Left Main Intervention
- General description of procedure, equipment, technique
- Indications and patient selection
- Details of how the procedure is performed
Interpretation of results
- Outcomes (applies only to therapeutic procedures)
- Alternative and/or additional procedures to consider
- Complications and their management
General description of procedure, equipment, technique
Left main disease
A significant stenosis (>50%) of the left main is present in 4% to 6% of patients who have undergone coronary angiography. If there is no patent bypass graft to the left anterior descending (LAD) or circumflex artery, the disease is then referred to as unprotected left main disease (UPLM). When it is protected by a graft, the outcome is more related to the extent of the remaining disease. Left main equivalent refers to one stenosis that serves a territory equal to a left main lesion (usually proximal LAD or circumflex, with an occlusion of the other major left coronary artery that is filled by collaterals from the patent LAD or circumflex coronary vessel).
Left main disease occurs with other coronary artery disease in 80% of patients. Stenosis occurs in the distal left main artery in two thirds of cases and proximally in the other third. Bifurcation stenoses were more likely to be more severe than other types of left main lesions. An isolated ostial stenosis is present in less than 1% of patients. These patients tend to be younger females, and are less likely to have risk factors for atherosclerosis.
The mortality of patients with LM is 2- to 3-fold greater than patients with one- or two-vessel disease, averaging 30% to 50% over 5 years in those treated medically. The more severe the stenosis (e.g., >90%) the greater the risk.
Most patients present with effort angina or acute coronary syndromes (ACS), but they may also be asymptomatic. As with other patients with coronary artery disease, elderly patients and women more commonly present with shortness of breath without chest pain.
LM should be suspected in patients with diffuse resting ST segment depression on the EKG in the absence of a conduction abnormality or severe left ventricular hypertrophy or in those with a high-risk stress test or marked ST depressions with spontaneous chest pain.
For non-invasive detection of left main disease, stress echocardiography and myocardial perfusion imaging have comparable sensitivities but the former has a higher specificity. A large perfusion defect involving the anterior, anterior-lateral and lateral walls is suggestive of either left main or multivessel disease. Other tests such as computed tomography (CT) angiography, magnetic resonance angiography, and Doppler echocardiography of the left main may be useful, but angiography is needed to confirm the diagnosis and plan treatment..
A definitive diagnosis is best made by coronary angiography. Care needs to be taken in the intubation of the left main to avoid plaque disruption and coronary dissection in patients who have an ostial left main lesion and a sinus injection of contrast is recommended to exclude an ostial lesion before entering the left main with the diagnostic catheter.
Lesions that are borderline significant (40% to 70% diameter on angiography) should be assessed with additional non-invasive or invasive diagnostic techniques, such as fractional flow reserve (FFR) or intravascular ultrasound (IVUS). An FFR measurement of <0.80% after intravenous adenosine or an IVUS cross-sectional area of <7 cm2 are considered significant. FFR has a much better correlation with noninvasive imaging and should be the preferred assessment technique.
Left main lesions frequently involve bifurcation, and these lesions should be described by one of the available classification schemes. The most commonly used is the Medina classification of coronary bifurcation lesions. It involves assigning a binary value (1, 0) to each of the three components of a bifurcation (proximal region of main branch, distal region of main branch, and the side branch), depending on whether they have more than (1) or less than (0) a 50% lesion. If only the proximal segment of the main branch has a significant lesion, it becomes Medina 1, 0, 0. If the distal segment of the main branch alone is involved, it becomes 0, 1, 0. Sole involvement of the side branch is designated 0, 0, 1 and involvement of all the three is designated 1, 1, 1.
Other bifurcation classifications, Movahed, Lefevere, and the newly proposed DINO (Descriptive Intelligible and Ordered), are more complex, although they describe the angle of the side branch and may be useful from a technical standpoint.
The procedure is done in a manner similar to coronary angiography using guiding catheters that have a larger lumen permitting passage of the angioplasty balloon and stent. As in stenting in general, the left main lesion is first crossed with a guidewire and positioned in the distal vessel. A predilatation balloon less than the estimated diameter of the reference diameter of the left main is used to enlarge the lumen, followed by placement of a properly sized stent (preferably a drug-eluting stent) in the lesion. Postdilation with a noncompliant balloon, to ensure complete expansion of the stent against the vessel wall and no residual stenosis, completes the procedure.
In the setting of a bifurcation lesion involving the ostium of the circumflex or left anterior descending artery or both, double wiring of both branches is used to allow a double stent technique to be done. In general a single stent technique is preferred given the superior long-term outcome. The specifics of the technique are described in a later section.
Indications and patient selection
Patient selection for interventions
Symptomatic or asymptomatic patients with left main disease and either demonstration of ischemia on a noninvasive test or significant symptoms should be considered for revascularization given the demonstrated improvement on survival from randomized trials conducted in the 1980s and 1990s.
Coronary artery bypass surgery is the gold standard for revascularization, but to determine optimal revascularization, the method requires an assessment of the operative risks of either procedure, the degree of revascularization achieved, and the long-term outcomes after the procedure. Percutaneous coronary interventions (PCI) and coronary artery bypass surgery (CABG) risk scores have been shown to assist in this assessment.
The SYNTAX score was designed to predict outcomes related to anatomic characteristics of patients with multivessel disease. The score is calculated from 11 variables, including the extent of disease, degree and characteristics of the lesions, and the diffuseness of the disease (
www.syntaxscore.com). Risk is divided into low (1 to 22), intermediate (23 to 33), and high (>34). In the SYNTAX trial, the benefit of CABG vs. PCI was directly related to the SYNTAX score. In left main disease, major adverse coronary and cerebral events (MACCE) was noninferior in the low and intermediate group but lower for surgery in the high-risk group.
The SYNTAX score alone only modestly predicts long-term outcome. Adding clinical variables such as age, creatinine, and ejection fraction (clinical SYNTAX score) can improve the predictive accuracy. Though high clinical SYNTAX predicts poor outcomes in patients undergoing PCI, it is a poor predictor of outcomes after CABG.
Two scores are commonly used to assess outcomes after CABG, the Euro Score and the Society of Thoracic Surgeons' (STS) score. The Euro Score assesses 17 risk factors (patient related, cardiac related, and surgery related). There are two available methods of calculating predicated outcomes, the original additive model and the logistic model. Both tend to overestimate risk in low and high-risk patients. A score ≥ 6 was found in registry data to be an independent predictor of mortality in patients who undergo PCI and CABG.
The STS score also accurately predicts the perioperative and 30-day mortality and morbidity in patients undergoing all types of open-heart surgery. In contrast to the Euro Score, the STS score slightly underestimates the preoperative risk but better predicts the long-term outcome. All risk scores have an easy to use online calculator. (
The combination of the Euro Score (>7.5%) and SYNTAX score (>25) provide incremental predictive value for risk stratification of unprotected left main coronary artery (ULMCA) patients. High-risk patients have a 7.1-fold higher risk for MACE after 1 year.
The decision to either perform elective or urgent CABG or PCI is best made by a joint conference with the cardiac surgeon, interventional cardiologist and clinicians. This heart team approach was demonstrated to be valuable in the SYNTAX trial and is a class I recommendation in the ACCF/AHA guidelines.
The 2011 ACCF/AHA/SCAI guidelines do not give UPLM PCI a class I indication for any clinical or anatomic situation. In patients with unstable angina/non ST elevation myocardial infarction (UA/NSTEMI) the guidelines recommend PCI as a class IIa indication if the patient is not a surgical candidate and recommend PCI for STEMI patients with normal blood flow (Thrombolysis in Myocardial Infarction (TIMI) flow grade 3) when the LM PCI can be done safely and swiftly. In stable angina patients, PCI for a UPLM is a class 11a indication for patients with a SYNTAX score of < = 22 and when surgery has an increased operative risk (e.g., an STS predicted mortality of >5%).
The ESC guidelines recommend PCI for a UPLM for ostial or body lesions as a class IIa indication and for more complex LM disease (e.g., bifurcation lesions) or those with one- or two-vessel disease with a SYNTAX score of <33) as a class IIb indication.
Left main PCI should not be done in stable patients if the anatomy is unfavorable or when the SYNTAX score is consistent with extensive CAD (e.g., a SYNTAX score of >33).
Details of how the procedure is performed
Technical consideration for left main PCI
Access can be obtained from a femoral or a radial approach using standard techniques. Radial access in patients with left main has been shown to have the same mortality or MACCE as a femoral approach, with a reduction of bleeding events and a hospital stay. It should be considered in patients who are at higher risk for procedural bleeds.
Optimal choice of guiding catheters to provide adequate support and visualization are critical. Care needs to be taken to avoid oversized guides that intubate the LM excessively and can cause left main dissection.
Left ventricular support devices, such as intraaortic balloon counter-pulsation or an Impella pump, are generally not needed unless the patient is having hemodynamic instability. These support devices should be considered in patients with a high-grade left main lesion and a chronically occluded dominant right coronary artery (RCA). Alternatively, femoral arterial access with a small sheath can be placed prior to PCI that allows rapid deployment of a support device in a high-risk patient.
In cases of an intermediate LM lesion (40% to 70% stenosis on angiography), either FFR or IVUS can provide useful additional information to determine if revascularization is indicated.
FFR is the ratio of distal coronary pressure to aortic pressure measured during maximal hyperemia. The normal FFR for all vessels under all hemodynamic conditions, regardless of the status of the microcirculation, is 1.0. FFR values less than 0.80 are associated with abnormal stress tests. Lesions that are responsible for reversible ischemia have FFRs less than 0.80.
IVUS provides a tomographic 360-degree scan of the vessel from the lumen through the media. IVUS measurements of arterial dimensions (minimal and maximal diameters, cross-sectional area, and plaque area) provide important additional information compared with angiography alone. IVUS is more sensitive in detecting calcification and significant LMCA stenosis compared to angiography. A significant stenosis by intravascular ultrasound (IVUS) is defined as a minimal lumen area of less than 5 mm2 (in some studies a significant stenosis is less than 7 mm2). IVUS is less accurate than FFR in determining the significance of a stenosis.
Bifurcation disease involving the origin of the LAD or circumflex is much more difficult to treat with stenting and has a worse outcome than PCI of lesion of the body or trunk. Two approaches are employed: a single stent technique or a two-stent technique. The size of the branches and the location of the stenosis before, after the bifurcation and in the main artery, or one or both branches are important considerations for the optimal strategy. The Medina classification is frequently used to describe the extent of disease.
The one stent is the preferred technique when possible given the superior long-term outcomes. Placement of a nonhydrophilic wire in the provisional branch not only protects the vessel but adds additional guide support. The kissing balloon technique with balloon inflation of the main and the side branch simultaneously has not been shown to improve outcomes and is not recommended routinely.
The two-stent technique is needed when the lesions in the ostium of both branches is severe or diffuse, or during a single stent technique when the vessel occludes or a localized dissection occurs that compromises the side branch. The technique places one stent in the side branch and the other in the main branch with crossing of the stents in the proximal main branch. The stents can be delivered using a t-stent, crush, culotte configuration, usually with kissing balloon inflations at the end to ensure optimal dilation of both stents. When the angle of the circumflex is 90 degrees, a t-stent or variant (T and protrusion (TAP) technique is preferred. Large observational studies have found no difference at 2 years of MACCE when a T stent, crush, or V stent are used.
Drug-eluting stents (DES) are preferred over bare-metal stents (BMS) due to the higher restenosis rate (18% to 31%) and the repeat revascularization rate in patients treated with bare-metal stents. It’s prudent to consider BMS when there is concern for long-term compliance with dual antiplatelet therapy or very high bleeding risk since dual antiplatelet therapy is recommended for 6 to 1 year after placement of a DES, while it is recommended for only one month with BMS in stable angina patients. In patients with ACS dual antiplatelet therapy is recommended for one year regarless of he stent type.
There is no sufficient data to support the use of one type of second generation DES versus another. The ISAR LEFT MAIN 2 trial showed no difference between everolimus- and zotarolimus-eluting stents..
Calcification of the left main is an independent predictor of all case mortality and is highly associated with bifurcation disease compared to ostial or midshaft disease. When extensive calcification is present, rotational atherectomy should be considered prior to stenting to ensure adequate stent expansion.
Interpretation of results
Angiography is usually sufficient to determine if the procedure has been successful. Use of orthogonal views of the LM can usually disclose if inadequate stent expansion has occurred or dissection proximal or distal to the stent or acute stent thrombosis is present.
Treatment for inadequate stent expansion is redilation with a larger balloon, whereas dissection not covered by the stent requires additional stenting.
IVUS has been successfully used as an adjunct to stent implantation, and there is evidence for a reduction in restenosis rates with routine use of IVUS as a tool for improving outcome in PCI. It should be considered whenever there is uncertainty about the success of the procedure and the adequacy of stent expansion and in the treatment of instent restenosis of the left main.
Outcomes (applies only to therapeutic procedures)
Therapy after coronary intervention
A dual antiplatelet with aspirin and a P2Y12 antagonist are recommended by the ACCF/AHA guidelines for at least 1 year in all patients with ACS regardless of stent type.
When a BMS is placed in stable angina, dual antiplatelet therapy should be given for a minimum of 1 month but preferably for 1 year as well. The guidelines indicate that continuation of dual therapy beyond 1 year is reasonable in high risk patients.
Current ACCF/ACC guidelines recommend either clopidogrel or ticagrelor be given prior to the procedure, with clopidogrel, ticagrelor, or prasugrel to be used at the time of the procedure. Prasugrel should not be used in patients with prior stroke or transient ischemic attack, and the dose should be reduced in patients with a weight less than 60 kg.
Routine follow-up coronary angiography for patients with left main stenting was initially recommended, but extensive experience has not shown not to be needed in most patients except those at high risk.
Routine follow-up exercise testing is not recommended but should be done when symptoms suggestive of ischemia occur.
As with all patients undergoing revascularization, aggressive risk factor modification is essential and can significantly improve long-term outcome.
Alternative and/or additional procedures to consider
Alternative procedures - CABG
Revascularization for left main disease- CABG versus PCI
Coronary artery bypass grafting has been the gold standard for treatment of left main disease. A meta-analysis of the seven major studies conducted between 1972-1984 by Yusuf et al, showed a 68% reduction in mortality at 5 years in patients who underwent CABG compared to medical therapy for left main disease. Even though these trials are outdated due to subsequent advances in medical management, these studies are the basis for the current ACCF/AHA guidelines that list CABG as a class Ib indication for revascularization of unprotected left main disease regardless of the presence or severity of symptoms.
The SYNTAX trial is the first major adequately powered randomized trial comparing CABG to PCI in left main disease or three-vessel disease who were eligible to both procedures. It was designed as a noninferiority study. The final overall results showed that there was an increase in MACCE, death, MI, stroke, and repeat revascularization at 1 year in the patients randomized to PCI (17.8% in PCI patients compared to 12.4% in CABG patients) primarily due to an increase in repeat revascularization with equivalent rates of death and MI.
Left main disease was present in 38.8% of patients in the CABG group and 39.5% of those in the PCI group in the SYNTAX trial. The majority of the left main group had additional diseased vessels (20% with one-vessel disease, 31% with two-vessel disease, and 37% with three-vessel disease).
In the SYNTAX LM subgroup, no significant difference in MACCE was demonstrated between CABG and PCI at 1 year (13.7% vs. 15.8%, respectively). Since the overall trial failed to demonstrate noninferiority, the analysis of the LM subgroup was considered exploratory. In a further subgroup analysis of those with no other diseased vessels or with only one-vessel disease, the trial showed a trend toward a better outcome with PCI while in those with 2- or 3-vessel disease, CABG was better. One has to be mindful that the LM subgroups were not powered for mortality. The 4-year follow-up continued to confirm no difference between the two groups (27.8% vs. 33.2%) in the primary endpoint.
When the SYNTAX LM subgroup was stratified by SYNTAX score, the equivalent outcome was only seen in those with a SYNTAX score of less than 33 (intermediate or lower).
Two additional randomized trials have been reported, EXCEL and NOBLE. EXCEL showed non-inferiority while NOBLE showed a worse composite endpoint (death, non-fatal MI, and repeat revascularization) at 5 years. Death and MI were not different in either trial.
Analysis of all five randomized controlled studies that involved unprotected LM in the modern era (4,594 pts) showed no difference at 3-5 years in MACCE, while revascularization rates were significantly lower in the CABG and strokes lower in the PCI group.
On the basis of the SYNTAX trial and the smaller trials, the ACCF/AHA guidelines recommend CABG as the treatment of choice for UPLM in patients who are good candidates for surgery. PCI is considered as an alternative only in those with higher surgical risk with a SYNTAX score of less than 33. The ESC guidelines likewise recommend CABG as the treatment of choice. When considering all trials reported, PCI appears to be safe and an alternative to CABG.
Complications and their management
Complications of LM PCI
Procedural complications are similar to PCI of other lesions and include coronary dissection, embolization, myocardial infarction, stroke, death, major bleeding, and local vascular access complications. The most serious complications of death, MI, and stroke occur in 1% but range from 0.3% to 3%.
Left main PCI due to the large area of myocardium served, can lead to hemodynamic compromise. When severe ischemia is anticipated, hemodynamic support should be considered with intraaortic balloon counter-pulsation or the Impella device. When extreme risk is present or when shock occurs during the procedure, percutaneous cardiopulmonary bypass or left ventricular assist devices may be needed.
Restenosis in UPLM occurs at a rate similar to other coronary lesions occurring in 7% to 12%, depending on the complexity of the left main disease and the number of additional lesions treated. In the past there was concern that restenosis could lead to a sudden and unexpected death; however, recent trials have not shown this to be the case. Treatment of restenosis is similar to other coronary lesions with the majority successfully treated percutaneously with additional DES. When the restenosis is severe, diffuse, or recurrent, bypass surgery should be considered.
Definite, probable, and possible stent thrombosis occur at a similar frequency to other coronary stenosis with a reported incidence of <1% due to high dynamic flow of the left main. Management is similar to stent thrombosis in other vessels with emergent PCI if the patient presents with an STEMI and urgent PCI in any patient with suspected stent thrombosis. If redilation and additional stenting is needed but fails to successfully restore flow, emergent CABG should be considered.
Copyright © 2017, 2013 Decision Support in Medicine, LLC. All rights reserved.
No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.