Introduction
The indications for coronary revascularization among patients with chronic coronary heart disease evolve as the scientific information accumulates and technology advances. The benefits associated with prompt coronary revascularization in reducing cardiovascular death and nonfatal myocardial infarction (MI) in patients presenting with acute coronary syndrome (ACS) are widely accepted. However, there is little clinical evidence to demonstrate a reduction in hard clinical endpoints in patients with stable symptoms undergoing coronary revascularization.
Clinical data generally support and clinicians have traditionally accepted that coronary revascularization in patients with chronic stable ischemic heart disease (SIHD) is appropriate in patients at high risk for future cardiovascular events or with lifestyle-limiting symptoms. Both the clinical guidelines and the coronary revascularization appropriate use criteria provide specific recommendations for clinicians considering a revascularization strategy among patients with SIHD.
Despite a nearly 70% decline in the age-standardized heart disease-related mortality over the last 40 years, coronary vascular disease remains the leading cause of death since 1900 in the United States. It has been estimated that approximately half of this reduced risk is directly attributable to improved medical treatment and coronary revascularization procedures. The majority of this risk reduction is associated with improved medical therapy rather than coronary revascularization procedures. This improvement in estimated cardiovascular risk has been attenuated in recent years because of an increase in societal obesity and a concomitant increase in the prevalence of diabetes.
Goals of Treatment
The major goals when treating patients with chronic coronary artery disease (CAD) are to reduce symptoms, improve the quality of life (QoL), and reduce risk of death and MI. Risk factor modification and optimal medical therapy are foundational strategies. Despite the use of these evidence-based medical therapies, patients often have persistent symptoms and residual cardiovascular risk. Data from a large, multinational, longitudinal registry of outpatients estimate that 1 in 3 patients have active symptoms and 1 in 4 have objective evidence of ischemia, and both are associated with future cardiovascular risk.
The decision to recommend coronary revascularization in patients with SIHD should be considered carefully. The discussion with the patient and family should include a transparent discussion of all treatment options, the anticipated benefits, and the risks of potential complications. In general, a discussion of an initial medical therapy approach should be discussed with the patient. In today’s healthcare environment, it is widely accepted that patients with chronic stable angina should be initially offered evidence-based medical therapy that can be optimized over time. This would include pharmacologic antianginal therapy, lifestyle intervention, and therapies to mitigate future cardiovascular risk. When patients undergo cardiac catheterization and the anatomy is appropriate for percutaneous coronary intervention (PCI), the trend in the last decade has been to proceed with ad hoc PCI rather than deferring it. Ad hoc PCI occurs approximately 86% of the time in the United States. For elective indications, deferring PCI allows one to consider alternative treatment strategies especially in the setting of high-risk multivessel CAD. Additionally, certain clinical scenarios are better suited for PCI whereas others are better suited for coronary artery bypass grafting (CABG). For example, patients who are medically noncompliant or who have recent gastrointestinal bleeding related to peptic ulcer disease may not be optimal candidates for long-term dual antiplatelet therapy and would be appropriate candidates to undergo CABG. On the other hand, patients with high clinical comorbidities or high frailty and three-vessel CAD coupled with impaired left ventricular (LV) dysfunction would be expected to have a survival benefit with CABG but may be too high risk and more appropriately referred for multivessel PCI. These and various other issues can often be better vetted with a heart team approach . This requires an interruption in the care process to enable a multidisciplinary team discussion and the willingness of busy practitioners to meet and discuss clinical cases. The heart team approach, especially in complex cases, is preferable.
Shared Decision-Making
Current guidelines recommend the use of a multidisciplinary heart team to facilitate decisions regarding coronary revascularization, percutaneous aortic valve replacement, and other high-risk cardiac procedures. This decision-making process is best shared across a wide variety of individuals including the patient, the patient’s family, the interventional cardiologist, the cardiac surgeon, the general cardiologist, and the primary care physician. It has been demonstrated that this clinical care paradigm is not only feasible but also appealing. Physicians who have known the patient over a prolonged period of time, often the general cardiologist and/or primary care physician, should play an important role in recommending therapies regarding coronary revascularization, whether it be CABG or PCI. Ideally the risk-benefit discussion during these heart team meetings would be provided by both the interventional cardiologist and the cardiac surgeon, who are best suited to provide an individualized risk-benefit assessment.
Coronary Revascularization: General Comments
Coronary revascularization procedures are common and costly, associated with an annual cost of $3.2 billion to Medicare. The rates of coronary revascularization procedures in the United States are decreasing. Between 2001 and 2008 there was a 14% decrease in the annual rate of coronary revascularization procedures, which was principally driven by a 28%, reduction in CABG procedures and an unchanged PCI rate. After 2008, PCI rates in the United States decreased by 28%, and much of this decline was in patients with SIHD ( Fig. 23.1 ).
Following publication of US national rates of “appropriate” coronary revascularization, data from the Washington State Clinical Outcomes Assessment Program between 2010 and 2013 showed that the number of PCI procedures decreased substantially, which was driven by a 43% decline in the number of PCI procedures performed for elective indications. The majority of this decline occurred following the onset of state-wide measurement of PCI appropriateness.
An estimated 500,000 PCI procedures are performed annually in the United States, and an equal number is performed in the rest of the world. The safety and efficiency of procedural techniques have evolved over the past 40 years. PCI is now considered a mature medical procedure. The physician-reported success rates are 99%, and the observed morbidity and mortality rates are approximately 1% to 2%. The complication rates for PCI procedures remain low despite an increasing complexity of lesions and higher-risk clinical demographics.
In 2010, 1488 facilities in the United States were enrolled in the American College of Cardiology National Cardiovascular Data Registry (ACC NCDR) CathPCI Registry, which captured approximately 85% of the PCI procedures performed. A 2012 analysis from this registry includes 1.1 million PCI patients and provides a cross-sectional description of the current state of coronary angiography and PCI in the United States. Institutional and operator volumes for PCI procedures are often used as proxies for PCI quality. Both vary considerably across institutions, with approximately half of all ACC NCDR institutions performing fewer than 400 PCI procedures and 13% performing more than 1000 PCI procedures annually. Currently, 40% of patients are older than age 65 and 12% are older than 80, 80% of patients are overweight, 45% are obese, and 36% have a history of diabetes mellitus. Approximately half the patients underwent a preprocedural stress test, with myocardial perfusion imaging being most commonly performed. Fig. 23.2 shows the percent of patients by indications for diagnostic coronary angiography and PCI. Approximately 18% of patients underwent PCI for stable angina symptoms, while an additional 9% were asymptomatic. Radial artery access in the 2010 report was used in approximately 7% of PCI procedures. This percentage has steadily increased over the last 7 years, with transradial access rates now approaching 20% in the United States (but higher in many parts of the world). The observed in-hospital mortality rate was 0.72%, and in-hospital stroke rate was 0.17%. The most common noncardiac complication continues to be periprocedural bleeding. With the emergence of third- and fourth-generation drug-eluting stents (DESs), target vessel revascularization rates are low at 5% to 7%, and stent thrombosis rates are measurably lower with newer-generation DES platforms.
Lesion progression is a well-recognized factor accounting for future cardiovascular risk following PCI. Approximately 20% of patients undergo repeat PCI within 3 years of the index PCI because of nonculprit lesion progression. Similarly, the Providing Regional Observations to Study Predictors of Events in the Coronary Tree (PROSPECT) trial demonstrated that 50% of all major adverse cardiovascular events at 3 years occur solely as a function of nonculprit lesion progression. Current PCI practices are designed to identify and effectively treat culprit lesions. Because of this, optimal medical therapy for secondary prevention is requisite among patients undergoing PCI.
Two major limitations of PCI remain: a higher than desirable frequency of incomplete revascularization related to complex disease and residual cardiovascular risk related to morbid and mortal events due to disease progression.
CABG is also very common and now performed approximately 400,000 times annually in the United States. There has, however, been a steady decline in the frequency of CABG procedures performed in the United States. Until recently this decline has been associated with an increase in percutaneous coronary revascularization procedures. The most commonly used conduits are the left internal mammary artery (LIMA) and the greater saphenous veins. The use of the LIMA is now considered a quality indicator for CABG and has long been linked to higher long-term patency than saphenous venous grafts. Moreover clinical outcomes are improved with use of the LIMA. Other arterial conduits such as the radial artery, the right internal mammary artery, and the gastroepiploic artery have been used and demonstrate improved patency rates compared with the saphenous venous grafts but are not routinely used in clinical practice.
In general, a CABG procedure takes 3 to 4 hours, and the patient remains hospitalized for 5 to 7 days and recuperates for 6 to 12 weeks following discharge. The risk for perioperative morbidity and mortality has decreased over time. There is now nearly universal participation of CABG centers in the Society of Thoracic Surgeons National Adult Cardiac Surgery Database. Although the predicted risk of mortality has not changed over time, there has been a measurable reduction in the adjusted mortality rates in the last 10 years ( Fig. 23.3 ). This has been similarly true for perioperative stroke rates.
Stroke remains a serious complication following CABG. Risk factors include increasing age, concomitant peripheral or cerebrovascular disease, diabetes, and aortic atherosclerosis. Neurocognitive decline has also been described in the post-CABG population and specifically linked to cardiopulmonary bypass. These studies have not been randomized controlled trials and the results are heterogeneous. Thus the link between CABG and cognitive decline remains uncertain. The current belief is that neurocognitive dysfunction is related to a number of factors including the impact of major surgery coupled with long-term effects in patients with cardiovascular risk factors and concomitant coronary artery and cerebrovascular disease.
There remains a risk for native disease progression and bypass graft failure following CABG. Thus it is vitally important that clinicians ensure patients remain on appropriate medical therapy. A 2016 American Heart Association (AHA) scientific statement clarifies appropriate secondary prevention therapy following CABG ( Box 23.1 ). In addition to these recommendations, studies have suggested that P2Y12 receptor inhibition following CABG may be associated with improved graft patency.
Aspirin, 81 mg
P2Y12 receptor inhibitors if they were indicated prior to CABG (i.e. for acute coronary syndrome or prior percutaneous coronary intervention)
β-blocker use in patients with prior myocardial infarction, left ventricular systolic dysfunction
Lifelong high-intensity statin therapy
Angiotensin-converting enzyme inhibitors in patients with diabetes and/or left ventricular systolic dysfunction
Aldosterone antagonists in patients with left ventricular systolic dysfunction and heart failure symptoms or signs
Consideration for participating in short-term cardiac rehabilitation program
There has been significant interest in maintaining the benefits of CABG with the use of less-invasive approaches. These techniques require specialized training and are limited in that complete revascularization is achieved less often. The sustainability of these techniques remains uncertain. Off-pump surgery has also been studied; however, the results have been inconsistent with no clear advantage over on-pump surgery. Hybrid surgical and percutaneous revascularization strategies have been explored in recent years. With this strategy patients undergo minimally invasive surgery with a LIMA usually to the left anterior descending artery and then subsequently undergo PCI to either the left circumflex or right coronary artery. Long-term data comparing this hybrid approach to conventional CABG are lacking, but this approach makes sound clinical sense in specific patient populations. Common appropriate use criteria for PCI and CABG for patients with multivessel disease are shown in Table 23.1 .
| CABG | PCI | |
|---|---|---|
| Two-vessel CAD with proximal LAD stenosis | A | A |
| Three-vessel CAD with low CAD burden (i.e., 3 focal stenoses, low SYNTAX score) | A | A |
| Three-vessel CAD with intermediate to high CAD burden (i.e., multiple diffuse lesions, presence of CTO, or high SYNTAX score) | A | U |
| Isolated left main stenosis | A | U |
| Left main stenosis and additional CAD with low CAD burden (i.e., one to two vessel additional involvement, low SYNTAX score) | A | U |
| Left main stenosis and additional CAD with intermediate to high CAD burden (i.e., three vessel involvement, presence of CTO, or high SYNTAX score) | A | I |
Indications for Coronary Revascularization in Patients with Stable Ischemic Heart Disease
The goals of coronary revascularization, whether PCI or CABG, are to relieve symptoms, improve QoL, and mitigate the risk of future complications related to atherosclerotic CAD such as cardiovascular mortality and nonfatal MI. These goals have not changed over the last several decades. However, the evidence base for coronary revascularization and medical therapy has greatly expanded over the past decade. These newer data provide clinicians with additional insights when recommending therapy to patients with chronic stable heart disease.
Relief of Symptoms
Relief of symptoms and improvement in QoL are central benefits of coronary revascularization in the SIHD population. Clinical data supporting sustained improved QoL following PCI in SIHD patients are varied. Study designs range from prospective nonrandomized data using pre-post analytic strategies with broad inclusion criteria to highly selected randomized controlled trials comparing PCI with medical and surgical therapies. Numerous noncomparative cohort studies demonstrate that PCI improves QoL and exercise capacity compared with pre-PCI assessment. Many (but not all) of these studies suggest PCI reduces both angina and need for antianginal medications and improves both exercise capacity and health status. A meta-analysis including 7818 patients demonstrated that PCI was superior to medical therapy in reducing angina. There was heterogeneity across these trials with shorter follow-up and earlier trials favoring PCI, whereas higher use of evidence-based therapies favored medical therapy. In general, factors associated with improved post-PCI QoL include increased frequency of baseline angina, greater extent of baseline myocardial ischemia, cardiac rehabilitation, and nonsmoking status. Lower socioeconomic status, unemployment, and numerous clinical comorbidities are associated with lower QoL after PCI.
In general, both CABG and PCI improve QoL. In the early weeks following coronary revascularization, PCI trends better than CABG, but this difference attenuates by 3 to 5 months. In a collaborative meta-analysis of 10 randomized clinical trials comparing CABG with PCI, CABG was superior to PCI in angina relief at 1 year in patients with multivessel disease. Most studies suggest CABG is superior to PCI in reducing angina and improving QoL, although the benefits of CABG diminish over many years of follow-up; this is likely related to vein graft failure and disease progression.
Coronary Revascularization to Reduce the Risk of Death and Nonfatal Myocardial Infarction
There are limited data to support the hypothesis that PCI reduces death and nonfatal MI in patients with low-risk SIHD. Selected trials have evaluated surgical revascularization for high-risk patients with three-vessel disease, left main disease, severe LV dysfunction, or severe ischemia. Both clinicians and practice guidelines extrapolate these early clinical trial findings to justify coronary revascularization (both PCI and CABG) in high-risk SIHD patients. Risk is often estimated following a noninvasive stress test. Features of low-, intermediate-, and high-risk stress test findings are shown in Table 23.2 .
| High risk (> 3% annual mortality) | Intermediate risk (1%–3% annual mortality) | Low risk (< 1% annual mortality) |
|---|---|---|
| Resting LVEF < 35% | Mild to moderate resting LVEF 35%–49% | Duke treadmill score ≥ 5 |
| Duke treadmill score ≤ −11 | Duke treadmill score 11–5 | Normal or small myocardial perfusion defect |
| Exercise LVEF < 35% | Stress-induced moderate perfusion defect without LV dilation or increased lung uptake | Normal stress echocardiographic wall motion during stress testing |
| Stress-induced large perfusion defect | Limited exercise capacity. Echocardiographic ischemia or wall motion abnormality at higher doses of dobutamine | |
| Stress-induced multiple perfusion defects | ||
| Large, fixed perfusion defect with LV dilation or increased lung uptake (thallium 201) | ||
| Stress-induced moderate perfusion defect with LV dilation or increased lung uptake | ||
| > 2 segments wall motion abnormality on dobutamine echocardiography | ||
| Stress echocardiographic evidence of extensive ischemia |
Low-Risk Patients
Many clinical trials have compared medical therapy with coronary revascularization in low-risk patients with chronic stable angina. Without exception these trials were neutral with respect to hard cardiovascular endpoints. Although these trials were largely conducted in the 1970s and 1980s and there have been major advances in medical, percutaneous, and surgical therapies since then, the overall findings are thought to be relevant today. In short, CABG was associated with improved symptoms but no measurable difference in survival or nonfatal MI.
Although many prior studies have failed to demonstrate a long-term benefit on death or MI following PCI or CABG in stable patients, the Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial is frequently cited as the landmark trial comparing medical therapy with PCI in SIHD patients. COURAGE evaluated the efficacy of PCI versus medical therapy in 2287 patients. Patients with either a greater than 70% diameter stenosis of at least one coronary artery with objective evidence of ischemia or greater than 80% stenosis coupled with typical angina symptoms were eligible. Exclusions included Canadian Cardiovascular Society (CCS) class IV angina, markedly abnormal stress tests, significant congestive heart failure symptoms, LV ejection fraction less than 30%, or revascularization within the prior 6 months. At 5 years, there was no difference in the primary endpoint of death or MI between PCI and the medical therapy group (odds ratio [OR] ratio 1.05, 95% confidence interval [CI] 0.87–1.27, p = 0.62). PCI patients had a lower rate of repeat revascularization and a lower need for antianginal medications than the medical therapy group during the first year.
There are a number of limitations to COURAGE that restrict generalizability to a broader SIHD population. COURAGE enrolled a lower-risk population than had been expected, with an annual cardiovascular mortality rate of 0.4%. There was an exceptionally high screen failure rate: of the nearly 36,000 patients screened, only 2287 enrolled. Other limitations included a low burden of baseline angina and a 30% crossover rate from medical therapy to PCI. The 5-year medical adherence rate in COURAGE was exceptional; 94% for aspirin, 93% for statins, and 86% for β-blocker therapy. Moreover 70% of patients achieved the low-density lipoprotein (LDL) target and almost half of the patients with diabetes achieved the hemoglobin A 1c target of less than 7%. These rates exceed those seen in usual clinical practice.
Substudy data from COURAGE support prior clinical studies and the evolving hypothesis that higher baseline myocardial ischemia is associated with higher future cardiovascular risk and that this risk may be attenuated with adequate coronary revascularization. In a single-center study preceding COURAGE of 5183 consecutive patients undergoing rest-stress single photon emission computed tomography (SPECT), the severity of stress perfusion abnormalities was associated with higher rates of MI and cardiovascular death. A summed stress score of greater than 13 was associated with an annual MI risk of 4.2% and a mortality rate of 2.9% when compared with patients with low-risk scans, who had an annual event rate of less than 1%. This same group of investigators also demonstrated that early revascularization following stress testing was associated with a lower adjusted rate of cardiovascular mortality in selected high-risk patients (4.6% vs 1.3%, p < 0.01). The benefits associated with early coronary revascularization appeared to be limited to those patients with the baseline ischemic burden of greater than 10%.
The COURAGE nuclear substudy ( n = 314 of 2287) also explored the ischemic burden hypothesis. Patients underwent sequential rest-stress myocardial perfusion imaging studies with SPECT at baseline and then 6 to 18 months following randomization. The baseline ischemic burden in these subjects was modest at 8%. Although both PCI and medical therapies reduced ischemia, the reduction was not robust, 2.7% and 0.5% in the PCI and medical therapy groups, respectively. Resolution of significant baseline ischemia, defined as greater than 5% reduction in ischemic burden and at least over 10% baseline ischemia, was associated with greater relief of angina and lower rates of death or nonfatal MI (13.4% vs 24.7%). Patients randomized to PCI were more likely to experience a 5% or greater reduction in ischemic myocardium (33% vs 19%, p = 0.0004). Lastly there was a graded relationship between the extent and severity of residual ischemia and the risk of future cardiovascular events. The rate of death or nonfatal MI ranged from 0% among those patients without residual ischemia to 39% for patients who after treatment had a 10% or greater residual ischemic burden. These data support the hypothesis that baseline ischemic burden is associated with risk and that this risk may be diminished with complete coronary revascularization if associated with a reduction in residual ischemic burden.
The ischemic burden hypothesis is being formally tested in the International Study of Comparative Health Effectiveness with Medical and Invasive Approaches (ISCHEMIA) (NCT01471522), which is designed to determine whether an invasive strategy is superior to a conservative strategy in reducing cardiovascular death or MI among patients with at least moderate ischemia on noninvasive imaging. ISCHEMIA is projected to enroll 8000 patients at 500 sites. In order to formally test the ischemia burden hypothesis, ISCHEMIA will need to have a very high frequency of complete revascularization and minimal residual ischemia in patients randomized to early angiography.
Cardiac troponin concentrations are used to risk stratify ACS patients for prompt revascularization but have not routinely been used to assess risk in patients with SIHD. High-sensitivity assays are now available to detect very low cardiac troponin levels in patients with stable heart disease. These low-level elevations are strongly associated with death, MI, and stroke in patients with SIHD. In the Bypass Angioplasty Revascularization Investigation in Type 2 Diabetes Trial (BARI 2D), elevated troponin T concentrations (≥ 14 ng/L) at baseline were common (approximately 40% of people had an increased baseline value) and associated with increased 5-year risk of death, MI, and stroke (hazard ratio [HR], 1.85; 95% CI, 1.48–2.32; p < 0.001). However, increased levels did not associate with improved outcome following coronary revascularization. Presently, whereas elevated troponins are associated with risk in SIHD patients, the elevated risk does not appear to be modified by routine coronary revascularization.
Complex Patients and Lesions
CABG has been shown to be superior to medical therapy in patients with severe multivessel CAD. These clinical trials date back to the 1970s and 1980s. When the early trials including the Veterans Administration Cooperative Study, the European Coronary Surgery Study, and the Coronary Artery Surgery Study are included in a large clinical database of over 2600 patients, CABG was associated with lower mortality at 5 years (10% vs 16%, p < 0.001), which extended to 10 years (26% vs 31%, p = 0.003) compared with medical therapy. The reduction in mortality was in general consistent across a variety of subgroups. However, the absolute benefit was greatest for patients at the highest risk including those with extensive CAD and those with moderate to severe LV systolic dysfunction. The clinical benefit of CABG is not realized in the first 1 to 2 years following surgery. Beyond this the benefit of CABG is apparent and divergent over time. Inclusion of these earlier trials is limited in that the background medical therapy is very different from current practices.
PCI is favored over medical therapy and CABG in patients presenting with an ACS, including non-ST segment elevation and ST segment elevation MI and in SIHD patients with physiologically important disease that is not extensive. The extent of disease has traditionally been quantified by number of diseased vessels. In patients with very complex coronary anatomy and acceptable clinical risk, CABG has been the preferred strategy. In this later group, the evidence base has expanded in recent years with publication of landmark trials, many of which have included patients with relatively complex coronary anatomy.
There are many strengths and limitations to both PCI and CABG. These treatment options are neither curative nor mutually exclusive when treating SIHD patients over the course of an individual’s lifetime. CABG is considered the gold standard for treating complex multivessel CAD and is associated with a higher rate of complete revascularization and lower likelihood of repeat revascularization than is PCI. The major advantage of CABG is a higher likelihood of complete revascularization and the ability to bypass a significant portion of the proximal epicardial coronary artery, mitigating the likelihood of future cardiovascular morbidity and mortality if the patient were to experience nonculprit proximal disease progression. The major limitation for CABG is the risk of vein graft degeneration over time and the perioperative complication rate, especially among those patients with multiple clinical comorbidities. The major advantages of PCI are its less invasive nature, the efficacy of PCI in the ACS population, and the ability to selectively treat the culprit lesion. The major limitations of PCI are a higher frequency of incomplete revascularization, the inability to prevent the clinical implications of nonculprit disease progression, a higher rate of repeat revascularization even in the DES era, and the technical challenges in very complex lesion subsets. In short, angiographic burden and lesion complexity limit the effectiveness of PCI, whereas clinical comorbidities limit the safety of CABG.
Comparisons of PCI and CABG in Patients with Multivessel Coronary Artery Disease
The Arterial Revascularization Therapies Study (ARTS), Argentine Randomized Trial of Coronary Angioplasty with Stenting Versus Coronary Bypass Surgery in Patients with Multiple Vessel Disease (ERACI II), Medicine Angioplasty or Surgery Study for Multi-Vessel Coronary Artery Disease (MASS-II), and Stent or Surgery (SoS) compared CABG to bare metal stenting and were each included in a pooled patient level meta-analysis of 3051 patients. At 5 years the incidence of death, MI, and stroke was similar (16.7% vs 16.9%; HR, 1.04, 95% CI, 0.86–1.27; p = 0.69), yet incomplete (89.4% vs 62.0%; p < 0.001) and repeat (29.0% vs 7.9%; p < 0.001) revascularizations were more frequent in the PCI population. There was a trend for improved mortality among diabetic patients randomized to CABG (12.4% vs 7.9%; p = 0.09). A larger meta-analysis spanning the percutaneous transluminal coronary angioplasty (PTCA) and stent era included 59,014 patients and demonstrated that risk-adjusted survival rates were improved in patients treated with CABG compared with those undergoing PCI.
The very large and more contemporary Synergy Between Percutaneous Coronary Intervention with TAXUS and Cardiac Surgery (SYNTAX) trial compared PCI with DESs to CABG in 1800 patients with previously untreated complex three-vessel and/or left main CAD. This trial incorporated a local heart team approach, which consisted of a cardiac surgeon and an interventional cardiologist. Each member reviewed the clinical case and coronary anatomy and each needed to agree that revascularization could be achieved with CABG or PCI. SYNTAX was powered as a noninferiority comparison of CABG and PCI with a primary endpoint of all-cause mortality, stroke, MI, or repeat revascularization. The 12-month primary endpoint occurred more commonly following PCI (17.8% vs 12.4%; p = 0.002), and thus noninferiority of PCI was not met. The difference was principally driven by higher repeat revascularization rates in the PCI-treated patients. The 12-month MI rates were similar; however, CABG was associated with a nearly fourfold increase in stroke (2.2% vs 0.6%; p = 0.003). At 5 years, PCI was associated with higher rates of death, stroke, MI, and repeat revascularization. Nonrandomized studies from 2012 have suggested lower mortality rates for CABG patients with multivessel disease.
A number of post hoc subgroup analyses from SYNTAX have been informative and have impacted both clinical practice and design of ongoing clinical trials.
Baseline SYNTAX Score
The SYNTAX score ( Fig. 23.4 ) was designed to predict the postprocedural risk associated with PCI or surgical revascularization. It is a visual estimate of CAD burden and complexity. The SYNTAX score takes into account complex lesions including bifurcations, chronic total occlusions, thrombus, calcification, and small diffuse disease. The score ranges from 0 to greater than 60 in very complex coronary anatomy lesions. In the SYNTAX trial, higher SYNTAX scores differentiated outcomes between CABG and PCI. There was no difference in endpoints for patients randomized to CABG across the entire range of SYNTAX scores. This was not the case for patients randomized to PCI where there was a significant interaction in the cumulative event rates between patients with low, intermediate, and higher SYNTAX scores. In patients with baseline SYNTAX scores of 0 to 22 the outcomes were comparable for CABG and PCI patients. CABG was associated with lower event rates with SYNTAX scores greater than 22. The SYNTAX calculator is online at www.syntaxscore.com .
Unprotected Left Main Disease
CABG has been the preferred strategy for treating patients with significant, unprotected left main disease. This treatment paradigm has evolved in recent years. PCI has moved from a class IIb to a class IIa indication for the treatment of left main disease among patients with lower SYNTAX scores. In the SYNTAX trial, there were 705 patients with unprotected left main stenosis. This was a prespecified subgroup and the 5-year major adverse coronary and cerebrovascular events (MACCE) rate was 36.9% for PCI and 31.0% for CABG (HR, 1.23; 95% CI, 0.95–1.59; p = 0.12). The composite of death, stroke, and MI was similar for PCI (19.0%) and CABG (20.8%). The rates of graft occlusion (4.4%) and stent thrombosis (5.1%) were also similar. When the patients were dichotomized based on SYNTAX scores, the 5-year MACCE rates were similar for PCI and CABG in left main patients with SYNTAX scores 0 to 32. In this low SYNTAX score group, the major adverse cardiac events (MACE) rate was numerically lower (14.8% vs 19.8%) and the death rate significantly lower (17.9% vs 15.1%) in the PCI group. However, in patients with SYNTAX scores of 33 or higher, CABG was associated with significantly lower cardiovascular death (15.8% vs 5.9%) and revascularization (34.1% vs 11.6%) rates. Based largely on these data, it is commonly believed that the extent and complexity of disease downstream from the left main is more related to clinical outcomes than the extent and/or complexity of the left main disease itself.
The Evaluation of Everolimus Eluting Stent System Versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization (EXCEL) clinical trial is projected to randomize 2600 patients with unprotected left main disease with low or intermediate SYNTAX scores to either PCI or CABG. The primary outcome will be the composite of death, MI, and stroke (EXCEL clinicaltrials.gov identifier NCT01205776. http//clinicaltrials.gov/ct2/show/study/NCT01205776 ).
Diabetes Mellitus
Patients with diabetes (see Chapter 24 ) are often thought to have very complex coronary anatomy with a greater number of diseased vessels, lesion complexity, and smaller more diffusely diseased coronary arteries. Diabetes is very often a predictor of poor outcomes following both PCI and CABG. In general, clinical trial data suggest that CABG is superior to PCI in patients with type 2 diabetes mellitus with multivessel disease. This recommendation dates back to the original BARI report.
The BARI 2D trial compared coronary revascularization, either CABG or PCI, with medical therapy among patients with type 2 diabetes mellitus and CCS class I or II angina. Whether patients underwent CABG or PCI was at the investigator’s discretion. Patients with perceived need for immediate revascularization, left main disease, significant heart failure, or revascularization within 1 year of study entry were excluded. The 5-year all-cause mortality and MACE rates were similar between the medical therapy and coronary revascularization groups. Crossover approached 40%. CABG but not PCI was associated with a significant reduction in nonfatal MI.
In the SYNTAX trial there were 296 patients with medically treated diabetes. In this substudy, the death, MI, stroke, and repeat revascularization rate was considerably higher in the PCI group (45.5% vs 23.6%; p < 0.001). This was driven by a twofold increase in all-cause mortality and a threefold increase in both repeat revascularization and MI rates. Stroke rates were comparable.
The Future Revascularization Evaluation in Patients with Diabetes Mellitus: Optimal Management of Multivessel Disease (FREEDOM) trial was a dedicated large-scale randomized clinical trial comparing CABG with PCI in patients with diabetes mellitus. It randomized 1900 patients with diabetes and multivessel disease to undergo either PCI with DESs or CABG. The 5-year rate of death, MI, or stroke occurred more frequently in the PCI-DES group (26.6% vs 18.7%; p = 0.005). CABG patients had significantly lower rates of nonfatal MI and all-cause mortality, whereas PCI patients had significantly fewer strokes. The 2- and 5-year results are shown in Fig 23.5 .
FREEDOM and SYNTAX trials suggest greater treatment benefit of CABG in patients with increasing complexity of CAD. The treatment effect of CABG for patients with SYNTAX scores greater than 22 was numerically greater. A similar numerical trend was seen in the 3-year results of the SYNTAX diabetes substudy. There is a consistent trend in the medical literature suggesting that, among patients with diabetes and complex multivessel CAD as assessed with the SYNTAX score, CABG lowers cardiovascular complications. On the basis of these clinical data the American College of Cardiology Foundation(ACCF)/AHA guidelines for the treatment of patients with SIHD give a class I recommendation favoring CABG over PCI for patients with multivessel CAD and concomitant diabetes mellitus.
Ischemic Cardiomyopathy
Ischemic cardiomyopathy is a morbid condition with a 10-year mortality rate of 60%. These patients have a multitude of comorbidities including LV systolic dysfunction, impaired coronary hemodynamics, abnormal myocardial energetics, increased myocardial oxygen consumption, and altered myocardial lactate metabolism even in the absence of significant CAD. In such patients, the decision whether or not to perform coronary revascularization is difficult.
The Surgical Treatment for Ischemic Heart Failure (STICH) trial randomized 1212 patients with ischemic cardiomyopathy with an ejection fraction of less than 35% to either CABG or medical therapy. This was a complex patient population and included a high prevalence of patients with diabetes, previous MI, and New York Heart Association (NYHA) class II-III heart failure. At a median follow-up of 56 months, there was no significant difference in cardiovascular deaths between CABG and medical therapy. The STITCH Extension Study was a 5-year extension that reported 10-year outcomes. The primary outcome of all-cause mortality occurred significantly less frequently in the CABG group (58.9% vs 66.1%; p = 0.02). This was driven by a significant reduction in both cardiovascular (40.5% vs 49.3%; p = 0.006) and noncardiovascular mortality (76.6% vs 87.0%; p < 0.001). The median survival advantage was 1.44 years in the CABG group with a number needed to treat of 14 to prevent 1 death.
Patients with Chronic Kidney Disease
Patients with chronic kidney disease (CKD) are often excluded from clinical trials, yet this population has grown substantially in the past 10 years. Although the risk of future cardiovascular events is substantial in the CKD population, there is a lower tendency to prescribe medical therapy and to perform stress tests, cardiac catheterizations, and revascularization procedures. The 2014 European Society of Cardiology and the European Association for Cardio Thoracic Surgery guidelines on myocardial revascularization recommend CABG over PCI (class IIa) in patients with moderate-severe CKD and multivessel disease when surgical risk is acceptable. These recommendations are based on nonrandomized large cohort studies. A large study from the New York State Registry (2015), however, suggests outcomes may be better following PCI. Using propensity-score matching methodology involving 5920 patients with CKD, PCI was associated with a lower risk of death (HR, 0.55; 95% CI, 0.35–0.87) and stroke (HR, 0.2; 95% CI, 0.12–0.42) yet a higher risk of repeat revascularization (HR, 2.42; 95% CI, 2.05–2.85). In a small substudy of 243 patients on renal replacement therapy PCI was associated with higher rates of death (HR, 2.02; 95% CI, 1.40–2.93) and higher rates of repeat revascularization (HR, 2.44; 95% CI, 1.50–3.96) compared with CABG. These nonrandomized studies suggest PCI may be preferable in patients with CKD with the possible exception of patients on chronic renal replacement therapy.
Patients with Prior Bypass Surgery
Approximately 20% of PCI patients in the United States have a history of previous CABG. There are a number of complex issues when intervening in patients with prior CABG ranging from the need to treat complex lesion subsets, including saphenous vein grafts, severe calcification, diffuse disease, bifurcation lesions, and chronic total occlusions. Patients with prior CABG are also high risk given advanced age, multivessel disease, and depressed LV systolic ejection fraction. Atheroembolization occurs more often following PCI of saphenous venous grafts because, in general, the atherosclerotic plaque in these grafts is more diffuse and friable, and patients have very thin fibrous caps and little or no calcification relative to native coronary atherosclerosis. This results in a higher frequency of atheroembolization, microvascular obstruction, no reflow, and postprocedural death or MI. No reflow can be treated with arterial vasodilators such as nitroprusside, verapamil, or adenosine. Embolic protection devices are generally considered effective in reducing the complications of atheroembolization in the setting of saphenous venous graft PCI. Many operators now preferentially treat obstructive coronary disease in the native coronary circulation rather than performing PCI on degenerated vein grafts, as it may be a safer, more effective, and more durable option than intervening on degenerated and/or occluded saphenous venous grafts.
Revascularization of Total Occlusions
Recent Occlusions Following ST Elevation Myocardial Infarction
Primary PCI for ST elevation myocardial infarction (STEMI) is the standard of care. However, approximately 1 in 3 patients does not receive prompt reperfusion therapy. Observational studies have described an association between patent infarct-related artery after late presentation and improved outcomes, whereas others have not.
The Occluded Artery Trial (OAT) enrolled 2166 patients with recent MI, ejection fraction less than 50%, and persistently occluded artery who were also beyond the traditional time period for myocardial salvage. The primary endpoint was death, MI, or NYHA class IV heart failure. There was no difference between the PCI and medical therapy groups (17.2% vs 15.6%; HR, 1.16; 95% CI, 0.92–1.45; p = 0.20) respectively with a mean follow-up of 3 years. The lack of benefit was consistent across subgroups. There was a numerical increase in the reinfarction rate in the PCI group that was greatest in the first 30 days but persisted through 5 years. Thus elective PCI does not reduce death or reinfarction in patients 3 to 28 days post MI with a persistently occluded infarct-related artery.
The Total Occlusion Study of Canada (TOSCA-2) was the angiographic substudy of OAT. The primary purpose was to establish the 1-year angiographic patency in OAT, which was approximately 82.7% in the PCI group compared with 25.2% in the medical group ( p < 0.0001). Patency of the infarct-related artery did not translate to an improvement in LV function or remodeling indices. The OAT nuclear substudy also showed no influence of baseline viability and the extent of LV remodeling as measured by end-systolic and diastolic volumes at 1 year. The ACC/AHA STEMI guidelines provide a class III recommendation for PCI of a totally occluded infarct artery greater than 24 hours after STEMI in patients who are asymptomatic and have one- or two-vessel disease if patients are stable and have no evidence of inducible cardiac ischemia.
Chronic Total Occlusions
OAT did not evaluate the efficacy of treating chronic total occlusions (CTOs). CTOs are occluded coronary arteries after thrombolysis with myocardial (TIMI) flow of 0 for 3 months or more. CTOs are found in 15% to 40% of diagnostic heart catheterizations. There has been a significant increase in the number of CTO PCIs being performed worldwide. Expert operators are now achieving procedural success rates approaching 80% to 90% using what is now called the hybrid approach . This approach utilizes both antegrade and retrograde approaches and subintimal techniques such as antegrade dissection reentry and reverse controlled antegrade and retrograde tracking to optimize successful crossing of the occlusion, while attempting to minimize procedural complications.
One of the major barriers to CTO PCI is the very high lesion complexity that is associated with prolonged, costly procedures that have a higher failure rate than standard PCI. The Japanese CTO score is commonly used to assess lesion complexity ( Table 23.3 ). A higher score is associated with higher technical failure rates and poor long-term outcomes.
