Long-term outcomes after percutaneous coronary intervention with drug-eluting stents versus coronary artery bypass grafting for left main coronary artery disease: a prospective analysis of the multicentre IRIS-MAIN registry
Source: View publication →
In a large, prospective Asian registry of patients with unprotected left main coronary artery disease, there was no statistically significant difference in 10-year all-cause mortality between PCI with drug-eluting stents and CABG.
Key Findings
Study Design
Study Limitations
Clinical Significance
These findings provide reassuring real-world evidence that PCI with drug-eluting stents and CABG offer comparable 10-year survival rates for selected patients with unprotected left main coronary artery disease. The virtually identical long-term mortality data (HR 1.00) reinforce the safety of PCI in experienced centers, while highlighting that the optimal revascularization strategy depends on individual patient factors and requires multidisciplinary Heart Team discussions.
Historical Context
The optimal revascularization strategy for unprotected left main coronary artery disease has been a topic of intense clinical debate for over a decade. Historically, CABG was the unquestioned gold standard. However, advancements in PCI techniques and drug-eluting stents prompted large randomized trials—like EXCEL and NOBLE—which showed conflicting results regarding composite ischemic outcomes at 3 to 5 years, sparking controversy particularly over definitions of periprocedural myocardial infarction. Furthermore, long-term (10-year) mortality data have historically been limited. The IRIS-MAIN registry contributes critical 10-year, real-world survival data using modern target trial emulation methodology, addressing a major gap in prolonged follow-up evidence for left main revascularization.
Guided Discussion
High-yield insights from every perspective
What does the term 'unprotected' left main coronary artery disease mean, and why has CABG historically been the gold standard treatment for this anatomical lesion?
Key Response
'Unprotected' refers to a left main coronary artery that has significant stenosis without a patent bypass graft or adequate collateral circulation supplying the left anterior descending (LAD) or circumflex arteries. Because the left main supplies the vast majority of the left ventricular myocardium, acute occlusion can cause massive infarction or sudden cardiac death, making durable surgical bypass (CABG) the historical gold standard to ensure long-term myocardial perfusion.
When evaluating a patient with unprotected left main coronary artery disease, how do we currently decide between PCI and CABG, and how does the IRIS-MAIN 10-year mortality finding influence this management decision?
Key Response
The decision relies on the Heart Team approach using anatomical complexity tools like the SYNTAX score, alongside patient comorbidities and surgical risk (e.g., STS score). CABG is generally preferred for highly complex anatomy (SYNTAX >32), while PCI is an acceptable alternative for low-to-intermediate complexity. The IRIS-MAIN finding that 10-year mortality is similar reinforces that PCI is a safe, viable long-term option when patients are appropriately selected based on anatomical suitability.
Although the IRIS-MAIN registry showed no significant difference in 10-year all-cause mortality between PCI and CABG, how should interventional cardiology fellows counsel patients regarding secondary endpoints such as repeat revascularization and spontaneous myocardial infarction?
Key Response
While mortality may be similar, both observational data and major RCTs (like EXCEL and NOBLE) consistently demonstrate that PCI is associated with higher long-term rates of repeat target lesion revascularization and non-procedural myocardial infarction compared to CABG. Fellows must balance the upfront morbidity and longer recovery of CABG against the higher likelihood of needing subsequent procedures with PCI during shared decision-making.
As an attending leading a multidisciplinary Heart Team, how do you synthesize and communicate the findings of 'real-world' registries like IRIS-MAIN when they contrast with specific subgroup signals from randomized controlled trials (such as the late mortality debate in the EXCEL trial)?
Key Response
Registries provide critical external validity by reflecting real-world clinical practice, patient selection, and operator techniques, which often differ from the rigid inclusion criteria of RCTs. An attending should use RCTs to establish baseline efficacy and use registry data to contextualize long-term safety in broader populations, emphasizing to the patient that highly tailored, anatomy-specific procedural selection by the Heart Team minimizes the survival gap between PCI and CABG.
Scholarly Review
Critical appraisal through the lens of expert reviewers and guideline development
In a 10-year prospective registry comparing surgical versus percutaneous interventions, how do methods like propensity score matching address treatment allocation bias, and what are the limitations of these methods regarding unmeasured confounding over a decade?
Key Response
Propensity score matching balances measured baseline covariates to mimic randomization, but it cannot account for unmeasured confounders such as patient frailty, operator skill, or evolving medical therapies over 10 years. A robust methodological critique would examine whether the authors used instrumental variables or sensitivity analyses (like the E-value) to quantify how strong an unmeasured confounder would need to be to overturn the study's null finding on mortality.
As a peer reviewer evaluating this 10-year registry data, what are the primary threats to validity when interpreting a 'no statistically significant difference' outcome in all-cause mortality, and how can authors prevent readers from incorrectly assuming clinical equivalence?
Key Response
Primary threats include selection bias (healthier or lower-risk patients preferentially receiving PCI), survivorship bias, and differential loss to follow-up over a decade. An editor must ensure the authors clearly distinguish 'failure to reject the null hypothesis' from 'non-inferiority,' requiring adequate statistical power and predefined non-inferiority margins before concluding that the two treatments are clinically equivalent.
Given that current ACC/AHA/SCAI guidelines give CABG a Class I recommendation and PCI a Class IIa recommendation for left main disease with low-to-intermediate anatomical complexity, does the 10-year IRIS-MAIN data warrant an upgrade of PCI to a Class I recommendation?
Key Response
The 10-year observational data from IRIS-MAIN provides strong Level of Evidence B-NR (non-randomized) supporting long-term safety, but it is unlikely to warrant upgrading PCI to a Class I recommendation over CABG. Guidelines heavily weight randomized data (Level of Evidence A) and consistently point to higher repeat revascularization rates with PCI. However, this robust registry data solidifies the current Class IIa recommendation, assuring guideline committees that PCI is a safe, durable alternative when CABG is not preferred.
Clinical Landscape
Noteworthy Related Trials
SYNTAX Trial
Tested
PCI with paclitaxel-eluting stents
Population
Patients with left main coronary disease or three-vessel disease
Comparator
CABG
Endpoint
Major adverse cardiac or cerebrovascular events (MACCE) at 12 months
EXCEL Trial
Tested
PCI with everolimus-eluting stents
Population
Patients with left main coronary artery disease and low or intermediate SYNTAX score
Comparator
CABG
Endpoint
Composite of death, stroke, or myocardial infarction
NOBLE Trial
Tested
PCI with biolimus-eluting stents
Population
Patients with unprotected left main coronary artery disease
Comparator
CABG
Endpoint
Major adverse cardiac or cerebrovascular events (MACCE)
Tailored to your role
Want this tailored to you?
Add your specialty or training stage to get role-specific takeaways and more questions.
Personalize this analysis