Revascularization in patients with stable coronary artery disease: Coronary artery bypass graft surgery versus percutaneous coronary intervention

INTRODUCTION — Patients with stable coronary artery disease should be assessed periodically to determine whether medical therapy or medical therapy with revascularization is a more appropriate strategy. (See "Chronic coronary syndrome: Indications for revascularization", section on 'Summary and recommendations'.)

This topic will present our approach to choosing between percutaneous coronary intervention (PCI) with stenting and coronary artery bypass graft surgery (CABG), the two principal options for revascularization. The recommended approach to choosing between CABG and PCI in patients with stable one, two, and three vessel disease who are candidates for intervention and who have an ejection fraction greater than about 35 to 40 percent will be reviewed here. Patients with ischemic cardiomyopathy, and who typically have a left ventricular ejection fraction ≤35 to 40 percent, are discussed separately. (See "Treatment of ischemic cardiomyopathy", section on 'Approach to revascularization'.)

The approach to revascularization in patients with left main disease, acute coronary syndromes, or diabetes is discussed in separate topic reviews. (See "Left main coronary artery disease" and "Acute ST-elevation myocardial infarction: Selecting a reperfusion strategy" and "Non-ST-elevation acute coronary syndromes: Selecting an approach to revascularization" and "Coronary artery revascularization in stable patients with diabetes mellitus".)

Regardless of which method of revascularization is used, aggressive risk-factor modification is necessary in all patients. (See "Prevention of cardiovascular disease events in those with established disease (secondary prevention) or at very high risk" and "Cardiac rehabilitation: Indications, efficacy, and safety in patients with coronary heart disease".)

INDICATIONS FOR REVASCULARIZATION — The decision to proceed with revascularization with either coronary artery bypass graft surgery or percutaneous coronary intervention (PCI), as opposed to continuing medical therapy, is made in three groups of stable patients:

●Patients with activity-limiting symptoms despite maximum medical therapy.

●Active patients who want PCI for improved quality of life compared to medical therapy, such as those who are not tolerating medical therapy well, or who want to increase their activity level.

●Patients with anatomy for which revascularization has a proven survival benefit such as significant left main coronary artery disease (greater than 50 percent luminal narrowing) or multivessel coronary artery disease with a reduction of left ventricular ejection fraction and a large area of potentially ischemic myocardium.

The supportive data are presented separately. (See "Chronic coronary syndrome: Indications for revascularization".)

OUR APPROACH — We generally avoid ad hoc PCI in persons with complex or multivessel disease, favoring a multidisciplinary team-based approach [1]. Our approach for patients with left main coronary artery disease is discussed separately. (See "Left main coronary artery disease".)

For coronary artery disease patients (excluding those with left main coronary artery disease) who need revascularization, the following is our approach to deciding between PCI and CABG:

●Prior to choosing PCI or CABG, the patient should be informed about periprocedural risks of death, stroke, and the need for repeat revascularization.

●In patients with single-vessel disease, we usually perform PCI with drug-eluting stents. (See 'Single vessel disease' below.)

●In patients with two-vessel disease involving the right and circumflex coronary arteries, we usually perform PCI with drug-eluting stents. For patients with disease of the left anterior descending and either right or circumflex coronary arteries, we prefer CABG in those with diabetes or those with a large amount of myocardium supplied by the diseased vessels. (See 'Outcomes based on lesion severity' below.)

●For patients with three-vessel disease, we prefer CABG. For these individuals with a low SYNTAX score and no diabetes, PCI is reasonable. (See 'Outcomes based on lesion severity' below.)

●For patients with two- or three-vessel disease in whom complete revascularization cannot be accomplished by PCI, we prefer CABG.

●In patients who refuse CABG or for whom it is not an option based on comorbidities or poor functional status, PCI may be reasonable to improve the quality of life irrespective of disease complexity.

●The STS mortality score can be useful to identify which patients will be at high risk for mortality with CABG. If a patient has an STS score >8, CABG is usually avoided regardless of anatomy. (See "Estimating the risk of valvular procedures", section on 'Commonly used models'.)

●In some patients, after considering these factors, it occasionally is not clear whether PCI or CABG should be preferred. We believe it is reasonable to use the SYNTAX score to help guide decision making. However, no studies have shown that patients managed using this score have better outcomes than those who are not. (See 'Outcomes based on lesion severity' below.)

LEFT MAIN CORONARY ARTERY DISEASE — The approach to these patients is discussed separately. (See "Left main coronary artery disease".)

SINGLE VESSEL DISEASE — In general, the left ventricular myocardium supplied by the left anterior descending (LAD) coronary artery is greater than that for the right or circumflex coronary artery. Our approach to patients with LAD coronary artery disease is discussed separately. In general, we prefer PCI to CABG in these patients. (See "Management of significant proximal left anterior descending coronary artery disease".)

For most patients with single vessel coronary artery disease, defined as at least one stenosis with ≥70 percent diameter stenosis or a 50 to 70 percent diameter stenosis with a positive fractional flow reserve (<0.80) in either the proximal or mid portion of the artery, we prefer percutaneous coronary intervention (PCI) to coronary artery bypass graft surgery (CABG). (See 'Indications for revascularization' above.)

There are little data on which to formulate recommendations for patients with single vessel disease involving the right or circumflex coronary artery. Our preference for PCI is based on the absence of evidence to suggest outcomes are better with CABG. Also, CABG is associated with a higher rate of early stroke. Indirect evidence to support this recommendation comes from our preference for PCI in patients with single vessel disease of the LAD coronary artery.

Studies of morbidity and mortality associated with disease in each of the three coronary arteries are over 30 years old and therefore are not directly applicable to patients on current optimal medical therapy. Overall risk seems to be associated with the amount of territory in jeopardy and this is in line with the findings in population studies where the morbidity and mortality are highest for proximal LAD lesions, lowest for right coronary artery disease, and intermediate for non-proximal LAD and circumflex disease. However, in the individual patient, the right or circumflex coronary artery may be associated with a large left ventricular territory.

Patients for whom it is reasonable to prefer CABG include those with a large amount of potentially ischemic myocardium, those who are not candidates for long-term dual antiplatelet therapy, or those who have failed prior PCI. Patients with ischemic cardiac arrest and anatomically complex disease may also be considered for CABG.

MULTIVESSEL DISEASE — For many patients with multivessel coronary artery disease, relatively well-preserved left ventricular systolic function, low complexity coronary anatomy, and no diabetes, we have a weak preference for CABG, as opposed to PCI. For these patients with complex anatomy (see 'Outcomes based on lesion severity' below) or diabetes, we have a strong preference for CABG.

The choice between CABG and PCI in patients with multivessel (two- and three-vessel) disease is influenced by a number of factors, including the number of vessels involved, the amount of myocardium supplied by the affected vessels, the anatomic complexity of the lesions requiring revascularization, likelihood of complete revascularization, patient comorbidities such as diabetes, and patient preference. At the time of decision making, the patient should be informed about the relative risks of death, stroke, and the need for repeat revascularization.

When the coronary anatomy is equally well suited for both PCI and CABG, the decision should be made jointly by the patient, referring physician, cardiologist, and a heart team including a cardiothoracic surgeon. Attention must be paid to the patient's expectations and willingness to undergo repeat procedures, if necessary. Some patients remain fearful of open heart surgery with increased periprocedural morbidity and mortality and the ensuing 4- to 12-week convalescence, while others are unwilling to undergo repeat interventions if necessary. Other patients are strongly stroke averse and are willing to accept an increased risk of revascularization after PCI, in order to decrease the increased risk of stroke with CABG. The importance of clinical decision making after consideration of all of these factors should not be underestimated and cannot be easily replaced even by the best evidence from observational or randomized clinical trials.

Multivessel PCI should not be undertaken by low volume operators who have limited experience with complex coronary lesions. Furthermore, in high-risk situations, multivessel PCI may require two revascularization procedures performed at separate times (staged revascularization). Thus, the patient must be willing to return to the catheterization laboratory on several occasions.

Outcomes with coronary artery bypass graft surgery and percutaneous coronary intervention — The risks of death, stroke, and the need for repeat revascularization are the most important outcomes in patients who require coronary revascularization with either CABG or PCI with drug eluting stents (DES). For most non-elderly patients, the most important deciding factors between the two are the relative rates of death and stroke. Individual studies generally have combined end points, and no single study was adequately powered to assess the relative risks of mortality, or stroke. Pooling of data from observational studies or randomized trials attempts to address this issue. Based on the evidence presented below, we conclude that the risk of stroke is small at 30 days and slightly higher with CABG. The risk of death is higher at five years with PCI.

The SYNTAX trial compared outcomes after either CABG or a first generation DES in patients with left main and three-vessel coronary artery disease (CAD) (movie 1). In SYNTAX, 1800 patients with three-vessel or left main CAD (approximately 60 and 40 percent, respectively) were randomly assigned to CABG or PCI with the paclitaxel-eluting stent [2]. All patients were eligible for either procedure and were treated with the intention of complete revascularization. (See 'SYNTAX I and II scores' below.)

The following findings were noted:

●After 12 months of follow-up, the composite primary endpoint (death from any cause, stroke, myocardial infarction [MI], or repeat revascularization) was significantly higher in the PCI group (17.8 versus 12.4 percent). This result was driven primarily by more frequent revascularization with PCI (13.5 versus 5.9 percent).

●Although the secondary end point of the rate of death, stroke, or MI was comparable (7.6 for PCI versus 7.7 for CABG), the stroke rate was significantly lower in the PCI group (0.6 versus 2.2 percent).

●More complete revascularization was achieved with CABG (63 versus 57 percent).

●After three and five years, the composite primary end point remained significantly higher with PCI (28 versus 20.2 percent and 37.3 versus 26.9 percent, respectively). These results were attributable primarily to higher rates of repeat revascularization and MI [3,4].

●After three years, in comparing PCI with CABG, there was no significant difference in the rates of the safety end points of death/stroke/MI (14.1 versus 12.0 percent), stroke (2.0 versus 3.4 percent), or all-cause death (8.6 versus 6.7 percent). After five years, there was no difference in the rates of all-cause death (13.9 versus 11.4 percent) or stroke (2.4 versus 3.7 percent) [3,4].

●At 10 years, there was no significant difference between PCI and CABG in all-cause death (27 versus 24 percent, respectively; hazard ratio [HR] 1.19, 95% CI 0.99-1.43) [5]. However, among patients with three-vessel disease, the rate was higher with PCI (28 versus 21 percent; HR 1.41, 95% CI 1.10-1.80).

The results of the SYNTAX trial need to be interpreted in the context of its limitations, including male predominance (78 percent), suboptimal use of medical therapy in the CABG cohort, and use of a first generation DES, which has a higher rate of angiographic and clinical restenosis than second generation DES. Newer generation stents, including everolimus-eluting stents (EES), reduce the risks of death, MI, and stent thrombosis compared with bare metal and first generation stents. (See "Intracoronary stents: Stent types", section on 'Summary and recommendations'.)

The BEST non-inferiority trial, which was terminated early due to slow enrollment, randomly assigned 880 patients with two- and three-vessel CAD to PCI with EES or CABG; patients with left main coronary disease were excluded [6]. The primary end point (a composite of death, MI, or target-vessel revascularization at two years) occurred more often in the PCI group (11.0 versus 7.9 percent; absolute risk difference 3.1 percentage points, 95% CI -0.8 to 6.9 [criteria for non-inferiority not met]). At 4.6 years, the primary end point occurred more often with PCI (15.2 versus 10.6 percent; HR 1.47, 95% CI 1.01-2.13). No differences were seen in the occurrence of a composite safety end point of death, MI, or stroke, but the rates of any repeat revascularization and spontaneous MI were significantly higher after PCI.

In an observational registry study of 34,819 patients with two- and three-vessel CAD (left main disease patients were excluded), 9223 patients who underwent multivessel PCI with EES were compared with 9223 patients who received CABG [7]. There was a similar risk of death at 2.9 years (3.1 versus 2.9 percent per year; HR 1.04, 95% CI 0.93-1.17). Significantly higher risks of MI (1.9 versus 1.1 percent per year) and repeat revascularization (7.2 versus 3.1 percent) and a lower risk of stroke (0.7 versus 1.0 percent per year) were seen with EES. This difference in the risk of MI was seen only among patients with incomplete revascularization. (See 'Our approach' above.)

We believe that the best available evidence regarding the individual outcomes of death and stroke come from a 2018 pooled analysis of individual patient data from 11 randomized trials that compared PCI with CABG. In this meta-analysis, subgroups were prespecified. Five-year, all-cause mortality among patients with multivessel disease (n = 7040) was higher with PCI than CABG (11.5 versus 8.9 percent; HR 1.28, 95% 1.09-1.49) [8]. The risk of death was significantly higher with PCI in those with diabetes (15.5 versus 10.0 percent; HR 1.48, 95% CI 1.19-1.84) but not in those without (8.7 versus 8.0 percent; HR 1.08, 95% CI 0.86-1.36). A higher SYNTAX score had a significant effect on the difference between PCI and CABG in patients with multivessel disease. The findings for the risk of stroke were presented in a separate publication [9]. Among 11,158 patients in these trials, the rate of stroke at 30 days was 0.4 percent after PCI and 1.1 percent after CABG (HR 0.33, 95% CI 0.20-0.53). At five years, stroke remained significantly lower after PCI (2.6 versus 3.2 percent). In the 7040 patients with multivessel disease, the rate of stroke was significantly lower after PCI at 30 days (0.4 versus 1.2 percent [HR 0.36, 95% CI 0.20-0.65]) and after five years (2.7 versus 3.6 percent; HR 0.75, 95 CI 0.56-0.99).

Most [10-14], but not all [15], studies of patients enrolled in registries have concluded that patients treated with PCI with DES have a higher long-term mortality compared with CABG.

As noted in the studies above, both CABG and PCI are associated with a small but important risk of stroke and many of these studies suggest a higher risk in patients receiving CABG.

Outcomes based on lesion severity — In formulating recommendations for revascularization, we have taken lesion complexity into account (see 'Our approach' above).

Two methods to assess different aspects of lesion severity are the SYNTAX score, a semi-quantitative tool based on the results of coronary angiography [16,17], and a hemodynamic measurement often done with angiography via fractional flow reserve (FFR) or instantaneous wave-free ratio (iFR) [18]. (See "Clinical use of coronary artery pressure flow measurements".)

Fractional flow reserve — FFR is primarily used to avoid unnecessary stenting at the time of angiography/PCI for coronary lesions that appear visually significant but that are not hemodynamically significant. (See "Clinical use of coronary artery pressure flow measurements".)

In the FAME 3 trial, 1500 patients with three-vessel coronary artery disease were randomly assigned to CABG or FFR-guided PCI [19]. Patients were followed for development of major adverse cardiac or cerebrovascular event (MACCE), defined by any death, myocardial infarction, stroke, or repeat revascularization. At one year following revascularization, the incidence of MACCE was higher in those assigned to FFR-guided PCI compared with CABG (10.6 versus 6.9 percent; HR 1.5, 95% CI 1.1-2.1). The secondary endpoint of death, myocardial infarction, or stroke at one year was also numerically higher for PCI (7.3 versus 5.2 percent; HR 1.4, 95% CI 0.9-2.1). However, procedural complications and rehospitalizations were higher in the CABG versus PCI groups. Major bleeding (3.8 versus 1.6 percent), arrhythmia (14.1 versus 2.4), acute kidney injury (0.9 versus 0.1 percent), and rehospitalization within 30 days (10.2 versus 5.5 percent) were higher in the CABG group. The FAME 3 trial results were in keeping with and extended findings of prior trials that showed less MACCE with CABG versus PCI for severe three-vessel coronary artery disease [2,6].

SYNTAX I and II scores — The SYNTAX score was developed during the SYNTAX trial to classify patients according to the severity of their CAD [20,21]. The score incorporates characteristics of lesion complexity, lesion location, and the number of lesions. Lower scores predict better PCI outcomes. Other aspects of the SYNTAX trial are discussed above. (see 'Outcomes with coronary artery bypass graft surgery and percutaneous coronary intervention' above)

In the SYNTAX trial, patients with lower SYNTAX scores had lower event rates than those with higher scores at 12 months [2]. In addition, among patients with low (0 to 22) and intermediate (23 to 32) SYNTAX scores, the clinical outcomes were comparable with PCI and CABG (13.6 versus 14.7 and 16.7 versus 12.0, respectively), whereas in those with a high score (≥33), outcomes were better with CABG (23.4 versus 10.9 percent, respectively) at 12 months.

After five years [4]:

●For patients with a low SYNTAX score, there was no difference in the composite primary end point (32.1 and 28.6 percent).

●For patients with an intermediate SYNTAX score, outcomes between PCI and CABG were similar if unprotected left main disease was present (32.7 and 32.3 percent, respectively). In patients with three-vessel disease, the primary composite outcome was higher with PCI (37.9 versus 22.6 percent; p = 0.0008).

●For patients with a high SYNTAX score, the primary composite outcome occurred more frequently in patients with and without unprotected left main disease (46.5 versus 29.7 [p = 0.003] and 41.9 versus 24.1 percent [p = 0.0005], respectively).

●The end points of all-cause death and MI, which were components of the primary composite outcome, were lower with CABG (11.4 versus 19.2 percent; HR 1.84, 95% CI 1.19-2.83 and 3.9 versus 10.1 percent; HR 2.57; 95% CI 1.31-5.06, respectively).

Subgroup analyses based on the SYNTAX score, a measure of lesion and disease complexity, showed the benefit of CABG was greatest among patients with intermediate (22 to 32) or complex (SYNTAX score ≥33) disease. During longer-term follow-up (three and five years), CABG was associated with a mortality benefit in the complex SYNTAX score subgroup and lower major adverse cardiovascular and cerebrovascular event for both the intermediate and complex SYNTAX score subgroups. The difference in the lowest SYNTAX score subgroup (≤22), however, was not significant and was related to a higher risk for repeat procedures for DES. Of note, more patients in the CABG group withdrew consent or were lost to follow-up, which should be considered in the overall interpretation of data. Other observational data also suggest that disease complexity is an important effect modifier, with similar outcomes for CABG and PCI in those with low-complex disease, and superiority for CABG in patients with more-complex disease.

Since the overall findings of the SYNTAX trial (which did not enroll patients with two vessel disease) favor CABG and as analyses by SYNTAX score and by individual end points were not prespecified in the SYNTAX trial, the finding of comparable outcomes in the subgroup with a low SYNTAX score must be interpreted with some caution. We are not confident that we know the optimal strategy in such patients. We suggest CABG for many of these patients, but some of our authors and reviewers frequently perform PCI, after discussing the pros and cons of both approaches with the patient. The risk of surgery, patient preference, technical difficulty for PCI, and compliance with dual antiplatelet therapy become especially important factors in this group.

One of the criticisms of the SYNTAX score (later referred to as SYNTAX I) has been that it does not include any clinical variables. In an attempt to improve the predictive capability of the SYNTAX I score, the SYNTAX score II was developed using baseline patient data in the SYNTAX I trial [22]. It contains eight predicators: the anatomical SYNTAX I score, age, creatinine clearance, left ventricular ejection fraction, presence of unprotected left main CAD, peripheral artery disease, female sex, and chronic obstructive pulmonary disease. The SYNTAX II score was externally validated using a large all comers (DELTA) registry of patients with three-vessel or complex CAD. The SYNTAX score II was better able to predict long-term mortality in patients with complex CAD than the angiographic SYNTAX I score. While promising, further validation of this new score is needed before we recommend its use.

Hybrid coronary revascularization — Hybrid coronary artery revascularization (HCR) refers to complete or near complete revascularization using the combination of single vessel CABG with the left internal mammary artery (LIMA) placed to the left anterior descending coronary artery (LAD) and PCI of other significant coronary lesions [23]. While it has achieved some measure of popularity, HCR has not been evaluated in randomized trials comparing it with PCI or CABG. Until evidence from randomized trials supporting its use is available, we believe it is a reasonable choice at centers with expertise.

In HCR, the LIMA is anastomosed to the LAD using minimally invasive direct CABG (MIDCAB); MIDCAB is applied to procedures that use alternative incisions to standard median sternotomy. Hybrid procedures are either staged or performed at the same time in a specially designed operative suite [24]. (See "Off-pump and minimally invasive direct coronary artery bypass graft surgery: Clinical use", section on 'MIDCAB'.)

The rationale for performing HCR, as opposed to complete revascularization with either CABG or PCI, is that HCR offers the possibility of decreasing the risk of stroke associated with CABG and the risk of vein graft failure while maintaining excellent long-term patency with a LIMA graft to the LAD. (See "Coronary artery bypass graft surgery: Long-term clinical outcomes", section on 'Vein graft failure'.)

Most observational studies comparing HCR with PCI have been small [25-28]. These small studies suggest that outcomes may be better with HCR, particularly with regard to durability, symptom relief, and survival compared with stenting. The largest observational study compared 200 patients who underwent HCR and 98 patients who underwent multivessel PCI [29]. Staged procedures were expected to be completed within six weeks of the first procedure. The rate of primary outcome of major adverse cardiac and cerebrovascular events (ie, death, stroke, MI, and repeat revascularization) within 12 months of the procedure was similar between the two groups after adjustment for baseline risk (0.142 versus 0.119 percent, respectively; hazard ratio 1.063; p = 0.80).

PATIENTS WITH LEFT VENTRICULAR DYSFUNCTION — Optimal therapy for patients with multivessel coronary disease with left ventricular dysfunction remains controversial and was not formally addressed in the large randomized trials [30]. The decision to proceed with either CABG or PCI in these patients is affected by the same variables used in decision making for the broad population of patients who need revascularization. (See 'Multivessel disease' above.)

PATIENTS WITH PREVIOUS CABG — Studies evaluating percutaneous coronary intervention versus repeat coronary artery bypass graft surgery (CABG) in patients with prior CABG are limited. The factors that influence the choice between these procedures are discussed in detail elsewhere. (See "Late recurrent angina pectoris after coronary artery bypass graft surgery", section on 'Management'.)

RECOMMENDATIONS OF OTHERS — Our recommendations regarding the choice between coronary artery bypass graft surgery (CABG) and percutaneous coronary intervention (PCI) are generally consistent with those found in the 2014 update of the American College of Cardiology American Heart Association guideline for the diagnosis and management of patients with stable ischemic heart disease [31].

Guidelines on myocardial revascularization were published by the European Society of Cardiology and the European Association for Cardio-Thoracic Surgery in 2018 [32]. The following summarizes recommendations for which there is a preference for one procedure or the other in patients with stable angina who have lesions suitable for both procedures and a low predicted surgical mortality:

●PCI is favored to CABG in patients with one or two vessel disease not involving the proximal left anterior descending coronary artery.

●CABG is favored to PCI in patients with three vessel disease and SYNTAX score ≥23; in patients with lower SYNTAX scores, PCI is considered a reasonable alternative.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Coronary artery bypass graft surgery" and "Society guideline links: Percutaneous coronary intervention" and "Society guideline links: Chronic coronary syndrome".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, “The Basics” and “Beyond the Basics.” The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on “patient info” and the keyword(s) of interest.)

●Beyond the Basics topics (see "Patient education: Angina treatment — medical versus interventional therapy (Beyond the Basics)" and "Patient education: Stenting for the heart (Beyond the Basics)" and "Patient education: Coronary artery bypass graft surgery (Beyond the Basics)")

●Basics topic (see "Patient education: Coronary artery bypass graft surgery (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Indications for revascularization – Either coronary artery bypass graft surgery (CABG) or percutaneous coronary intervention (PCI) is indicated in the following groups of patients with stable angina pectoris (see 'Indications for revascularization' above):

•Patients with activity-limiting symptoms despite optimal medical therapy

•Active patients who prefer PCI for improved quality of life compared to medical therapy

•Those with anatomy for which revascularization has a proven survival benefit.

●Stable angina and left ventricular ejection fraction ≥35 percent – Our recommendations for these patients assume that local expertise is satisfactory for both PCI and CABG:

•For patients with single vessel disease involving the right or circumflex coronary arteries, we suggest PCI rather than CABG (Grade 2C).

Our approach to revascularization of the left anterior descending artery is presented separately. (See "Management of significant proximal left anterior descending coronary artery disease", section on 'Summary and recommendations'.)

•For patients with multivessel coronary artery disease, relatively well preserved left ventricular systolic function, low complexity coronary anatomy, and no diabetes, we suggest CABG, as opposed to PCI (Grade 2B).

In these patients, the combined outcome of death, myocardial infarction, and stroke appears similar in the first year with either CABG or PCI. However, revascularization rates are higher with PCI. (See 'Multivessel disease' above.)

Patients with two-vessel disease involving the right and circumflex coronary arteries may reasonably choose PCI. (See 'Multivessel disease' above.)

•For patients with multivessel coronary artery disease and moderate or high complexity anatomy or with diabetes, we recommend CABG as opposed to PCI (Grade 1B). (See 'Outcomes based on lesion severity' above.)

●Lesion anatomic complexity – This may influence the relative benefit of CABG and PCI. For patients with intermediate anatomic complexity, we advise seeking the input of a heart team or cardiothoracic surgeon prior to proceeding directly to PCI at the time of diagnostic coronary angiography. (See 'Multivessel disease' above.)

●Risk factor modification – Regardless of which method of revascularization is used, aggressive risk-factor modification is necessary in all patients. (See "Prevention of cardiovascular disease events in those with established disease (secondary prevention) or at very high risk".)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Dr. Julian M. Aroesty for his past contributions as an author to prior versions of this topic review.