Methods

Top Methods Author & Article Info References

We reviewed relevant articles on the management of chronic stable angina published in English-language medical journals. We searched the MEDLINE database for articles published between 1976 and 1996 by using the keywords angioplasty, coronary artery bypass, and angina. To identify additional studies, we reviewed the abstracts presented at recent meetings of the American Heart Association and the American College of Cardiology. We also searched the reference lists of identified articles.

We selected original research articles, meta-analyses, and abstracts that were related to the management of chronic stable angina and that emphasized comparisons of medical therapy, PTCA, and CABG. Randomized trials that compared PTCA with medical therapy (n = 3), compared CABG with medical therapy (n = 7), or compared PTCA with CABG (n = 9) were then grouped. Two meta-analyses compared PTCA with CABG, and one meta-analysis compared CABG with medical therapy; thus, a new meta-analysis was not done. These articles were used to develop our recommendations for the management of chronic stable angina.

Data Synthesis

Three large randomized trials done in the 1970s established a role for CABG in the treatment of multivessel coronary artery disease (Table 1). The Veterans Administration Cooperative Study (VACS), the European Coronary Surgery Study (ECSS), and the Coronary Artery Surgery Study (CASS) compared initial CABG with initial medical therapy [1-3]. Patients with the highest risk (as defined by severity of angina and ischemia, extent of coronary artery disease, and presence of left ventricular dysfunction) had the greatest benefit in terms of survival when they were randomly assigned to a surgical strategy [4]. Surgical revascularization thus became the treatment of choice for multivessel coronary artery disease, particularly in patients with left ventricular dysfunction.

Gruntzig introduced PTCA in 1979 as a catheter-based technique for the treatment of simple, concentric lesions involving a single coronary artery [5]. However, over the past two decades, the use of PTCA has been expanded to more complex lesions and to patients with multivessel coronary artery disease. This has resulted in an explosion in the number of PTCA procedures performed: A greater than 10-fold increase was seen during the past decade, and the number of procedures may have approached 500 000 in the United States in 1996 [6]. Despite this growth, however, data from randomized trials on the role of PTCA in the treatment of coronary artery disease were lacking until about 5 years ago.

Percutaneous Transluminal Coronary Angioplasty Compared with Medical Therapy

Only three randomized trials have compared PTCA with medical therapy, and all three studied patients with single-vessel coronary artery disease. A multicenter Veterans Affairs trial (ACME [Angioplasty Compared to Medicine]) compared the effects of PTCA on angina and exercise tolerance with those of medical therapy in 212 patients with stable single-vessel coronary artery disease [7]. After 6 months of follow-up, more of the patients who received PTCA were free of angina (64% compared with 46%; P < 0.01) (Figure 1). Patients in the PTCA group also had a greater improvement in exercise duration and psychological well-being than did medically treated patients [8]. However, patients who received PTCA had more cardiac procedures and spent more days in the hospital; this increased the initial costs of this strategy. Of note, the mortality rates and rates of myocardial infarction did not differ between the two groups.

View larger version (50K): [in this window] [in a new window]   Figure 1. Percentage of patients who were free of angina after randomization in the ACME (Angioplasty Compared to Medicine) study. White bars indicate medical therapy; striped bars indicate percutaneous transluminal coronary angioplasty. Adapted from Parisi and colleagues [7] with permission.

A second trial, published in abstract form, compared medical therapy with PTCA in 88 patients with asymptomatic single-vessel coronary artery disease [9]. After 2 years of follow-up, no difference was seen in exercise tolerance or the development of symptoms between the two groups. Death, myocardial infarction, and need for revascularization procedures did not differ between the groups.

The Medicine, Angioplasty or Surgery Study (MASS) evaluated the roles of medical therapy, PTCA, and CABG done by using an internal mammary artery graft in 214 patients with stable angina, normal left ventricular function, and stenosis of the proximal left anterior descending coronary artery [10]. The primary end point of cardiac death, myocardial infarction, or refractory angina requiring revascularization was significantly decreased in patients undergoing surgical revascularization compared with those receiving PTCA or medical therapy. In addition, both revascularization strategies resulted in greater relief of symptoms and less inducible ischemia during treadmill testing than did medical therapy. However, no difference was seen in mortality rates, rates of myocardial infarction, or rates of functional class III or IV angina among the three treatment groups.

The results of these three trials are consistent. Patients treated with PTCA had greater symptomatic benefit than medically treated patients, and no change in mortality rates or rates of myocardial infarction was seen. However, the limitations of these studies cloud the conclusions that can be drawn from them. Each involved a small number of patients and thus did not have the power to evaluate meaningful differences in mortality rates or rates of myocardial infarction. Furthermore, clinical practice has changed significantly since the completion of these trials. The role of aggressive lipid-lowering therapy in the primary and secondary prevention of cardiac events, respectively, has been shown in the West of Scotland Coronary Prevention Study and the Scandinavian Simvastatin Survival Study [11, 12]. In addition, there is a growing body of data on the safety and efficacy of intracoronary stents in the treatment of stable and unstable ischemic chest pain syndromes [13].

Despite these limitations, the results of these trials suggest that a strategy of initial pharmacologic therapy is a reasonable approach in patients with stable single-vessel coronary artery disease. Nearly half of medically treated patients become asymptomatic, and no excess in morbidity or mortality is seen. If medical therapy does not relieve angina or if it causes adverse effects, PTCA is usually the preferred alternative. However, a large randomized study comparing PTCA, CABG, and medical therapy in patients with single-vessel coronary artery disease is clearly warranted. This study should use intracoronary stents, internal mammary artery grafts, and modern medical therapy. In addition, "minimally invasive" CABG should be included if its initial promise is sustained.

Coronary Artery Bypass Graft Surgery Compared with Medical Therapy

Shortly after the introduction of CABG, three large, multicenter, randomized trials began enrolling patients to compare a strategy of initial CABG with a strategy of initial medical therapy [1-3]. In VACS, 686 male patients who had had stable angina for 6 months were randomly assigned to treatment [14]. In ECSS, 768 men with multivessel disease who had had angina for more than 3 months were randomly assigned to treatment [15]. Finally, CASS enrolled 780 patients with single- or multivessel coronary artery disease [16]. Few women or elderly patients were included in these trials (Table 1).

Much has been made about the differences among these trials, but the consistencies among them deserve greater emphasis. Yusuf and colleagues [4] performed a meta-analysis of seven randomized trials that compared a strategy of initial surgical therapy with a strategy of initial medical therapy. This analysis involved 2649 patients, of whom 2233 (84%) were from VACS, ECSS, or CASS. Surgical revascularization resulted in substantially lower mortality rates at 5, 7, and 10 years compared with medical treatment (Figure 2). In addition, the greatest absolute benefit for CABG was seen in patients with the highest preoperative risk. Specifically, patients with the most severe symptoms, ischemia, extent of coronary artery disease, and left ventricular dysfunction had the greatest comparative benefit. In each trial, left main coronary artery disease and three-vessel coronary artery disease with moderately impaired left ventricular function were the clearest indications for CABG. However, patients with three-vessel coronary artery disease and normal left ventricular function and patients with significant stenosis in the proximal left anterior descending coronary artery also had benefit from surgical revascularization [4]. In addition, several nonrandomized registry studies emphasized the benefit of surgery over medical therapy in patients with multivessel disease and severe angina pectoris, regardless of left ventricular function [17]. In contrast, in patients with single-vessel disease (excluding a proximal stenosis of the left anterior descending coronary artery) and normal left ventricular function, surgery conferred no benefit with respect to mortality.

View larger version (15K): [in this window] [in a new window]   Figure 2. Cumulative mortality data from the meta-analysis of coronary artery bypass graft surgery (CABG) (1324 patients) compared with medical therapy (1325 patients). Solid line indicates medical therapy; dashed line indicates CABG. Adapted from Yusuf and colleagues [4] with permission.

The relevance of these trials to current practice has been questioned because the trials antedate current surgical and medical management. They compared medical and surgical treatment of coronary artery disease before internal mammary artery grafts, which have clearly improved long-term graft patency, came into widespread use [18]. The importance of risk factor modification and platelet-inhibitor therapy in preventing saphenous vein graft attrition was not fully realized until these studies had been completed [19, 20]. In addition, preoperative evaluation, cardiac anesthesia, and postoperative care have improved since the 1970s.

Medical therapy has also improved substantially since these trials were conducted. Antiplatelet agents were used in only approximately 20% of patients in these trials, but they are used almost universally today. Public awareness of the importance of treating hypercholesterolemia in the secondary prevention of cardiac events has increased considerably [12]. Although ß-blockers were available when these studies were done, their full benefit had not been documented. Finally, the implications of nitrate tolerance on dosing schedules had not yet been established.

Despite the limitations of these studies, their major conclusions can be applied to current clinical practice. Surgical revascularization clearly relieves symptoms and improves survival in high-risk patients (risk is defined by the severity of ischemia, the number of diseased vessels, and the presence of left ventricular dysfunction). This benefit continued for at least 10 years in the meta-analysis, even though 41% of patients who were randomly assigned to medical therapy underwent CABG and almost 7% of the surgical cohort never had CABG [4]. On the other hand, in low-risk patients, surgical revascularization is usually not indicated until medical therapy or PTCA has failed.

Percutaneous Transluminal Coronary Angioplasty Compared with Coronary Artery Bypass Graft Surgery

The trials comparing CABG with medical therapy have established a role for revascularization in the treatment of multivessel coronary artery disease [1-3]. However, the preferred method of revascularization remains controversial and is a major issue in the 1990s. Several randomized studies comparing PTCA with CABG have recently been completed [21-26]. The Randomised Intervention Treatment of Angina trial (RITA), an Argentine trial of PTCA and CABG (ERACI), the German Angioplasty Bypass Surgery Investigation (GABI), the Emory Angioplasty versus Surgery Trial (EAST), the Coronary Angioplasty Bypass Revascularization Investigation (CARBI), and the Bypass Angioplasty Revascularization Investigation (BARI) together enrolled almost 4800 patients (Table 2). The Lausanne Trial and MASS [10, 27] both limited their investigations to patients with isolated proximal left anterior descending coronary artery stenoses.

The patients included in these trials were highly selected, and this must be emphasized because the data should not be extrapolated to other subsets of patients. Only 5.2% of 91 730 patients screened were randomly assigned to a study group. Nearly two thirds of patients were excluded for appropriate angiographic reasons, including left main coronary artery disease, chronic total occlusion, diffuse disease, or inability to achieve "functional" complete revascularization. In addition, most of the patients had well-preserved left ventricular systolic function with mean ejection fractions ranging from 0.56 to 0.63. Patients with left main coronary artery disease, a recent myocardial infarction, or previous PTCA or CABG were excluded from all of the trials. As a result, patients who had previously been shown to benefit most from surgical revascularization were often not included.

Despite differences in study design as well as inclusion criteria, the results of these six trials were remarkably consistent. Two meta-analyses that include each of the above trials except BARI have been completed [28, 29]. No differences in mortality or in the combined end point of death and nonfatal myocardial infarction were seen between patients treated with PTCA and those treated with CABG (Figure 3). However, the meta-analyses showed that a second revascularization procedure was much more likely to be required in the first year after PTCA than in the first year after CABG (33.7% compared with 3.3%) [28]. In the PTCA-treated patients, repeated revascularization procedures comprised almost equal numbers of PTCA and CABG procedures (Figure 4). In addition, patients treated with surgical revascularization were more likely to be free of angina after 1 year, but this difference had decreased by the end of the third year. As expected, initial costs were much lower in patients undergoing PTCA. However, the need for repeated revascularization procedures and hospitalizations in the angioplasty group contributed to an increase in post-discharge costs; thus, the gap between the overall costs of the two strategies narrowed after 5 years (Figure 5) [24, 30].

View larger version (19K): [in this window] [in a new window]   Figure 3. Overall combined risk for death and nonfatal myocardial infarction in five randomized trials comparing percutaneous transluminal coronary angioplasty (PTCA) with coronary artery bypass graft (CABG) surgery. CABRI = Coronary Angioplasty Bypass Revascularization Investigation; EAST = Emory Angioplasty versus Surgery Trial; ERACI = Estudio Randomizado Argentino de Angioplastia vs Cirugia; GABI = German Angioplasty Bypass Surgery Investigation; OR = odds ratio; RITA = Randomised Intervention Treatment of Angina. Reproduced from Sim and coworkers [29] with permission.

View larger version (36K): [in this window] [in a new window]   Figure 4. Risk for repeated percutaneous transluminal coronary angioplasty (PTCA) (white bars), coronary artery bypass graft (CABG) surgery (striped bars), or both (black bars) after randomization in six trials comparing PTCA with CABG. BARI = Bypass Angioplasty Revascularization Investigation; CABRI = Coronary Angioplasty Bypass Revascularization Investigation; EAST = Emory Angioplasty versus Surgery Trial; ERACI = Estudio Randomizado Argentino de Angioplastia vs Cirugia; GABI = German Angioplasty Bypass Surgery Investigation; RITA = Randomised Intervention Treatment of Angina.

View larger version (15K): [in this window] [in a new window]   Figure 5. Cumulative costs during follow-up of patients randomly assigned to percutaneous transluminal coronary angioplasty (solid line) or coronary artery bypass graft surgery (dashed line) in the Bypass Angioplasty Revascularization Investigation. Adapted from Hlatky and colleagues [30] with permission.

The results of BARI, the only trial powered to detect differences in mortality, were consistent with those of previous studies [26]. Of 25 200 patients screened, 1829 (7.3%) patients with multivessel coronary artery disease were randomly assigned to study groups. Forty-one percent had three-vessel coronary artery disease and 64% had unstable angina; the mean left ventricular ejection fraction was 0.57. No statistically significant difference in 5-year survival was seen between the CABG and PTCA groups (89.3% compared with 86.3%; P = 0.19). In addition, no difference was seen in the rate of survival free of Q-wave myocardial infarction after 5 years. A repeated revascularization procedure was required in 8% of patients assigned to CABG and 54% of patients assigned to PTCA (subsequent CABG in 20% of patients, subsequent PTCA in 23%, and both in 11%).

A subset analysis of 353 patients in BARI who were being treated for diabetes mellitus with insulin or an oral hypoglycemic agent has provided new and controversial information (Figure 6). In this cohort, the 5-year survival rate was significantly better with CABG than with PTCA (80.6% compared with 65.5%; P = 0.003). These data are supported by a subgroup analysis from CABRI that showed 2-year mortality rates of 15.6% in PTCA-treated diabetic patients and 3.5% in CABG-treated diabetic patients (Bertrand M. Long-term follow-up of European revascularization trials. Presented at the 68th Scientific Sessions, American Heart Association, Anaheim, California, 16 November 1995). In addition, Faxon and colleagues [31] compared 280 diabetic patients with 1833 nondiabetic patients from the PTCA registry at the National Heart, Lung, and Blood Institute [31]. The diabetic patients included more women and had an increased incidence of hypertension, congestive heart failure, and severe concomitant noncardiac disease. Just as in BARI and CABRI, late mortality was increased in diabetic patients. The 8-year mortality rate was 31% in diabetic patients and 15% in nondiabetic patients (P < 0.001). In addition, diabetic patients had an increased rate of myocardial infarction at 8 years (30% compared with 18%; P < 0.001). This may be partly explained by the angiographic results in the diabetic patients, which showed more multi-vessel disease, significant lesions, and distal lesions. Several surgical series have also reported an increased late mortality rate in diabetic patients compared with nondiabetic patients [26, 32].

View larger version (13K): [in this window] [in a new window]   Figure 6. Five-year survival rates comparing patients being treated for diabetes mellitus with all other patients in the Bypass Angioplasty Revascularization Investigation. Diabetic patients assigned to coronary artery bypass surgery (CABG) are indicated by the heavy solid line; all other patients assigned to CABG are indicated by the light solid line; diabetic patients assigned to percutaneous transluminal coronary angioplasty (PTCA) are indicated by the heavy dashed line; and all other patients assigned to PTCA are indicated by the light dashed line. Adapted from [26] with permission.

These trials provide objective data, obtained over a 1- to 5-year period, that strongly suggest that in selected, eligible patients with multivessel disease, PTCA is a reasonable initial strategy. This approach is not accompanied by any increase in mortality rates or rates of myocardial infarction; however, the price to be paid is a 32% to 54% chance of a repeated revascularization procedure. Approximately half of these procedures will be bypass surgical procedures; the rest will be repeated angioplasty. The late follow-up of these trials will be of interest because restenosis after PTCA is a relatively early phenomenon, whereas graft attrition after CABG produces its major effect several years after the initial operation.

Crucial to the understanding of these trials and their effect on clinical practice is an appreciation of the characteristics of the patient populations entered into these trials. Most had two-vessel disease, well-preserved left ventricular function, and "suitable" coronary anatomy. The definition of the latter will change over time as new information about intracoronary stents and other techniques widens or expands the pool of patients in whom transcatheter technologies may be used. We are also likely to see the "complementary" use of both methods of revascularization in the future, whereas the current trend is to consider them competitive. Over time, PTCA will be increasingly used in patients who have previously undergone CABG, and patients initially treated with PTCA will often need subsequent CABG. Joint revascularization procedures done by using transcatheter technologies in some vessels and "minimally invasive" surgery in others is an attractive and increasingly feasible option for achieving "complete" revascularization in some patients.

An accumulating body of evidence points toward CABG as the preferable therapy in treated diabetic patients. Nonetheless, the data are preliminary and do not justify broad extrapolation to the management of all diabetic patients with symptomatic multivessel coronary artery disease. We need more information about this subgroup of patients and the extent to which the apparent superiority of surgery over PTCA reflects a higher proportion of patients with triple-vessel disease, diffuse disease, and abnormal left ventricular function. The combination of diffuse multivessel disease and left ventricular dysfunction categorizes the group that would be expected to benefit from surgery, with its potential for complete revascularization [33]. It is relevant, however, that the lack of any difference between PTCA and CABG in BARI seemed to be independent of the extent of coronary disease and left ventricular function.

Therapeutic Strategies

Randomized trials define broad therapeutic principles that need to be placed in context when they are applied to a specific patient. Extrapolation of data to different subsets of patients, however, must be done with great caution. For example, in the trials comparing CABG with medical therapy, only 3% of patients were women and patients older than 65 years of age were often excluded. Nonetheless, the trials have yielded a consistent body of objective evidence that provides us with at least a logical basis for determining the "best" therapy for the individual patient.

In this review, patients with chronic stable angina are divided into low-, moderate-, and high-risk groups. Low-risk patients have single-vessel coronary artery disease, no or mild ischemia, and normal left ventricular function. High-risk patients have multivessel or left main coronary artery disease, impaired left ventricular function, and moderate to severe ischemia. Moderate-risk patients usually have two-vessel coronary artery disease and normal ventricular function. In defining risk, the best way to quantify the extent of ischemia is by using a nuclear or echocardiographic stress test (exercise or pharmacologic).

Single-Vessel Coronary Artery Disease

No trial to date has shown a definitive survival advantage for revascularization compared with medical therapy in patients with single-vessel disease and chronic stable angina. Therefore, in patients with single-vessel disease and no or mild ischemia, it is logical to use initial medical therapy. In ACME [7], nearly half of medically treated patients were free of angina at 6 months. The use of PTCA is an excellent strategy for patients in whom medical therapy fails.

Among patients with single-vessel coronary artery disease and moderate to severe ischemia or angina, PTCA is currently the preferred therapy. However, definitive evidence based on randomized trials is unavailable. Nonetheless, a particularly strong case can be made for PTCA in preference to medical therapy in patients with single-vessel disease involving the proximal left anterior descending coronary artery. Trials show improved quality of life when PTCA is compared with medical therapy [8]. In addition, information from the Duke University database suggests that a subset of patients with single-vessel disease involving the proximal left anterior descending coronary artery have a better 5-year survival rate with PTCA than with medical therapy [34]. This important issue can be resolved only by a large trial that takes into account new catheter technologies and "optimal" medical therapy, which must include aggressive lipid-lowering therapy using both diet and drugs.

Trials comparing CABG with medical therapy in this group of patients with single-vessel disease have not shown a survival advantage with CABG; however, the results of MASS, which was confined to patients with proximal left anterior descending coronary artery stenosis, are provocative [10]. In that study, CABG resulted in a significant reduction in angina and ischemia compared with PTCA and medical therapy. The evolving technique of "minimally invasive" bypass surgery is intriguing; if it lives up to its initial promise, a larger trial comparing it with medical therapy and "state-of-the-art" trans-catheter technologies for proximal left anterior descending coronary artery disease will probably be needed in the near future.

Multivessel Coronary Artery Disease

In most patients with multivessel coronary artery disease, revascularization is usually indicated. However, in selected patients with mild ischemic symptoms and normal left ventricular function, an initial trial of medical therapy is reasonable. The goal is to relieve angina and reduce objective measures of ischemia. If this is successful, continued medical therapy is reasonable; coronary revascularization is indicated for patients in whom medical therapy fails.

It should again be emphasized that crucial components of both therapeutic strategies include aggressive lipid-lowering therapy, antiplatelet therapy, and modification of risk factors. Recent trials have conclusively shown the benefits of lowering low-density lipoprotein and total cholesterol levels in patients with symptomatic coronary artery disease, patients with risk factors for coronary artery disease, and patients who have undergone CABG [11, 12, 35]. In addition, aspirin and ß-blockers have been shown to decrease the incidence of subsequent cardiac events in patients with known coronary artery disease. In contrast, percutaneous and surgical revascularization have not been shown to decrease rates of subsequent myocardial infarction. As a result, medical therapy (which is often thought to be less definitive than revascularization) may actually be the more definitive treatment, whereas revascularization (which is often considered more definitive) may be useful only to relieve symptoms.

When revascularization is considered for the treatment of multivessel coronary artery disease, the selection of PTCA or CABG depends on the coronary anatomy, left ventricular function, need for complete revascularization, and patient preference. In high-risk patients who have left main coronary artery disease or three-vessel coronary artery disease with impaired left ventricular function, current data support surgical revascularization as the treatment of choice to achieve complete revascularization [33]. In low- and moderate-risk patients with multivessel coronary artery disease and normal or mildly impaired left ventricular function, PTCA and CABG are both reasonable options when the coronary anatomy is suitable for either technique. As long as the patient understands the likelihood of the need for a repeated revascularization procedure within 1 to 3 years, using initial PTCA is a rational therapeutic strategy. However, it seems that for most patients, the price of undergoing initial PTCA is not an increase in mortality or late myocardial infarction.

In summary, a series of clinical trials completed in the past 20 years have allowed us to design a logical algorithm for the treatment of patients with coronary artery disease. In low-risk patients with single-vessel coronary artery disease and normal left ventricular function, initial medical therapy consisting of anti-anginal agents, aspirin, and modification of risk factors should be attempted. In moderate-risk patients (most of whom have two-vessel coronary artery disease and normal left ventricular function), PTCA and CABG result in similar mortality rates and rates of nonfatal myocardial infarction. However, CABG provides better relief of angina and results in fewer repeated revascularization procedures. The choice of revascularization method depends on coronary anatomy and patient preference, although initial PTCA is usually preferred. In high-risk patients with left main coronary artery disease or three-vessel disease with left ventricular dysfunction, CABG is the most reasonable strategy. Finally, aggressive lipid-lowering therapy, antiplatelet therapy, and modification of risk factors are critical to the success of each therapeutic option. This strategy has clearly been shown to decrease the rate of cardiac events in medically treated patients and after revascularization.

Dr. Gersh: Georgetown University Medical Center, Division of Cardiology, 5 PHC, 3800 Reservoir Road, NW, Washington, DC 20007.