Academic Editors: George Dangas and Christian Hengstenberg
Revascularization therapy significantly improves the outcomes of patients with left main coronary artery disease (LMCAD), compared with medical therapy alone. For many years, coronary artery bypass grafting (CABG) has been the primary and standard treatment strategy. However, with advances in percutaneous coronary intervention (PCI) techniques and improvements in patients’ outcomes, there is growing evidence supporting PCI for LMCAD. In this review, we aim to integrate the available evidences on advances in PCI treatment for LMCAD and provide guidance for further clinical practice.
Significant stenosis of the left main coronary artery (LMCA) is present in around 7% of patients undergoing coronary angiography . Since approximately 2/3 of the cardiac blood flow is supplied by the left main artery, patients with significant left main coronary artery disease (LMCAD) often have poor prognosis . Revascularization therapy greatly improves the outcomes of patients with left main disease, compared with medical therapy alone . Traditionally, coronary artery bypass grafting (CABG) has been the primary treatment strategy, however, with advances in percutaneous coronary intervention (PCI) techniques and improvements in patients’ outcomes, more and more patients with LMCAD are receiving PCI, and there is growing evidence supporting PCI for LMCAD. In this review, we aim to integrate the available evidence on advances in PCI treatment for left main coronary artery disease and provide guidance for further clinical practice.
There has been controversy for many years over whether PCI in the era of drug-eluting stents (DESs) can replace or even be superior to CABG in treating patients with LMCAD.
The latest clinical findings in recent years, especially the 5-year follow-up
results of the four most representative large randomized controlled trials (RCT),
including SYNTAX, PRECOMBAT, NOBLE, and EXCEL, have brought more conclusive
evidence and deeper understanding on how to choose the best revascularization
strategy for LMCAD. The detailed 5-year follow-up results from these RCTs are
shown in Table 1. The 5-year follow-up results from the SYNTAX (Synergy Between
PCI With Taxus and Cardiac Surgery) trial  which enrolled 705 patients with
LMCAD (PCI 357 vs. CABG 348) showed that no significant difference was noted
between PCI and CABG in overall major adverse cardiac and cerebrovascular events
(MACCEs) (36.9% vs. 31.0%, p = 0.12) nor in mortality (12.8% vs.
14.6%, p = 0.53). A lower incidence of stroke (1.5% vs. 4.3%,
p = 0.03) was seen in patients who received PCI instead of CABG, while a
significant increase of repeat revascularization was observed in the PCI group
(26.7% vs. 15.5%, p
|Primary outcomes at 5-year follow-up (PCI vs. CABG)|
|No. of patients||705||600||1905||1184|
|357 vs. 348||300 vs. 300||948 vs. 957||598 vs. 603|
|SYNTAX score (mean)||30||25||20||22|
|Overall MACCEs||36.9% vs. 31.0%||17.5% vs. 14.3%||-||28% vs. 19%*|
|All death/stroke/MI||19.0% vs. 20.8%||8.4% vs. 9.6%||22.0% vs. 19.2%||-|
|All-cause mortality||12.8% vs. 14.6%||5.7% vs. 7.9%||13.0% vs. 9.9%*||9% vs. 9%|
|Cardiac death||8.6% vs. 7.2%||3.8% vs. 6.9%||5.0% vs. 4.5%||4% vs. 4%|
|Stroke||1.5% vs. 4.3%*||0.7% vs. 0.7%||2.9% vs. 3.7%||4% vs. 2%|
|Myocardial infarction||8.2% vs. 4.8%||2.0% vs. 1.7%||10.6% vs. 9.1%||8% vs. 3%*|
|Total repeat revascularization||26.7% vs. 15.5%*||13.0% vs. 7.3%*||17.2% vs. 10.5%*||17% vs. 10%*|
The SYNTAX score is an angiographic assessment of the coronary vasculature and
reflects complexity of coronary artery disease, with scores
In recent years, 10-year follow-up studies on PCI and CABG have also been released, revealing a longer-term comparison between efficacy of the two treatment strategies. The detailed 10-year follow-up results are shown in Table 2. The PRECOMBAT 10-year follow-up data  showed no significant difference in death of any cause nor composite endpoint events including death, MI, and stroke in ten years between the two groups, while ischemia-driven target vessel revascularization was higher after PCI by 2-fold (16.1% vs. 8.0%). However, due to the unexpectedly low event rates of the PRECOMBAT trial, the study was underpowered. Thus, the lack of yearly clinical follow-up between 5 and 10 years after 5 years of follow-up may have led to an underestimation of clinical event rates. The LE MANS (Left Main Coronary Artery Stenting) trial  also showed no difference between PCI and CABG in terms of mortality rate and MACCEs rate, but the study was extremely limited as it only included 105 patients. The MAIN-COMPARE (Revascularization for Unprotected Left Main Coronary Artery Stenosis: Comparison of Percutaneous Coronary Angioplasty Versus Surgical Revascularization) observational registry  enrolled 2240 patients from the real world and showed that the composite endpoints such as death, MI and stroke were similar in both groups, but repeat TVR was significantly higher in the PCI group. Finally, in 10-year follow-up data from the SYNTAX extended survival study , no statistically significant difference was discovered in mortality rate between PCI and CABG. As in the above-mentioned follow-up studies, the stents used for PCI were former bare-metal stents (BMSs) or first-generation DESs, but the long-term comparison revealed that PCI was not inferior to CABG in terms of mortality.
|Primary outcomes at 10-year follow-up (PCI vs. CABG)|
|No. of patients||1800||600||105||2240|
|903 vs. 897||300 vs. 300||52 vs. 53||1102 vs. 1138|
|SYNTAX score (mean)||29||25||25||-|
|Overall MACCEs||-||29.8% vs. 24.7%||51.1% vs. 64.4%||-|
|All death/stroke/MI||-||18.2% vs. 17.5%||-||25.0% vs. 24.6%|
|All-cause mortality||27% vs. 24%||14.5% vs. 13.8%||21.6% vs. 30.2%||22.2% vs. 21.4%|
|Cardiac death||-||7.8% vs. 8.7%||-||-|
|Stroke||-||1.9% vs. 2.2%||4.3 vs. 6.3%||-|
|Myocardial infarction||-||3.2% vs. 2.8%||8.7 vs. 10.4%||-|
|Total repeat revascularization||-||21.3% vs. 10.6%*||26.1% vs. 31.3%||22.6% vs. 5.4%*|
Some other meta-analyses [14, 15] showed that in low and intermediate complexity of LMCAD, risk of all-cause death and cardiac death after PCI were comparable to that after CABG, while risk of repeat revascularization was significantly higher after PCI.
Although there is not sufficient evidence to date that PCI can be a perfect substitute for CABG, an increasing number of studies have proved that with advances in PCI techniques [4, 5, 6, 7, 10, 11, 12, 13], the efficacy of PCI is approaching or even comparable to that of CABG regarding all-cause mortality and cardiac death for patients with low and moderate complexity of LMCAD, however, it is undeniable that ischemia-driven revascularization is still higher with PCI. Currently, PCI and CABG should be complementary treatment strategies for stable LMCAD, and how to choose the best treatment strategy should depend on a full evaluation of many factors including anatomy of the coronary artery, the SYNTAX score, overall medical conditions, and preference of the patient.
Left main coronary artery disease can be classified into ostial, midshaft, and distal lesions according to location of the stenosis. Compared with distal bifurcation LMCAD, ostial/midshaft lesions are often regarded less challenging with more favorable outcomes, and the efficacy of PCI and CABG were comparable in terms of ostial/midshaft lesions in 5-year follow-up study . The 12-year follow-up results of another observational study  also showed that the long-term outcomes of PCI and CABG were generally comparable for ostial/midshaft left main lesions. In addition, Yoon et al.  analyzed 2112 cases with ostial/midshaft LMCAD from the IRIS-MAIN (Interventional Research Incorporation Society-Left MAIN Revascularization) registry study (PCI 1329 vs. CABG 783) and stratified the patients by stent type availability as BMSs, first-generation DESs and second generation-DESs. The results revealed that with iterations of stenting techniques, the efficacy of PCI treatment for LMCAD improved greatly, approaching the efficacy of CABG. However, the distal bifurcation lesions are more commonly seen in patients with LMCAD, accounting for approximately 80% of all left main lesions , the prognosis of which is worse than ostial/midshaft lesions due to the involvement of bifurcation and complicated anatomy [20, 21]. The 12-year follow-up results of the MAIN-COMPARE trial  found that there was no difference in the rate of composite endpoint events and mortality between PCI and CABG in ostial/midshaft lesions, while the efficacy of PCI was significantly inferior to that of CABG in terms of distal LMCAD. A subgroup analysis of the EXCEL trial  showed that, as for low and moderate complexity of LMCAD with SYNTAX scores below 32, the composite endpoint events defined as death, MI and stroke at 3 years of follow-up were comparable after PCI and CABG with both distal LMCAD (PCI 15.6% vs. CABG 14.9%, p = 0.61) as well as ostial/midshaft LMCAD (PCI 12.4% vs. CABG 13.5%, p = 0.77), and no difference in ischemia-driven repeat revascularization was seen with ostial/midshaft lesions between the two treatment strategies (PCI 9.7% vs. CABG 8.4%, p = 0.68), while repeat revascularization with distal bifurcation lesions after PCI was significantly higher than that after CABG (13.0% vs. 7.2%, p = 0.0001).
As a result, the current available evidence suggests that the overall efficacy of PCI for ostial/midshaft LMCAD is close to or even comparable to that of the standard treatment CABG, whereas PCI for distal LMCAD is still slightly inferior to CABG, with the main weakness being the higher rate of ischemia-driven repeat revascularization. When evaluating and considering PCI as the treatment strategy for LMCAD, the 2018 ESC/EACTS Guidelines  suggested based on available evidence that PCI can be considered for patients with low to moderate SYNTAX scores (0–22, level of evidence: I; 22–32, level of evidence: IIa) regardless of the location of the diseased LMCA, whereas the AHA/ACC guidelines  gave different recommendations of PCI for LMCAD according to the location of the lesions, namely a class IIa recommendation for ostial/midshaft LMCAD and a class IIb recommendation for distal LMCAD (level of evidence: IIb).
When treating LMCAD with PCI, the first decision to make is whether to propose a
single-stenting or dual-stenting technique. According to current consensus, a
provisional single stent is the preferred strategy for LMCAD without significant
lesions at the side branch ostium . A secondary analysis of the EXCEL trial
 showed that the provisional strategy was better than upfront dual-stenting
for LMCAD without true distal bifurcation lesions in reducing the composite
endpoint events including death, MI and stroke at 3 years of follow-up (13.8%
vs. 23.3%, p = 0.03). In the case of true distal left main bifurcation
diseases, the 3-year follow-up results of the DKCRUSH-V trial (Double Kissing
Crush versus Provisional Stenting for Left Main Distal Bifurcation Lesions: The
DKCRUSH-V Randomized Trial)  revealed that double-kissing crush
(DK-Crush) dual-stenting was superior to the provisional
strategy in composite endpoints defined as target lesion revascularization (TLR),
target vessel myocardial infarction (TV-MI) and cardiac death (8.3% vs. 16.9%,
The distinguished results of the two studies [28, 29] might be due to the following
reasons. Firstly, the lesions included in the DKCRUSH-V trial were more complex
with a mean SYNTAX score of 31 and side branch lesions
In addition, the recent DEFINITION (Definitions and impact of complEx
biFurcation lesIons on clinical outcomes after percutaNeous coronary IntervenTIOn
using drug-eluting steNts) II study  with similar side branch lesion
enrollment criteria as DKCRUSH-V including LCX stenosis severity
These findings suggest that, for non-true distal bifurcation LMCA lesions, the
provisional single stenting is the preferred standard treatment strategy, and for
true distal bifurcation lesions, efficacy of provisional stenting and upfront
dual-stenting are equivalent, whereas for complex true bifurcation lesions with
side branch ostia stenosis
Currently, the most commonly used dual-stenting techniques are
T-stent, T and protrusion (TAP), Culotte, and DK-Crush. It is
important to note that T-stent, TAP, and Culotte techniques can be used in a
provisional strategy in a bailout indication such as dissection and flow
compromise etc. as well as in a planned dual-stent strategy. The BBK
(Bifurcations Bad Krozingen) II study  showed that the Culotte
technique was able to reduce risk of restenosis (6.5 vs. 17%, p =
0.006) and target lesion failure (6.7% vs. 12.0%, p = 0.069) compared
to TAP. The DEFINITION III study  revealed that the
DK-Crush technique was superior to the Culotte technique,
reducing rate of TVR (4.3% vs. 11.0%, p
The proximal optimization technique (POT) is a mandatory step for unprotected left main coronary artery lesions regardless of true bifurcation lesions. The POT involves post dilation with a short balloon larger than the stent size positioning with its distal end in front of the carina, thus adapting the stent frame to the bifurcation anatomical morphology and correcting the proximal segment malapposition or underexpansion. A recent clinical registry study  enrolling 4395 LMCAD patients with bifurcation lesions supported the application of POT to reduce the incidence of TLF in true bifurcation lesion after PCI. Another commonly used technique for bifurcation lesions is kissing balloon inflation (KBI), which is a standard step when using a dual-stenting technique, but whether it is necessary as a routine procedure in a single-stent strategy remains controversial. The COBIS (Korean Coronary Bifurcation Stenting) II registry study  showed that final KBI in single-stent approach is associated with favorable long-term clinical outcomes and reduced TLR. The recent AOI-LMCA (Assessing Optimal Percutaneous Coronary Intervention for LMCA) study , on the other hand, indicated that performing final KBI did not affect TLR and other clinical outcomes at 5 years of follow-up. In the recent EXCEL study, the 4-year composite end points such as death, MI and stroke were not significantly different with or without final KBI for either single or dual stenting. The result that there was no evident advantage of final KBI for dual-stenting was quite unexpected, since final KBI was regarded essential for dual-stenting techniques with much clinical evidence to support it . The recent RAIN-CARDIOGROUP VII study (Very Thin Stents for Patients With MAIN or BiF in Real Life: The RAIN, a Multicenter Study)  revealed that routine final KBI with dual-stenting reduced TVR and restenosis, and balloon with short proximal overlap contributed to better outcomes. However, the impact of the interaction between POT and KBI on procedural outcomes is still not clearly elucidated, and future studies are necessary to further clarify the role of the interaction between the two on procedural outcomes.
Transfemoral access (TFA) has been the traditional approach for PCI as it can provide easy access for catheters, stents or other devices through its larger caliber . Over the years, more and more evidence has supported the transradial access (TRA) owing to less bleeding, fewer vascular complications, and better patients experience following TRA [40, 41]. As there are limited studies reporting the access types used in PCI for LMCAD, along with the fact that LMCAD is often more complicated and associated with worse prognosis, a number of professionals still preferred TFA over TRA since they are more experienced with TFA and can better handle emergent situations during the procedure. A cohort study including 931 patients from the EXCEL trial , with 248 (26.6%) patients undergoing TRA, have revealed that TRA and TFA are comparable in the risk of all-cause death, MI, TVR and stent thrombosis. A recent meta-analysis involving 12 observational studies  have indicated that, compared with TFA, TRA is associated with fewer bleeding outcomes and lower in-hospital mortality. Therefore, for the majority of patients with LMCAD, TRA is at least as safe as TFA, however, for patients with true distal bifurcations and complex anatomical characteristics, TFA might be the more appropriate approach.
Intracoronary imaging techniques such as intravascular ultrasound (IVUS) and
optical coherence tomography (OCT) during PCI can provide assessment and guidance
that is essential to achieve good long-term outcomes for LMCAD. Intracoronary
imaging can quantify lesion severity, guide lesion preparation with morphologic
data, select appropriate stent size by accurate vessel sizing, identify the
landing zone, diagnose acute vessel complication and define procedural success.
Studies have proved that IVUS-guided PCI for LMCAD improves
patient outcomes [44, 45, 46, 47]. OCT has a shorter wavelength and can provide higher
resolution images (10–20
Current practice guideline recommends applying dual anti-platelet therapy
(DAPT), the standard anti-platelet strategy consisting of aspirin and oral P2Y12
inhibitors, for at least 12 months after PCI with DES for ACS patients . For
patients with LMCAD, there lacks sufficient data and universal consensus on DAPT
duration after PCI in order to reduce the risk of ischemic events while minimize
bleeding. Results of an observational study including 1827 patients who received
PCI with DES for LMCAD from two large multicenter registries IRIS-MAIN and KOMATE
(Korean Multicenter Angioplasty Team Study)  revealed that DAPT
To date, several studies have analyzed the effect of operators’ experience or
PCI volume on procedural outcomes. Xu et al.  analyzed 1948
unprotected left main (uLM) PCIs performed by 25 operators and showed that
although experienced operators performed procedures with more complexity, the
3-year follow-up results still showed that patients undergone procedures
performed by experienced operators (
Upgraded intracoronary imaging techniques and coronary functional assessment methods integrating hemodynamic and anatomical assessment of coronary artery will be more prevalent in the future, which can contribute to precise preoperative planning, timely intraoperative guidance and ambulatory monitoring of postoperative recovery. The innovation of stents (including the next generation of DES and bioresorbable scaffold) as well as drug-coated balloon (DCB) may further improve the efficacy, safety and long-term outcomes after PCI for LMCAD.
As devices and techniques advance rapidly, more patients with unprotected left main coronary artery disease are undergoing PCI, and in low and moderate complexity lesions with SYNTAX score of 32 or less, the long-term outcomes after PCI are comparable to those after CABG, although ischemia-driven repeat revascularization is still higher after PCI. For those intended to treat with PCI, whether to perform a provisional single-stent or a planned dual-stent strategy needs to be evaluated. Current available evidence supports the choice of a single-stent strategy through transradial access in most cases, that is, the simpler, the better. For complex distal true bifurcation lesions, an upfront dual-stent strategy with transfemoral access may be required, and the DK-Crush may be the preferred procedure for experienced operators. The guidance of intracoronary imaging has been proven to help improve the outcomes of PCI. Duration of dual anti-platelet therapy should be decided after full clinical evaluation for individual patients and is recommended to be no less than 12 months. Unprotected left main percutaneous coronary intervention is a relatively complex procedure and is recommended to be performed in large centers by experienced operators. In the future, with the help of upgraded intracoronary imaging techniques and overall functional assessment, along with application of new generation stents, PCI for LMCAD will be more precise with advanced efficacy and better prognosis in long term.
XJ and KN designed and drafted the manuscript. CS provided help and advice on the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.
This research received no external funding.
The authors declare no conflict of interest.
Publisher’s Note: IMR Press stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.