Left Main Coronary Artery Disease: Current Updates on CABG versus PCI

Abstract

Most patients of LMCA disease are symptomatic and at high risk of cardiovascular (CV) events, since occlusion compromises flow, and it is associated with >20% mortality at 1 year. Coronary artery by-pass graft (CABG) is the main mode of revascularization procedure for significant left-main coronary artery (LMCA) disease unless contraindicated or unsuitable for surgery, and in patients with complex coronary anatomy. Percutaneous coronary intervention (PCI) of left-main (LM) is emerging as an alternative to CABG especially in patients with low syntax score with suitable coronary anatomy for PCI, and life-saving emergency situations like acute coronary syndrome (ACS) with hemodynamically unstable, and high risk group patients who are unsuitable coronary anatomy for grafting or due to associated co-morbidities.

1. Introduction

LMCA arises from the left coronary sinus in majority of patients, most often divides into two major branches; 1) Left anterior descending (LAD) is the larger vessel in majority of patients, supplies anterior aspect of left ventricle and anterior portion of septum. 2) left circumflex (LCX) supplies left lateral and posterior aspect of left ventricle (LV). In some patients, it trifurcates into LAD, LCX and Ramus branches, and in 1% of the population, may present like atretic segment or both branches may arise directly from the aorta via separate ostia [1]. LMCA has an average length of 10.8 ± 5.2 mm (range 2–23 mm), an average diameter of 4.9 ± 0.8, and supplies more than 75% of the blood supply to the LV in a right dominant system and almost 100% supply to the LV in a left dominant system [2].

LM has higher elastic component that may lead to stent under expansion & recoiling which necessitates use of stents with sufficient radial strength. LM disease more often associated with hard fibro- calcific plaques with tapering of the vessel [3]. Conventionally, an angiographic cut off of >50% diameter stenosis (equivalent to >75% area stenosis) has been used to indicate hemodynamic significance, as suggested by early work in an animal model by Gould that demonstrated a reduction in hyperaemic flow. Significant LMCA disease defined as >50% narrowing is found in 4 to 6% of patients who undergo coronary angiography, and it involves ostium in 15%, mid segment in 20%, ostium to Proximal shaft in 15%, and LM bifurcation lesions in 50% of patients. Figure 1 across lesions beyond 50% degree of stenosis of LM [4].

Figure 1.

Type of lesions involving left main.

Clinical Presentation: LM disease most often presents with ACS in >63%, and stable ischemic heart disease in 37% of cases, and sometimes with life threatening arrhythmias and sudden death. Ostial stenosis is more often seen in women (44 vs. 20%) compared to men [5]. Oveido et al. demonstrated that LM lesion extending into the proximal LAD, LCX or both may be seen in 90%, 60.4% and 62% of patients, respectively, whereas isolated ostia of LAD & LCx lesions without extending to LM were seen in 9.3% and 17.1% of patients, respectively [6].

1.1 Etiology

Atherosclerosis is the most common cause of LMCA disease, and other causes of LMCA disease are rare. Diseases involving ascending aorta may also cause LMCA obstruction such as aortic dissection, aortic aneurysm, Takayasu arteritis, systemic vascular disorders, thromboembolism to LM, and sometimes it might be due to iatrogenic causes like trans-catheter valve implants in a lower coronary origin or shallow sinuses of Valsalva, and iatrogenic catheter induced traumatic dissection or spasm of LM [1].

2. Replace this text with your section heading

2.1 Non-invasive predictors of LMCA disease

Myocardial perfusion scan indicative of significant LM disease is presence of multiple large perfusion defects in the LAD and LCX territories mainly associated with exercise induced transient ischemic dilation (TID) of left ventricle, and increased lung uptake of tracer. Probability of LM disease more likely in the presence of stress-induced sustained ventricular tachyarrhythmia or non-sustained ventricular tachyarrhythmia >30 seconds or ST-segment elevation, exercise LV ejection fraction ≤35%, and appearance of new regional wall motion abnormality (RWMA) involving >2 segments at a low-dose dobutamine stress test (≤10 mg/kg per minute), inducible ischemia at a low heart rate (<120 beats per minute) or at low level of exercise test, and exercise induced fall in systolic BP [7].

2.2 Assessment of LM with Computed Tomography (CT) and Magnetic Resonance Imaging (MRI)

Multi-slice spiral computed tomography (MSCT) is accurate in identifying LM lesions with >50% narrowing with a sensitivity of 97% and specificity of 86% compared with angiography, and CMRI detection of coronary lesions in heavily calcified coronary segments is more reliable than by cardiac CT. Overall, the accuracy of MSCT for detection of angiographic in-stent restenosis (ISR) of LM was 93% with 100% sensitivity, 91% specificity, and 100% negative predictive values [8]. The DISCOVER FLOW trial demonstrated that Fractional flow reserve (FFR) CT could dramatically improve the diagnostic accuracy of CT imaging without the need for invasive FFR imaging [9].

2.3 Angiographic assessment of LMCA

Coronary angiography remains the gold standard diagnostic technique for the diagnosis of clinically important LMCA disease. Angiography is poor in assessing characterization of tissue or plaque (except for calcium, aneurysms, coarse ulcerations, or large dissections) and features associated with suboptimal stent deployment. In order to avoid precipitating myocardial ischemia in patients with severe LMCA disease perform angiography with careful manipulations of catheters, limited angiographic images with minimal contrast dose to avoid procedure related sudden events. Sometimes, Ostial LMCA stenosis is very difficult to make out on angiography and it should be suspected if there is any pressure damp and absence of reflex of dye in to the coronary sinus. Disease involving entire LMCA may be underestimated due to lack of reference segment, and in such cases indirect assessment of LMCA diameter and size can be estimated by using Finet’s and Murray’s law using main branch (MB) and side branch (SB) diameters [10, 11]. Angio is also poor at assessing lesion calcification due to its low sensitivity in detecting calcium (45–50%), and this may lead to underestimation of calcium contributing to procedure delay, failure of PCI due to absence of plaque modification resulting in under expansion, dissection, failure to cross lesions with balloons and stents, and dislodgement of stents. Intravascular ultrasound (IVUS) and Optical coherence tomography (OCT) have better sensitivity 80% vs. 50% in detecting calcium compared to angiography.

2.4 Usefulness of IVI and Functional testing to assess LM disease

Angiographic assessment of borderline LM lesions (30–70%) is inaccurate with significant inter-observer variability whereas the reproducibility and accuracy of the angiographic evaluation of LM lesions <30% and ≥ 70% is excellent Thus, revascularization strategies of borderline lesions based solely on the angiography may lead to incorrect revascularization strategies due to improper assessment of LMCA severity which might adversely affects clinical outcomes because of low graft patency rates and up to a 6-fold higher rate of atherosclerotic disease progression of bypassed native coronary vessels [12]. Intravascular imaging is helpful in assessing severity and to decide revascularization strategies particularly in patients with angiography showing doubtful, inaccurate, ambiguous lesions, intermediate lesions without any noninvasive evaluation of inducible ischemia or and whenever no correlation between angiography lesion severity and clinical symptoms. OCT is not considered as ideal imaging option for ostio-proximal LM lesions and might be limited in case of large vessel (>5.5 mm diameter) but the technique superior to IVUS in identification of thrombus, stent under expansion, struts mal apposition and edge dissection thrombus, due to its better spatial resolution. Lesions involving mid and distal LM can be adequately visualized by OCT imaging modality with high resolution images and comparable results with IVUS. OCT can reveal more detail, whereas IVUS provides more insight in deeper layers of the coronary arteries. The expert consensus group stated that IVUS and OCT are equivalent and both superior to coronary angiogram (CAG) guidance [13]. However, an extensive Random clinical trial (RCT) that addresses superiority of OCT guidance is currently still lacking.

Based on the findings of LITRO study, patients with intermediate LM stenosis between 25% and 60% lesions with minimal luminal area (MLA) of ≥6 mm² revascularization of LM (Figure 2) can be safely deferred with favorable outcomes at 2 years of follow-up (cardiac death-free survival of 97.7%). Nearly 30% of patients with mild disease of LMCA with less than 30% narrowing had an MLA of <6 mm², whereas 43% of patients with angiographic LM stenosis ≥50% had a prognostically favorable MLA of ≥6 mm² [14].

Figure 2.

Mean luminal area of LMCA by IVUS from various studies.

S J Park group suggested the MLA cut-off for FFR < 0.80 was 4.5 mm² for south Asians with a sensitivity of 77% and a negative predictive value of 75% [15]. Based on results of DEFINE-FLAIR and iFR-SWEDEHEART studies cut off value of instantaneous wave free ratio (iFR) for deferring revascularization of lesions is >0.89. However, if the FFR is between 0.81 and 0.85, then the hemodynamic significance of the LM lesion cannot be accurately determined if the combined FFR of the LM and the downstream disease is ≤0.45. In such situations, IVUS or OCT imaging assessment of mean luminal areas will be helpful to decide whether revascularization is required or not, IVUS MLA < 4.5MM² needs revascularization, and IVUS MLA is between 4.5mm² and 6mm² requires FFR assessment after treating downstream vessel to decide revascularization of LM is required or not based on FFR value. > 0.80 or < 0.80, and revascularization should be avoided if IVUS measured LM MLA > 6 mm² [16].

2.5 Early experience of PCI

Most patients of LMCA disease are symptomatic and at high risk of CV events, since occlusion compromises flow to at least 75% of the LV, and it is associated >20% mortality at 1 year. Presently four management strategies recommended for LMCA disease: medical therapy, PCI, or surgical revascularization (CABG) either off pump or on-pump, and hybrid (CABG + PCI) procedures. For all practical purposes, CABG is the main mode of revascularization procedure for significant LMCA disease unless contraindicated or unsuitable for surgery due to better long term results particularly in diabetics, and in patients with complex coronary anatomy, and it was based on superior results observed in 3 randomized trials conducted in 70’s and 80’s - VA study [17], ECSS study [18], and CASS study [19]. CABG improved survival and symptoms mainly in patients with triple Vessel Disease and LMCA disease associated with severe LV dysfunction, and positive exercise induced ischemia. 150 patients of Left Main Disease in VA and EU RCT study showed 5 years Mortality was 36.5% v/s 16% in Medical treated v/s CABG group. CASS registry consisting of 1484 patients showed significantly improved survival rate at the end of 10 years and 15 years follow up in CABG treated patients v/s Medically treated patients [20]. Until recently CABG was the only option considered for significant LMCA disease and PCI was regarded as a harmful procedure with poor acute and long term results. However, PCI of LM is emerging as an alternative to CABG especially in patients with low syntax score with suitable coronary anatomy for PCI, and life-saving emergency situations like ACS with hemodynamically unstable, and high risk group patients who are unsuitable coronary anatomy for grafting or due to associated co-morbidities.

Andreas Gruntzig was the 1st person to perform first plain balloon angioplasty of LMCA in 1976. Later, O’Keefe et al. reported 127 angioplasty procedures of LMCA lesions with a procedural mortality of 9.1% and 3-year survival rate of 36% [21]. The beginning era of LM interventional management with plain balloon angioplasty was associated with a high mortality and morbidity due to abrupt vessel closure and acute stent thrombosis. Subsequently, Era of Bare-metal stents showed high restenosis and repeat revascularization rates with an increased incidence of sudden cardiac deaths.

In 1994 a meta-analysis of 7 studies that randomized a total of 2649 patients to medical therapy or CABG, showed survival advantage of surgery over medical therapy for patients with LMCA or three-vessel disease [22]. Later S J Park reported series of 42 patients with stenting of LMCA with immediate and late outcomes in 1998 and suggested stenting of unprotected LMCA as a safe and effective alternative to CABG in a carefully selected patient with normal LV Function with a 22% restenosis at 6 months follow up [23]. Erglis et al. analyzed results of PCI with stenting of LM, 103 patients with stable angina treated with either paclitaxel-eluting stent or bare-metal stent, IVUS and CB were used prior to stenting, which resulted in binary restenosis in 11 (22%) bare-metal stent and in 3 (6%) paclitaxel-eluting stent patients (p = 0.021) [24]. ISAR-LM randomized trial, comparing sirolimus-eluting stent vs. Paclitaxel-eluting stent, revealed no significant differences were reported in the composite outcome of death, MI, and TLR at 1 year follow-up, and no difference seen in TLF and 2-year LM-specific revascularization [25].

2.6 Early experience of PCI of LMCA with 1st generation DES

Acute and long term out comes of PCI of LMCA compared with CABG studied in 4 major trials (Table 1) using the 1st generation DES – LEMANS [26], SYNTAX left main [27], BOUDRIOT [28], and PRECOMBAT trails [29].

Trail	Recruitment period	N (PCI /CABG)	Longest follow-up, Y	Primary end point	Key findings	Strength	Weakness
LEMANS [26]	2001–2004	52/53	10	Change in left ventricular ejection fraction (LVEF)	Improvement in LVEF only with PCI, comparable rates of death, MI, stroke or TVR at 1 & 5 years	1st RCT comparing PCI & CABG for LM disease	Very small number of pts., surrogate primary end point, DES used only in 35%
SYNTAX-LM [27]	2005–2007	357/348	5	Death, MI, stroke or RR	PCI was non-inferior to CABG at 1 & 5 years	1st moderate sized RCT, mainly used for the current guideline recommendation	Sub-group analysis, only hypothesis generating
BOUDRIOT [28]	2003–2009	100/101	1	Cardiac death, MI or TVR	PCI was inferior to CABG at 1 year	1st RCT comparing sirolimus-eluting stents and CABG for LM disease	Limited sample size Lack of long-term follow-up, Stroke was not included in end pint
PRECOMBAT [29]	2004–2009	300/300	5	Death, MI, stroke or RR	PCI was non-inferior to CABG at 1 & 5 years	1st LM specific moderate sized RCT comparing DES & CABG for disease	Non-inferiority margin was wide, Routine angiographic follow-up in the PCI group

Table 1.

Prior trials of PCI vs CABG for LMCA in the era of the 1st generation DES.

LEMANS [26] trial is the first RCT showed PCI comparable rates of death, myocardial infarction (MI), stroke and target vessel revascularization (TVR) at 1 and 5 years with CABG for LM disease. BOUDRIOT [28] study of 100 PCI patient’s v/s 101 CABG patients showed PCI was inferior to CABG at one year. In PRECOMBAT [29] trial, 600 patients randomized to LMCA PCI with first-generation drug eluting stent (DES) or CABG which showed 17.5% major adverse cerebro-cardiovascular events (MACCE) in PCI group compared to 14.3% in CABG group at 5 years of follow-up, and no significant differences in the all cause death, MI and stroke with an increased target vessel revascularization in PCI group (11.4% vs. 5.5%). At 10 years, MACCE was 29.8% in the PCI group and 24.7% in the CABG group. No significant differences were found with respect to death, stroke, or MI. However, the incidence of TVR was significantly higher in the PCI group.

A subsequent study from the French Left Main Taxus (FLM Taxus) and the Left Main with Xience (LEMAX) registries [30], comparing 2-year outcomes using either everolimus eluting stent (EES) or paclitaxel eluting stent (PES), demonstrated a reduction in target lesion failure (TLF) – a composite of cardiac death, target vessel MI, and clinically driven TLR – with PES, by 53% at 2 years (EES: 7.6% vs. PES: 16.3%,). Significant differences in target vessel MI (PES: 9.9% vs. EES: 4.1%,) and target vessel failure (PES: 16.3% vs. EES: 7.6%) were associated with EES at 2 years. Furthermore higher SYNTAX Score groups (intermediate-high) demonstrated a trend towards improved clinical benefit in patients treated with EES compared to PES.

The SYNTAX trial [27] was initially published in 2009, and it remains the landmark study for decision-making and risk stratification of complex coronary artery disease (CAD), compared TAXUS (paclitaxel coated stent) v/s CABG consisting of 1800 patients randomized to PCI with TAXUS v/s CABG in Triple Vessel Disease and LM Disease (705 patients) involving 62 EU sites and 23 US sites.

Results were analyzed according to different sub groups based on SYNTAX score (Figure 3) - Low SYNTAX score < 22, intermediate score > 22 - <32, and high syntax score > 32. Study showed PCI was inferior to CABG for the composite primary end points of death, MI, stroke and unplanned revascularization [31]. SYNTAX study follow up results of 1 year and 5 years showed that CABG is associated with fewer major adverse cardiac and cerebrovascular events compared with PCI. The SYNTAX study emphasized the heart team concept and the SYNTAX score to assess the risk status by grading of the patients’ coronary disease burden. SYNTAX LM subset consisting of low and intermediate score 0–32 showed similar cumulative event rate 32.1% and 31.3% in CABG group v/s TAXUS group at 5 years follow-up and higher cumulative event rate at 5 years follow up in CABG compared to PCI 46.5% v/s 29.7% in LM subset with high SYNTAX SCORE > 33, and results demonstrated that surgery remains the gold standard for patients with complex multi vessel disease. 2009 American college of cardiology (ACC)/American heart association (AHA) guidelines recommended PCI of LMCA class II–b for noncomplex LMCA disease based on meta-analysis of multiple trails among patients with 1st generation DES showed that death, MI and stroke (major adverse cardiac events - MACE) were starting to show similarity in PCI and CABG patients at 1 year.

Figure 3.

Coronary segment weighting derived from Leamon score.

SYNTAX score guides in assessing severity and extent of CAD and provides information to take proper decision in planning appropriate revascularization strategy. The clinical SYNTAX score is a combination of age, creatinine and ejection fraction (ACEF) model and SYNTAX scores, and subsequent development of a logistic model has provided better risk assessment [31]. The SYNTAX II score is useful to predicts long-term mortality in patients with severe triple-vessel or LMCA disease, and it is assessed by considering anatomical and clinical factors (age, creatinine clearance, LV function, gender, chronic obstructive pulmonary disease, and peripheral vascular disease) along with SYNTAX score. It was found to be superior to the conventional SYNTAX score in guiding decision-making between CABG and PCI in the risk assessment. The STS (The Society of Thoracic Surgeons) score is a risk-prediction model, validated in patients undergoing cardiac surgery, with a specific model for CABG and combined CABG and valve surgery. It can be used to predict in-hospital or 30-day mortality and in-hospital morbidity [32]. ACCF/AHA guideline suggests that calculation of the SYNTAX and STS scores is reasonable in patients with unprotected LM and complex CAD (Class IIa recommendation, level of evidence; B).

The DELTA registry (n = 2,775), [33] a multi-center, multinational registry of LM PCI with first generation DES (n = 1,874, PES or SES) against CABG (n = 901) for ULMCA disease, no differences in the primary composite endpoint of all-cause death, CVA and MI were seen (HR 1.11; 95% CI 0.85–1.42; p = 0.47).

2.7 Experience of PCI of LMCA with 2nd generation DES

EXCEL [34] and NOBLE [35] are the two major trials to see efficacy of 2nd generation DES v/s CABG based on the evidence of superior results of 2nd Generation Everolimus drug eluting (XIENCE) in reducing MACE rate against TAXUS in SPRIT III and CIPHER in ISAR-test 4 trail with a lower mortality rate (22%) at 10 years. Unfavorable results reported in some trails of PCI of LMCA lesions may be due to use of earlier generation stents and to various technical issues such as catheter induced dissections, under expansion, uncovered diseased segments, multiple mal apposed stent layers, jailing of Ostium multiple struts, longitudinal compression of struts and accidentally crushed stent which will contribute to worse out comes .Similarly, many advances took place in the field of CABG by total arterial revascularization using left internal mammary artery (LIMA), right internal mammary artery (RIMA), and bilateral internal mammary artery (BIMA), on pump v/s off pump Bypass surgery, minimally invasive techniques, sternal sparing, endovascular harvesting, and hybrid philosophy.

The EXCEL trial [34] was a prospective randomized open-label, non-inferiority trial undertaken at 126 centers in 17 countries around the world. Study included 948 patients XIENCE group v/s CABG 957 patients with unprotected LMCA with >70% DS, or > 50–70% with either 1) Non-invasive evidence of LM ischemia, 2) IVUS MLA < 6.0 mm², or 3) FFR: <0.80 with SYNTAX SCORE < 32. SYNTAX score was ≤22 in 60.5%, >23 – <32 in 39.5%, and distal LMCA was present in 80.5% of the patients. IVUS guidance was used in nearly 80% of the patients in the PCI group. It included both stable and unstable angina but excluded patients of ST elevation myocardial infarction (STEMI). Distal bifurcating lesions were treated with a two-stent strategy using various techniques. CABG was performed both on- and off-pump, with the aim of complete revascularization for vessels with 50% stenosis. There was no difference between the two groups in respect to the primary composite end-point event of death, stroke, or myocardial infarction at 3 years (15.4% of the patients in the PCI group and in 14.7% of the patients in the CABG group), ischemia-driven revascularization was more frequent after PCI compared to CABG (in 12.6% vs. 7.5% of the patients, p < 0.001), and Stent thrombosis occurred in only 0.7% of patients within 3 years which was less common than symptomatic graft occlusion. EXCEL 5 years follow-up data showed primary end point - All cause death, stroke or MI 22.9% in PCI group v/s 19.2% in CABG group. In EXCEL study [34] IVUS use was not mandatory which was used in only 77% of patients, and no specific bifurcation technique was followed which was left to the operator discretion, and the use of proximal optimization technique (POT) and final kissing balloon (FKB) were also not specified (Table 2), these would have influenced study outcomes.

The Nordic-Baltic-British Left Main Revascularization Study [35] is a prospective, randomized, open-label, non-inferiority trial done at 36 centers in Europe. Patients with LMCA visually assessed with diameter ≥ 50% or fractional flow reserve ≤0.80 in different segments of the left main coronary artery were randomized to CABG or PCI with Biolimus eluting stent. SYNTAX score was calculated and all patients with low, medium, and high score were included. Patients were treated with the intention of achieving complete revascularization (CR). Distal LM bifurcation was treated with “Culotte” technique in majority of patients, enrolled 592 patients in each group, on-pump CABG was performed in 84% of patients, and LIMA graft was used in 96% of patients. Results showed higher MACCE rate in PCI group (28%) compared to CABG (18%) group due to higher MI and repeat revascularization in PCI group, but without significant difference in overall mortality and stroke rate. PCI group experienced lower stroke rate at 30 days of follow up compared to CABG group, but this difference was not seen at 1- and 5-year follow-up. In NOBLE study, only 75% underwent IVUS, FKB was performed in 55% of patients, 8% were implanted with first generation DES, and these factors might have influenced study outcomes to some extent.

2.8 Meta-analysis of PCI vs CABG

Individual patient data analysis from 11 PCI v/s CABG trails consisting of 11,518 randomized patients out of which, 4,394 (38.9%) (Figure 4) patients with LM disease showed all-cause mortality of 10.7% in CABG patient v/s 10.7% in PCI patients at follow-up of 5 years’ period which also showed mortality after LM DES v/s CABG 12.8 v/s 14.6% in SYNTAX, 5.7% v/s 7.9% in PRECOMBAT, 13.0% v/s 9.9% in EXCEL and 9.4% vs. 8.7% in NOBLE studies. Higher all-cause mortality was observed in PCI group v/s CABG in LM subset patients with diabetes 16.5% v/s 13.5% and patients with syntax score > 32–15% v/s 12.4% [35].

Figure 4.

Meta-analysis of PCI vs CABG.

2.9 Long term evidence of PCI of LMCA with DES

LM registry analysis of a total of 913 patients who underwent LM PCI at FU- WAI hospital between 2004 to 2008 revealed ten year outcomes of unprotected LM PCI in selected patients had acceptable results, though majority were implanted with 1st generation DES, and reduced 10 years mortality of PCI of LM was observed in lower risk patients stratified by SYNTAX Score and SYNTAX II, IVUS guidance, and usage of DES can significantly reduce 10 years mortality, stroke, and MI. Study suggested Age, LV EF, and incomplete revascularization are independent predictors of 10 years death or MI [36].

MAIN COMPARE Study of LMCA stenting v/s CABG showed The rate of target vessel failure (TVF), risk of death, and serious composite outcomes higher in PCI compared to CABG after 5 years [37] & 10 years [38] follow up results showed no significant difference in the rates of death and composite end points of death, Q wave MI and stroke between PCI and CABG groups.

The SYNTAXES study (Synergy Between Percutaneous Coronary Intervention with TAXUS and Cardiac Surgery Extended Survival) is the 10-year follow-up of the original SYNTAX trial [39], comprising 72% of the syntax 10 years’ data (No: 1301), showed a comparable survival rate between CABG group 26.7% and PCI group 26.1% at 10 years (Figure 5).

Figure 5.

SYNTAX left Main at 10 years: Mortality.

2.10 Evidence of PCI vs CABG for ostial and shaft LMCA

Meta-analysis of studies comparing the clinical outcome (MACE) in 3291 patients receiving PCI with DES stenting of Ostial & Mid shaft showed favorable outcome compared with distal LM lesions. Excel study also revealed better 3 years’ outcome (Death, MI or Stroke) after Ostial and shaft lesions of LM (CABG 13.5% v/s PCI 12.4%) compared with LM bifurcation v/s CABG (CABG 14.9% v/s 15.6% PCI) [34].

2.11 General principles of PCI of LMCA

PCI of LM is a high risk interventional procedure requires meticulous planning, adequate skills and experience to produce best possible results. PCI of ostium and shaft carries better results compared to LM bifurcation. PCI of distal LM associated with more risk and increased MACE due to requirement of more number of stents in complex lesions, and tendency for increased restenosis at ostium of LCX.

High risk may be due to associated co-morbidities, complex coronary anatomy and hemodynamic compromise status. Most of the patients undergoing LM PCI do not require hemodynamic support, but the operator should consider it if he anticipates or encounters any hemodynamic compromise, slow or no reflow or other procedural complications (Figure 6). During pre-procedure evaluation of LMCA revascularization screen for its association with carotid artery disease, cerebrovascular disease and peripheral artery disease, aortic aneurysm, and porcelain aorta which make the surgical procedure more challenging. Compared to unprotected lesions, complications of PCI (abrupt closure and restenosis) of protected LMCA lesions are more often well tolerated because of continued flow to the protected territory [40]. Significant LMCA disease more often associated with Carotid artery disease which is seen in nearly 40% of patients undergoing angiography for angina. The AHA guidelines recommend screening of all patients undergoing bypass surgery for left main stem disease to identify carotid artery disease [41].

Figure 6.

Challenges for LM PCI for interventional cardiologists in high-risk procedures.

Take precautions to reduce contrast volume to avoid contrast induced nephropathy (CIN) and exposure to radiation while performing high-risk, complex LM PCI. Use appropriate devices to expedite the procedure by using guide extension catheters, adequate guide catheter backs up for good support, micro catheters, guide wires, balloons to avoid complications related to procedure delay, and use standard current generation DES for best possible results. It would be preferable to keep thin profile balloons to cross critical lesions, high pressure balloons, and scoring balloons to tackle tough un-dilatable lesions which are difficult to dilate with regular NC balloons to improve procedural success and long term outcomes.

2.12 Advantages of PCI over CABG

PCI is less invasive with fewer peri -procedural complications, fewer 30 day MACE, early rapid recovery with better quality of life (QOL) and earlier angina relief. It is preferable in patients who require urgent revascularizations mainly in ACS setting, coexisting serious co-morbidities and comes under high surgical risk (ie., chronic lung disease, advanced age, history of previous stroke, and prior Bypass surgery).

2.13 Revascularization strategy of angiographically significant LMCA disease during Acute STEMI-PCI vs CABG

Revascularization of Acute STEMI patients with significant LM lesions depends upon the culprit vessel, type of MI, dominant vessel, complexity of coronary disease and anatomy, and hemodynamic status. In patients with Acute Inferior STEMI with significant LM (non-culprit) with multi vessel disease, and culprit vessel is right coronary artery (RCA), recommended to perform primary PCI of culprit vessel only with optimal medical therapy for bystander lesions and PCI/CABG of the non-culprit arteries only for spontaneous angina or myocardial ischaemia on stress testing or LM with multi-vessel (MV) PCI guided by angiography or FFR after finishing culprit PCI during same sitting. Primary PCI of Culprit vessel only, followed by angiography or FFR-driven staged PCI of non-culprit arteries during the index hospitalization or after hospital discharge.

Acute STEMI with cardiogenic shock – Recommended to perform culprit vessel PCI initially, then PCI of LM and other vessels is reasonable option if no improvement in hemodynamic status but should be differed if hemodynamic status improves after culprit PCI (Figure 7). If culprit vessel is LAD or LCX in STEMI patients with significant LM disease, consider PCI of LM along with PCI of LAD or LCX. If the patient is having STEMI with LM and multi-vessel disease (MVD), perform Primary PCI of culprit vessel, and if the non-culprit vessel or vessels are having significant disease (>70% stenosis), complete the revascularization of all diseased vessels during original hospital stay. If the non-culprit lesion has intermediate lesion (40–70% stenosis), perform PCI of non-culprit lesions under FFR/ iFR guidance or 5–7 days after massive MI.

Figure 7.

A systemic algorithm for LMCA disease in patients with ACS.

2.14 Early vs Delayed CABG in Acute STEMI

Early CABG is associated with mortality rate (MR) in acute MI, preferable to postpone surgery for 3 to 5 days in the absence of absolute indications for CABG due to high mortality rate with very early surgery. Multicenter study of 32,099 cases reported by Lee and colleagues, showed mortality rate of 14.2% if operated within 6 hours, vs. 2.7% if operated beyond 15 days [42]. Thielmann et.al reported mortality rate of 23.8% if operated between 7 and 24 hours, and 2.4% if operated between 8 to 14 days period [43]. Early CABG in STEMI is associated with reduction in the size of infarct, and reduces the potential for mechanical complications whereas late CABG associated with reperfusion injury and increased systemic inflammatory response syndrome (SIRS).

In STEMI and multi vessel disease associated with cardiogenic shock short-term mechanical support device (e.g., percutaneous cardiopulmonary support, extra corporeal membrane oxygenation (ECMO), or ventricular assist device) with / without intra-aortic balloon pump (IABP) may be considered as a rescue therapy in patients with refractory circulatory support.

2.15 Hemodynamic support during LM PCI

Mechanical circulatory support devices should be considered while performing high risk LM interventions particularly in the presence of severe LV dysfunction, unstable status hemodynamic status reflected by left ventricular end diastolic pressure (LVEDP) >20 mm Hg, systolic BP <100 mm Hg, or mixed venous oxygen saturation < 55%, and complex procedures requiring longer time specially while handling diffusely calcified multi vessel disease, or single surviving vessel to avoid sudden hemodynamic collapse which might result in stoppage of procedure in midway, or might not get enough time to carry effective cardiopulmonary resuscitation (CPR) to revive the patient.

2.16 Influence of stenting technique and optimization on LM PCI outcomes

Provisional stenting is the technique of choice in bifurcation lesions, as it is technically simpler with improved clinical outcomes to a systematic 2-stent strategy. Many Bifurcation trials like NORDIC [44], BBC ONE [45], BBK [46], CACTUS [47] have not shown any benefit associated with systematic two-stent strategies, and EBC TWO study [48] also showed worse outcomes with systematic dual stenting even in patients with larger, true bifurcations. Ample evidence from non-randomized trials showing worse outcomes for two-stent techniques. However, randomized data from Dr. Shao-Liang Chen et al. support Double Kissing (DK)-crush in left main bifurcations Patients with true bifurcation should be treated with two stents preferably with DKC, because of recent evidence of better results, compared to provisional stenting, and other two stent techniques [49]. DKCRUSH II showed that a 2-stent strategy using the double kissing (DK) crush technique is superior to provisional stenting particularly in more complex lesions [50]. DKCRUSH-III, showed superior results of DK crush over Coulotte technique at 3 years with lower MACCE rate (8.2% vs. 23.7%) and stent thrombosis (0% vs. 3.7%) [51]. DKCRUSH-V study showed superior results DK crush technique compared to provisional stenting in distal LM bifurcation in terms of lower TLF at 1 year (5% versus 10.7%) and stent thrombosis (0.4% versus 3.3%) [52]. Recently, published EBC MAIN was designed to examine clinical outcomes in patients are treated equally well with a stepwise layered provisional approach, starting with a single stent, as with a more complex dual stent implant, and Only one-fifth of patients in provisional group required second stent showed Procedure time, X-ray dose and consumables were less, fewer adverse events systematic provisional (n-230) vs. systematic dual approach (n − 237) with a MACE rate of 14.7% vs. 17.7%, Death 3%vs. 4.2%, MI 10% vs10.1%, TLR 6.1% vs. 9.3%, and stent thrombosis (ST) 1.7% vs. 1.3%, they concluded that The stepwise provisional strategy should remain the approach of choice for the majority of left main bifurcation interventions [53]. Angio graphic ISR more frequent in lesions with under-expansion than without (24.1% Vs 5.4%), and proper use of kissing balloon and POT is essential to get the good expansion and proper opposition of stent struts. In the 2 stent group, the lesions with complete expansion of all sites showed a restenosis of only 6%, similar to that in the single stent group (6.3%). It would be preferable to achieve post bifurcation PCI MSA of LM > 8 mm², LM confluence >7mm², LAD ostium >6mm² and LCX ostium >5mm² or aim to achieve mean reference diameter of stented area at least >80% compared to proximal and distal reference areas. Post-stenting under-expansion was an independent predictor of 2-year MACE, especially repeat revascularization (Figure 8).

Figure 8.

Criteria for stent under-expansion at the distal LMCA bifurcation.

Intravascular imaging and FFR should be used to optimize DES results, to stent only physiologically significant lesions, to avoid unnecessary stenting thereby reduces number of stents. Kang et al. evaluated IVUS predictors of ISR after LM bifurcation stenting, and post-stenting IVUS mean stent area (MSA) cut-offs that best predicted ISR on a segmental basis were 5.0 mm² (ostial LCX), 6.3 mm² (ostial LAD), 7.2 mm² (POC, confluence zone of LAD and LCX), and 8.2 mm² (LM above the POC). A smaller IVUS-MSA within any one of these segments was responsible for a higher rate of angiographic ISR and clinical major adverse cardiovascular events (MACE) [54].

IVUS guided PCI of LM stenting associated with trend towards decreased mortality that is 13.6% v/s 6.0%. ADAPT – DES [55] study and ULTIMATE study [56], and recent IVUS meta-analysis also highlighted the same, implement the optimal stenting techniques and optimize DES implantation with IVI to provide best acute and long term results.

2.17 CABG vs PCI of LMCA in diabetics

CABG is the standard revascularization strategy in patients with diabetes mellitus (DM) and multi-vessel or complex CAD with long term favorable outcomes. Recent evidence suggests that PCI is a safe and effective modality for patients with LMCA disease with <22 and > 22 to <32 syntax score as compared with CABG, no significant difference in the 10-year risks of mortality and serious composite outcome after PCI or CABG in patients both with and without DM, but the risk of TVR was consistently higher after PCI. In a recent pooled analysis of individual patient data, the presence of DM was reported to have a significant interaction effect for 5-year mortality favoring CABG over PCI in patients with multi-vessel CAD, but not in those with LMCA disease. These findings also confirmed the impact of DM with respect to the primary composite end point and mortality in the subgroup analysis of the EXCEL trial with low-to- intermediate SYNTAX scores.

The FREEDOM trial 8 years’ follow-up data showed that CABG leads to lower all-cause mortality than PCI in patients with DM with multi vessel disease [57]. The benefit of CABG in patients with DM might be attributed to complete revascularization in more diffuse and complex multi vessel CAD. By contrast, moving on to the DES era, DM did not appear to modify the treatment effects of PCI and CABG for LMCA disease.

2.18 Influence of Hospital and operator volume on LM PCI outcomes

Hospital and operator volume also impacts the outcomes of LMCA PCI, and results are better in a center with a high volume and operated by high volume operator (Figure 9), all cause death (0.5% v/s 2.1%) and cardiac death (0.5 v/s 2.1%). Study revealed that results are better with the operators who were performing at least 15 PCI of LMCA per year in 3 consecutive years [58].

Figure 9.

Impact of operator volume on outcomes of LMCA PCI.

2.19 Heart team approach – Risk assessment of CABG vs PCI

LM PCI out comes can be improved with the proper selection of patients after assessing risk v/s benefit by involving Heart team with adequate counseling and education about decease nature and various modes of treatment options. All patients with LM disease should be assessed with SYNTAX, Functional SYNTAX, SYNTAX II, EURO & STS scores (Table 3) to assess risk and mortality rate and to decide plan of treatment involving Heart team. The most widely used surgical risk score is the Society of Thoracic Surgeons score. It classifies operative risk based on predicted risk of mortality into low (<4%), intermediate (4% to <8%), high (8% to <12%), or extreme (≥12%). The newly incorporated “Functional SYNTAX score” (Functional SXscore) essentially incorporates FFR measurements into the SYNTAX Score calculation, and was recently shown potentially to improve the stratification of low and high risk patients, when compared to the conventional visual-based angiographic approach [59]. The heart team should assess risks and benefits of surgery in the high- and extreme-risk population, and careful evaluation of clinical history, physical examination for co-morbidities with necessary investigations to plan for appropriate revascularization strategies after proper and repeated counseling of the patient and his or her relatives. Special attention should be focused on frailty, cognitive status, acute and long term results, importance of life style modification including stopping of smoking and long term usage and adherence of drug therapy to control risk factors and to prevent recurrence of angina and MI from progression of atheroma in grafts and native vessels.

Recommendations according to extent of CAD	CABG		PCI
	Class	Level	Class	Level
Left main CAD
Left main disease with low SYNTAX score (0–22)	I	A	I	A
Left main disease with intermediate SYNTAX score; (23–32)	I	A	IIa	A
Left main disease with high SYNTAX score (≥ 33)	I	A	III	B

Table 2.

Recommendations for the type of revascularization with left main disease.

Currently, in the US guidelines (Table 2), PCI has a class IIa recommendation (“is reasonable”) in select patients with isolated LM stenosis involving the ostium or shaft and without coexisting multi vessel disease and the risk of surgical bypass is increased. PCI has a class IIb recommendation (“may be reasonable”) in patients with LM stenosis involving the distal bifurcation or with less complex coexisting multi vessel disease as defined by a low or intermediate SYNTAX score (≤33) and who have an elevated surgical risk. The current US guidelines recommend against PCI in patients who are good candidates for surgical bypass with coexisting complex multi vessel disease as defined by highest tertile of the SYNTAX score (≥33). Hybrid bypass is another revascularization approach m that combines coronary bypass using a minimally invasive direct coronary artery bypass approach of grafting the LIMA to LAD artery and PCI to the remaining vessels in an attempt to achieve the most desired aspects of each revascularization strategy. Always aim for complete revascularization because major adverse cardiovascular events including mortality are higher in patients with incomplete revascularization than those with complete revascularization regardless of the revascularization strategy.

Recommendations:	Class	Level
Assessment of surgical risk
It is recommended that the STS score is calculated to assess in-hospital; or 30 day mortality, and in-hospital mortality after CABG	I	B
Calculation of the EuroSCORE II score may be considered to assess in-hospital mortality after CABG	IIb	B
Assessment of CAD complexity
In patients with LM or multi-vessel disease, it is recommended that the SYNTAX score is calculated to assess the anatomical complexity of; CAD and the long-term risk of mortality and morbidity after PCI	I	B
When considering the decision between CABG and PCI, completeness; of revascularization should be prioritized	IIa	B

Table 3.

Criteria for the choice between PCI and CABG.

PCI should be considered, if the Heart Team is concerned about the surgical risk or if the patient refuses CABG after adequate counseling by the Heart Team.

2.20 Who should be treated with CABG?

CABG has more durable long term outcome with fewer adverse events beyond 30 days particularly MI due to protection against future events, improved long term relief of angina and repeat revascularization procedures particularly in more complex anatomies more often due to complete revascularization. It is preferred in patients with poor LV function, long standing DM, concomitant cardiac surgery, high bleeding risk patients who are unable to comply with dual anti-platelet therapy (DAPT). CABG scores over PCI in patients with LM with addition Triple Vessel Disease, unsuitable for PCI due to complex anatomy, severely calcified and tortuous of coronaries, chronic total occlusion (CTO), Multiple diffuse long segment lesions, and complex ISR lesions. Patients with critical LMCA disease with severe symptoms or with life-threatening ventricular arrhythmias, which is believed to be ischemic in origin should be subjected for early surgery because of increased incidence of sudden death. CABG surgery associated with in hospital mortality rate of 1%, and < 3% perioperative MI in low-risk patients Surgery associated with increased incidence of Peri -operative MI, bleeding and transfusions, arrhythmias, renal failure, increased incidence of sternal dehiscence, and Repeat revascularization. Predictors of increased mortality after CABG are emergency procedure, extreme age, past history of cardiac surgery, female gender, LV dysfunction, severity of LM stenosis, and number of vessels with significant stenosis.

2.21 On pump vs off pump CABG

CABG with on-pump surgery considered as a preferred and standard revascularization procedure for 80% of CAD patients after seeing successful results of series of surgical cases following first CABG in the late 1960s. On-pump surgery has problems related to manipulation of the ascending aorta leading cerebrovascular accidents particularly in patients with aotic-atheroma and porcelain aorta, myo-necrosis due to aortic occlusion, cognitive dysfunction, renal failure, and systemic inflammatory response syndrome. Whereas, off-pump surgical technique overcomes these limitations, and is more often associated with hemodynamic instability, mainly in patients with recent MI, LV dysfunction, dilated ventricles, and while grafting the branches of the LCx in patients with significant mitral insufficiency, and less complete revascularization. Compared to traditional on-pump CABG with LIMA to LAD, irrespective of SVG or arterial grafts to other vessels neither off-pump CABG nor the use of bilateral internal mammary arteries has been shown to improve CABG outcomes in RCTs.

2.22 Influence of arterial vs venous grafts on long term outcomes of CABG

Advantages of CABG over PCI - PCI treats an isolated lesion in the proximal vessel, complexity of the lesion affects clinical outcome, CABG by passes the proximal 2/3 of the vessel, where current lesions and future threatening lesions can occur. This advantage will persist, even if stent restenosis is zero. The LIMA is the ideal graft of choice to bypass the LAD artery (Class I recommendation for ACCF/AHA guideline for CABG surgery) due to graft patency of LIMA is >90% after 10 years. LIMA is resistant to atherosclerosis, and release prostacyclin and nitric oxide contributing to vasodilation, inhibition of platelet function, and improved survival rate which is independent of the patient’s sex, age, extent of CAD, and LV systolic function. LIMA to LAD graft decreases the occurrence of late MI, reoperation, recurrence of angina, and repeated hospitalizations. Radial artery graft patency results are better when it is grafted to Lcx with >70% narrowing and worst when it is used to graft the RCA with a stenosis of only moderate severity. Patency of Radial artery grafts better when used for >90% lesions, and choose the radial artery of non-dominant upper-limb, and radial artery with a > 2 mm diameter after testing modified Allen test for ulnar dominance. Combination of LIMA, RIMA, radial artery, and or gastroepiploic artery may be used for full arterial revascularization processes. Reversed saphenous vein grafts are also routinely used in combination with LIMA, RIMA, and other arterial grafts in patients undergoing CABG surgery depending upon suitability of grafts. Saphenous vein grafts (SVG) have a track record of poor long-term patency with a closure rate of about 10–25% during 1st year post CABG period, an additional closure rate of 1–2% each year during the 1–5 years of post-surgery, and 4–5% occlude each year between 6 and 10 years postoperatively, with an overall 10 years’ patency of SVGs is about 50–60%. Major determinants of graft selection are age, severity of narrowing of the vessel and hemodynamic status of the patient. LIMA-LAD graft should be offered to all patients with LMCA disease undergoing CABG either with on-pump or off-pump with aim of total revascularization.

2.23 Post CABG progression of atherosclerosis

Post CABG patients develops progression of atherosclerosis in native vessels which was accelerated by vein grafts and observed in over 50% of the native vessels, 35% of native coronary arteries bypassed with a venous grafts progressed to total occlusion in 35% of SVG compared to 8% of LIMA. Vein graft failure associated with increased death, MI and revascularization.

2.24 Adjunctive Therapy and supportive measures for LM disease

All patients with LMCA with or without revascularization should be emphasized about the need to continue adequate guideline directed therapies such as DAPT, anti-Hypertensive (calcium channel blockers (CCB), angiotensin converting enzyme inhibitors (ACEI), angiotensin receptor blockers (ARB), and beta blockers (BB)), anti-Diabetic drugs (SGL2 inhibitors, and Metformin), high dose statin therapy, and drugs for LV Dysfunction (ARNIs, SGL2 inhibitors, aldosterone antagonists, and beta blockers) to provide better long term results of revascularization. Patients with poly vascular disease should be recommended with Rivaroxaban 2.5 mg bid, and aspirin 100 mg, diabetics with renal dysfunction and heart failure should be supplemented with SGL2 inhibitors to reduce repeat hospitalizations, and MACE rates. During follow up all patients should be monitored for any recurrence of symptoms which requires appropriate evaluation with necessary investigations and adequate control of risk factors and life style modification.

3. Conclusion

Angiographic assessment of borderline LM lesions (30–70%) is inaccurate with significant inter-observer variability.1 Intravascular imaging is helpful in assessing severity and to decide revascularization strategies particularly in patients with angiography showing lesions of uncertain severity, and recent evidence is in more favor of image guided PCI over angio guided PCI with improved clinical outcomes.

Acknowledgments

The author would like to thank Mr. Manikandhar Pendyala and Mr. Chandrashekar Challa for their assistance in the manuscript preparation.