medical therapy with versus without revascularization in stable … · 2016-11-13 · high-risk cad...

Listen to this manuscript’s

audio summary by

JACC Editor-in-Chief

Dr. Valentin Fuster.

J O U R N A L O F T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y VO L . 6 7 , N O . 1 , 2 0 1 6

ª 2 0 1 6 B Y T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y F O UN DA T I O N I S S N 0 7 3 5 - 1 0 9 7 / $ 3 6 . 0 0

P U B L I S H E D B Y E L S E V I E R h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j a c c . 2 0 1 5 . 0 9 . 0 5 6

THE PRESENT AND FUTURE

STATE-OF-THE-ART REVIEW

Medical Therapy With Versus WithoutRevascularization in Stable PatientsWith Moderate and Severe Ischemia
The Case for Community Equipoise
Gregg W. Stone, MD,* Judith S. Hochman, MD,y David O. Williams, MD,z William E. Boden, MD,xT. Bruce Ferguson, JR, MD,k Robert A. Harrington, MD,{ David J. Maron, MD{

ABSTRACT

Fro

Re

ve

ch

Alb

No

tria

do

As

Ar

no

Sto

Ma

Am

Vid

Sm

an

rel

Ma

All patients with stable ischemic heart disease (SIHD) should be managed with guideline-directed medical therapy

(GDMT), which reduces progression of atherosclerosis and prevents coronary thrombosis. Revascularization is also

indicated in patients with SIHD and progressive or refractory symptoms, despite medical management. Whether a

strategy of routine revascularization (with percutaneous coronary intervention or coronary artery bypass graft surgery

as appropriate) plus GDMT reduces rates of death or myocardial infarction, or improves quality of life compared to

an initial approach of GDMT alone in patients with substantial ischemia is uncertain. Opinions run strongly on both

sides, and evidence may be used to support either approach. Careful review of the data demonstrates the limitations

of our current knowledge, resulting in a state of community equipoise. The ongoing ISCHEMIA trial (International

Study of Comparative Health Effectiveness With Medical and Invasive Approaches) is being performed to determine

the optimal approach to managing patients with SIHD, moderate-to-severe ischemia, and symptoms that can be

controlled medically. (International Study of Comparative Health Effectiveness With Medical and Invasive Approaches

[ISCHEMIA]; NCT01471522) (J Am Coll Cardiol 2016;67:81–99) © 2016 by the American College of Cardiology

Foundation.

m the *Department of Medicine, Columbia University Medical Center, New York Presbyterian Hospital and the Cardiovascular

search Foundation, New York, New York; yDepartment of Medicine, Cardiovascular Clinical Research Center, New York Uni-

rsity School of Medicine, New York, New York; zDepartment of Medicine, Brigham and Women’s Hospital, Boston, Massa-

usetts; xDepartment of Medicine, Samuel S. Stratton VA Medical Center, Albany Medical Center and Albany Medical College,

any, New York; kDepartment of Cardiovascular Sciences, East Carolina Heart Institute, East Carolina University, Greenville,

rth Carolina; and the {Department of Medicine, Stanford University School of Medicine, Stanford, California. The ISCHEMIA

l, which is discussed in this article, is supported by National Heart, Lung, and Blood Institute grant U01HL105907, in-kind

nations from Abbott Vascular; Medtronic, Inc., St. Jude Medical, Inc., Volcano Corporation, Arbor Pharmaceuticals, LLC,

traZeneca Pharmaceuticals, LP, Merck Sharp & Dohme Corp., and Omron Healthcare, Inc.; and by financial donations from

bor Pharmaceuticals LLC and AstraZeneca Pharmaceuticals LP. The content is solely the responsibility of the authors and does

t necessarily represent the official views of the National Heart, Lung, and Blood Institute or the National Institutes of Health. Dr.

ne is a consultant for Reva Corporation; and has received NIH grant support for the ISCHEMIA trial. Drs. Hochman, Boden, and

ron have received NIH grant support for the ISCHEMIA trial. Dr. Harrington has received consultant fees or honoraria from

gen Inc., Adverse Events, Daiichi-Lilly, GILEAD Sciences, Janssen, Medtronic, Merck & Co., Novartis, The Medicines Company,

a Health, Vox Media, and WebMD; has received research grants from AstraZeneca, Bristol-Myers Squibb, CSL Behring, Glaxo-

ithKline, Merck & Co., Portola, Regado, Sanofi, and The Medicines Company; has equity in Element Science and MyoKardia;

d is an officer, director, or trustee of Evidint and Scanadu. All other authors have reported that they have no relationships

evant to the contents of this paper to disclose.

nuscript received July 3, 2015; revised manuscript received September 10, 2015, accepted September 22, 2015.

https://www.clinicaltrials.gov/ct2/show/NCT01471522

https://s3.amazonaws.com/ADFJACC/JACC6701/JACC6701_fustersummary_07




http://crossmark.crossref.org/dialog/?doi=10.1016/j.jacc.2015.09.056&domain=pdf

http://dx.doi.org/10.1016/j.jacc.2015.09.056

ABBR EV I A T I ON S

AND ACRONYMS

ACS = acute coronary

syndrome

BMS = bare-metal stent(s)

CABG = coronary artery

bypass graft

CAD = coronary artery disease

CCTA = coronary computed

tomographic angiography

CI = confidence interval

DES = drug-eluting stent(s)

FFR = fractional flow reserve

GDMT = guideline-directed

medical therapy

HR = hazard ratio

LVEF = left ventricular ejection

fraction

MACE = major adverse cardiac

events

MI = myocardial infarction

MT = medical therapy

OMT = optimal medical therapy

OR = odds ratio

PCI = percutaneous coronary

intervention

QALY = quality-adjusted

life-year

QoL = quality of life

RR = risk ratio

SIHD = stable ischemic

heart disease

SPECT = single-photon

emission computed

tomography

Stone et al. J A C C V O L . 6 7 , N O . 1 , 2 0 1 6

Ischemia Revascularization Equipoise J A N U A R Y 5 / 1 2 , 2 0 1 6 : 8 1 – 9 9

82

P atients with obstructive atheroscle-rotic coronary artery disease (CAD)may be asymptomatic (with or with-

out ischemia), or present with symptomsranging from stable angina, to acute coronarysyndromes (ACS) (unstable angina, non–ST-segment elevation myocardial infarction,or ST-segment elevation myocardial infarc-tion), to sudden cardiac death. All patientswith established CAD should be prescribedguideline-directed medical therapy (GDMT)to mitigate progression of atherosclerosisand to prevent myocardial infarction (MI)and cardiovascular death (1,2). In patientswith biomarker-positive ACS, it is widelyaccepted that routine revascularization, inaddition to GDMT, reduces the short- andlong-term rates of death and MI comparedwith a more conservative approach (3–5). Bycontrast, the extent to which routine revas-cularization reduces death or MI, or improvesquality of life (QoL) in patients with stableischemic heart disease (SIHD) representsone of the greatest uncertainties in contem-porary cardiology. Given that an estimated15.5 million Americans have CAD, and thatrevascularization is performed in more than1.3 million patients per year in the UnitedStates alone (6), the appropriate (but judi-cious) application of revascularization hasenormous implications for the medical andeconomic health of the nation and the globalcommunity.

Early randomized trials of coronary arterybypass graft (CABG) surgery versus conser-

vative care in patents with SIHD performed severaldecades ago suggested a survival benefit for CABG inpatients with extensive anatomic disease, in whom alarge amount of myocardium was at risk (left maindisease, 3-vessel disease, and possibly 2-vessel dis-ease involving the proximal left anterior descendingcoronary artery) (7). Ischemia on an exercise stresstest also identified patients in whom mortality wasreduced with CABG compared with medical therapy(MT) (7). These earlier randomized trials of CABGversus MT, however, antedated the more contem-porary use of “disease-modifying” pharmacologicalinterventions, including statins, inhibitors of therenin-angiotensin-aldosterone axis, and antiplateletagents that individually have been shown to reducedeath and MI in placebo-controlled trials. The aggre-gate use of such secondary prevention therapies,along with lifestyle interventions, such as cigarette

smoking cessation, diet, and regular exercise, hasbeen referred to as optimal medical therapy (OMT),or GDMT (1,2).

More recently, the benefits of routine revasculari-zation in SIHD have been questioned by the similarrates of death and MI observed in OMT-treatedpatients with and without percutaneous coronaryintervention (PCI) in the COURAGE (Clinical Out-comes Utilizing Revascularization and AggressiveDrug Evaluation) trial, and with and without PCI orCABG in the BARI 2D (Bypass Angioplasty Revas-cularization Investigation 2 Diabetes) trial (8,9).It may be argued that revascularization in SIHDmay not be beneficial because not all anatomicallyobstructive coronary stenoses produce ischemia,or because not all high-grade coronary stenosesresult in cardiac death and/or MI, or conversely,because most cases of cardiac death and/or MI arisefrom angiographically mild coronary lesions, whichare not revascularized. However, some observa-tional studies and hypothesis-generating substudydata from randomized trials suggest that themagnitude of ischemia is associated with adverseoutcomes and that alleviation of ischemia mayimprove prognosis. Conversely, credible studiesdrawn from different (or even the same!) data-sets have cast doubt on this premise. And im-portantly, often lost in this discussion is theextent to which revascularization improves QoL, aworthwhile goal, assuming noninferior rates of“hard” adverse event endpoints and reasonablecost-effectiveness.

Recent clinical practice guidelines from theUnited States and Europe, as well as U.S. appro-priate use criteria, endorse GDMT for all patientswith SIHD, but recommend (with variable levels ofcertainty) consideration of revascularization in pa-tients with significant ischemia or symptoms thatpersist despite MT (10–14). Despite this uncertainty,highly enthusiastic proponents of both routine andselective revascularization for SIHD patients withischemia may be found, and nearly everyone has anopinion. Indeed, attitudes run so strongly on thistopic that it may be questioned whether clinicalequipoise exists, although, when pushed, nearlyall agree that definitive trials addressing the roleof revascularization in optimally treated SIHDpatients with substantial ischemia have not yetbeen performed.

The purpose of this review is to describe theevidence supporting the initial strategies of rou-tine revascularization plus GDMT versus GDMTalone, with revascularization reserved for MT failure

J A C C V O L . 6 7 , N O . 1 , 2 0 1 6 Stone et al.J A N U A R Y 5 / 1 2 , 2 0 1 6 : 8 1 – 9 9 Ischemia Revascularization Equipoise

83

(e.g., progressive or refractory symptoms or thedevelopment of ACS), in patients with SIHD andmoderate or severe ischemia. Each perspective issummarized in Table 1. The evidence is presented asan internal debate, and the reader may naturallygravitate to one position or the other. We recommendthat those with a strong preconceived preference payparticular attention to the data supporting the oppo-site view. The authors believe that on the basis of thepresent level of evidence, a justifiable case can bemade for either an initial strategy of upfront revas-cularization (with PCI or CABG, as determined by thelocal heart team) plus GDMT, or an initial strategy ofGDMT alone. The equipoise expressed in this docu-ment served as the impetus for the collaborativeISCHEMIA trial (International Study of ComparativeHealth Effectiveness with Medical and Invasive Ap-proaches), as described later.

THE CASE FOR ROUTINE REVASCULARIZATION

IN PATIENTS WITH SIHD AND ISCHEMIA

OVERVIEW. Revascularization may benefit SIHD pa-tients by preventing death, MI, and unstable angina

TABLE 1 Supportive Evidence for an Initial Routine Revascularization

Favors an Initial Strategy of

1. Most data from large observational studies and from within randomizeddeath and MI.

2. Revascularization relieves ischemia to a greater extent than the best me

3. Several large-scale observational studies and some randomized data havrevascularization compared with medical therapy in SIHD.

4. Both PCI and CABG revascularization strategies are continually being imdocumented to be more effective in relieving ischemia and reducing dea

5. Nearly all studies suggest that routine revascularization more effectivelyQoL more than GDMT for at least several years. The rates converge overrequired revascularization for progressive symptoms.

6. Adherence to GDMT (especially multiple daily medications) is difficult anreduction in symptoms and avoidance of antianginal medications achieva

7. An initial strategy of routine revascularization does not preclude patientthrombosis. However, by reducing the requirement for antianginal medicto reduce MI and death.

Favors an Initial Con

1. GDMT is necessary in all SIHD patients, whether or not revascularizationan alternative to antianginal medication, not an alternative to GDMT. Paoutcomes.

2. Recent studies in the current GDMT era have not demonstrated a link be

3. PCI, the most common form of revascularization, has never been shown

4. Medical therapy is continually being improved. Prolonged dual antiplatedeath, MI, and stroke following remote MI, and stent thrombosis and Mor without concomitant statin treatment, and may reduce death and MI.

5. The QoL benefit from revascularization appears to be time limited, moreabsence of sham procedure control groups.

6. Because revascularization has not been proven to reduce death or MI into see if symptom relief is adequate. An initial trial of GDMT without rev

7. Routine revascularization in SIHD is more costly to the patient and socie

CABG ¼ coronary artery bypass graft; CAD ¼ coronary artery disease; GDMT ¼ guideline-PCI ¼ percutaneous coronary intervention; PCSK9 ¼ proprotein convertase subtilisin/ke

(which, even if successfully managed with urgenthospitalization and treatment, may be more dis-tressing for patients than a controlled elective pro-cedure), and by improving QoL. More than 25 yearsago, Ellis et al. (15) reported that the incidence ofanterior MI increases with the severity of untreatedleft anterior descending coronary artery lesions.The extent of incomplete anatomic revasculariza-tion after PCI and CABG is strongly associatedwith subsequent death, MI, and recurrent anginarequiring rehospitalization (16,17). Furthermore,large-scale, nonrandomized studies have reportedimproved prognosis with revascularization in SIHD.In a propensity-matched observational analysis in39,131 patients with SIHD, early revascularizationwas associated with fewer deaths and MI during4-year follow-up, compared with initial MT alone(Figure 1) (18). Similarly, in a study of 15,223 patientswith newly diagnosed SIHD by coronary computedtomographic angiography (CCTA), patients withhigh-risk CAD in whom revascularization was per-formed within 90 days had significantly reducedmortality at a median 2.1-year follow-up com-pared with those treated medically (2.3% vs. 5.3%,

Strategy Plus GDMT Versus GDMT Alone in Patients With SIHD

Routine Revascularization

trials suggest a relationship between ischemia and subsequent

dical therapy.

e demonstrated a reduction in spontaneous MI and death with

proved, with contemporary stent technologies and surgical techniquesth and MI than prior approaches.

reduces angina and the need for antianginal medications, and improvestime only in part because many GDMT patients have “crossed over” and

d not routinely achieved. Many patients prefer the more immediateble with revascularization compared with GDMT.

s benefitting from GDMT that inhibit coronary atherosclerosis andations, revascularization may facilitate compliance with GDMT proven

servative Strategy

is performed, because it prevents MI and death. Revascularization istients who undergo PCI and CABG must still take GDMT to optimize

tween ischemia and death or MI.

in randomized trials to improve survival in SIHD.

let therapy results in reductions in the composite rate of cardiovascularI after stents. PCSK9 inhibitors reduce LDL by $50% in patients with

so for PCI than for CABG. There are challenges to assessing QoL in the

SIHD, a trial of GDMT should be tried before elective revascularizationascularization does not increase the risk of death or MI.

ty.

directed medical therapy; LDL ¼ low-density lipoprotein; MI ¼ myocardial infarction;xin type 9; QoL ¼ quality of life; SIHD ¼ stable ischemic heart disease.

FIGURE 1 Propensity-Matched Analysis in 39,131 Canadian Patients With SIHD

Undergoing Early Revascularization (n ¼ 23,992) or Treated Conservatively (n ¼ 15,139)

1.000.980.960.940.920.900.880.860.84 0.820.800.780.760.74

Surv

ival

Time (Years)0 1 2 3 4

1.000.980.960.940.920.900.880.860.84 0.820.800.780.760.74

Free

dom

Fro

m M

yoca

rdia

l Inf

arct

ion

Time (Years)0 1 2 3 4

Number at risk:Revascularization

Medical therapy1236212362

1208211961

84337987

43014092

340316

Number at risk:Revascularization

Medical therapy1236212362

1156211412

83687979

43774281

692804

Medical therapyRevascularization

Medical therapyRevascularization

4–year event rates8.6% vs 12.7%

HR [95%CI] = 0.75 [0.69—0.81]P<0.001

4–year event rates11.7% vs 14.4%

HR [95%CI] = 0.84 [0.77—0.93]P<0.001

A

B

Over 4 years of follow-up, significant reductions in death (A) and myocardial infarction

(B)wereobservedwith revascularization.Adaptedwithpermission fromWijeysunderaet al. (18).

CI¼ confidence interval; HR¼ hazard ratio; SIHD¼ stable ischemic heart disease.



84

p ¼ 0.008; adjusted hazard ratio [HR]: 0.38; 95%confidence interval [CI]: 0.18 to 0.83) (19). No sig-nificant difference in mortality with revasculari-zation was observed in lower-risk patients withouthigh-risk CAD. Finally, in a propensity-matchedstudy of 1,866 patients in New York State withSIHD undergoing cardiac catheterization, PCI com-pared with MT was associated with reduced 4-yearrates of death or MI (16.5% vs. 21.2%; adjusted HR:1.49; 95% CI: 1.16 to 1.93) (20). Although not ran-domized, these real-world studies included higher-risk patients than the COURAGE and BARI 2D trials

(8,9), and with many more events had substantiallygreater power to detect reductions in death and MIbetween treatments.

These studies reported outcomes with anatomi-cally driven revascularization. Targeting revasculari-zation to lesions causing substantial ischemia mayfurther improve results. The angiogram is a poordiscriminator of physiological lesion significance.Many lesions that appear angiographically severemay not produce ischemia, and conversely, ischemiamay be present despite a benign angiographicfootprint (21,22).

OBSERVATIONAL DATA DEMONSTRATE A STRONG

RELATIONSHIP BETWEEN THE EXTENT OF ISCHEMIA

AND SUBSEQUENT DEATH AND/OR MI, AND A POSSIBLE

BENEFIT FROM REVASCULARIZATION. Most single-photon emission computed tomography (SPECT)myocardial perfusion imaging studies have demon-strated a strong relationship between the extent ofischemia and prognosis. In 1,137 patients with chestpain or suspected CAD in whom thallium-201 SPECTwas performed, a strong graded association was pre-sent between the number of abnormal segmentsand the 6-year rate of cardiac death and MI (23). In astudy of 205 patients with angiographically provenCAD, freedom from cardiac death or MI at 10 yearswas greater in patients with a normal comparedwith an abnormal thallium-201 SPECT (83% vs. 58%;p ¼ 0.005) (24). In a study of 1,126 asymptomaticpatients, the presence of $10% ischemia by SPECTwas independently associated with death or MI atmedian follow-up of 6.9 years (HR: 2.67; 95% CI:1.31 to 5.44; p ¼ 0.007) (25). Among 10,627 patientsin whom a quantitative stress SPECT study wasperformed (671 of whom were treated with earlyrevascularization), the mean 1.9-year rate of cardiacmortality in nonrevascularized patients increasedmonotonically, from 0.7% in those with no ischemiato 6.7% in those with >20% ischemia (Figure 2) (26).After accounting for baseline variables and the pro-pensity for revascularization, a strong relationshipwas present between the percentage of myocardialischemia and cardiac mortality.

In the COURAGE serial nuclear substudy, 314patients underwent rest/stress SPECT before treat-ment and at 6 to 18 months (mean 374 � 50 days),with the amount of ischemia assessed at a blindedcore laboratory (27). A strong graded relationship waspresent between the amount of residual ischemiaon the 6- to 18-month test and subsequent death orMI (p ¼ 0.001), which was attenuated after adjust-ment for baseline variables and treatment (p ¼ 0.09).Similarly, long-term freedom from death or MI was

FIGURE 2 Cardiac Mortality as a Function of Total Ischemic Myocardium at a Mean Follow-Up Time of 1.9 � 0.6 Years

10%

8%

6%

4%

2%

0%

Card

iac

Deat

h

7110 16 1331 56 718 109 545 243 252 2670% 1–5% 5–10% 11–20% >20%

% Total Ischemic Myocardium

0.7%

6.3%

1.0%1.8%

2.9%3.7%

4.8%

3.3%

6.7%

2.0%

P<0.0001*

Medical therapy Revascularization

6

5

4

3

2

1

0

Log

Haza

rd R

atio

0 12.5% 25% 32.5% 50%% Total Ischemic Myocardium

Model P<0.0001Pinteraction = 0.03 Medical therapy*

Revascularization*

A

B

Cardiac mortality as a function of total ischemic myocardium is shown at a mean follow-up time of 1.9 � 0.6 years in 10,627 patients at

Cedars-Sinai Medical Center without previous myocardial infarction or revascularization who underwent exercise or adenosine thallium 201

single-photon emission computed tomography myocardial perfusion scintigraphy. Outcomes are shown according to whether elective

revascularization was versus was not performed within 60 days. Patients with a greater percentage of ischemic myocardium had a greater

survival benefit with revascularization. (A) Unadjusted analysis. (B) Log hazard ratio of cardiac mortality for revascularization versus medical

therapy as a function of % ischemic myocardium from a Cox proportional hazards regression model. *p < 0.001. Adapted with permission from

Hachamovitch et al. (26).


85

greater in patients achieving versus not achieving$5% reduction in ischemia (whether attained byPCI or OMT alone), especially in those with at leastmoderate ($10%) ischemia at baseline. Of note, PCIresulted in a substantial reduction in quantitativeischemia from the baseline to the follow-up SPECTstudy (8.2% vs. 5.5%; p < 0.0001). By contrast,despite excellent compliance with GDMT in theCOURAGE trial, no reduction in ischemia from base-line to follow-up occurred with MT (8.6% vs. 8.1%;p ¼ 0.93) (Figure 3).

The adverse prognostic implications of ischemiahave been observed with other noninvasive imagingtechniques and with invasive physiological lesionassessment. Among 7,061 patients at 4 centers inwhom a clinically indicated rest/stress rubidium-82positron emission computed tomography was per-formed, the adjusted risk of cardiac death at a medianfollow-up of 2.2 years increased 84% for each10% of ischemic myocardium (p < 0.0001) (28). Inanother study, stress echocardiography was perfor-med in 14,140 patients at 2 Italian institutions (29).

FIGURE 3 Reduction in Inducible Ischemia From Baseline to 6 to 18 Months

40

35

30

25

20

15

10

5

0

40

35

30

25

20

15

10

5

0Index 6–18 months Index 6–18 months

Mean change (95%CI) =-2.7% (-3.8%, -1.7%)

P<0.0001P<0.0001

Mean change (95%CI) =-0.5% (-1.6%, 0.6%)

P=0.93

8.2%(7.2%–9.3%)

p=0.63 5.5%(4.7%–6.3%)

8.6%(7.5%–9.8%) 8.1%

(6.9%–9.4%)P<0.0001

Percutaneous Coronary Intervention + Optimal Medical Therapy(n=159)

Optimal Medical Therapy(n=155)

33.3% with ≥5% ischemia reduction(p=0.0004)

18.9% with ≥5% ischemia reduction

Reduction in inducible ischemia from baseline to 6 to 18 months is shown in patients treated with OMT with versus without a strategy of routine

upfront PCI, as assessed by myocardial perfusion scintigraphy. Left graph: with routine upfront PCI; right graph: without. The reduction

in ischemia in the PCI arm was significant, whereas there was no significant reduction in ischemia with OMT. Adapted with permission from Shaw

et al. (27). CI ¼ confidence interval; OMT ¼ optimal medical therapy; PCI ¼ percutaneous coronary intervention.



86

At median 2.5-year follow-up, the presence ofischemia was a strong independent predictor ofmortality in both diabetic patients (HR: 1.71; 95% CI:1.34 to 2.18; p < 0.0001) and nondiabetic patients(HR: 1.54; 95% CI: 1.32 to 1.80; p < 0.0001).Finally, identification of hemodynamically significantflow-limiting lesions during hyperemia with adeno-sine in the cardiac catheterization laboratory (frac-tional flow reserve [FFR]) has been correlated withdeath and MI in medically treated patients. In acollaborative meta-analysis, the rates of death or MIand major adverse cardiac events (MACE) wereinversely related to FFR at a median 16-month follow-up (Figure 4) (30).

Observational studies further support the potentialrole of routine revascularization to improve prog-nosis in patients with moderate or severe ischemia.In a propensity-controlled multivariable analysisfrom Cedars-Sinai Medical Center, 2-year rates ofcardiac death were lower with revascularizationwithin 60 days in patients with moderate or severeischemia (cutoff w12.5% of the left ventricle) bySPECT (Figure 2) (26). In a subsequent investigationfrom the same center in 13,969 patients with a mean8.7-year follow-up, early revascularization in patients

with $10% inducible ischemia by SPECT was associ-ated with improved late survival, unless a large fixeddefect was present (31). Early revascularization inpatients with $10% myocardial ischemia was associ-ated with improved survival, even in those $75 yearsof age (32). In a Mayo Clinic study, a survival benefitamong diabetic patients with CABG compared withMT or PCI was noted only in those with a high-riskSPECT scan (33). Finally, in an adjusted study-levelmeta-regression of 90 cohorts in which FFR wasassessed, revascularization was associated with alower 1-year rate of death or MI when FFR was <0.90(30).

RANDOMIZEDTRIAL DATA SUGGEST REVASCULARIZATION

IS SAFE AND REDUCES DEATH AND/OR MI AND/OR

IMPROVES QoL, ESPECIALLY WHEN SUBSTANTIAL

ISCHEMIA IS PRESENT. In the MASS II trial (Medicine,Angioplasty or Surgery Study), 611 patients withproximal multivessel disease and documentedischemia were randomly assigned to CABG, PCI,or OMT (34). The 10-year mortality rates in the 3groups were 25.1%, 24.9%, and 31.0%, respectively(p ¼ 0.09). The 10-year MI rates were 10.3%, 13.3%,and 20.7%, respectively (p < 0.01). Freedom from

FIGURE 4 Relationship Between FFR and 1-Year MACE According to Whether or Not Revascularization Was Performed

40%

30%

20%

10%

0%

Norm

aliz

ed 1–

Year

MAC

E (%

)

40%

30%

20%

10%

0%Co

x M

odel

1–Ye

ar M

ACE

(%)

0.4 0.5 0.6 0.7 0.8 0.9 1.0Mean Cohort FFR

0.4 0.5 0.6 0.7 0.8 0.9 1.0Fractional Flow Reserve (FFR)

PCI/CABG medically treated

medically treated

PCI/CABG

FFR 0.75–0.80

0.67

0.75

8,418 patients from90 cohorts with a total of 458 deaths 235 non–fatal MI 326 revascularizations

5,979 patients with a total of 258 deaths 163 non–fatal MI 552 revascularizations

A B

(A) Study-level meta-regression in 8,418 patients from 90 cohorts. Revascularization was associated with a lower rate of MACE when the FFR was <0.75 in

an unadjusted random effects model, and <0.90 in an adjusted model. (B) Patient-level meta-analysis in 5,979 patients. Revascularization was associated

with a lower rate of MACE when the FFR was <0.67 in an unadjusted Cox model, and <0.76 in an adjusted model. Reprinted with permission from

Johnson et al. (30). CABG ¼ coronary artery bypass graft; FFR ¼ fractional flow reserve; MACE ¼ major adverse cardiac events; MI ¼ myocardial infarction;

PCI ¼ percutaneous coronary intervention.


87

angina at 10 years was 64% with CABG, 59% with PCI,and 43% with OMT (p < 0.001).

In the COURAGE trial, randomization of 2,287patients to PCI plus OMT versus OMT alone did notreduce the long-term rate of death or MI (8). Nor,however, did PCI worsen prognosis, and crossover toPCI for progressive symptoms or ACS was required in32% of OMT patients during a median 4.6-year follow-up. Moreover, patients randomized to PCI had lessdocumented angina, were more likely to be angina-free (despite requiring fewer nitrates and calcium-channel blockers), and had improved QoL for up to3 years (8,35). The reduction in angina (assessed bythe Seattle Angina Questionnaire) with PCI versusOMT was most evident in those with the greatestlevel of baseline angina (35).

In the BARI 2D trial, 2,368 patients with type 2diabetes (90% with SIHD) were randomized to promptrevascularization with intensive MT or to intensiveMT alone, with randomization stratified by intendedPCI versus CABG (9). The 5-year rates of death(the primary endpoint) and MACE (death, MI, orstroke) were not significantly different with eitherstrategy. However, patients in the CABG stratum hadmore advanced CAD (including more triple-vessel andproximal left anterior descending coronary arterydisease) than those in the PCI stratum, and patientsrandomized to CABG versus intensive MT had lower

5-year rates of MACE (22.4% vs. 30.5%; p ¼ 0.01),driven by less MI (7.4% vs. 14.6%). In patients withless extensive CAD, there was no difference in MACEwith PCI versus intensive MT. Compared with inten-sive MT, prompt revascularization resulted in signif-icantly greater freedom from angina for up to 4 years(36). Most measures of QoL through the 4-year follow-up were also improved with routine revascularizationcompared with intensive MT only (37), and revascu-larization was ultimately required in 42% of MTpatients during follow-up (9).

Meta-analyses have also shown reduced mortalityand greater angina relief with early routine revascu-larization versus MT. In a systematic review andmeta-analysis of 12 randomized trials comparing PCIversus MT in 7,182 SIHD patients, Pursnani et al. (38)found a strong trend for lower mortality with PCI(risk ratio [RR]: 0.85; 95% CI: 0.71 to 1.01; p ¼ 0.07).Others have reported similar trends (39,40). From 17trials of PCI versus OMT in 7,513 stable patients withischemia (some with recent MI), Schömig et al. (41)reported that PCI was associated with a significantreduction in all-cause mortality (odds ratio [OR]:0.80; 95% CI: 0.64 to 0.99). Jeremias et al. (42) per-formed the largest meta-analysis to date from 28trials of revascularization versus MT in 13,121 patientswith nonacute CAD. Revascularization was associatedwith a reduction in mortality (OR: 0.74; 95% CI: 0.63



88

to 0.88), a difference that was significant for bothCABG (OR: 0.62; 95% CI: 0.50 to 0.77) and PCI(OR: 0.82; 95% CI: 0.68 to 0.99). PCI was also asso-ciated with greater freedom from angina comparedwith MT in a meta-analysis of 12 randomized trialsin SIHD (RR: 1.20; 95% CI: 1.06 to 1.37), a benefitthat was present at follow-up durations #1 year, 1 to5 years, and $5 years (38).

Of note, bare-metal stents (BMS) were used in mostPCI versus MT trials to date (including the MASS II,COURAGE, and BARI 2D studies). First-generationdrug-eluting stents (DES) markedly reduce recur-rent ischemia compared with BMS (43), resulting infewer hospitalizations for repeat revascularization(44). Compared with BMS and first-generation DES,second-generation DES may further reduce deathand MI, and enhance event-free survival (39,45,46).In a comprehensive network meta-analysis of revas-cularization versus MT in SIHD in which stent typewas considered (100 trials, 93,553 patients, 262,090patient-years of follow-up), PCI with everolimus-eluting stents compared with MT was associatedwith significant 25%, 22%, and 73% reductions indeath, death or MI, and repeat revascularization,respectively (Figure 5) (47). Similar benefits werenoted with CABG compared with MT, and in analysesthat excluded trials of patients with recent MI orstabilized ACS.

Outcomes may be further improved by a routinerevascularization strategy in patients with moreextensive ischemia. Only w33% of patients inCOURAGE had at least moderate ($10%) ischemia atbaseline, either measured quantitatively (27) or siteassessed (48); w40% of patients had <5% ischemia(27). Ischemia, as evidenced by FFR, may predictwhich patients and lesions are likely to benefit fromroutine revascularization versus GDMT alone. In theDEFER trial, 91 patients with FFR >0.75 were ran-domized to balloon angioplasty versus MT (49). The2-year rates of death, MI, or repeat revascularizationwere similar in the deferral and PCI groups (89% vs.83%, respectively). In the FAME (Fractional FlowReserve versus Angiography for Multivessel Evalua-tion) trial, FFR was performed in 1,005 patients withmultivessel disease undergoing PCI with first-generation DES. Patients were randomized to treat-ment of all angiographically significant lesions,versus only lesions with FFR #0.80. Compared withanatomically based revascularization, FFR-guided PCIresulted in lower rates of the primary endpoint ofdeath, MI, or repeat revascularization at 1 year (13.2%vs. 18.3%; p¼ 0.02), and of death or MI at 2 years (8.4%vs. 12.9%; p ¼ 0.02) (50,51). These data, consistentwith earlier observational studies, suggest that the

prognosis of non–ischemia-producing lesions withoutrevascularization is favorable.

By contrast, the FAME-2 trial demonstrated worseoutcomes with an initial MT approach in flow-limitinglesions. In the FAME-2 trial, 888 of a planned 1,632patients with SIHD and 1 or more lesions with anFFR #0.80 (i.e., ischemia) were randomized to PCIwith second-generation DES plus GDMT versus GDMTalone (52). The trial was stopped early because ofexcess events in the GDMT arm. The 2-year primaryendpoint of death, MI, or urgent revascularizationoccurred in 8.1% of PCI patients versus 19.5% ofGDMT patients (HR: 0.39; 95% CI: 0.26 to 0.57;p < 0.001), driven by fewer urgent revascularizations(4.0% vs. 16.3%; p < 0.001), including those triggeredby MI or severe unstable angina (3.4% vs. 7.0%;p ¼ 0.01). Class II to IV angina was also significantlymore frequent in the GDMT arm during follow-up,and PCI was required in 40.6% of GDMT-assignedpatients within 2 years for refractory symptoms orACS. Although there was no significant difference inthe composite rate of death or MI, landmark analysisshowed that periprocedural MI rates were increasedwith PCI, whereas MI occurred less frequently in thePCI group compared with the GDMT group between8 days and 2 years. Similarly, in a meta-analysisof 12 randomized trials of PCI versus OMT with37,548 patient-years of follow-up, PCI comparedwith MT was associated with a significant 23%reduction in spontaneous MI, which paralleled atrend toward reduced mortality (53). SpontaneousMI has been strongly associated with subsequentmortality, whereas most periprocedural MIs are notclinically relevant (54).

Finally, because most periprocedural MIs do nothave major clinical consequence (54), their inclusionin a primary composite death or MI endpoint of priortrials may have statistically masked an isolated mor-tality benefit of revascularization in SIHD patientswith documented ischemia. Three randomized trialsof PCI versus MT in SIHD have been performed inwhich 1,557 patients had evidence of ischemia onnoninvasive stress imaging or FFR before randomi-zation (the COURAGE baseline nuclear substudy, theFAME-2 study, and the SWISSI [Silent Ischemia AfterMyocardial Infarction] II trial [which enrolled stabi-lized patients after acute MI]) (48,52,55). In a meta-analysis of these 3 trials, routine PCI compared withMT was associated with lower 3-year mortality(HR: 0.52; 95% CI: 0.30 to 0.92; p ¼ 0.02), with noheterogeneity between studies (56).

MEDICATION DISUTILITY AND PATIENT PREFERENCE.

Adherence to biology-altering medications, such

FIGURE 5 Network Meta-Analysis Comparing Different Revascularization Modalities With MT

Rate ratio (95% CrI)All cause mortality

Rate ratio (95% CrI)

(95 trials; 93,553 patients)

CABG v medical treatment

PTCA v medical treatment

BMS v medical treatment

PES v medical treatment

SES v medical treatment

E–ZES v medical treatment

R-ZES v medical treatment

EES v medical treatment

















0.80 (0.70 to 0.91)

0.85 (0.68 to 1.04)

0.92 (0.79 to 1.05)

0.92 (0.75 to 1.12)

0.91 (0.75 to 1.10)

0.88 (0.69 to 1.10)

0.65 (0.42 to 1.00)

0.75 (0.59 to 0.96)

0.79 (0.63 to 0.99)

0.88 (0.70 to 1.11)

1.04 (0.84 to 1.27)

1.18 (0.88 to 1.54)

0.94 (0.71 to 1.22)

0.80 (0.56 to 1.10)

0.82 (0.52 to 1.26)

0.75 (0.55 to 1.01)

0.81 (0.70 to 0.94)

0.83 (0.70 to 0.97)

0.99 (0.85 to 1.12)

1.06 (0.87 to 1.27)

0.96 (0.79 to 1.13)

0.85 (0.67 to 1.05)

0.81 (0.59 to 1.10)

0.78 (0.63 to 0.96)

Myocardial infarction


Death or myocardial infarction


0.1 0.3 1 3FavorsRevascularization

FavorsMedical Treatment

Estimated risk rate ratios (95% credible intervals) are shown for death, MI, and the composite of death or MI. Compared with MT, a reduction in

death was observed with EES, R-ZES, and CABG; a reduction in MI was observed with CABG; and a reduction in the composite of death or MI was

observed with EES and CABG. Adapted with permission from Windecker et al. (47). BMS ¼ bare-metal stent(s); CrI ¼ credible interval;

EES ¼ everolimus-eluting stent(s); E-ZES ¼ fast-release zotarolimus-eluting stent(s); MT ¼ medical therapy; PES ¼ paclitaxel-eluting stent(s);

PTCA ¼ percutaneous transluminal coronary angioplasty; R-ZES ¼ slow-release zotarolimus eluting stent(s); SES ¼ sirolimus-eluting stent(s);

other abbreviations as in Figure 4.


89

as aspirin and statins, reduces death and MI inSIHD patients treated with or without revas-cularization (1,2,57). In addition to potentially im-proving prognosis, routine revascularization reducesthe requirement for antianginal medications, as

observed in the COURAGE trial (8). Recent studieshave emphasized the cost and inconvenience(disutility) to the patient of taking medications.Participants of 2 large surveys were willing to sacri-fice 4 and 6 months of life, respectively, to avoid



90

taking a daily pill (58,59). A meta-analysis from 20studies and 376,162 patients examining usage of 7drug classes that prevent cardiovascular disease re-ported average 2-year adherence rates of 50% forprimary prevention and 66% for secondary preven-tion, with few differences between drug classes (60).Recent reports have emphasized the importance ofpatient-centered care and shared decision making,taking into account patient goals and preferenceswhen choosing between therapies, especially whenthe prognosis of alternative approaches are roughlysimilar (61). Beyond considerations of whetherrevascularization reduces death and MI in SIHD,many patients favor the more immediate reductionin symptoms achievable with PCI and CABG, andavoidance of antianginal medications.

Finally, as the disutility of taking medications in-creases with the number of daily pills, reducing therequirement for antianginal medications may facili-tate compliance with GDMT that reduces MI anddeath, such as statins. Revascularization may serve asa “wake-up call” to patients, and stimulate compli-ance with secondary prevention. In fact, GDMTadherence may be increased after PCI, compared withmore conservative care (62).

THE CASE FOR INITIAL GDMT IN

PATIENTS WITH SIHD AND ISCHEMIA

OVERVIEW. GDMT should be initiated in allSIHD patients, whether or not revascularization is

FIGURE 6 Effect of OMT on Clinical Outcomes in Patients Undergoin

2.0

1.5

1

0.5

0.01–month 6–month

Favo

rs N

on–O

MT

Favo

rs O

MTHa

zard

Rat

io

OMT significantly lowered the risk of death throughout the 5-year follow

was independently associated with improved survival throughout follow-

patients (HR: 0.65; 95% CI: 0.49 to 0.86; pinteraction ¼ 0.44). Adapted

Percutaneous Coronary Intervention With TAXUS and Cardiac Surgery; o

performed. GDMT consists of lifestyle and pharma-cological interventions that lower the risk of deathand MI; revascularization is not a substitute for this.Although revascularization might be considered areasonable alternative to antianginal medicationsto treat angina, disease-modifying medications arestill necessary to improve survival (Figure 6) (57,63).Comprehensive risk factor control, although notcommonly achieved, is associated with a 50% reduc-tion in mortality over 5 years in SIHD patientswith diabetes, with or without revascularization(64). Adherence to GDMT following revascularizationin the SYNTAX (Synergy Between Percutaneous Cor-onary Intervention With TAXUS and Cardiac Surgery)trial was suboptimal in the PCI group, and even worsein the CABG group (57). Clearly, strategies arerequired to improve GDMT adherence, whether ornot early revascularization is performed.

Recent studies suggest that, in the contemporaryGDMT era, the extent of ischemia is not related tothe risk of death or MI in patients with SIHD, andthat revascularization does not improve prognosis,regardless of the extent of ischemia. If true, then theonly reason to routinely revascularize SIHD patientswould be to improve QoL. If an initial trial of GDMTdoes not increase the risk of death or MI, it isdifficult to justify revascularization in patients withabsent or mild symptoms not treated with anti-anginal therapy, because all procedures haveinherent risks. Furthermore, as MT improves andevent rates decrease, possible early harm from

g PES Implantation and CABG in the SYNTAX Trial

1–year 3–year 5–year

-up. In a Cox regression model, OMT as a time-dependent covariate

up both in PCI patients (HR: 0.70; 95% CI: 0.48 to 0.998) and CABG

with permission from Iqbal et al. (57). SYNTAX ¼ Synergy Between

ther abbreviations as in Figures 1, 3, 4, and 5.

FIGURE 7 Baseline Severity of Ischemia and Outcomes by Treatment Group

in the COURAGE Trial

0.5

0.4

0.3

0.2

0.1

0.0De

ath

or M

I0 1 2 3 4 5 6 7

Time in Study (Years)

Log Rank P=0.87Wilcoxon P=0.61

OMT

PCI + OMT

Log Rank P=0.47Wilcoxon P=0.20

OMTPCI+OMT

0 1 2 3 4 5 6 7Time in Study (Years)

0.5

0.4

0.3

0.2

0.1

0.0

Deat

h or

MI

PCI + OMTOMT

459454

406403

381383

334335

259258

175172

8980

PCI+OMTOMT

223245

192223

183207

166187

127143

8296

4140

A

B

Cumulative rates of all-cause death or MI are shown according to randomized treatment

assignment in patients with no-to-mild ischemia and moderate-to-severe ischemia on

myocardial perfusion imaging at baseline in the COURAGE (Clinical Outcomes Utilizing

Revascularization and Aggressive Drug Evaluation) trial. (A) No-to-mild ischemia.

(B) Moderate-to-severe ischemia. Ischemia severity was determined by the participating

sites, and was not associated with risk of death or MI. PCI did not reduce death or MI in

patients with moderate or severe ischemia. Adapted with permission from Shaw et al. (48).

Abbreviations as in Figures 3 and 4.


91

routine revascularization—death, MI, and stroke—become relatively more important in weighing thepotential benefits and risks of performing PCI orCABG.

OBSERVATIONAL DATA SUGGEST THE EXTENT OF

ISCHEMIA IS NOT RELATED TO DEATH AND MI IN

THE CONTEMPORARY GDMT ERA. In the COURAGEserial nuclear substudy, a strong relationship wasfound between the extent of core laboratory–assessedresidual ischemia at 6 to 18 months and subsequentdeath or MI in 314 patients, but this was not signifi-cant after multivariable adjustment (27). This sub-study included w14% of enrolled patients and maynot be representative of the entire COURAGE popu-lation. Although there was a signal that ischemiareduction was associated with better outcomes, andthat PCI was associated with a greater degree ofischemia reduction than OMT, the critical link ofwhether PCI was causally associated with fewer car-diac events was not analyzed. Thus, this substudywas underpowered and hypothesis-generating. In aseparate, larger COURAGE report, in which site in-vestigators evaluated the extent of ischemia in 1,381patients with baseline SPECT imaging, no significantrelationship was present between the rates of deathand MI in those with $3 versus <3 ischemic segments(Figure 7) (48). In a follow-up COURAGE core labora-tory study in 621 patients, the extent of baselineischemia did not correlate with the rate of death, MI,or ACS after a mean follow-up of 4.7 years, whereasthe angiographic extent of atherosclerosis correlatedwith events (Figure 8) (65). These COURAGE trial an-alyses assessed death and MI rates differently. In theserial nuclear substudy, outcomes after repeat SPECTdid not include events between randomization and6 to 18 months, or events in patients who did notreturn for the follow-up stress test, and events werenot reported according to treatment group. Bycontrast, in the substudies that evaluated prognosisusing baseline ischemia, all events during follow-upwere ascertained, and event rates were reported ac-cording to treatment group.

Other recent studies have also failed to demon-strate a link between ischemia and prognosis. In apost-hoc analysis from the BARI 2D trial, the prog-nostic impact of SPECT performed 1 year afterrandomization was examined in 1,505 patients, asassessed by a core laboratory (66). Increasing severityof ischemia did not predict greater risk of death orcardiovascular events. In an observational study,Cleveland Clinic investigators compared survival ofasymptomatic patients with previous revasculariza-tion and documented ischemia on SPECT 5 years after

the initial revascularization with a mean 5.7-yearfollow-up. The severity of ischemia was not associ-ated with all-cause death, regardless of whetherearly repeat revascularization was performed or MTcontinued (67). In the STICH (Surgical Treatment for

FIGURE 8 Ischemia Severity, Anatomic Burden, and Outcomes in COURAGE

1.0

0.8

0.6

0.4

0.2

0.0


Free

dom

Fro

m D

eath

, MI,

or N

STE–

ACS

1.0

0.8

0.6

0.4

0.2

0.0


Free

dom

Fro

m D

eath

, MI,

or N

STE–

ACS

p=0.75

p=0.001

Amount of ischemia

0% – <5% myocardium, n=2415% – <10% myocardium, n=192≥10% myocardium, n=188

Burden of angiographic atherosclerosis0–5, n=2306–13, n=24214–17, n=149

Number at risk:0% – <5% myocardium

5% – <10% myocardium≥10% myocardium

241192188

213163159

196145151

173128146

137108144

906783

625051

Number at risk:0–5

6–1314–17

230242149

208205120

195191106

18016496

14913668

998945

574730

A

B

Freedom fromdeath,MI, or NSTE-ACS is shown by percent ischemicmyocardium and burden

of angiographic atherosclerosis. (A) Percent ischemic myocardium. (B) Burden of angio-

graphic atherosclerosis. Thenumber of patients pertaining to each colored curve is shownper

year of follow-up. The Cedars-Sinai Nuclear Core Laboratory determined the percent

ischemicmyocardium at baseline. Atherosclerotic burdenwas determined by a custom score,

accounting for the number and specific diseased coronary arteries, and proximal versus

nonproximal location. Ischemic burden was not associated with events, in contrast to

anatomic burden, which was. PCI did not reduce the event rate in patients with high ischemic

or anatomic burden. Adapted with permission from Mancini et al. (65). NSTE-ACS ¼ non–

ST-segment elevation acute coronary syndrome; other abbreviations as inFigures 3,4, and 7.



92

IsChemic Heart Failure) trial of CABG versus MT inSIHD patients with left ventricular ejection fraction(LVEF) #35%, at a median follow-up of 4.7 years,the presence of inducible ischemia at baselinedid not identify patients with a worse prognosis(68). The CLARIFY (Prospective Observational

Longitudinal Registry of PatientsWith Stable CoronaryArtery Disease) registry enrolled 32,105 outpatientswith prior MI, of whom 20,291 had undergone anoninvasive test for ischemia within 12 months ofenrollment (69). Anginal symptoms (with or withoutischemia), but not silent ischemia, were associatedwith an increased 2-year rate of cardiovascular deathor MI. These studies, performed in an era of moreintensive MT, are distinct from earlier reports in whichstrong associations between ischemia and outcomeswere observed.

RANDOMIZED TRIALS SUGGEST REVASCULARIZATION

DOES NOT REDUCE DEATH OR MI, EVEN IF ISCHEMIA

IS PRESENT, AND DOES NOT PRODUCE A DURABLE

IMPROVEMENT IN QoL. In the COURAGE trial, after amedian 4.6-year follow-up, there was no significantdifference between PCI plus OMT versus OMT alonein the primary endpoint of death or MI (19.0% vs.18.5%, respectively; HR: 1.05; 95% CI: 0.87 to 1.27;p ¼ 0.62) (8). Although high-risk subgroups wereidentified, none had an improved prognosis with PCI.Similarly, in the BARI 2D study, 5-year survival ratesin type 2 diabetic patients did not differ significantlybetween the revascularization and MT groups (88.3%vs. 87.8% respectively; p ¼ 0.97); MACE rates werealso similar (77.2% vs. 75.9% respectively; p ¼ 0.70)(9). A lower rate of MI was reported in the BARI 2DCABG stratum with revascularization compared withMT, although CABG-related MI was not ascertained,and this subgroup finding should be consideredhypothesis-generating. BARI 2D patients who ach-ieved excellent risk factor control had one-half thedeath rate of those with poor risk factor control (64).

Meta-analyses of SIHD strategy trials that excludepatients with recent ACS do not support a differencein prognosis between routine revascularization andMT only. Before the COURAGE trial, Katritsisand Ioannidis (70) performed a meta-analysis of11 randomized trials with 2,950 patients comparingPCI (with balloon angioplasty or BMS) to conservativetreatment. PCI did not reduce the rates of death orMI. Addition of the results from the COURAGE trial tothis meta-analysis did not alter the estimate of PCIon mortality (71). Schömig et al. (41) reported, in ameta-analysis of 17 trials, that PCI was associatedwith a significant reduction in all-cause mortality.However, in 4 trials, patients were treated with CABGas well as PCI, 1 trial included patients with MI within3 months, and 4 trials included MI patients within4 weeks. If the studies that included CABG andrecent MI patients are excluded from the analysis, asignificant mortality difference between PCI andmedical therapy is no longer present (OR: 0.91;

FIGURE 9 Meta-Analysis of PCI and MT Versus MT Alone in Patients With Documented Myocardial Ischemia

Source OR (95% CI) P Value OR (95% CI) Source OR (95% CI) P Value OR (95% CI)

0.01 0.1 1 10 100Favors PCI Favors MT

0.01 0.1 1 10 100Favors PCI Favors MT

HambrechtMASS IICOURAGEBARI 2DFAME 2Overall

1.02 (0.02–52.43)0.76 (0.27–2.16)0.84 (0.61–1.18)1.06 (0.71–1.58)0.33 (0.03–3.16)0.90 (0.71–1.16)

.99

.60

.32

.78

.33

.42

HambrechtMASS IICOURAGEBARI 2DFAME 2Overall

3.12 (0.12–78.45)1.24 (0.40–3.88)1.24 (0.94–1.65)1.29 (0.82–2.04)1.06 (0.51–2.22)1.24 (0.99–1.55)

.49

.71

.13

.27

.88

.06

A B

(A) Death. (B) Nonfatal MI. This meta-analysis required 50% statin use and only included studies of SIHD that directly compared the 2 randomized groups.

The addition of PCI to MT did not reduce the incidence of death or MI as compared with MT alone. Adapted with permission from Stergiopoulos et al. (73).

OR ¼ odds ratio; other abbreviations as in Figures 3, 4, and 5.


93

95% CI: 0.74 to 1.12). Trikalinos et al. (72) performeda network meta-analysis of 61 randomized trials(n ¼ 25,388) comparing balloon angioplasty, BMS, andDES with each other and with MT in SIHD. None of thePCI modalities was associated with reduced death orMI. The RR for comparisons between DES and MT was0.96, (95% CI: 0.60 to 1.52) for death, and 1.15(95% CI: 0.73 to 1.82) for MI. Stergiopoulos et al. (73)performed a meta-analysis of the subset of patientsfrom 5 SIHD strategy trials (5,286 patients) in whichstents and statins were used in more than 50% ofpatients, with ischemia diagnosed by stress testing orFFR. There were no significant differences in eventrates for PCI plus GDMT versus GDMT alone for deathor nonfatal MI, respectively (Figure 9). By contrast,early randomized trials (before the establishment ofcurrent standards for secondary prevention) demon-strated reduced rates of death, MI, and ACS withCABG compared with MT in patients with ischemia(7,34). For example, in the MASS II study, 1-yearusage rates of aspirin (77%), statins (63%), beta-blockers (58%), and angiotensin-converting enzymeinhibitors (27%) were substantially less than inrecent strategy trials (74).

Recent randomized trials with greater use ofsecondary prevention measures have not shown areduction in death or MI after PCI or CABG in patientswith documented ischemia. In a COURAGE trialanalysis of 1,381 patients, including 468 with at leastmoderate ischemia on site-interpreted SPECT studies,PCI was not associated with a reduction in the 5-yearrate of death or MI (48) (Figure 7). Similarly, in aseparate COURAGE substudy of 621 patients with corelaboratory–interpreted SPECT data, ischemic burdenat baseline did not predict cardiovascular events,nor did it identify a subgroup with an improvementin prognosis after PCI (65). Although the extent of

atherosclerosis was associated with the compositeendpoint of death, MI, or ACS in this study (Figure 8),there was no benefit from PCI across the spectrum ofdisease. In the BARI 2D SPECT substudy, a greaterextent of ischemia 1 year after randomization did notpredict reduced death or cardiovascular events fromrevascularization (66). In the STICH trial, inducibleischemia did not identify patients with benefit fromCABG compared with MT in patients with reducedLVEF (68). In the FAME-2 trial, FFR-guided PCI didnot significantly reduce the 2-year risk of death or MIcompared with MT (52). FAME-2 reported an increasein periprocedural MI (defined by stringent criteriathat required $10� creatine kinase-myocardial band[CK-MB] elevation or new Q waves with $5-foldCK-MB elevation), but a reduction in spontaneousMI with PCI compared with GDMT between 8 daysand 2 years. In light of recent evidence that prolongeddual antiplatelet therapy (which is more likely to beused after PCI than MT) reduces spontaneous MI andthe composite rate of cardiovascular death, MI, andstroke (75,76), the extent to which late benefits of PCIin FAME-2 and reductions of death and MI in somemeta-analyses with PCI (53) are attributable to dualantiplatelet therapy, rather than revascularization,is uncertain.

The impact of revascularization on angina-relatedQoL has more evidence to support its practice,although its effects are time-limited, and intensiveMT results in a high proportion of patients becomingangina-free. In the COURAGE trial, using the SeattleAngina Questionnaire, patients randomized to PCIreported greater freedom from angina for 24 monthsafter the procedure (35). In the BARI 2D study,greater freedom from angina with revascularizationcompared with intensive MT lasted beyond 1 yearonly in the CABG stratum (36). Finally, a report from



94

10 randomized trials of PCI versus MT in 6,762patients with SIHD found no difference in angina re-lief between the 2 approaches at the end of studyfollow-up (RR: 1.10; 95% CI: 0.97 to 1.26) (40). Theincremental benefit of PCI observed in older trials(OR: 3.38; 95% CI: 1.89 to 6.04) was substantially lessor absent in recent trials (OR: 1.13; 95% CI: 0.76 to1.68), possibly due to greater use of evidence-basedtherapies. It should also be recognized that a chal-lenge in the assessment of QoL in all of these trialsis the placebo effect, given the absence of shamPCI/CABG procedures in the comparator groups.COST-EFFECTIVENESS AND INFORMED CONSENT. Thecost-effectiveness of routine revascularization inSIHD must be considered. If revascularization doesnot reduce cardiovascular death or MI when addedto GDMT, it is difficult to justify the cost of revascu-larization for asymptomatic patients or those withmild symptoms who have not had a trial of anti-anginal therapy. Furthermore, if revascularizationimproves QoL (but does not reduce death or MI),patients and payers should understand what itcosts to buy a period of freedom from angina.

In the COURAGE trial, the addition of PCI to OMTwas not cost-effective (77). The added cost of PCIwas approximately $10,000, without significant gainin life-years or quality-adjusted life-years (QALYs).The incremental cost-effectiveness ratio varied fromjust over $168,000 to just under $300,000 per life-year or QALY gained with PCI. In the BARI 2D study,cost-effectiveness also favored MT over promptrevascularization (78). Lifetime projections of cost-effectiveness found MT to be cost-effective ($600per life-year added) compared with PCI, but sug-gested that CABG may be cost-effective ($47,000per life-year added). In the FAME-2 trial, the incre-mental cost-effectiveness ratio of PCI was $36,000per QALY (79), and the authors concluded that PCIwas economically attractive compared with the bestavailable MT. However, this analysis was limited,because 12-month QoL data was available for only 11%of patients in the economic substudy.

In the era of patient-centered care, patients shouldbe educated regarding the potential risks and benefitsof undergoing versus not undergoing revasculariza-tion. Interviews of patients suggest that they are nottruly knowledgeable after receiving “informed con-sent” for elective PCI. Rothberg et al. (80) found thatmost patients believe elective PCI will prevent deathor MI. Better informed patients are less likely tochoose angiography and possible PCI (81). Decisionsupport tools are needed to assess procedural riskand educate patients about the probability of deathor MI and improved QoL with or without PCI or

CABG. Part of the challenge is that providers are notskilled at explaining risks in probabilistic terms topatients. Providers need training to accurately andeffectively communicate potential risks and benefitsto patients.

ANALYSIS, DATA GAPS, AND

FUTURE DIRECTIONS

ACKNOWLEDGING EQUIPOISE. Credible observa-tional data support the position that ischemia is apredictor of adverse events, or alternatively, theposition that systemic inflammation and throm-bosis of vulnerable plaques (either with or withoutischemia) are responsible for most cases of cardio-vascular death, MI, and ACS (82,83). The optimalapproach to patients with SIHD remains unsettledbecause all prior randomized trials, either by designor execution, have limitations. Indeed, varyingCOURAGE nuclear substudy results can be used tosupport either routine revascularization or a moreconservative approach. Similarly, the case can bemade from large-scale observational studies and ran-domized trial substudy data that either initial strategyis appropriate. The strengths and limitations of ran-domized trials versus observational studies must beweighed. For example, death and MI was not reducedin 2,287 patients with SIHD randomized to PCI plusOMT versus OMT in the randomized COURAGE trial(8), but revascularization (PCI or CABG) comparedwith MT was associated with reduced death and MIamong 9,676 propensity-matched “real-world” pa-tients meeting COURAGE eligibility criteria (18). Arethese discordant results explained by residual con-founding in the observational study analysis (e.g.,factors contributing to patient unsuitability forrevascularization after cardiac catheterization), by theinclusion of a higher-risk population, revasculariza-tion by CABG, as well as PCI, and more endpointsyielding greater power to elicit differences betweenthe groups, or by better MT in the clinical trial setting?Frequent crossovers in strategy trials, although ar-guably representing an essential aspect of eachapproach, reduce power to demonstrate differences.For example, in the STICH trial, CABG reduced mor-tality in both the per-protocol and crossover patientpopulations (although whether ischemia was amodulating influence in these cohorts was not re-ported) (84). It should also be acknowledged that therisk profile of patients with SIHD may vary greatly,independent of the degree of ischemia, and thatoptimal tools for risk stratification are lacking.

When data are conflicting or individual trials areunderpowered, meta-analysis is often relied upon to

TABLE 2 Current Guideline Indications for Revascularization in SIHD

To Improve Survival To Improve Symptoms

Class ofRecommendation

Level ofEvidence

Class ofRecommendation

Level ofEvidence

2012/2014 ACCF/AHA/ACP/AATS/PCNA/SCAI/STS guidelines (10,11)

Left main disease

CABG I B — —

PCI IIa, IIb or III* B or C* — —

3-vessel disease

CABG I B — —

PCI IIb B — —

2-vessel disease with PLAD disease

CABG I B — —

PCI IIb B — —

2-vessel disease without PLAD disease

CABG IIa B — —

CABG IIb C — —

PCI IIb B — —

1-vessel disease with PLAD disease

CABG IIa B — —

PCI IIb B — —

1-vessel disease without PLAD disease

CABG III B — —

PCI III B — —

Left ventricular dysfunction

CABG IIa or IIb† B — —

CABG IIb B — —

PCI —‡ —

Unacceptable angina despite GDMT‡

PCI or CABG — — I A

PCI — — IIa§ C

CABG — — IIb§ C

Unacceptable angina with GDMT noncompliance or side effects,or patient preferencek

PCI or CABG — — IIa C

2013/2014 ESC/EACTS guidelines (12,13)

Left main disease I A I A

Proximal LAD disease I A I A

Multivessel disease with LVEF <40% I B IIa B

Large area of ischemia (>10% LV) I B I B

Single remaining vessel I C I A

Limiting symptoms or symptoms not responsive/intolerant to GDMT — — I A

Heart failure with >10% ischemia/viability IIb B IIa B

No limiting symptoms with GDMT and with none of the aboveor with FFR $0.80

III A III C

*Varies according to associated clinical conditions and anatomic complexity. †IIa for LVEF 35% to 50%; IIb for LVEF <35% without left main disease. ‡Insufficient data for arecommendation. §In patients with prior CABG. kIn patients with significant anatomic (>50% left main or >70% non–left main disease) or physiological (FFR <0.80) coronaryartery stenoses.

ACCF/AHA/ACP/AATS/PCNA/SCAI/STS ¼ American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines/American College ofPhysicians/American Association for Thoracic Surgery/Preventive Cardiovascular Nurses Association/Society for Cardiovascular Angiography and Interventions/Society ofThoracic Surgeons; ESC ¼ European Society of Cardiology; EACTS ¼ European Association for Cardio-Thoracic Surgery; FFR ¼ fractional flow reserve; LAD ¼ left anteriordescending coronary artery; LV ¼ left ventricular; LVEF ¼ left ventricular ejection fraction; PLAD ¼ proximal left anterior descending artery; other abbreviations as in Table 1.


95

provide a consensus direction. However, the results ofmeta-analyses may vary tremendously, depending onthe studies selected for inclusion. For example, somemeta-analyses included patients with recent MI;others did not. Meta-analyses share the limitations oftheir component trials, and have other drawbacks that

may not be immediately obvious. For example, in the100-trial network meta-analysis of revascularizationversus MT (47), only 15 trials had a MT arm that re-ported all-cause mortality. Most of the conclusionsfrom this study were thus drawn from indirect com-parisons across studies, and should be considered

CENTRAL ILLUSTRATION Ischemia Revascularization Equipoise:The ISCHEMIA Trial Design

+

Patients with stable ischemic heart disease (SIHD) and moderate or severe ischemia

Refer patients with eGFR <30 or on dialysis to ISCHEMIA-CKD trial

Blinded coronary CT angiogram* to exclude patients

Randomize patients into ISCHEMIA trial approaches

After average 3-year follow-up, compare rates of cardiovascular deathand myocardial infarction. Also compare quality of life (secondary endpoint)

(Based on the present level of evidencecan be made for either initial approach)

Optimal medical therapy (OMT) with cardiac catheterization (Cath)

reserved for failure of OMT

CONSERVATIVE

OMT+

Cath+

Optimal revascularization

INVASIVE

Stone, G.W. et al. J Am Coll Cardiol. 2016; 67(1):81–99.

CCTA may not be performed in participants with estimated glomerular filtration rate

<60 ml/min. Participants in whom CCTA shows significant left main disease ($50%

stenosis) or no obstructive disease are excluded. CCTA results are otherwise kept blinded.

Cath ¼ catheterization; CCTA ¼ coronary computed tomographic angiography; CKD ¼chronic kidney disease; eGFR ¼ estimated glomerular filtration rate; OMT ¼ optimal

medical therapy; SIHD ¼ stable ischemic heart disease.



96

hypothesis-generating. Moreover, patient-level datawere not available, and thus comparisons were unad-justed. Numerous confounders were likely present,which varied between studies, including differences inbaseline patient and lesion characteristics, and the typeof and adherence to GDMT and other pharmacother-apies. Endpoint definitions, such as periprocedural MI,varied between studies. Subgroup analysis was notpossible, and outcomes according to the extent of CADand ischemia are undetermined. In addition, thesestudies were performed over several decades, andcontrolling for evolution in general medical practice isnot possible. Indeed, many of these studies are ofquestionable relevance to contemporary practice, givenadvances in GDMT and revascularization techniquesand devices. These and other factors likely explain whymany of the 16 meta-analyses described in the presentreport reached diametrically opposite conclusions.

Beyond death and MI, the extent to which revas-cularization incrementally improves QoL may alsobe debated, and the frequency and importance ofrevascularization-related complications, as well asthe impact of drug-related side effects on QoL andadherence must be considered. And although bothemphasize the central role of GDMT in all patientswith SIHD, current U.S. and European guidelinesdiffer regarding the strength of the evidence forrevascularization in SIHD, and the relative weightthat should be accorded to anatomy versus ischemiawhen considering whether revascularization is useful(10–13) (Table 2). Reasonable physicians may variablyevaluate the strength of the evidence presentedherein; thus, a lucid argument can be made for eithermanagement approach, emphasizing the importanceof shared decision making between the treatingcardiologist, the referring physician, and the patient.Thus, despite strongly held individual beliefs, theoverall balance of data supporting each approachreflects a state of “community equipoise” (85).Exemplifying this equipoise, only 35% to 65% ofSIHD patients with documented moderate or severeischemia undergo cardiac catheterization within90 days (86,87).

An adequately powered, randomized trial ofcontemporary conservative versus invasive ap-proaches is greatly needed to provide guidance forthe optimal approach in patients with SIHD, moderateor severe ischemia, and symptoms that can becontrolled with MT. Despite frequently held ferventbeliefs favoring either early revascularization orinitial GDMT in SIHD, 80% of 499 cardiologists sur-veyed stated that they would enroll their patientswith stable anginal symptoms, >10% myocardialischemia, and normal LVEF in a randomized trial witha 50% chance of being conservatively managedwithout cardiac catheterization, provided their pa-tients did not have very high-risk features on stressimaging (88). Following this positive response, theISCHEMIA trial was formulated.THE ISCHEMIA TRIAL. The primary aim of theISCHEMIA trial (NCT01471522) is to determinewhether an initial invasive strategy of cardiac cathe-terization and optimal revascularization (with PCI orCABG, as determined by the local heart team) plusOMT will reduce the primary composite endpoint ofcardiovascular death or nonfatal MI in SIHD patientswith moderate or severe ischemia and medicallycontrollable or absent symptoms, as compared withan initial conservative strategy of OMT alone,with catheterization reserved for failure of OMT(Central Illustration). The major secondary endpoint isangina-related QoL. Other important secondary

https://clinicaltrials.gov/ct2/show/NCT01471522?term=NCT01471522%26rank=1


97

endpoints are health resource utilization, costs, andcost-effectiveness. The National Heart, Lung, andBlood Institute is sponsoring the trial, with oversightby a National Heart, Lung, and Blood Institute–appointed independent data and safety monitoringboard. In addition to the main trial, patients withadvanced chronic kidney disease (estimated glomer-ular filtration rate <30 ml/min or on dialysis) arerandomized in a parallel substudy. Blinded CCTA isperformed before randomization in participants withnormal renal function to exclude those with signifi-cant left main disease and no obstructive CAD.Enrollment began in late 2012 and is projected to endin 2017. Average follow-up will be approximately3 years. There are currently w300 participating sitesin more than 30 countries. The ISCHEMIA study thusaims to address limitations of previous strategytrials by: 1) enrolling patients before catheterization,so that anatomically high-risk patients are notexcluded; 2) enrolling a higher-risk group with atleast moderate ischemia; 3) minimizing crossovers; 4)using contemporary DES and physiologically guideddecision making (FFR) to achieve complete ischemic(rather than anatomic) revascularization; and 5) beingadequately powered to demonstrate whether routine

revascularization reduces cardiovascular death ornonfatal MI in patients with SIHD and at leastmoderate ischemia.

The results of the ISCHEMIA trial will haveimportant implications regarding global guidelinesfor performance and reimbursement of revasculari-zation procedures in patients with SIHD. To date,over 2,000 patients have been randomized, with nosafety concerns expressed by the data and safetymonitoring board. Given the clear state of communityequipoise that exists regarding how best to managepatients with SIHD, it is our hope that primary careproviders, cardiologists, and cardiac surgeons aroundthe world will enthusiastically support enrollment oftheir patients into the ISCHEMIA trial so that wemay provide much needed prospective evidence toinform the optimal management of patients withSIHD, substantial myocardial ischemia, and anginasymptoms that are controlled or absent.

REPRINT REQUESTS AND CORRESPONDENCE: Dr.Gregg W. Stone, Columbia University Medical Center,Cardiovascular Research Foundation, 111 East 59thStreet, 11th Floor, New York, New York 10022. E-mail:[email protected].

RE F E RENCE S

1. Smith SC Jr., Benjamin EJ, Bonow RO, et al.AHA/ACCF secondary prevention and risk reduc-tion therapy for patients with coronary and otheratherosclerotic vascular disease: 2011 update: aguideline from the American Heart Association andAmerican College of Cardiology Foundation. J AmColl Cardiol 2011;58:2432–46.

2. Piepoli MF, Corrà U, Benzer W, et al. Secondaryprevention through cardiac rehabilitation: fromknowledge to implementation: a position paperfrom the Cardiac Rehabilitation Section of theEuropean Association of Cardiovascular Preven-tion and Rehabilitation. Eur J Cardiovasc PrevRehabil 2010;17:1–17.

3. Keeley EC, Boura JA, Grines CL. Primary angio-plasty versus intravenous thrombolytic therapy foracute myocardial infarction: a quantitative reviewof 23 randomised trials. Lancet 2003;361:13–20.

4. Mehta SR, Cannon CP, Fox KA, et al. Routinevs selective invasive strategies in patients withacute coronary syndromes: a collaborative meta-analysis of randomized trials. JAMA 2005;293:2908–17.

5. Fox KA, Clayton TC, Damman P, et al.Long-term outcome of a routine versus selectiveinvasive strategy in patients with non-ST-segmentelevation acute coronary syndrome a meta-analysis of individual patient data. J Am CollCardiol 2010;55:2435–45.

6. Mozaffarian D, Benjamin EJ, Go AS, et al.American Heart Association Statistics Committeeand Stroke Statistics Subcommittee. Heart disease

and stroke statistics—2015 update: a report fromthe American Heart Association. Circulation 2015;131:e29–322.

7. Yusuf S, Zucker D, Peduzzi P, et al. Effect ofcoronary artery bypass graft surgery on survival:overview of 10-year results from randomisedtrials by the Coronary Artery Bypass Graft Sur-gery Trialists Collaboration. Lancet 1994;344:563–70.

8. Boden WE, O’Rourke RA, Teo KK, et al., for theCOURAGE Trial Research Group. Optimal medicaltherapy with or without PCI for stable coronarydisease. N Engl J Med 2007;356:1503–16.

9. BARI 2D Study Group. A randomized trial oftherapies for type 2 diabetes and coronary arterydisease. N Engl J Med 2009;360:2503–15.

10. Fihn SD, Gardin JM, Abrams J, et al. 2012ACCF/AHA/ACP/AATS/PCNA/SCAI/STS guidelinefor the diagnosis and management of patientswith stable ischemic heart disease: a report ofthe American College of Cardiology Foundation/American Heart Association Task Force on PracticeGuidelines, and the American College of Physi-cians, American Association for Thoracic Surgery,Preventive Cardiovascular Nurses Association,Society for Cardiovascular Angiography andInterventions, and Society of Thoracic Surgeons.J Am Coll Cardiol 2012;60:e44–164.

11. Fihn SD, Blankenship JC, Alexander KP, et al.2014 ACC/AHA/AATS/PCNA/SCAI/STS focusedupdate of the guideline for the diagnosis andmanagement of patients with stable ischemic

heart disease: a report of the American Collegeof Cardiology/American Heart Association TaskForce on Practice Guidelines, and the AmericanAssociation for Thoracic Surgery, PreventiveCardiovascular Nurses Association, Society forCardiovascular Angiography and Interventions,and Society of Thoracic Surgeons. J Am CollCardiol 2014;64:1929–49.

12. Montalescot G, Sechtem U, Achenbach S, et al.2013 ESC guidelines on the management of stablecoronary artery disease: the Task Force on themanagement of stable coronary artery disease ofthe European Society of Cardiology. Eur Heart J2013;34:2949–3003.

13. Windecker S, Kolh P, Alfonso F, et al. 2014ESC/EACTS guidelines on myocardial revasculari-zation. Eur Heart J 2014;35:2541–619.

14. Patel MR, Dehmer GJ, Hirshfeld JW, et al.ACCF/SCAI/STS/AATS/AHA/ASNC/HFSA/SCCT 2012appropriate use criteria for coronary revasculariza-tion focused update: a report of the AmericanCollege of Cardiology Foundation Appropriate UseCriteria Task Force, Society for CardiovascularAngiography and Interventions, Society of ThoracicSurgeons, American Association for Thoracic Sur-gery, American Heart Association, American Societyof Nuclear Cardiology, and the Society of Cardio-vascular Computed Tomography. J Am Coll Cardiol2012;59:857–81.

15. Ellis S, Alderman E, Cain K, et al. Prediction ofrisk of anterior myocardial infarction by lesionseverity and measurement method of stenoses in

mailto:[email protected]

http://refhub.elsevier.com/S0735-1097(15)06614-0/sref1


































































































98

the left anterior descending coronary distribution:a CASS Registry study. J Am Coll Cardiol 1988;11:908–16.

16. Garcia S, Sandoval Y, Roukoz H, et al. Outcomesafter complete versus incomplete revascularizationof patients with multivessel coronary artery dis-ease: ameta-analysis of89,883patients enrolled inrandomized clinical trials andobservational studies.J Am Coll Cardiol 2013;62:1421–31.

17. Head SJ, Mack MJ, Holmes DR Jr., et al.Incidence, predictors and outcomes of incom-plete revascularization after percutaneous coro-nary intervention and coronary artery bypassgrafting: a subgroup analysis of 3-year SYNTAXdata. Eur J Cardiothorac Surg 2012;41:535–41.

18. Wijeysundera HC, Bennell MC, Qiu F, et al.Comparative-effectiveness of revascularizationversus routine medical therapy for stable ischemicheart disease: a population-based study. J GenIntern Med 2014;29:1031–9.

19. Min JK, Berman DS, Dunning A, et al. All-causemortality benefit of coronary revascularization vs.medical therapy in patients without knowncoronary artery disease undergoing coronarycomputed tomographic angiography: results fromCONFIRM (COronary CT Angiography EvaluatioNFor Clinical Outcomes: An InteRnational Multi-center Registry). Eur Heart J 2012;33:3088–97.

20. Hannan EL, Samadashvili Z, Cozzens K, et al.Comparative outcomes for patients who do and donot undergo percutaneous coronary interventionfor stable coronary artery disease in New York.Circulation 2012;125:1870–9.

21. Tonino PAL, Fearon WF, De Bruyne B, et al.Angiographic versus functional severity of coro-nary artery stenoses in the FAME study: fractionalflow reserve versus angiography in multivesselevaluation. J Am Coll Cardiol 2010;55:2816–21.

22. Park SJ, Kang SJ, Ahn JM, et al. Visual-functional mismatch between coronary angiog-raphy and fractional flow reserve. J Am CollCardiol Intv 2012;5:1029–36.

23. Vanzetto G, Ormezzano O, Fagret D, et al.Long-term additive prognostic value of thallium-201 myocardial perfusion imaging over clinicaland exercise stress test in low to intermediaterisk patients: study in 1137 patients with 6-yearfollow-up. Circulation 1999;100:1521–7.

24. Pavin D, Delonca J, Siegenthaler M, et al.Long-term (10 years) prognostic value of a normalthallium-201 myocardial exercise scintigraphy inpatients with coronary artery disease documentedby angiography. Eur Heart J 1997;18:69–77.

25. Chang SM, Nabi F, Xu J, et al. The coronaryartery calcium score and stress myocardial perfu-sion imaging provide independent and comple-mentary prediction of cardiac risk. J Am CollCardiol 2009;54:1872–82.

26. Hachamovitch R, Hayes SW, Friedman JD, et al.Comparison of the short-term survival benefitassociated with revascularization compared withmedical therapy in patients with no prior cor-onary artery disease undergoing stress myocar-dial perfusion single photon emission computedtomography. Circulation 2003;107:2900–7.

27. Shaw LJ, Berman DS, Maron DJ, et al., for theCOURAGE Investigators. Optimal medical therapy

with or without percutaneous coronary interven-tion to reduce ischemic burden: results fromthe Clinical Outcomes Utilizing Revascularizationand Aggressive Drug Evaluation (COURAGE) trialnuclear substudy. Circulation 2008;117:1283–91.

28. Dorbala S, Di Carli MF, Beanlands RS, et al.Prognostic value of stress myocardial perfusionpositron emission tomography: results from amulticenter observational registry. J Am CollCardiol 2013;61:176–84.

29. Cortigiani L, Borelli L, Raciti M, et al. Predic-tion of mortality by stress echocardiography in2835 diabetic and 11 305 nondiabetic patients.Circ Cardiovasc Imaging 2015;8:e002757.

30. Johnson NP, Tóth GG, Lai D, et al. Prognosticvalue of fractional flow reserve: linking physio-logic severity to clinical outcomes. J Am CollCardiol 2014;64:1641–54.

31. Hachamovitch R, Rozanski A, Shaw LJ, et al.Impact of ischaemia and scar on the therapeuticbenefit derived from myocardial revascularizationvs. medical therapy among patients undergoingstress-rest myocardial perfusion scintigraphy. EurHeart J 2011;32:1012–24.

32. Hachamovitch R, Kang X, Amanullah AM, et al.Prognostic implications of myocardial perfusionsingle-photon emission computed tomography inthe elderly. Circulation 2009;120:2197–206.

33. Sorajja P, Chareonthaitawee P, Rajagopalan N,et al. Improved survival in asymptomatic diabeticpatients with high-risk SPECT imaging treatedwith coronary artery bypass grafting. Circulation2005;112:I311–6.

34. Hueb W, Lopes N, Gersh BJ, et al. Ten-yearfollow-up survival of the Medicine, Angioplasty, orSurgery Study (MASS II): a randomized controlledclinical trial of 3 therapeutic strategies for multi-vessel coronary artery disease. Circulation 2010;122:949–57.

35. Weintraub WS, Spertus JA, Kolm P, et al., forthe COURAGE Trial Research Group. Effect of PCIon quality of life in patients with stable coronarydisease. N Engl J Med 2008;359:677–87.

36. Dagenais GR, Lu J, Faxon DP, et al., for theBypass Angioplasty Revascularization Investiga-tion 2 Diabetes (BARI 2D) Study Group. Effectsof optimal medical treatment with or withoutcoronary revascularization on angina and subse-quent revascularizations in patients with type2 diabetes mellitus and stable ischemic heartdisease. Circulation 2011;123:1492–500.

37. Brooks MM, Chung SC, Helmy T, et al., for theBypass Angioplasty Revascularization Investiga-tion 2 Diabetes (BARI 2D) Study Group. Healthstatus after treatment for coronary artery diseaseand type 2 diabetes mellitus in the Bypass An-gioplasty Revascularization Investigation 2 Dia-betes trial. Circulation 2010;122:1690–9.

38. Pursnani S, Korley F, Gopaul R, et al. Percuta-neous coronary intervention versusoptimalmedicaltherapy in stable coronary artery disease: a sys-tematic review and meta-analysis of randomizedclinical trials. CircCardiovasc Interv2012;5:476–90.

39. Bangalore S, Toklu B, Amoroso N, et al. Baremetal stents, durable polymer drug eluting stents,and biodegradable polymer drug eluting stents

for coronary artery disease: mixed treatmentcomparison meta-analysis. BMJ 2013;347:f6625.

40. Thomas S, Gokhale R, Boden WE, et al.A meta-analysis of randomized controlled trialscomparing percutaneous coronary interventionwith medical therapy in stable angina pectoris. CanJ Cardiol 2013;29:472–82.

41. Schömig A, Mehilli J, de Waha A, et al. A meta-analysis of 17 randomized trials of a percutaneouscoronary intervention-based strategy in patientswith stable coronary artery disease. J Am CollCardiol 2008;52:894–904.

42. Jeremias A, Kaul S, Rosengart TK, et al.The impact of revascularization on mortality inpatients with nonacute coronary artery disease.Am J Med 2009;122:152–61.

43. Zellweger MJ, Kaiser C, Brunner-La Rocca HP,et al., for the BASKET Investigators. Value andlimitations of target-vessel ischemia in predictinglate clinical events after drug-eluting stent im-plantation. J Nucl Med 2008;49:550–6.

44. Kirtane AJ, Gupta A, Iyengar S, et al. Safetyand efficacy of drug-eluting and bare metal stents:comprehensive meta-analysis of randomized trialsand observational studies. Circulation 2009;119:3198–206.

45. Valgimigli M, Sabaté M, Kaiser C, et al. Effectsof cobalt-chromium everolimus eluting stents orbare metal stent on fatal and non-fatal cardio-vascular events: patient level meta-analysis. BMJ2014;349:g6427.

46. Palmerini T, Biondi-Zoccai G, Della Riva D,et al. Clinical outcomes with bioabsorbablepolymer- versus durable polymer-based drug-eluting and bare-metal stents: evidence from acomprehensive network meta-analysis. J Am CollCardiol 2014;63:299–307.

47. Windecker S, Stortecky S, Stefanini GG, et al.Revascularisation versus medical treatment inpatients with stable coronary artery disease:network meta-analysis. BMJ 2014;348:g3859.

48. Shaw LJ, Weintraub WS, Maron DJ, et al.Baseline stress myocardial perfusion imaging re-sults and outcomes in patients with stable ischemicheart disease randomized to optimal medicaltherapy with or without percutaneous coronaryintervention. Am Heart J 2012;164:243–50.

49. Bech GJ, De Bruyne B, Pijls NH, et al. Fractionalflow reserve to determine the appropriateness ofangioplasty in moderate coronary stenosis: a ran-domized trial. Circulation 2001;103:2928–34.

50. Tonino PA, De Bruyne B, Pijls NH, et al., for theFAME Study Investigators. Fractional flow reserveversus angiography for guiding percutaneous cor-onary intervention.NEngl JMed2009;360:213–24.

51. Pijls NH, Fearon WF, Tonino PA, et al., for theFAME Study Investigators. Fractional flow reserveversus angiography for guiding percutaneouscoronary intervention in patients with multivesselcoronary artery disease: 2-year follow-up of theFAME (Fractional Flow Reserve Versus Angiog-raphy for Multivessel Evaluation) study. J Am CollCardiol 2010;56:177–84.

52. De Bruyne B, Pijls NH, Kalesan B, et al., for theFAME 2 Trial Investigators. Fractional flowreserve-guided PCI versus medical therapy in










































































































































































































99

stable coronary disease. N Engl J Med 2012;367:991–1001.

53. Bangalore S, Pursnani S, Kumar S, et al.Percutaneous coronary intervention versus optimalmedical therapy for prevention of spontaneousmyocardial infarction in subjects with stable is-chemic heart disease. Circulation 2013;127:769–81.

54. Moussa ID, Klein LW, Shah B, et al. Consider-ation of a new definition of clinically relevantmyocardial infarction after coronary revasculari-zation: an expert consensus document from theSociety for Cardiovascular Angiography and In-terventions (SCAI). J Am Coll Cardiol 2013;62:1563–70.

55. Erne P, Schoenenberger AW, Burckhardt D,et al. Effects of percutaneous coronary in-terventions in silent ischemia after myocardialinfarction: the SWISSI II randomized controlledtrial. JAMA 2007;297:1985–91.

56. Gada H, Kirtane AJ, Kereiakes DJ, et al. Meta-analysis of trials on mortality after percutaneouscoronary intervention compared with medicaltherapy in patients with stable coronary heartdisease and objective evidence of myocardialischemia. Am J Cardiol 2015;115:1194–9.

57. Iqbal J, Zhang YJ, Holmes DR, et al. Optimalmedical therapy improves clinical outcomes inpatients undergoing revascularization with percu-taneous coronary intervention or coronary arterybypass grafting: insights from the Synergy Be-tween Percutaneous Coronary Intervention withTAXUS and Cardiac Surgery (SYNTAX) trial at the5-year follow-up. Circulation 2015;131:1269–77.

58. Hutchins R, Viera AJ, Sheridan SL, et al.Quantifying the utility of taking pills for car-diovascular prevention. Circ Cardiovasc QualOutcomes 2015;8:155–63.

59. Fontana M, Asaria P, Moraldo M, et al. Patient-accessible tool for shared decision making incardiovascular primary prevention: balancinglongevity benefits against medication disutility.Circulation 2014;129:2539–46.

60. Naderi SH, Bestwick JP, Wald DS. Adherenceto drugs that prevent cardiovascular disease:meta-analysis on 376,162 patients. Am J Med2012;125:882–7.e1.

61. Walsh MN, Bove AA, Cross RR, et al. ACCF 2012health policy statement on patient-centered care incardiovascular medicine: a report of the AmericanCollege of Cardiology Foundation Clinical QualityCommittee. J Am Coll Cardiol 2012;59:2125–43.

62. Kocas C, Abaci O, Oktay V, et al. Percutaneouscoronary intervention vs. optimal medicaltherapy—the other side of the coin: medicationadherence. J Clin Pharm Ther 2013;38:476–9.

63. Spertus JA, Kettelkamp R, Vance C, et al.Prevalence, predictors, and outcomes of prema-ture discontinuation of thienopyridine therapy af-ter drug-eluting stent placement: results from thePREMIER registry. Circulation 2006;113:2803–9.

64. Bittner V, Bertolet M, Felix RB, et al., BARI 2DStudy Group. Comprehensive cardiovascular riskfactor control improves survival: the BARI 2D trial.J Am Coll Cardiol 2015;66:765–73.

65. Mancini GB, Hartigan PM, Shaw LJ, et al.Predicting outcome in the COURAGE trial (Clinical

Outcomes Utilizing Revascularization and Aggres-sive Drug Evaluation): coronary anatomy versusischemia. J Am Coll Cardiol Intv 2014;7:195–201.

66. Shaw LJ, Cerqueira MD, Brooks MM, et al.Impact of left ventricular function and the extentof ischemia and scar by stress myocardial perfu-sion imaging on prognosis and therapeutic riskreduction in diabetic patients with coronary arterydisease: results from the Bypass AngioplastyRevascularization Investigation 2 Diabetes (BARI2D) trial. J Nucl Cardiol 2012;19:658–69.

67. Aldweib N, Negishi K, Hachamovitch R, et al.Impact of repeat myocardial revascularization onoutcome in patients with silent ischemia afterprevious revascularization. J Am Coll Cardiol 2013;61:1616–23.

68. Panza JA, Holly TA, Asch FM, et al. Induciblemyocardial ischemia and outcomes in patientswith coronary artery disease and left ventriculardysfunction. J Am Coll Cardiol 2013;61:1860–70.

69. Steg PG, Greenlaw N, Tendera M, et al., forthe Prospective Observational Longitudinal Reg-istry of Patients With Stable Coronary Artery Dis-ease (CLARIFY) Investigators. Prevalence ofanginal symptoms and myocardial ischemia andtheir effect on clinical outcomes in outpatientswith stable coronary artery disease: data from theInternational Observational CLARIFY Registry.JAMA Intern Med 2014;174:1651–9.

70. Katritsis DG, Ioannidis JP. Percutaneous coro-nary intervention versus conservative therapy innonacute coronary artery disease: a meta-analysis.Circulation 2005;111:2906–12.

71. Wijeysundera HC, Ko JT. Does percutaneouscoronary intervention reduce mortality in patientswith stable chronic angina: are we talking aboutapples and oranges? Circ Cardiovasc Qual Out-comes 2009;2:123–6.

72. Trikalinos TA, Alsheikh-Ali AA, Tatsioni A, et al.Percutaneous coronary interventions for non-acute coronary artery disease: a quantitative20-year synopsis and a network meta-analysis.Lancet 2009;373:911–8.

73. Stergiopoulos K, Boden WE, Hartigan P, et al.Percutaneous coronary intervention outcomes inpatients with stable obstructive coronary arterydisease and myocardial ischemia: a collaborativemeta-analysis of contemporary randomized clin-ical trials. JAMA Intern Med 2014;174:232–40.

74. Hueb W, Soares PR, Gersh BJ, et al. TheMedicine, Angioplasty, or Surgery Study (MASS-II):a randomized, controlled clinical trial of threetherapeutic strategies for multivessel coronaryartery disease one-year results. J Am Coll Cardiol2004;43:1743–51.

75. Mauri L, Kereiakes DJ, Yeh RW, et al., for theDAPT Study Investigators. Twelve or 30 months ofdual antiplatelet therapy after drug-elutingstents. N Engl J Med 2014;371:2155–66.

76. Bonaca MP, Bhatt DL, Cohen M, et al., for thePEGASUS-TIMI 54 Steering Committee andInvestigators. Long-term use of ticagrelor inpatients with prior myocardial infarction. N EnglJ Med 2015;372:1791–800.

77. Weintraub WS, Boden WE, Zhang Z, et al., forthe Department of Veterans Affairs CooperativeStudies Program No. 424 (COURAGE Trial)

Investigators and Study Coordinators. Cost-effec-tiveness of percutaneous coronary intervention inoptimally treated stable coronary patients. CircCardiovasc Qual Outcomes 2008;1:12–20.

78. Hlatky MA, Boothroyd DB, Melsop KA, et al.,for the Bypass Angioplasty RevascularizationInvestigation 2 Diabetes (BARI 2D) Study Group.Economic outcomes of treatment strategies fortype 2 diabetes mellitus and coronary artery dis-ease in the Bypass Angioplasty RevascularizationInvestigation 2 Diabetes trial. Circulation 2009;120:2550–8.

79. Fearon WF, Shilane D, Pijls NH, et al., for theFractional Flow Reserve Versus Angiography forMultivessel Evaluation 2 (FAME-2) Investigators.Cost-effectiveness of percutaneous coronary in-tervention in patients with stable coronary arterydisease and abnormal fractional flow reserve.Circulation 2013;128:1335–40.

80. Rothberg MB, Sivalingam SK, Ashraf J, et al.Patients’ and cardiologists’ perceptions of thebenefits of percutaneous coronary interventionfor stable coronary disease. Ann Intern Med 2010;153:307–13.

81. Rothberg MB, Sivalingam SK, Kleppel R, et al.Informed decision making for percutaneous coro-nary intervention for stable coronary disease.JAMA Intern Med 2015;175:1199–206.

82. Hansson GK. Inflammation, atherosclerosis,and coronary artery disease. N Engl J Med 2005;352:1685–95.

83. Stone GW, Maehara A, Lansky AJ, et al., forthe PROSPECT Investigators. A prospectivenatural-history study of coronary atherosclerosis.N Engl J Med 2011;364:226–35.

84. Doenst T, Cleland JG, Rouleau JL, et al., for theSTICH Investigators. Influence of crossover onmortality in a randomized study of revascularizationin patients with systolic heart failure and coronaryartery disease. Circ Heart Fail 2013;6:443–50.

85. Freedman B. Equipoise and the ethics of clin-ical research. N Engl J Med 1987;317:141–5.

86. Hachamovitch R, Nutter B, Hlatky MA, et al.Patient management after noninvasive cardiacimaging results from SPARC (Study of MyocardialPerfusion and Coronary Anatomy Imaging Roles inCoronary Artery Disease). J Am Coll Cardiol 2012;59:462–74.

87. Shaw LJ, Hachamovitch R, Berman DS, et al.,for the Economics of Noninvasive Diagnosis (END)Multicenter Study Group. The economic conse-quences of available diagnostic and prognosticstrategies for the evaluation of stable angina pa-tients: an observational assessment of the value ofprecatheterization ischemia. J Am Coll Cardiol1999;33:661–9.

88. Maron DJ, Stone GW, Berman DS, et al. Iscardiac catheterization necessary before initialmanagement of patients with stable ischemicheart disease? Results from a Web-based survey ofcardiologists. Am Heart J 2011;162:1034–43.

KEY WORDS angina pectoris, coronaryartery bypass, coronary artery disease,guideline-directed medical therapy,percutaneous coronary intervention



































































































































































































medical therapy with versus without revascularization in stable … · 2016-11-13 · high-risk cad...

Documents