Douglass A Morrison

KEY POINTS

■ The relative roles of medical therapy, coronary artery bypass grafting (CABG), and percutaneous coronary intervention (PCI) in the treatment of patients with coronary artery disease (CAD) are evolving.

■ Traditionally, CABG has been recommended, largely on anatomic and functional bases, over a broad clinical and/ or physiologic spectrum. Anatomic features such as greater than 70% epicardial narrowing of all three major coronary arteries; greater than 50% narrowing of the left main coronary artery; or greater than 70% narrowing of two major coronary arteries, if one of them is the left anterior descending artery, as well as functional criteria such as left ventricular ejection fraction (LVEF) less than .55, have been used to select patients with stable or unstable angina or myocardial infarction (MI) for CABG.

■ Although some structural features, such as "diffuse disease" or "small-caliber targets," are used to deny patients CABG on a regular basis, these features are hard to define and therefore have not generally found their way into practice guidelines.

■ Reperfusion therapy has revolutionized the care of patient with acute MI; it has spawned several major shifts in the clinical choice between CABG and PCI, all of which are primarily a matter of physiology, not anatomy. Randomized studies of patients with ST-segment elevation myocardial infarction (STEMI), comparing fibrinolytic therapy with emergency PCI, are concordant in showing improved survival, reduced reinfarction, reduced stroke, and reduced late revascularization with PCI. (In part because of the significance of time to complete reperfusion, very few patients in these trials were sent to CABG, regardless of anatomy.) Randomized trials of patients with non-ST-

segment elevation myocardial infarction (NSTEMI), comparing aggressive with conservative treatment, are also concordant in demonstrating advantage to early revascularization. (Again, only a minority of these patients were revascularized by CABG.) Moreover, every CABG registry has demonstrated that operating between day 1 and day 7 after an MI is an extremely high-risk endeavor.

■ Almost independent of the extent of anatomic CAD and the resting LVEF, emergent or urgent revascularization in any of the following settings, indicates PCI, rather than CABG, if at all technically achievable:

■ cardiogenic shock,

■ hemodynamic instability,

■ acute pulmonary edema,

■ recurrent or refractory NSTEMI.

■ Comorbidities, which render CABG high risk for mortality and/or morbidity, can also shift emphasis toward PCI and away from CABG, as in the following examples:

■ chronic obstructive pulmonary disease, which increases the risks of mediastinitis, prolonged mechanical ventilation, and/or atrial fibrillation;

■ ascending aortic disease or carotid vascular disease, either of which increases the risks of postoperative stroke and/or mental impairment; and

■ I iver disease, which greatly increases the risk with several anesthetic agents for post-CABG liver failure.

■ In the future, the objective, evidence-based choice between CABG and PCI is more a matter of clinical setting and patient physiology than of coronary anatomy and/or resting left ventricular global systolic performance.

The fundamental point of this chapter is to synthesize from four decades of disparate trials and registries a shift in emphasis from anatomic to physiologic thinking about (1) whether to revascularize, (2) how to revascularize, and (3) how to measure success of revascularization.1,2 An important caveat is that the physiologic approach emphasizes that the decision whether to revascularize, based on potential clinical benefit, should be answered before trying to answer the question of how to revascularize (i.e., CABG versus PCI).2 Focusing the decision of whether to revascularize on relief of medically refractory myo-

*None of these trials enrolled patients with acute MI, hemodynamic instability, severe LV dysfunction, prior CABG, or age >70 years. Medical therapy consisted of antianginals (nitrates, ^-blockers, or calcium channel blockers) on an as-needed basis; aspirin was not routine, and statins, ACE inhibitors, and angiotensin receptor blockers were not yet available. The subset of LMCA disease was included only in the first VA study. The world literature of patients with left main stenosis, randomly allocated between the medical therapy of the early 1970s and CABG, consists of 91 U.S. veterans. Percutaneous coronary intervention was not yet available at the time of these studies. Use of the internal mammary artery was not common, and cardioplegia has evolved since these trials.

ACE, angiotensin-converting enzyme; CABG, coronary artery bypass grafting; CASS, Coronary Artery Surgery Study; LMCA, left main coronary artery; LV, left ventricular; LVEF, left ventricular ejection fraction; Ml, myocardial infarction; NHLBI, National Heart, Lung and Blood Institute; NIH, National Institutes of Health; PCI, percutaneous coronary intervention; VA, Veterans Administration.

Modified from Morrison DA, Sacks J: Balancing benefit against risk in the choice of therapy for coronary artery disease (CAD): Lessons from prospective, randomized, clinical trials (RCTs) of percutaneous coronary intervention (PCI) and coronary artery bypass graft surgery (CABG). Minerva Cardioangiol 2003;51:585-597, with permission of Minerva Cardioangiologica.

Table 4-1. Randomized Trials of CABG versus Medical Therapy*

Major Outcome in

Primary Result in

Trial

N

Years

Exclusions

First Report

First Report

VA Stable Angina

686

1972-1974

Recent MI

Survival at 5 yr

No difference

Heart failure

Severe LV dysfunction

European Stable Angina

768

1973-1976

LVEF <.50

Survival at 7 yr

CABG better

LMCA disease

CASS Stable Angina

780

1975-1979

LMCA disease

Survival at 5 yr

No difference

NHLBI Unstable Angina

288

1972-1976

LMCA disease

Survival at 30 mo

No difference

VA Unstable Angina

468

1976-1978

LMCA disease

Survival at 5 yr

No difference

*None of these trials enrolled patients with acute MI, hemodynamic instability, severe LV dysfunction, prior CABG, or age >70 years. Medical therapy consisted of antianginals (nitrates, ^-blockers, or calcium channel blockers) on an as-needed basis; aspirin was not routine, and statins, ACE inhibitors, and angiotensin receptor blockers were not yet available. The subset of LMCA disease was included only in the first VA study. The world literature of patients with left main stenosis, randomly allocated between the medical therapy of the early 1970s and CABG, consists of 91 U.S. veterans. Percutaneous coronary intervention was not yet available at the time of these studies. Use of the internal mammary artery was not common, and cardioplegia has evolved since these trials.

ACE, angiotensin-converting enzyme; CABG, coronary artery bypass grafting; CASS, Coronary Artery Surgery Study; LMCA, left main coronary artery; LV, left ventricular; LVEF, left ventricular ejection fraction; Ml, myocardial infarction; NHLBI, National Heart, Lung and Blood Institute; NIH, National Institutes of Health; PCI, percutaneous coronary intervention; VA, Veterans Administration.

Modified from Morrison DA, Sacks J: Balancing benefit against risk in the choice of therapy for coronary artery disease (CAD): Lessons from prospective, randomized, clinical trials (RCTs) of percutaneous coronary intervention (PCI) and coronary artery bypass graft surgery (CABG). Minerva Cardioangiol 2003;51:585-597, with permission of Minerva Cardioangiologica.

cardial ischemia places optimal medical therapy before any revascularization. It also serves to emphasize the importance of medical management and risk factor modification after revascularization. Focusing on medically refractory ischemia as the primary indication for revascularization also serves to refine the definition of success. A "successful" CABG or PCI would not be a procedure in which the patient only survives and has a patent conduit; rather, it would require relief of the ischemia and/or symptoms for which the patient underwent the procedure in the first place.

The optimal roles of medical therapy and risk factor modification alone, compared with percutaneous coronary intervention (PCI) plus medical therapy, or with coronary artery bypass graft surgery (CABG) plus medical therapy, have not yet been fully defined. With technologic advances and changing operator experience, their relative roles are changing.12

CABG was introduced in the late 1960s, after the advent of the heart-lung bypass machine made it possible to operate on a still heart. Despite the complexity and significant potential morbidity of the procedure, CABG was shown to be accompanied by relief of acute ischemia and resolution of anginal symptoms in most patients. There was an almost immediate call to develop randomized clinical trials to determine whether the relief of ischemia and symptoms led to even more favorable outcomes, such as prevention of coronary events and improved survival. By the mid-1970s, three trials, of approximately 800 patients each, were completed that compared the medical therapy of that period with CABG (Table 4-1) in relatively low-risk patients with stable angina. These were the Veterans Affairs (VA) Cooperative Study; the European Cooperative Study; and the National Heart, Lung and Blood Institute (NHLBI) Coronary Artery Surgery Study (CASS).1-4 They were followed by two trials (VA and NHLBI) of medical therapy versus CABG in patients with unstable angina (myocardial infarction [MI] within 30 days, as defined by the World Health Organization criteria, was one exclusion criteria).

None of these five trials demonstrated any evidence to suggest prevention of subsequent MI by CABG. Four of the five trials failed to demonstrate a survival difference between CABG and medical therapy in their first reports, which included follow-up of from 3 to 5 years.1-4 Subsequent analyses and reports of longer follow-up demonstrated survival benefits for CABG among selected anatomic subsets: (1) left main coronary artery stenosis greater than 50%; (2) three-vessel stenoses greater than 70% (proximal left anterior descending [LAD], circumflex, and posterior descending, or anterior, lateral, and inferior territories); and (3) two-vessel stenoses greater than 70% (provided that LAD was one of the two), especially if (4) mild to moderately reduced left ventricular function, as measured by the ejection fraction (LVEF) is present. Meta-analysis of long-term results demonstrated an overall survival benefit with initial CABG. It is important to recognize which clinical categories of patients were systematically excluded from these trials, so as to avoid overgeneralizing the results. Another significant limitation of these trials derives from the very limited medical therapy in use at the time they were completed.

Through consecutive versions of the American College of Cardiology/American Heart Association (ACC/AHA) Guidelines for CABG, MI, and angina, the relative role of CABG has continued to be defined primarily in terms of numbers of major epicardial coronary arteries with greater than 50% (left main) or greater than 70% (LAD, circumflex, posterior

THE EVOLUTION OF REVASCULARIZATION FOR THE TREATMENT OF MYOCARDIAL ISCHEMIA AND THE ORIGINS OF THE "ANATOMIC PARADIGM"

Table 4-2. Randomized Clinical Trials of Medical Therapy versus PCI in Patients with Stable Coronary Artery Disease*

Trial

Years

Major Outcome

Result

Caveats and Comments

ACME (One-vessel) ACME (Two-vessel) RITA-2 AVERT

ACIP

MASS (LAD) COURAGE*

1987-1990 212 Exercise tolerance

1987-1990 101 Exercise tolerance

1992-1996 1018 Death or MI

1995-1996 341 Any ischemic event

1991-1993 558 Mortality at 2 yr

1994 214 Event-free survival

1999-2004 2287 Death or MI

PCI better No difference PCI better No difference

No difference CABG

No difference

Pre-stent era Pre-stent era 9% with stents 39% of PCI = stents 25% of PCI = no lipid Both CABG or PCI Both CABG or PCI PCI with stents

*For the most part, these studies included only low-risk patients and demonstrated relatively little benefit from revascularization. Medical management was not optimal at the time most were conducted. Medical therapy has improved significantly since all of these trials were completed. PCI outcomes have improved since these trials were completed.

'Boden WE, O'Rourke RA, Teo KK, et al; COURAGE Trial Research Group: Optimal medical therapy with or without PCI for stable coronary disease. N Engl J Med 2007;356:1503-1516.

ACIP, Asymptomatic Cardiac Ischemia Pilot Study; ACME, Angioplasty Compared to Medicine; AVERT, Atorvastatin versus Revascularization Treatments; CABG, coronary artery bypass grafting; COURAGE, Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation; LAD, left anterior descending coronary artery; MASS, Medicine, Angioplasty or Surgery Study; MI, myocardial infarction; PCI, percutaneous coronary intervention; RITA-

2, Randomized Intervention Treatment of Angina-2.

Modified from Morrison DA, Sacks J: Balancing benefit against risk in the choice of therapy for coronary artery disease (CAD): Lessons from prospective, randomized, clinical trials (RCTs) of percutaneous coronary intervention (PCI) and coronary artery bypass graft surgery (CABG). Minerva Cardioangiol 2003;51:585-597, with permission of Minerva Cardioangiologica. Compare with Table 11 in Smith SC Jr, Feldman TE, Hirshfeld JW Jr, et al: ACC/AHA/ SCA&I guideline update for percutaneous coronary intervention: Summary article. Cath Cardiovasc Intervent 2006;67:87-112.

descending) narrowings and LVEF greater than or less than .55—virtually regardless of symptoms, documented ischemia, or medical therapy that the patients are receiving.3 The trial results and guideline recommendations constitute the core of the "anatomic paradigm" for revascularization decision-making.

PCI, using only a balloon, was introduced in the late 1970s.1 Because general anesthesia, sternotomy, and heart-lung bypass were not a part of early PCI, the procedure had much less morbidity than CABG but was applied only to stable patients with relatively simple, single-vessel (>70%) disease. An important lesson was that ballooned lesions could suddenly occlude, leading to acute MI and even death. Surgical teams were held on standby so as to emergently perform CABG on patients with acute occlusion from PCI. Only proximal lesions and relatively "simple" lesions could be approached with early wire and balloon technology. The largest part of early PCI courses and textbooks focused on the myriad selection factors, and in those years, most practitioners speculated that PCI would not be applied to more than 5% to 10% of patients with symptomatic coronary artery disease (CAD). Anatomic characteristics of individual lesions (length, diameter of reference or "normal" segment, presence of thrombus, presence of calcification, presence of tortuosities, ostial location, bifurcation, saphenous vein graft lesions) were used to further refine the patient selection process for PCI, in terms of likelihood of success and likelihood of complication.1,2

The early randomized comparisons of PCI with medical therapy were even more limited in numbers of patients, categories of patients, and outcomes than the trials of medical therapy versus CABG (Table 4-2).125-7 A small VA trial (ACME) and a larger trial from the United Kingdom (RITA-2) documented relief of ischemia and relief of anginal symptoms. A small pilot trial funded by the NHLBI (ACIP) sup ported revascularization by either CABG or PCI for patients with documented myocardial ischemia, even in the absence of symptoms. None of these trials, nor any large registry series, supported either prevention of MI or survival benefit with balloon-only PCI. All of these trials were limited to clinically and angio-graphically "simple" cases.

Despite the lack of evidence supporting a survival benefit, use of PCI grew at an exponential rate during this period, primarily by application to low-risk patients with single-vessel disease, many of whom had neither symptoms nor documented ischemia. There was little emphasis on trying to optimize medical therapy for relief of ischemia or improving outcomes before performing PCI.1,2,5-7 In essence, PCI staked out a territory that was largely distinct from that of CABG on anatomic grounds; this lent further support to the "anatomic paradigm" of revasculariza-tion decision-making.

As technology improved and operator experience widened, PCI was applied to a larger spectrum of the CAD population and began to be considered as an alternative to CABG for patients with single-vessel CAD and for some patients with two-vessel CAD, provided that the anatomic features were favorable. The clinical spectrum approached with PCI remained clinically low risk, and surgical standby remained a routine part of the procedure. Because PCI was being applied to some patients with multivessel disease (mostly two-vessel disease excluding the LAD) who previously had been considered for CABG, there was a call to compare PCI with CABG. Over the 1980s and 1990s, nine trials of balloon-only PCI versus CABG were completed and reported: BARI, EAST, GABI, CABRI, RITA, ERACI I, MASS I, and the studies from Lausanne and Toulouse (Table 4-3).1-5,8 More than 90% of the screened patients were excluded before enrollment in any of these trials, mostly for clinical or angiographic features that made them unfavorable for balloon-only PCI. As in the previous

BARI had both the largest enrollment (1829) and the largest proportion of screened patients enrolled (15%). None of these studies had stents, glycoprotein Ilb/IIIa inhibitors, or "dual antiplatelet therapy" available. All studies excluded patients with MI within the previous 5 days, those with prior CABG, and those who required emergency revascularization. Some studies excluded patients with total occlusions (GABI), more than two total occlusions (EAST, ERACI-I), or greater than 2-cm length (GABI). Functional exclusions included, in some cases, ">50% of LV potentially involved in the event of abrupt occlusion" (GABI).

BARI, Bypass Angioplasty Revascularization Investigation; CABG, coronary artery bypass grafting; CABRI, Coronary Angioplasty versus Bypass Revascularization Investigation; EAST, Emory Angioplasty versus Surgery; ERACI I, Argentine Randomized Trial of Percutaneous Transluminal Coronary Angioplasty Versus Coronary Artery Bypass Surgery in Multivessel Disease I; GABI, German Angioplasty Bypass Surgery Investigation; LVEF, left ventricular ejection fraction; MASS I, Medicine, Angioplasty or Surgery Study I; MI, myocardial infarction; PCI, percutaneous coronary intervention; RITA, Randomized Intervention Treatment of Angina.

Modified from Morrison DA, Sacks J: Balancing benefit against risk in the choice of therapy for coronary artery disease (CAD): Lessons from prospective, randomized, clinical trials (RCTs) of percutaneous coronary intervention (PCI) and coronary artery bypass graft surgery (CABG). Minerva Cardioangiol 2003;51:585-597, with permission of Minerva Cardioangiologica. Compare with Table 10 in Smith SC Jr, Feldman TE, Hirshfeld JW Jr, et al: ACC/AHA/ SCA&I guideline update for percutaneous coronary intervention: Summary article. Cath Cardiovasc Intervent 2006;67:87-112.

Table 4-3. Randomized Trials Comparing Balloon-Only PCI with CABG

No.

No.

Mean

Mean

Prior

Diabetes

Angina Grade

PCI

CABG

Trial

Years

Screened

Enrolled

Age (Yr)

LVEF

MI (%)

(%)

3/4 (%)

Mortality

Mortality

BARI

1988-1991

12,530

1,829

61

.57

55

25

64

14 at 5 yr

11 at 5 yr

CABRI

1988-1993

23,047

1,054

60

.63

42

12

62

4 at 1 yr

2 at 1 yr

EAST

1987-1990

5,118

392

62

.61

41

23

80

7 at 3 yr

6 at 3 yr

GABI

1986-1991

8,981

359

59

.56

47

13

2 at 1 yr

5 at 1 yr

RITA

1989-1991

27,975

1,011

57

43

6

59

3 at 2.5 yr

4 at 2.5 yr

ERACI-I

1988-1990

1,409

127

57

.61

50

11

100

5 at 1 yr

5 at 1 yr

MASS

1988-1991

142

56

.75

0

23

1 at 3 yr

1 at 3 yr

Lausanne

134

56

12

78

Toulouse

152

67

38

13

53

BARI had both the largest enrollment (1829) and the largest proportion of screened patients enrolled (15%). None of these studies had stents, glycoprotein Ilb/IIIa inhibitors, or "dual antiplatelet therapy" available. All studies excluded patients with MI within the previous 5 days, those with prior CABG, and those who required emergency revascularization. Some studies excluded patients with total occlusions (GABI), more than two total occlusions (EAST, ERACI-I), or greater than 2-cm length (GABI). Functional exclusions included, in some cases, ">50% of LV potentially involved in the event of abrupt occlusion" (GABI).

BARI, Bypass Angioplasty Revascularization Investigation; CABG, coronary artery bypass grafting; CABRI, Coronary Angioplasty versus Bypass Revascularization Investigation; EAST, Emory Angioplasty versus Surgery; ERACI I, Argentine Randomized Trial of Percutaneous Transluminal Coronary Angioplasty Versus Coronary Artery Bypass Surgery in Multivessel Disease I; GABI, German Angioplasty Bypass Surgery Investigation; LVEF, left ventricular ejection fraction; MASS I, Medicine, Angioplasty or Surgery Study I; MI, myocardial infarction; PCI, percutaneous coronary intervention; RITA, Randomized Intervention Treatment of Angina.

Modified from Morrison DA, Sacks J: Balancing benefit against risk in the choice of therapy for coronary artery disease (CAD): Lessons from prospective, randomized, clinical trials (RCTs) of percutaneous coronary intervention (PCI) and coronary artery bypass graft surgery (CABG). Minerva Cardioangiol 2003;51:585-597, with permission of Minerva Cardioangiologica. Compare with Table 10 in Smith SC Jr, Feldman TE, Hirshfeld JW Jr, et al: ACC/AHA/ SCA&I guideline update for percutaneous coronary intervention: Summary article. Cath Cardiovasc Intervent 2006;67:87-112.

trials of medical therapy versus CABG, medical therapy was neither standardized nor mandated.1,2,8 None of these trials demonstrated a difference in subsequent MI rate or survival, comparing patients initially treated with PCI versus CABG, on follow-up of 3 to 5 years. Alternatively, even in this low-risk and small subset of CAD patients, relief of angina and need for subsequent revascularization both favored CABG from the beginning. Longer follow-up, meta-analysis of multiple trials, and at least one subset (diabetes) have all allowed for demonstration of a survival advantage with CABG, even among these low-risk patients.1-4 None of these results challenged the "anatomic paradigm" that CABG is routinely indicated for patients with left main disease and/or three-vessel disease, whereas PCI is indicated for favorable patients with one-vessel disease and patients with two-vessel disease in an area of active conflict between the two revascularization methods.1,2

In the ensuing decades, PCI advanced rapidly and a host of moderate to large (>1000 patients) randomized trials documented the objective clinical advances. Specifically. the application of bare metal stents and the use of advanced antiplatelet pharmacologic therapies, including thienopyridines rather than Couma-din anticoagulation for stenting, and glycoprotein IIB/IIIa inhibitors have greatly reduced the problem of acute and subacute occlusive syndromes.9-19 PCI-associated rates of MI, death, and emergent CABG have all declined as a result of these advances, and these declines have occurred despite a far wider application, in both clinical and angiographic terms.9-19

Perhaps the most important broadening of clinical application of PCI has been to include patients with acute coronary syndromes and, more specifically, patients with acute MI. During the past 20 years, more than 20 randomized trials, including more than

8000 patients with acute ST-segment elevation myocardial infarction (STEMI), have compared revascu-larization with contemporary medical therapy, including fibrinolysis (Table 4-4).20-24 Taken together, these trials demonstrate a survival benefit, with reduced rates of cardiac events and stroke, for revas-cularization, predominantly by PCI.20-24 Similarly, more than 6000 patients with high-risk unstable angina, mostly non-ST-segment elevation myocardial infarction (NSTEMI), have been randomly allocated between a conservative strategy of medical therapy and an aggressive strategy using predominantly PCI (Table 4-5).25-27 These trials have documented an objective clinical advantage for PCI (FRISC II, TACTICS-TIMI 18, RITA-3).725-27 PCI can be justified for almost all STEMI and most high-risk NSTEMI patients for emergent or urgent relief of medically refractory ischemia.20-27 The higher the clinical risk of MI patients, the more likely they will also enjoy a survival benefit compared with optimal contemporary medical management ("revascularization paradox").1,2 It is primarily the application of PCI to high-risk clinical subsets, coupled with randomized trial demonstration of major clinical advantage of PCI relative to contemporary medical therapy, that has allowed for evidence-based application of PCI to anatomic subsets (two- and three-vessel disease) previously reserved for CABG. The rationale for this application has been predominantly physiologic (acute MI and/or medically refractory ischemia), and the definition of success has been almost completely physiologic (reliefof ischemia, termination of acute infarction, recovery of ventricular function, enhanced survival).

When CABG was introduced, medical therapy consisted of antianginal medications such as nitrates, P-blockers, and calcium channel blockers on an "as

Table 4-4. Trials of PCI versus Thrombolysis in Patients with STEMI

Table 4-4. Trials of PCI versus Thrombolysis in Patients with STEMI

First Author

PCI

Lytic

Lytic

Stent

Gp IIb/IIIa

PCI

Lytic

PCI

Lytic

Zilstra

152

1 49

SK

No

No

1

7

0.7

2

Riberio

50

50

SK

No

No

6

2

0

0

Grinfeld

54

58

SK

No

No

9

14

Zilstra

47

53

SK

No

No

2

2

2

4

Akhras

42

45

SK

No

No

0

9

Widimsky

1 01

99

SK

Yes

No

7

14

0

1

DeBoer

46

41

SK

Yes

No

7

22

2

7

Widimsky

429

421

SK

Yes

Yes

7

10

DeWood

46

44

Duteplase

No

No

7

5

Grines

195

200

TPA

3 hr

No

No

3

7

0

4

Gibbons

47

56

Duteplase

No

No

4

4

0

0

Ribichini

55

55

Acc.

tPA

No

No

2

6

0

0

Garcia

95

94

Acc.

tPA

No

No

3

11

GUSTO IIB

565

573

Acc.

tPA

No

No

6

7

1

2

LeMay

62

61

Acc.

tPA

Yes

Yes

5

3

2

3

Bonnefoy

421

41 9

Acc.

tPA

Yes

Yes

5

4

0

1

Schomig

71

69

Acc.

tPA

Yes

Yes

4

7

Vermeer

75

75

Acc.

tPA

Yes

No

7

7

3

3

Andersen

790

782

Acc.

tPA

Yes

NA

7

8

1

2

Kastrati

81

81

Acc.

tPA

Yes

Yes

3

6

1

1

Aversano

225

226

Acc.

tPA

Yes

Yes

5

7

1

4

Grines

71

66

Acc.

tPA

Yes

Yes

8

12

0

5

Hochman

152

1 50

Acc.

tPA

Yes

Yes

47

56

3

1

Acc. tPA, accelerated tissue plasminogen activator; GUSTO IIB, Global Use of Strategies to Open Occluded Arteries in Acute Coronary Syndromes IIB; PCI, percutaneous coronary intervention; SK, streptokinase; STEMI, ST-segment elevation myocardial infarction; TPA, tissue plasminogen activator. Adapted from Keeley EC, Boura JA, Grines CL: Lancet 2003;361:13-20.) Reproduced from Morrison DA, Sacks J: Balancing benefit against risk in the choice of therapy for coronary artery disease (CAD): Lessons from prospective, randomized, clinical trials (RCTs) of percutaneous coronary intervention (PCI) and coronary artery bypass graft surgery (CABG). Minerva Cardioangiol 2003;51:585-597, with permission of Minerva Cardioangiologica.

Table 4-5. NSTEMI/Unstable Angina Strategy Trials*

Death or MI At Death At

Table 4-5. NSTEMI/Unstable Angina Strategy Trials*

Trial

Years

N

Conserv.

Invas.

Conserv.

Invas.

Conserv.

Invas.

TIMI-IIIb (U.S.)

1989-1992

1473

50

63

12

11

VANQWISH (U.S. VA)

1993-1996

920

33

44

14

23

8

14

FRISC-II (Scandinavia)

1996-1998

2457

9

71

14

10

4

2

TACTICS-TIMI 18 (U.S.)

1997-1999

220

36

61

10

7

4

3

RITA-3 (U.K.)

1997-2001

1 81 0

10

44

8

8

4

5

ICTUS (Netherlands)*

2001-2003

1200

40

76

14.4*

19.9*

4.5*

4.4*

*The trials with the biggest differences between invasive and conservative strategies in their primary end points were trials that included more contemporary therapy and had the largest differences in proportions of patients receiving conservative versus invasive pharmacotherapy (FRISC-II, TACTICS-TIMi, and RITA-3). Multivariate analysis of TIMI-IIIb yielded four variables that identified a high-risk subset that did benefit from early invasive strategy; patients in this subset had age >65 yr, ST depression at presentation, "complicated angina," and elevated CK-MB. Both FRISC II and TACTICS-TIMI 18 demonstrated benefit in reduction of death and MI only among patients with elevated troponin on presentation, and the highest reductions occurred among older patients with diabetes, prior MI, and electrocardiographic ST depression as well as troponin elevation. All of these factors are also predictors of adverse outcome; accordingly, the "revascularization paradox" (i.e., patients with the most long-term benefit often also have the highest short-term risk) is supported by these data in aggregate.

'Hirsch A, Windhausen F, Tijssen JG, et al: Long-term outcome after an early invasive versus selective invasive treatment strategy in patients with non-ST-elevation acute coronary syndrome and elevated cardiac troponin T (the ICTUS trial): A follow-up study. Lancet 2007;369:827-835.

'Two-year, not 12-month results.

Conserv., conservative strategy; FRISC-II, Fragmin and Revascularization during InStability in Coronary Artery Disease II; ICTUS, Invasive versus Conservative Treatment in Unstable Coronary Syndromes; Invas., invasive strategy; MI, myocardial infarction; NSTEMI, non-ST-segment elevation myocardial infarction; RITA-3, Randomized Intervention Trial of unstable Angina-3; TACTICS, Treat Angina with Aggrastat and Determine Cost of Therapy with an Invasive or Conservative Strategy; TIMI, Thrombolysis in Myocardial Infarction; VANQWISH, Veterans Affairs Non-Q-Wave Myocardial Infarction Strategies In-Hospital.

Modified from Morrison DA, Sacks J: Balancing benefit against risk in the choice of therapy for coronary artery disease (CAD): Lessons from prospective, randomized, clinical trials (RCTs) of percutaneous coronary intervention (PCI) and coronary artery bypass graft surgery (CABG). Minerva Cardioangiol 2003;51:585-597, with permission of Minerva Cardioangiologica. Compare with Table 18 in Smith SC Jr, Feldman TE, Hirshfeld JW Jr, et al: ACC/AHA/ SCA&I guideline update for percutaneous coronary intervention: Summary article. Cath Cardiovasc Intervent 2006;67:87-112.

needed" basis.1-4 Not only had there been no documentation of survival benefit with any form of medical therapy (beyond blood pressure control), but the early CABG trials regarded the "need" for medication after CABG as an indication of "failed" CABG.

The mindset developed among clinicians and patients that revascularization "fixed" the problem of atherosclerosis, whereas medical therapy simply "put it off." Patients were encouraged to "get it over with". Many patients continue to be convinced that CABG

*All of these trials, except ERACI II, excluded unprotected left main coronary artery disease. All of these trials, except AWESOME, excluded prior CABG. Only AWESOME mandated any definition of "medically refractory." Average time between randomization and treatment in ARTS was 28 days for CABG and 11 days for PCI. Nine of 22 PCI deaths in SoS were cancer-related (statistical aberration).

ARTS, Arterial Revascularization Therapies Study; AWESOME, Angina With Extemely Serious Operative Mortality Evaluation; CABG, coronary artery bypass grafting; ERACI II, Argentine Randomized Trial of Percutaneous Transluminal Coronary Angioplasty Versus Coronary Artery Bypass Surgery in Multivessel Disease (Estudio Randomizado Argentino de Angioplastia vs. Cirugía) II; LAD, left anterior descending coronary artery; LVEF, left ventricular ejection fraction; MASS II, Medicine, Angioplasty or Surgery Study II; MI, myocardial infarction; PCI, percutaneous coronary intervention; SoS, Stent or Surgery Trial.

Modified from Morrison DA, Sacks J: Balancing benefit against risk in the choice of therapy for coronary artery disease (CAD): Lessons from prospective, randomized, clinical trials (RCTs) of percutaneous coronary intervention (PCI) and coronary artery bypass graft surgery (CABG). Minerva Cardioangiol 2003;51:585-597, with permission of Minerva Cardioangiologica. Compare with Table 10 in Smith SC Jr, Feldman TE, Hirshfeld JW Jr, et al: ACC/AHA/ SCA&I guideline update for percutaneous coronary intervention: Summary article. Cath Cardiovasc Intervent 2006;67:87-112.

Table 4-6.

Trials of PCI versus CABG in

the Bare Metal Stent Era*

Trial

Years

N

Age (Yr)

(%)

Two- Or Three-Vessel Disease (%)

Unstable Angina

(%)

Exclusions

Death

Follow-up

PCI

CABG

ARTS

1997-1998

1 205

61

.60

44

17

98

37

LVEF <.30

1 yr

6%

3%

SoS

1996-1 999

988

62

.57

45

14

100

24

MI within 48 hr

2 yr

5%

2%

MASS II

2000

611

60

.68

44

30

0

Only single-

1 yr

2%

1%

vessel LAD

included

ERACI II

1996-1998

5619

61

29

17

95

92

25 mo

1%

4%

AWESOME

1995-2000

2431

67

.45

71

33

82

1 00

Only "medically

3 yr

20%

21%

refractory"

cases

included

*All of these trials, except ERACI II, excluded unprotected left main coronary artery disease. All of these trials, except AWESOME, excluded prior CABG. Only AWESOME mandated any definition of "medically refractory." Average time between randomization and treatment in ARTS was 28 days for CABG and 11 days for PCI. Nine of 22 PCI deaths in SoS were cancer-related (statistical aberration).

ARTS, Arterial Revascularization Therapies Study; AWESOME, Angina With Extemely Serious Operative Mortality Evaluation; CABG, coronary artery bypass grafting; ERACI II, Argentine Randomized Trial of Percutaneous Transluminal Coronary Angioplasty Versus Coronary Artery Bypass Surgery in Multivessel Disease (Estudio Randomizado Argentino de Angioplastia vs. Cirugía) II; LAD, left anterior descending coronary artery; LVEF, left ventricular ejection fraction; MASS II, Medicine, Angioplasty or Surgery Study II; MI, myocardial infarction; PCI, percutaneous coronary intervention; SoS, Stent or Surgery Trial.

Modified from Morrison DA, Sacks J: Balancing benefit against risk in the choice of therapy for coronary artery disease (CAD): Lessons from prospective, randomized, clinical trials (RCTs) of percutaneous coronary intervention (PCI) and coronary artery bypass graft surgery (CABG). Minerva Cardioangiol 2003;51:585-597, with permission of Minerva Cardioangiologica. Compare with Table 10 in Smith SC Jr, Feldman TE, Hirshfeld JW Jr, et al: ACC/AHA/ SCA&I guideline update for percutaneous coronary intervention: Summary article. Cath Cardiovasc Intervent 2006;67:87-112.

or PCI will "fix" their problem but that medical therapy, which has many side effects and is costly and inconvenient, only temporizes.

In the 3 decades since PCI was introduced, medical therapy and risk factor modification have evolved dramatically.1,2 Numerous large trials, megatrials, and meta-analyses have provided support for prolonged survival with each of the following medical therapies, among subgroups with both stable and unstable CAD: aspirin,28,29 statins and other lipid-lowering drugs,30 P-blockers,31 and angiotensin-converting enzyme inhibitors (ACE) inhibitors and/or angioten-sin receptor blocking agents (ARBs).32

A fundamental implication of the advances in medical therapy over the last 3 decades is that the inference that CABG will lead to prolonged survival in purely anatomic subsets is much less likely in 2008 than it was in 1970.2 In addition, all trials and registries are concordant in suggesting that patients continue to derive multiple objective clinical benefits, including survival benefit, if they continue optimal medical management after either CABG or PCI.3,5 Patients who continue optimal medical management and risk factor modification will likely go on to more "successful" long-term CABG or PCI outcomes.2

Despite the advances in technology and experience, many practitioners and observers have continued to view the choice between CABG and PCI largely in anatomic terms. Some of this tendency derives from the limiting of comparative trials in the bare metal stent era to clinically low-risk patients, and the screening and enrolling of patients starting with a "multivessel" classification (Table 4- 6).33-37 Additionally, many interventionists have seen the failure to demonstrate a survival difference between CABG and PCI in the pre-stent trials (see Table 4-3), not as a function of either low risk/low event rates or modest sample sizes leading to limited power and type I error

(failure to detect a difference where there is truly a difference), but rather reflecting the reality of the situation. (For example, compare the interventional view in Klein38 with the surgical view in the CABG Guideline.3) From the interventionist's perspective, the only task remaining for PCI to supplant CABG for multivessel disease was to reduce the late repeat revascularization difference by reducing restenosis. By this (anatomic) thinking, drug-eluting stents could allow, by inference, the notion that stents "prolong life" in three-vessel disease.10-14

Was this article helpful?

0 0

Post a comment