Revascularization of Left Main Disease: Do we EXCEL at Stenting? Or is it More NOBLE to Treat With Surgery?

Editor's Note: This is Part Two article of a two-part Expert Analysis. Click here for Part One.


About 5% of patients undergoing coronary angiography have significant disease of the left main coronary artery. Left main diseases usually have extensive coronary artery disease (CAD), and the coexisting disease is often complex. The left main stem subtends a large amount of myocardium and may be associated with ventricular dysfunction, arrhythmia, or hemodynamic compromise, making it a potentially lethal condition. Early clinical trials showed better survival with coronary artery bypass graft surgery (CABG) compared with medical therapy in this subset, establishing CABG as the "gold standard" for treatment. Percutaneous coronary intervention (PCI) can be a safe and effective method for revascularization; however, it has typically been reserved for poor surgical candidates. Up until recently, there was a paucity of randomized data comparing outcomes of CABG versus PCI in left main disease.

What is Known About Left Main PCI?

Left main disease is a challenging subset for PCI. Lesions of the mid-shaft of the left main stem are fairly straightforward, but they are also uncommon. Instead, the disease more typically involves the aorto-ostium with heavy calcification. Or the disease involves the distal bifurcation with all of the associated challenges of any bifurcation lesion with the exception that the "side branch" is a major epicardial vessel. This is particularly problematic in left-dominant circulations (Figure 1). As noted, there is also extensive CAD elsewhere. Additionally, the large mass of myocardium subtended by the left main stem may cause hemodynamic collapse during intervention.

Figure 1: Bifurcation Distal Left Main Disease in a Patient With Left-Dominant Circulation

Figure 1

For many years, the only data regarding outcomes of left main PCI came from non-randomized and registry studies. The few randomized trials of left main PCI consisted of small cohorts. The SYNTAX (Synergy Between Percutaneous Coronary Intervention With TAXUS and Cardiac Surgery) trial was a large randomized study comparing PCI and CABG for patients with complex CAD and included a subset of 705 patients with left main disease.1 A major contribution of the SYNTAX trial was the creation of the SYNTAX score, a tool to quantify the extent and complexity of CAD. For the left main subset, the 5-year outcome data reported that patients with a SYNTAX score >33 had lower mortality and a lower rate of repeat revascularization with CABG compared with PCI, thus establishing CABG as the preferred revascularization method for patients with left main disease and high SYNTAX scores. Interestingly, the subgroup with low-to-intermediate SYNTAX scores (0-32) had a lower mortality and stroke rate with PCI with no difference in rate of repeat revascularization.1 Although this study was underpowered to draw definitive conclusions, these results provided equipoise for the design of larger randomized trials.2,3

Recognizing the complexity of this high-risk subgroup, our current guidelines4 advocate for use of the Heart Team approach for decision-making. This involves evaluation by both a surgeon and an interventional cardiologist who together determine the best revascularization strategy, taking into account patient comorbidities and the technical issues of PCI versus CABG for that patient. Currently, CABG has a Class I indication to improve survival in patients with significant left main disease. For PCI, the guidelines consider PCI in patients with a low (<22) SYNTAX score with ostial or mid-shaft left main disease who are at increased risk for CABG (estimated mortality of >5%) a Class IIA indication. Patients with acute coronary syndromes with left main disease as the culprit who are not surgical candidates and patients with acute ST-segment elevation myocardial infarction (MI) in whom the left main is the culprit also have Class IIA indications for PCI. PCI carries a Class IIB indication for patients with SYNTAX score <33 and/or bifurcation disease who are at increased risk for CABG (estimated mortality >2%). Finally, the guidelines state that PCI should not be performed for left main disease in patients with unfavorable anatomy (STS >33) who are good candidates for CABG.

The Recent Randomized Trials

The two recent randomized trials of PCI versus CABG for left main disease are interesting in that they enrolled similar patients yet came to different conclusions. The primary endpoints for both studies are shown in Figure 2.

Figure 2: Primary Endpoint Results From EXCEL and NOBLE

Figure 2

The EXCEL trial (Evaluation of XIENCE versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization) enrolled 1,905 patients with significant left main disease eligible for both PCI or CABG who had low or intermediate SYNTAX scores (<32) and randomized 948 patients to PCI using second-generation drug-eluting stents and 957 patients to CABG.2 The primary endpoint of the study was the combined endpoint of death from any cause, stroke, or MI at 3 years. Secondary endpoints were the combined endpoint at 30 days and the combined endpoint plus target vessel revascularization (TVR) at 3 years. The primary endpoint was achieved in 15.4% of PCI patients and 14.7% of CABG patients, achieving statistical non-inferiority of PCI. The pre-specified secondary endpoint of the combined endpoint at 30 days was lower in the PCI group (4.9% for PCI vs. 7.9% for CABG). The 3-year combined endpoint including TVR again showed non-inferiority for PCI (23.1% for PCI vs. 19.1% for CABG). Ischemia-driven revascularization was more common during follow-up for PCI patients (12.6% for PCI vs. 7.5% for CABG).

The EXCEL trial consisted primarily of patients with preserved ejection fraction. Of those patients, 60% had stable syndromes, and 30% were diabetic. The majority (80.5%) had disease involving the distal bifurcation. The site-designated average SYNTAX score was 20: 60.5% had low scores (<22), and 39.5% had intermediate scores (23-32). However, by core laboratory analysis, the average score was higher (26.5), fewer had low scores (35.8%), and 24.2% had scores >33, which should have been excluded. Little data were provided regarding technical aspects of PCI such as use of support devices, bifurcation strategy, or need for rotational atherectomy. An average of 1.9 lesions were treated per patient. The procedure was guided by intravascular ultrasound in 77% of cases. The study did not report on the rate of complete revascularization achieved.

NOBLE (Nordic-Baltic-British Left Main Revascularization Study) enrolled 1,201 patients with left main disease and no more than 3 additional non-complex lesions (defined as chronic total occlusions, bifurcations requiring 2-stent strategies, or extensive coronary calcification or tortuosity); the SYNTAX score was not used.3 Of those patients, 598 were randomized to PCI and 603 to CABG. The primary study endpoint was a combined endpoint of all-cause mortality, non-procedure related MI, stroke, and repeat revascularization at 5 years. The primary endpoint was achieved in 29% of PCI patients and 19% of CABG patients, indicating statistical superiority of CABG. Regarding the individual endpoints, all cause death was not different, but the individual rates of MI, revascularization, and stroke were each higher in PCI patients compared with CABG patients. The SYNTAX score was not related to outcomes, and even patients with low SYNTAX scores did better with CABG.

NOBLE patients had preserved ejection fraction; 82% had stable syndromes, and 15% were diabetic. Disease complexity appeared similar to EXCEL with an average SYNTAX score of 22.5 and 81% with distal left main disease. The distal bifurcation was treated using a 2-stent strategy in 35% of patients. The median number of lesions was 2.0, intravascular ultrasound was used in 74%, and complete revascularization was achieved in 92% of patients.

What can We Conclude and how do We Apply These Results to Our Practice?

These two trials included very similar patients who are representative of this subset. The most important difference is that the primary endpoints differed. The EXCEL trial's primary endpoint consisted of only death, MI, and stroke at 3 years. NOBLE's primary endpoint consisted of death, MI, stroke, and revascularization at 5 years. When you add revascularization to the EXCEL trial's endpoint, that trial starts looking more like NOBLE, and the p value for superiority of CABG approaches significance at 0.10 (Figure 3). The rate of repeat revascularization will likely increase in the PCI group from 3 to 5 years, and it is possible that by 5 years the rates of achieving the combined endpoint that includes repeat revascularization will be similar to NOBLE.

Figure 3: EXCEL Versus NOBLE: Inclusion of TVR in 3-Year EXCEL Combined Endpoint

Figure 3

Unanswered questions remain. When you exclude high SYNTAX scores and more complex disease, the role of the SYNTAX score is no longer clear. Note that in EXCEL, there was a discrepancy between the site-designated and core laboratory-designated SYNTAX scores. In fact, 24% of patients had high scores (>33) by core laboratory, and yet the population overall still did well with PCI. Subgroup analysis in the EXCEL trial based on SYNTAX score did not reveal differences in outcome. In NOBLE, outcomes did not relate to SYNTAX score, and low scores did better with CABG. We need to know more about how the technical aspects of PCI for left main disease related to outcomes, particularly for management of left main bifurcations. The higher stroke rate in the PCI arm of NOBLE was unexpected and occurred at 1 year; this coincided with cessation of dual antiplatelet therapy (DAPT). Thus, the optimal duration of DAPT in this group needs to be better understood. Finally, as noted above, I will be very interested in the 5-year outcomes from the EXCEL trial because this may shift the benefit to CABG.

As an interventional cardiologist, I draw several conclusions from these studies. First, these data support PCI as a reasonable method for the treatment of non-complex left main disease. Even in NOBLE, the mortality did not differ, and most of the benefit was in the lower rate of revascularization over follow-up. Many patients will chose that option because they do not perceive an additional catheter-based procedure as negatively as they do surgery. Second, the outcomes for both PCI and CABG when performed by experienced operators at experienced centers are quite good. I believe this is, in part, a testimony to the value of the Heart Team approach because those who felt to be better served by one or the other approach when evaluated jointly were not enrolled. Patient discussion should include the major risks and benefits of both options and include the important issue of long-term DAPT. Finally, it is important to note that the benefits of CABG in NOBLE took several years. At 1 year, the outcomes were the same; therefore, patients with limited life expectancy can be expected to do just as well with PCI.

As a final note, it should be understood that left main PCI is a high-risk subset and requires special skill sets that not all interventional cardiologists possess. This is especially true for low-volume operators and for those practicing without surgical back-up. The interventionalist contemplating left main PCI should be comfortable with the management of aorto-ostial disease, coronary calcification, and strategies for management of complex bifurcations and the use of rotational atherectomy, intravascular ultrasound, and hemodynamic support devices (Figure 4).

Figure 4: Post-Stent Result of Patient Shown in Figure 1 Using IMPELLA Support to Treat Bifurcation

Figure 4


  1. Morice MC, Serruys PW, Kappetein AP, et al. Five-year outcomes in patients with left main disease treated with either percutaneous coronary intervention or coronary artery bypass grafting in the synergy between percutaneous coronary intervention with taxus and cardiac surgery trial. Circulation 2014;129:2388-94.
  2. Stone GW, Sabik JF, Serruys PW, et al. Everolimus-Eluting Stents or Bypass Surgery for Left Main Coronary Artery Disease. N Eng J Med 2016;375:2223-35.
  3. Mäkikallio T, Holm NR, Lindsay M, et al. Percutaneous coronary angioplasty versus coronary artery bypass grafting in treatment of unprotected left main stenosis (NOBLE): a prospective, randomised, open-label, non-inferiority trial. Lancet 2016;388:2743-52.
  4. Levine GN, Bates ER, Blankenship JC, et al. 2011 ACCF/AHA/SCAI Guideline for Percutaneous Coronary Intervention. A report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines and the Society for Cardiovascular Angiography and Interventions. J Am Coll Cardiol 2011;58: e44-122.

Keywords: Angiography, Acute Coronary Syndrome, Arrhythmias, Cardiac, Atherectomy, Coronary, Comorbidity, Coronary Angiography, Coronary Artery Bypass, Coronary Artery Disease, Diabetes Mellitus, Myocardial Infarction, Myocardium, Percutaneous Coronary Intervention, Risk Assessment, Stroke, Ventricular Dysfunction, Surgeons

< Back to Listings