registry cohort. No deaths
occurred after 12 months in any of the groups. Two patients in the PCI group, none in the medical-therapy group, and one in the registry
cohort had a myocardial infarction after 12 months. One patient in the registry cohort, and none in the other two groups, had an urgent
revascularization performed after 12 months.
ease and hemodynamically significant stenoses. FFR-guided PCI with drug-eluting stents plus the In the FAME 2 trial, we compared the treatment best available medical therapy, as compared with strategy of PCI, performed according to current the best available medical therapy alone, resulted quality standards, plus the best available medical in significantly improved clinical outcomes. The therapy with the best available medical therapy difference between the two strategies was driven alone in patients with stable coronary artery dis- by an increase by a factor of 8 in the need for n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. Table 2. Clinical Events.
Registry Cohort
Randomly Assigned Groups
PCI plus Medical Primary end point 0.32 (0.19–0.53) Components of primary end point Death from any cause 0.33 (0.03–3.17) Myocardial infarction 1.05 (0.51–2.19) Urgent revascularization 0.13 (0.06–0.30) Death or myocardial infarction 0.61 (0.28–1.35) 0.96 (0.06–15.17) 0.14 (0.08–0.26) 0.17 (0.08–0.39) 0.49 (0.04–5.50) Definite or probable stent thrombosis 4.98 (0.59–42.25) urgent revascularization in the medical-therapy previous trials involving patients with stable coro- group. In the case of half of these urgent revas- nary disease.3,4 First, in previous trials in which cularizations, the need for the procedure was various revascularization methods were com- triggered by an increase in biomarker levels, pared with the best available medical therapy, ischemic changes on ECG, or both. When we patient enrollment was based primarily on an- performed a landmark analysis, we found that giographic findings, with or without noninva- the strategy of PCI plus the best available medical sive documentation of ischemia. It is likely that therapy was more beneficial 8 days or more after a sizable proportion of the patients had only randomization than 7 days or less after random- limited ischemia. Even in the Clinical Outcomes ization, with interactions between time and Utilizing Revascularization and Aggressive Drug treatment with respect to the primary end point, Evaluation (COURAGE) trial, in which noninvasive as well as with respect to death and myocardial testing was performed in 85% of the patients,3 infarction, suggesting that the benefit of PCI less than one third of the patients had more than plus the best available medical therapy might be- 10% ischemia on myocardial perfusion imag- come more pronounced with an increasing dura- ing.8 In daily clinical practice, less than half of tion of follow-up. The percentage of patients with patients undergo noninvasive stress testing be- angina of CCS class II to IV was markedly lower fore elective PCI.22 In the current trial, all the among patients in the PCI group than among patients who underwent randomization had at patients in the medical-therapy group. Moreover, least one functionally significant stenosis. More- in 25% of the patients in whom PCI was consid- over, a mean FFR value of 0.68 in large epicar- ered, none of the stenoses that were visible on an dial arteries suggests that there were large areas angiogram were hemodynamically significant as of myocardium that were at risk for ischemia. The assessed by means of the measurement of FFR. low-risk patients with nonischemic FFR values Among these patients, the strategy of providing were not randomly assigned to a study group but the best available medical therapy alone was as- were followed in a registry — a study design that sociated with a very low event rate.
was unlike that of previous trials.
Several factors may explain the differences be- Second, among patients in the PCI group who tween results in the present study and those in had several stenoses, PCI was performed only in n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. Fractional Flow Reserve in Stable Coronary Disease P Value for
End Point
Relative Risk (95% CI)
P Value Interaction
no. of events/total no. (%) Primary end point 2.24 (0.69–7.31) 8 days to maximum follow-up 0.17 (0.09–0.35) Death or myocardial infarction 7.99 (0.99–64.57) 8 days to maximum follow-up 0.42 (0.17–1.04) 8 days to maximum follow-up 0.33 (0.03–3.17) Myocardial infarction 7.99 (0.99–64.57) 8 days to maximum follow-up 0.52 (0.21–1.32) Urgent revascularization 0.49 (0.09–2.70) 8 days to maximum follow-up 0.10 (0.04–0.26) PCI Better Medical Therapy
Figure 2. Landmark Analysis of the Primary End Point and Its Components.
The relative risk of the primary end point of death, myocardial infarction, or urgent revascularization and of components of the primary
end point are shown, according to the time from randomization (7 days or less vs. 8 days or more). The solid boxes represent relative-
risk estimates for 7 days or less after randomization, and the open boxes represent relative-risk estimates for 8 days to the maximum fol-
low-up. Arrows indicate that the lower end of the confidence interval is less than 0.1. (The lower end of the confidence interval for urgent
revascularization at 8 days to maximum follow-up, which could not be shown on the plot, was 0.04.) P values were calculated with the
use of a log-rank test, except for the following, which were calculated with the use of Fisher's exact test: death or myocardial infarction at
7 days or less; death at 8 days to maximum follow-up; myocardial infarction at 7 days or less; and urgent revascularization at 7 days or
less. P values for the interaction between time and treatment with respect to the end points were calculated with the use of the Mantel–
Cox method. A total of 10 patients randomly assigned to PCI plus the best available medical therapy and 8 patients assigned to the best
available medical therapy alone underwent randomization during the week before January 15, 2012, and their data were censored for the
analysis of 8 days to maximum follow-up. In addition, 1 patient in each of those groups withdrew consent during the first week of follow-
up, and their data were also censored in the analysis of the subsequent period.
lesions with an FFR of 0.80 or less. This FFR- talizations with urgent revascularization among guided approach is associated with a better clini- patients randomly assigned to the best avail- cal outcome than that with PCI performed on able medical therapy alone than among those the basis of angiographic results alone.10 These assigned to PCI plus the best available medical features probably explain the similarity of event therapy.
rates between patients who were treated with Finally, the primary end point of the present PCI plus the best available medical therapy and study included not only death and myocardial patients with equivalent baseline characteristics infarction but also urgent revascularization, a but no functionally significant lesions who were component that was not included in the primary enrolled in the registry and treated with the best end point of previous trials. The definition of available medical therapy alone.
urgent revascularization was stringent in order Third, we used drug-eluting stents in patients to distinguish it from nonurgent — albeit clini- who underwent PCI, a strategy that resulted in a cally justified — revascularizations. Among pa- low number of repeat revascularizations.18-20 The tients who underwent urgent revascularization, use of anti-ischemic medication was similar to the clinical presentation met the criteria of an that reported in the COURAGE trial23 and was acute coronary syndrome as assessed by an inde- most likely much higher than that in routine pendent clinical events committee whose mem- clinical practice.24 Nevertheless, receipt of the best bers were unaware of the treatment assignments. available medical therapy did not preclude a sig- In half the patients who underwent an urgent nificantly higher number of unplanned hospi- revascularization, the unstable nature of the symp- n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. Randomly Assigned Groups
Randomly Assigned Groups
vs. Registry
1.04 (0.95–1.13) 1.08 (0.95–1.23) 1.04 (0.92–1.19) 0.39 (0.28–0.53) 0.72 (0.45–1.15) 1.87 (1.25–2.81) 0.46 (0.28–0.74) 0.62 (0.33–1.19) 1.36 (0.77–2.40) 12 Months
0.15 (0.02–1.19) 0.17 (0.01–2.42) 1.14 (0.16–8.03) Patients with CCS II to IV (%)
Figure 3. Patients with Angina Class II to IV and Corresponding Relative Risks.
The percentage of patients with angina of class II to IV on the Canadian Cardiovascular Society (CCS) scale (which ranges from I to IV,
with higher classes indicating greater limitations on physical activity owing to angina) and the corresponding relative risks are shown at
various time points for the group that was randomly assigned to PCI and the best available medical therapy (PCI), the group that was
randomly assigned to the best available medical therapy alone (medical therapy), and the group that did not undergo randomization and
was enrolled in a registry (registry). A total of 414 patients in the PCI group, 417 in the medical-therapy group, and 151 in the registry cohort
were eligible for follow-up at 30 days; the corresponding numbers at 6 months were 238, 241, and 92, and the corresponding numbers at
1 year were 41, 38, and 7.
toms was evidenced by ST-segment depression, data and safety monitoring board believed that biomarker elevation, or both. The occurrence of exposing more patients with functionally signifi- an acute coronary syndrome necessitates hospi- cant stenoses to the risk of urgent revasculariza- talization and is associated with an unfavorable tion was inappropriate. Second, although random- prognosis, and it should therefore be considered ization was concealed,29 it is possible that the to be a treatment failure. More important, revas- awareness of the presence of a stenosis influ- cularization has been shown to improve the rate enced decisions regarding revascularization. Third, of survival and decrease the risk of myocardial even though the adherence to medications was infarction among high-risk patients with an acute high, the best available medical therapy did not coronary syndrome.25-27 include interventions by nurse case managers that The trial has several limitations. First, because were aimed at lifestyle changes and risk-factor of the premature termination of enrollment, there reduction, interventions that were included as was an unusually short follow-up period — too part of the best available medical therapy in the short to see restenosis emerge as a complication COURAGE trial.23 Fourth, the strategic nature of of PCI. Differences in the rates of death and the trial meant that we followed contemporary myocardial infarction between the strategies of guidelines,30 which require dual antiplatelet treat- PCI and medical therapy alone that were seen in ment only for patients who undergo stenting. It one recent registry study28 could not be confirmed. is unlikely that this difference in drug regimen However, the difference in the primary outcome between the two groups could explain the mag- between the two treatment groups was large and nitude of the observed difference with respect to was steadily increasing over time; therefore, the the primary end point.
n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. Fractional Flow Reserve in Stable Coronary Disease In conclusion, among patients with stable coro- that were not functionally significant, the best nary artery disease and at least one stenosis with available medical therapy alone resulted in an an FFR of 0.80 or less, FFR-guided PCI with drug- excellent outcome, regardless of the angiographic eluting stents plus the best available medical ther- appearance of the stenoses.
apy, as compared with the best available medical therapy alone, decreased the rate of urgent re- Supported by St. Jude Medical.
Disclosure forms provided by the authors are available with vascularization. Among patients with stenoses the full text of this article at NEJM.org.
1. Mehta SR, Cannon CP, Fox KA, et al. myocardial, and collateral blood flow by based EQ-5D index score. August 2005
Routine vs selective invasive strategies in pressure measurements for assessing (http:/ www.ahrq.gov/rice/EQ5Dscore.htm).
patients with acute coronary syndromes: functional stenosis severity before and 22. Lin GA, Dudley RA, Lucas FL, Malenka
a collaborative meta-analysis of random- after percutaneous transluminal coronary DJ, Vittinghoff E, Redberg RF. Frequency ized trials. JAMA 2005;293:2908-17.
angioplasty. Circulation 1993;87:1354-67.
of stress testing to document ischemia 2. Katritsis DG, Ioannidis JPA. Percuta-
12. De Bruyne B, Pijls NH, Paulus WJ, prior to elective percutaneous coronary
neous coronary intervention versus con- Vantrimpont PJ, Sys SU, Heyndrickx GR. intervention. JAMA 2008;300:1765-73.
servative therapy in nonacute coronary Transstenotic coronary pressure gradient 23. Maron DJ, Boden WE, O'Rourke RA,
artery disease: a meta-analysis. Circula- measurement in humans: in vitro and in et al. Intensive multifactorial intervention vivo evaluation of a new pressure moni- for stable coronary artery disease: opti- 3. Boden WE, O'Rourke RA, Teo KK, et toring angioplasty guide wire. J Am Coll mal medical therapy in the COURAGE
al. Optimal medical therapy with or with- Cardiol 1993;22:119-26.
(Clinical Outcomes Utilizing Revascular- out PCI for stable coronary disease. N Engl 13. De Bruyne B, Baudhuin T, Melin JA, et ization and Aggressive Drug Evaluation)
J Med 2007;356:1503-16.
al. Coronary flow reserve calculated from trial. J Am Coll Cardiol 2010;55:1348-58.
4. The BARI 2D Study Group. A random-
pressure measurements in humans: vali- 24. Naderi SH, Bestwick JP, Wald DS. Ad-
ized trial of therapies for type 2 diabetes dation with positron emission tomogra- herence to drugs that prevent cardiovas- and coronary artery disease. N Engl J Med phy. Circulation 1994;89:1013-22.
cular disease: meta-analysis on 376,162 14. Pijls NHJ, de Bruyne B, Peels K, et al. patients. Am J Med 2012 June 27 (Epub
5. Mukherjee D, Moliterno DJ. Effective-
Measurement of fractional flow reserve to ahead of print).
ness of PCI for non-acute coronary artery assess the functional severity of coronary- 25. Wallentin L, Lagerqvist B, Husted S,
disease. Lancet 2009;373:870-2.
artery stenoses. N Engl J Med 1996;334: Kontny F, Ståhle E, Swahn E. Outcome at 6. Hachamovitch R, Hayes SW, Fried-
1 year after an invasive compared with a man JD, Cohen I, Berman DS. Compari- 15. Hamilos M, Muller O, Cuisset T, et al. non-invasive strategy in unstable coro-
son of the short-term survival benefit as- Long-term clinical outcome after frac- nary-artery disease: the FRISC II invasive sociated with revascularization compared tional flow reserve-guided treatment in randomised trial. Lancet 2000;356:9-16.
with medical therapy in patients with no patients with angiographically equivocal 26. Fox KA, Poole-Wilson PA, Henderson
prior coronary artery disease undergoing left main coronary artery stenosis. Circu- RA, et al. Interventional versus conserva- stress myocardial perfusion single photon lation 2009;120:1505-12.
tive treatment for patients with unstable emission computed tomography. Circula- 16. Muller O, Mangiacapra F, Ntalianis A, angina or non–ST-elevation myocardial
et al. Long-term follow-up after fractional infarction: the British Heart Foundation 7. Pijls NH, van Schaardenburgh P, flow reserve-guided treatment strategy in RITA 3 randomised trial: Randomized In-
Manoharan G, et al. Percutaneous coro- patients with an isolated proximal left an- tervention Trial of unstable Angina. Lan- nary intervention of functionally nonsig- terior descending coronary artery stenosis. cet 2002;360:743-51.
nificant stenosis: 5-year follow-up of the JACC Cardiovasc Interv 2011;4:1175-82.
27. Damman P, Hirsch A, Windhausen F,
DEFER Study. J Am Coll Cardiol 2007;49: 17. Puymirat E, Peace A, Mangiacapra F, Tijssen JG, de Winter RJ, ICTUS Investi-
et al. Long-term clinical outcome after gators. 5-Year clinical outcomes in the 8. Shaw LJ, Berman DS, Maron DJ, et al. fractional flow reserve-guided percutane-
ICTUS (Invasive versus Conservative Treat- Optimal medical therapy with or without ous coronary revascularization in pa- ment in Unstable coronary Syndromes) percutaneous coronary intervention to re- tients with small-vessel disease. Circ Car- trial: a randomized comparison of an duce ischemic burden: results from the diovasc Interv 2012;5:62-8.
early invasive versus selective invasive Clinical Outcomes Utilizing Revascular- 18. Windecker S, Serruys PW, Wandel S, management in patients with non-
ization and Aggressive Drug Evaluation et al. Biolimus-eluting stent with biode- ST-segment elevation acute coronary syn- (COURAGE) trial nuclear substudy. Circu- gradable polymer versus sirolimus-eluting drome. J Am Coll Cardiol 2010;55:858-64.
stent with durable polymer for coronary 28. Hannan EL, Samadashvili Z, Cozzens
9. Bech GJ, De Bruyne B, Pijls NHJ, et al. revascularisation (LEADERS): a ran-
K, et al. Comparative outcomes for pa- Fractional flow reserve to determine the domised non-inferiority trial. Lancet 2008; tients who do and do not undergo percu- appropriateness of angioplasty in moder- 372:1163-73.
taneous coronary intervention for stable ate coronary stenosis: a randomized trial. 19. Stone GW, Rizvi A, Newman W, et al. coronary artery disease in New York. Cir-
Everolimus-eluting versus paclitaxel-eluting culation 2012;125:1870-9.
10. Tonino PAL, De Bruyne B, Pijls NHJ, stents in coronary artery disease. N Engl 29. Jüni P, Altman DG, Egger M. Assess-
et al. Fractional flow reserve versus angi- J Med 2010;362:1663-74.
ing the quality of controlled clinical tri- ography for guiding percutaneous coro- 20. Serruys PW, Silber S, Garg S, et al. als. BMJ 2001;323:42-6.
nary intervention. N Engl J Med 2009;360: Comparison of zotarolimus-eluting and 30. Simoons ML, Windecker S. Chronic
everolimus-eluting coronary stents. N Engl stable coronary artery disease: drugs vs. 11. Pijls NHJ, van Son JA, Kirkeeide RL, J Med 2010;363:136-46.
revascularization. Eur Heart J 2010;31: De Bruyne B, Gould KL. Experimental ba- 21. Agency for Healthcare Research and 530-41.
sis of determining maximum coronary, Quality. Calculating the U.S. population- Copyright 2012 Massachusetts Medical Society. n engl j med nejm.org The New England Journal of Medicine Downloaded from nejm.org on August 30, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved.

Source: http://www.sabincor.com.br/astherlab/uploads/arquivos/artigos/Fractional_Flow_ReserveGuided_PCI_versus.pdf