New England Journal of Medicine April 28, 2011

Coronary-Artery Bypass Surgery in Patients with Left Ventricular Dysfunction (STICH)

Eric J. Velazquez, Kerry L. Lee, Marek A. Deja, et al.

Bottom Line

In patients with ischemic cardiomyopathy, adding CABG to guideline-directed medical therapy did not significantly reduce the primary endpoint of all-cause mortality at 5 years, though it did significantly reduce rates of cardiovascular death and hospitalization.

Key Findings

1. At a median follow-up of 56 months, death from any cause occurred in 36% (218 of 610) of the CABG group compared to 41% (244 of 602) of the medical-therapy group, a difference that did not reach statistical significance (Hazard Ratio [HR] 0.86; 95% CI 0.72-1.04; P=0.12).

2. Death from cardiovascular causes was significantly reduced in patients randomized to CABG compared to medical therapy alone (28% vs. 33%; HR 0.81; 95% CI 0.66-1.00; P=0.05).

3. The composite secondary endpoint of death from any cause or hospitalization for cardiovascular causes occurred significantly less often in the CABG arm (58% vs. 68%; HR 0.74; 95% CI 0.64-0.85; P<0.001).

4. There was a substantial crossover rate: 17% (100 patients) of the medical-therapy arm underwent CABG before the end of the follow-up period, while 91% (555 patients) of the CABG arm actually underwent the assigned surgery.

Study Design

Design

RCT

Open-Label

Sample

1,212

Patients

Duration

56 mo

Median

Setting

Multicenter, 22 countries

Population Patients with ischemic cardiomyopathy (LVEF ≤35%) and coronary artery disease amenable to coronary artery bypass grafting, excluding those with significant left main disease or severe (CCS class III or IV) angina.

Intervention Coronary artery bypass grafting (CABG) in addition to guideline-directed medical therapy.

Comparator Guideline-directed medical therapy alone.

Outcome Death from any cause.

Study Limitations

• A high crossover rate (17%) from the medical-therapy group to the surgical group likely diluted the intention-to-treat analysis and biased the primary endpoint toward the null.

• The median follow-up of 56 months may have been too short to demonstrate an overall survival advantage, given the early upfront periprocedural mortality risk associated with CABG (which was later confirmed by the 10-year STICHES extension study).

• Patients with left main coronary artery disease or severe angina (CCS Class III or IV) were excluded, limiting the generalizability of the findings since these populations already have established indications for revascularization.

• The inherent open-label nature of a surgical vs. medical trial introduces potential bias in the management and reporting of non-fatal events.

Clinical Significance

The initial 5-year results of the STICH trial narrowly missed statistical significance for reducing all-cause mortality, but the significant reductions in cardiovascular mortality and hospitalizations highlighted the durable benefits of revascularization for ischemic heart failure. The trial framed the clinical dilemma of CABG in reduced ejection fraction patients as an upfront procedural risk traded for long-term cardiovascular benefit. Notably, the subsequent 10-year follow-up (STICHES, 2016) eventually demonstrated a definitive all-cause mortality benefit (NNT = 14), firmly establishing CABG plus medical therapy as the standard of care for suitable patients with severe ischemic cardiomyopathy.

Historical Context

Prior to STICH, the recommendation for CABG in patients with severe ischemic left ventricular dysfunction (LVEF ≤35%) was based largely on subgroup analyses from landmark trials conducted in the 1970s and 1980s (e.g., CASS, the Veterans Administration Cooperative Study). Because these foundational trials predated modern guideline-directed medical therapy (beta-blockers, ACE inhibitors, aldosterone antagonists), the contemporary benefit of exposing high-risk heart failure patients to the operative morbidity of CABG was heavily debated. The STICH trial was designed to definitively test whether surgical revascularization still offered a survival advantage over modern optimal medical therapy alone.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

What is the pathophysiological difference between myocardial 'stunning' and myocardial 'hibernation,' and why is myocardial hibernation a key concept when considering CABG in patients with severe ischemic cardiomyopathy as studied in the STICH trial?

Key Response

Students need to understand that myocardial stunning is a temporary post-ischemic dysfunction following acute reperfusion, whereas hibernation is a chronic state of down-regulated myocardial contractility due to prolonged, persistent hypoperfusion. CABG aims to restore blood flow to hibernating myocardium, potentially recovering left ventricular ejection fraction and improving long-term outcomes in patients with ischemic cardiomyopathy.

Resident

Given that CABG did not significantly reduce the primary endpoint of 5-year all-cause mortality but did reduce cardiovascular death and hospitalizations, how should you counsel a 65-year-old patient with an LVEF of 30% and multivessel CAD regarding the upfront risks of surgery versus long-term medical management?

Key Response

Residents must learn how to navigate shared decision-making by balancing early perioperative risks against delayed benefits. Counseling should address the early hazard phase of surgical mortality (usually within the first 30 days) versus the reduction in recurrent ischemic events and cardiovascular admissions over the subsequent years, framing the decision around the patient's immediate surgical candidacy and long-term goals of care.

Fellow

The STICH viability sub-study evaluated whether myocardial viability testing predicted a survival benefit from CABG. Based on these findings, should routine myocardial viability testing dictate whether a patient with ischemic cardiomyopathy is referred for surgical revascularization?

Key Response

Fellows should know that the STICH viability sub-study challenged prevailing dogma. While the presence of viability was associated with lower overall mortality, viability testing did not identify a specific subgroup of patients who derived a differential survival benefit from CABG compared to medical therapy alone. Therefore, lack of viability on imaging should not be the sole reason to exclude a patient from potentially beneficial surgical revascularization.

Attending

The STICH trial at 5 years was technically a 'negative' trial for its primary endpoint of all-cause mortality, yet CABG remained heavily utilized for ischemic cardiomyopathy. How do the concepts of intention-to-treat analysis, patient crossover, and early surgical hazard inform how we teach junior doctors to contextualize this 'negative' result?

Key Response

Attendings must emphasize that intention-to-treat analyses can dilute surgical benefit when medical patients cross over to surgery (17% crossed over in STICH). Furthermore, the early surgical hazard masks the divergence of survival curves that occurs later. Teaching this prevents the premature dismissal of surgical therapies and anticipates the need for extended follow-up (which eventually proved the mortality benefit of CABG in the 10-year STICHES extension).

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The 5-year STICH results demonstrated an early hazard in the CABG arm followed by a later potential benefit, implying non-proportional hazards. How does the presence of non-proportional hazards compromise the standard Cox proportional-hazards model used in the primary analysis, and what alternative statistical methods could better capture this time-varying effect?

Key Response

PhDs focus on statistical assumptions. The Cox model assumes a constant hazard ratio over time, which is violated when comparing surgery (early perioperative risk) to medical therapy. Using alternative approaches, such as restricted mean survival time (RMST), time-dependent covariates, or landmark analyses, provides a much more robust and accurate methodological framework for evaluating therapies with intersecting survival curves.

Journal Editor

As an editor reviewing the 5-year STICH manuscript, how would you address the authors' emphasis on statistically significant secondary endpoints (cardiovascular death and hospitalization) in the presence of a null primary endpoint (all-cause mortality, p=0.12), and what editorial boundaries must be set to prevent 'spin'?

Key Response

Editors must rigorously guard against outcome reporting bias. When a primary endpoint is not statistically significant, drawing definitive conclusions from secondary endpoints can be misleading due to the lack of alpha-control for multiple testing. An editor must ensure the abstract and discussion clearly label secondary findings as hypothesis-generating or supportive, rather than definitive proof of efficacy, preventing 'spin'.

Guideline Committee

Despite the 5-year STICH results missing statistical significance for all-cause mortality, how do current guidelines (e.g., 2021 ACC/AHA/SCAI Guideline for Coronary Artery Revascularization) classify the recommendation for CABG in patients with severe LV dysfunction (EF < 35%) and multivessel CAD, and how does extended follow-up data justify this?

Key Response

Guideline committees synthesize evolving evidence. While the 5-year STICH data was technically negative for the primary outcome, the 10-year STICHES extension demonstrated a significant and durable overall survival benefit. Consequently, current guidelines award a Class 1 recommendation (Level of Evidence B-R) for CABG in patients with ischemic cardiomyopathy and multivessel CAD to improve survival, illustrating how guidelines rely on long-term data for surgical interventions.

Clinical Landscape

Noteworthy Related Trials

1983

CASS Trial

n = 780 · Circulation

Tested

Coronary artery bypass grafting (CABG)

Population

Patients with stable ischemic heart disease

Comparator

Medical therapy

Endpoint

Annual mortality rate

Key result: While there was no overall survival benefit for CABG, a significant survival benefit was observed in the subgroup of patients with 3-vessel disease and reduced ejection fraction.

2016

STICHES Trial

n = 1,212 · NEJM

Tested

CABG plus medical therapy

Population

Patients with CAD amenable to CABG and LVEF <= 35%

Comparator

Medical therapy alone

Endpoint

All-cause mortality

Key result: At 10 years, CABG significantly reduced all-cause mortality and cardiovascular mortality compared to medical therapy alone.

2022

REVIVED-BCIS2 Trial

n = 700 · NEJM

Tested

PCI plus optimal medical therapy

Population

Patients with severe ischemic cardiomyopathy (LVEF <= 35%) and viable myocardium

Comparator

Optimal medical therapy alone

Endpoint

All-cause death or hospitalization for heart failure

Key result: PCI did not reduce the incidence of all-cause death or heart failure hospitalization compared to medical therapy alone.

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