In 1912, James Herrick reported the first description of left main coronary artery (LMCA) occlusion when he described the case of a 55-year-old man who died of cardiogenic shock and was subsequently found on autopsy to have extensive necrosis of the left ventricle associated with thrombotic occlusion of the LMCA superimposed on an area of atherosclerotic narrowing.¹ It is now estimated that 4% to 6% of patients undergoing coronary angiography have significant LMCA disease defined as ≥50% stenosis, with more than two thirds noted to have concurrent multivessel disease.^2,3 Insofar that interventional cardiologists are usually the first in diagnosing and often determining treatment strategies for LMCA stenoses, it is important to have a thorough understanding of the disease process in order to make informed decisions for this complex disease process.

On angiography, the LMCA is designated as either protected or unprotected. The “protected” LMCA refers to the presence of patent coronary artery bypass graft (CABG) to either the left anterior descending or left circumflex artery or the presence of collateral vessels from the right coronary artery to an occluded major branch of the left coronary artery. This distinction is of particular importance as it directly influences the risk and, therefore, the approach to treatment. When faced with unprotected left main disease, the decision is complex and often is between percutaneous coronary intervention (PCI) or CABG. In this chapter, we will review data and topics concerning LMCA anatomy, assessment and significance of LMCA stenosis, prognosis, treatment, and factors that play a crucial role in successful unprotected left main PCI.

The LMCA originates from the left coronary sinus of Valsalva and gives origin to the left anterior descending coronary artery (LAD), the left circumflex coronary artery (LCx), and in 12% of cases, the ramus intermedius. The LMCA itself is a relatively short arterial segment arising from the aortic root in the area of the aortic valve leaflets and extending leftward for a short distance to its point of bifurcation.⁴ In adults, it ranges from 1 to 25 mm in length and between 2.0 and 5.5 mm in diameter.⁵ In the absence of coronary anomalies, the LMCA lies in a relatively fixed anatomic position in relation to the left ventricle and great arteries, bordered anteriorly by the right ventricular outflow tract, inferiorly by the superior margin of the left ventricle, superiorly by the descending portion of the pulmonary artery, and posteriorly by the left atrium.⁵ Although initially believed to have more of a funnel-shaped structure, more recent sophisticated imaging studies using multidetector computed tomography have shown 3 possible additional shapes: (1) biconcave morphology; (2) tapering morphology; or (3) a combined morphology involving a cone-shaped ostium and a tubular-shaped shaft (Figure 16-1).⁶

Assessing lesion severity by coronary angiography, long considered the gold standard, can often be difficult and unreliable, especially in lesions of intermediate severity in the setting of short or diffusely diseased LMCA. In lesions of ambiguous severity, use of fractional flow reserve (FFR) or assessment with intravascular ultrasound (IVUS) is helpful. FFR provides a physiologic evaluation of the significance of the left main lesion, while IVUS provides absolute anatomic dimensions of the vessel as well as morphologic characterizations of the lesion (Figure 16-2). A minimum luminal diameter of 2.8 mm and minimum luminal area (MLA) of 5.9 mm² by IVUS have been shown to correlate with a FFR value of <0.75.⁷ A prospective, multicenter trial involving 354 patients who had intermediate unprotected left main lesions demonstrated that deferring revascularization for lesions with an MLA >6 mm² was safe, as the 2-year cardiac death-free survival rate was 97.7% compared to 94.5% in patients with an MLA <6 mm² who were revascularized (P = .5).⁸ In a study involving a South Korean population, an MLA cutoff <4.8 mm² correlated best with a FFR <0.8 and MLA <4.1 mm² with an FFR <0.75.⁹

In the 1970s, multiple randomized trials and registry studies demonstrated the negative prognostic impact of the presence of an LMCA stenosis ≥50%, with the severity of stenosis being a strong predictor of outcomes. In a series of observational studies, long-term mortality for medically treated with LMCA stenosis was reported as 29% at 18 months,¹⁰ 39% to 48% at 2 years,^11,12 35% at 30 months,¹³ and 51% at 5 years.¹⁴ With the exception of the study by Lim et al,¹⁴ most of these studies included less than 70 patients. Conley et al¹⁵ analyzed 163 consecutive medically treated patients with LMCA disease divided into 2 subgroups: LMCA stenosis of 50% to 69% and LMCA stenosis ≥70%. The 3-year survival rate was 41% in the group with a stenosis ≥70% versus 66% in the group with a lesser stenosis (P <.05, Figure 16-3). When compared to medically treated patients with 3-vessel disease but an otherwise nondiseased LMCA, the group with 50% to 69% LMCA stenosis had a similar 3-year survival rate (68% and 66%, respectively). The 3-year survival rate for patients with 3-vessel disease and ≥70% LMCA stenosis was only 37% (Figure 16-4). Further attempts to identify predictors of 1-year survival in patients with ≥70% LMCA stenosis revealed that abnormal left ventricular (LV) function on ventriculogram was one of the strongest predictors of mortality at 1 year, with 61% survival for those with abnormal LV function versus 95% for those with normal LV function (P <.05).

In light of the grim prognosis associated with medically treated LMCA disease, revascularization became the preferred strategy for the management of patients with significant LMCA disease. In the initial studies (both randomized and observational), revascularization done primarily through CABG demonstrated a significant survival benefit.^16-18 The Collaborative Study in Coronary Artery Surgery (CASS) Registry observational study¹⁸ compared 1492 patients with LMCA disease (≥50% stenosis) who underwent aortocoronary CABG (79%) versus medical therapy (21%). Although some patients were treated medically because they were anatomically or hemodynamically unsuitable for surgery, most were considered operable candidates. Patients with LMCA disease were further divided into risk categories according to age, severity of stenosis, and degree of LV dysfunction as assessed on ventriculogram. The overall 3-year survival rate for patients with ≥50% LMCA disease was 91% for patients who underwent CABG compared to 69% for patients treated with medical therapy (P <.0001; Figure 16-5). The 3-year survival rate in patients treated with surgery compared to medical treatment with severe (score ≥15)¹, moderate (score 11-14), mild (score 6-10), and no (score 5) LV dysfunction was 75% versus 46%, 85% versus 60%, 94% versus 83%, and 95% versus 84%, respectively (Figure 16-6). The 3-year survival rate in patients undergoing CABG compared to medical therapy was 93% versus 81% with an LMCA stenosis of 50% to 59%, 91% versus 64% for an LMCA stenosis of 60% to 69%, 92% versus 68% for an LMCA stenosis of 70% to 79%, and 90% versus 49% for an LMCA stenosis ≥80%, respectively (P <.0001),¹⁸ pointing to the greater benefit of surgical revascularization as the severity of stenosis progresses.

In the Veterans Affairs Cooperative Study, 91 patients with LMCA disease (≥50% stenosis) were randomized to either medical treatment or CABG using saphenous vein grafts, with a 42-month follow-up conducted as an intent-to-treat analysis. At 42 months, there was a significant survival benefit in patients treated surgically (88% vs. 65%, P = .016), even after adjustments were made for 2 important differences in baseline characteristics between the 2 groups: duration of angina and high-risk angiographic findings (Figure 16-7). The benefit of surgery was even more pronounced in patients with >75% stenosis (83% vs. 48%, P = .036) and in patients with abnormal LV function (89% vs. 62%, P = .012).¹⁶ In the European Coronary Surgery Study Group,¹⁹ 768 men under the age of 65 with mild to moderate angina, at least 2-vessel disease (≥50% stenosis), and normal LV function were randomized to either medical therapy or surgery. Of the 768 patients included in the analysis, 59 were noted to have LMCA disease. In the subgroup of patients with LMCA disease, the survival rate at 60 months was 61.7% in patients treated medically versus 92.9% in patients who underwent CABG (P = .037).¹⁹

While these studies clearly make a strong case for revascularization, optimal medical therapy for the treatment of stable coronary disease has significantly improved over the ensuing years. The use of statins, angiotensin-converting enzyme inhibitors, and β-blockers was relatively low at the time of these studies. Nonetheless, surgical revascularization has remained the accepted treatment approach.

Prior to the introduction of stents, percutaneous interventions of the LMCA were primarily reserved for patients with a protected LMCA. High rates of restenosis, risk of hemodynamic instability during balloon inflation, and abrupt vessel closure were factors that discouraged the use of percutaneous transluminal coronary angioplasty (PTCA). In a retrospective analysis conducted by O’Keefe et al,²⁰ the immediate and long-term outcomes of PTCA were assessed in 127 consecutive patients. These patients were divided into 3 groups: (1) 84 patients with protected LMCA undergoing elective PTCA; (2) 33 patients with unprotected LMCA undergoing elective PTCA; and (3) 10 patients undergoing emergent LMCA PTCA in the setting of a myocardial infarction, 1 of whom had a protected LMCA. Long-term follow-up was available for 98% of patients. Results were striking, with a procedural mortality of 2.4% in the elective protected LMCA group, 9.1% in the elective unprotected LMCA group (P = .14), and 50% in the acute group. Three-year survival rates were 90% in the elective protected LMCA subgroup versus 36% in the elective unprotected LMCA subgroups (P <.0005). In the acute subgroup, only 3 of 10 patients were alive at follow-up, with all 3 requiring CABG (Figure 16-8).

Kelley et al²¹ reviewed the outcomes of 142 consecutive patients who underwent protected and unprotected LMCA stenting with bare metal stents (BMS). Ninety-nine patients (70%) underwent protected LMCA stenting and 43 patients (30%) underwent unprotected LMCA stenting. In the unprotected group, 86% were considered poor surgical candidates. Survival rate at 1 year was 88% for all patients, with the target lesion revascularization (TLR) rates (Table 16-1) of 18% in the protected versus 23% in the unprotected group (P = .50). Major adverse clinical event rates were 25% in the protected group compared to 49% in the unprotected group (P = .005), largely driven by a higher rate of death in the unprotected group. At 1 year, survival in the unprotected group was 72% compared to 95% in the protected group (P <.001; Figure 16-9). Despite the major discrepancy between the 2 groups, this study demonstrated that compared to historical controls, the introduction of BMS improved both the safety and feasibility of PCI in both protected and unprotected LMCA.

Use of drug-eluting stents (DES) in the setting of a protected LMCA has shown good outcomes. In a prospective analysis of patients who underwent PCI with DES (sirolimus or paclitaxel) in the DEScover Registry,²² the 1-year survival for patients who underwent PCI of a protected LMCA was 95.5% with an observed TLR rate of 9.5% and target vessel revascularization (TVR; see Table 16-1) rate of 10.6% at 1 year, nearly reducing in half the historical rates of restenosis associated with protected BMS in LMCA.

The introduction of BMS significantly reduced the rates of abrupt closure and restenosis and augmented greater acute gain in minimal lumen diameter in patients undergoing PCI, benefits that were also applicable to patients undergoing unprotected LMCA stenting. As such, PCI of the unprotected LMCA became a plausible alternative to CABG in patients at high or prohibitive risk for CABG (eg, liver cirrhosis, severe lung disease, redo sternotomy). Observational trials^21,23-25 in the BMS era showed excellent procedural success rates, often with rates exceeding 90%. Observed in-hospital and long-term mortality rates were considerably reduced compared to the rates obtained by PTCA (Table 16-2). Nonetheless, outcomes were still suboptimal for patients at high surgical risk,^23,26 including those with an LVEF <40% (9-month survival of 26.5% ± 10%) and those presenting with an acute coronary syndrome (9-month survival of 31.3% ± 12.1%).²⁷ Long-term event-free survival (survival from death, infarction, and bypass) was inversely related to presentation with progressive and rest angina (P <.001), implying that patients often presented with late sudden death in the setting of subclinical severe restenosis.²⁷ Rates of restenosis and revascularization (TLR or TVR) were frequently >20%, with a particularly high incidence within the first year (Figure 16-10). Although BMS improved the safety and efficacy profile in PCI of the unprotected LMCA, CABG with its more durable outcomes remained the preferred method of revascularization compared to BMS. Although there are no large randomized trials of unprotected LMCA revascularization comparing BMS to CABG, a meta-analysis of observational and randomized control trials that included 10,342 patients undergoing unprotected LMCA PCI demonstrated a lower crude event rate for DES compared to BMS at 3 years. Adjusted analysis of 5081 patients demonstrated lower mortality for DES-treated patients at both 2 and 3 years,²⁸ thus supporting DES as the primary choice for PCI of the unprotected LMCA.