New England Journal of Medicine October 10, 2019

Complete Revascularization with Multivessel PCI for Myocardial Infarction

Shamir R. Mehta, David A. Wood, Robert F. Storey, Roxana Mehran, Kevin R. Bainey et al.

Bottom Line

In hemodynamically stable patients with STEMI and multivessel coronary artery disease, routine staged percutaneous coronary intervention of nonculprit lesions significantly reduced the composite risk of cardiovascular death or myocardial infarction compared to culprit-lesion-only PCI.

Key Findings

1. The first coprimary outcome of cardiovascular death or myocardial infarction occurred in 7.8% of the complete-revascularization group compared to 10.5% of the culprit-lesion-only group (HR 0.74; 95% CI 0.60-0.91; P=0.004) over a median follow-up of 3 years.
2. The second coprimary outcome of cardiovascular death, myocardial infarction, or ischemia-driven revascularization occurred in 8.9% of the complete-revascularization group versus 16.7% of the culprit-lesion-only group (HR 0.51; 95% CI 0.43-0.61; P<0.001).
3. The clinical benefit of complete revascularization was consistent regardless of the planned timing for nonculprit-lesion PCI, whether it was performed during the index hospitalization or after discharge (up to 45 days).
4. There were no significant differences between the two groups regarding safety endpoints, including rates of major bleeding, stroke, or stent thrombosis.

Study Design

Design
RCT
Open-Label
Sample
4,041
Patients
Duration
3 yr
Median
Setting
Multinational
Population Hemodynamically stable patients presenting with ST-segment elevation myocardial infarction (STEMI) and multivessel coronary artery disease who had undergone successful PCI of the culprit lesion.
Intervention Complete revascularization consisting of routine staged percutaneous coronary intervention (PCI) of all angiographically significant nonculprit lesions, performed either during the index hospitalization or within 45 days after discharge.
Comparator Culprit-lesion-only PCI with no further revascularization of nonculprit lesions, managed exclusively with guideline-directed medical therapy.
Outcome Two coprimary outcomes evaluated at a median of 3 years: 1) The composite of cardiovascular death or myocardial infarction; and 2) The composite of cardiovascular death, myocardial infarction, or ischemia-driven revascularization.

Study Limitations

The trial focused primarily on staged revascularization (performed either during the index hospitalization or within 45 days post-discharge); thus, it did not directly evaluate the safety and efficacy of performing multivessel PCI simultaneously during the index primary PCI procedure.
Patients presenting with cardiogenic shock or those with prior coronary artery bypass grafting (CABG) were excluded, limiting the trial's generalizability to these higher-risk populations.
Nonculprit lesions were predominantly evaluated using visual angiographic estimation rather than routine physiological assessment (e.g., fractional flow reserve), which was primarily reserved for lesions with intermediate stenoses (50-69%).

Clinical Significance

The COMPLETE trial shifted clinical paradigms and international guidelines by providing definitive evidence that complete revascularization in STEMI with multivessel disease reduces 'hard' clinical endpoints (cardiovascular death or recurrent MI), rather than merely reducing the need for repeat revascularization.

Historical Context

Prior to COMPLETE, trials such as PRAMI, CvLPRIT, DANAMI-3-PRIMULTI, and Compare-Acute demonstrated the superiority of complete revascularization over culprit-only PCI in STEMI, but their composite primary endpoints were largely driven by reductions in ischemia-driven repeat revascularization. COMPLETE was the first adequately powered, large-scale randomized trial to definitively prove a significant reduction in the hard endpoints of cardiovascular death or new MI.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

In the context of a STEMI, what is the pathophysiologic rationale that supports treating 'non-culprit' coronary lesions after the initial culprit lesion is stabilized, rather than leaving them alone?

Key Response

STEMI represents a highly inflammatory, pro-thrombotic systemic state. While the 'culprit' lesion is responsible for the acute infarction, non-culprit plaques often possess vulnerable characteristics (e.g., thin fibrous caps, lipid-rich cores) that can destabilize due to this systemic inflammation. Treating these lesions prevents future plaque rupture and subsequent spontaneous myocardial infarctions.

Resident
Resident

The COMPLETE trial allowed for staged PCI of non-culprit lesions to occur either during the index hospitalization or after discharge (within 45 days). How should you weigh patient and procedural factors when deciding the optimal timing for this staged procedure?

Key Response

The decision involves balancing the risk of contrast-induced nephropathy (total contrast load from the index procedure), radiation exposure, patient hemodynamic stability, and the complexity of the remaining lesions. Since the trial showed clinical benefit regardless of whether the staged PCI was done in-hospital or post-discharge, clinicians have the flexibility to delay the procedure to allow for renal recovery and careful pre-procedural planning, provided the patient is compliant and can safely return within the 45-day window.

Fellow
Fellow

The COMPLETE trial randomized patients based primarily on angiographic severity (>70% stenosis) of non-culprit lesions, using FFR only for intermediate (50-69%) lesions. How does this angiographic strategy compare to routine physiology-guided (FFR/iFR) revascularization of non-culprit lesions seen in trials like DANAMI-3-PRIMULTI or COMPARE-ACUTE?

Key Response

While trials like DANAMI-3-PRIMULTI and COMPARE-ACUTE proved that an FFR-guided approach to non-culprit lesions reduces composite endpoints (mostly driven by repeat revascularization), COMPLETE demonstrated that a predominantly angiography-guided approach effectively reduced the hard endpoints of CV death and MI. Fellows must critically evaluate whether acute microvascular dysfunction during a STEMI alters FFR reliability in non-culprit territories, and recognize that treating anatomically severe (>70%) lesions, even without FFR, yields significant prognostic benefit.

Attending
Attending

Prior randomized trials evaluating complete revascularization in STEMI primarily showed reductions in urgent, ischemia-driven repeat revascularization, but failed to show a definitive reduction in hard endpoints. Why was the COMPLETE trial uniquely able to demonstrate a significant reduction in cardiovascular death or myocardial infarction?

Key Response

The COMPLETE trial's success in demonstrating a reduction in hard endpoints was largely due to its robust statistical power (enrolling over 4,000 patients, far more than previous trials) and an extended follow-up period (median 3 years). The survival curves for hard events like spontaneous MI take time to diverge. As an attending, highlighting this teaches that adequately powering a study and allowing sufficient time for subsequent plaque ruptures to occur are critical for detecting differences in hard clinical outcomes.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

The COMPLETE trial utilized two co-primary outcomes and tested them using a hierarchical (sequential) statistical strategy. What are the methodological advantages of using this alpha-spending approach for co-primary composite endpoints in a large-scale cardiovascular outcomes trial?

Key Response

Hierarchical testing protects the family-wise error rate without requiring a strict Bonferroni penalty that would heavily reduce statistical power. By testing the first co-primary outcome (CV death or MI) at the full alpha level, and only proceeding to the second (CV death, MI, or ischemia-driven revascularization) if the first is significant, investigators can rigorously claim success on the most objective, 'hardest' clinical endpoints before evaluating a broader composite that includes more subjective, physician-driven endpoints.

Journal Editor
Journal Editor

Because the COMPLETE trial was an open-label study without a sham-PCI arm, patients and physicians were aware of the treatment assignments. As a reviewer, how do you appraise the potential for ascertainment bias, and does this threaten the internal validity of the study's conclusions?

Key Response

An open-label design introduces a high risk of bias for subjective endpoints like 'ischemia-driven revascularization,' as physicians are more likely to re-cath and intervene on a patient they know received conservative therapy. However, the trial's first co-primary endpoint (CV death or new MI) relies on objective, hard clinical criteria that are highly resistant to ascertainment bias. A critical editor would note that the reduction in these hard endpoints preserves the study's internal validity despite the lack of a sham control.

Guideline Committee
Guideline Committee

Prior to the COMPLETE trial, ACC/AHA and ESC guidelines gave varying (Class IIb or IIa) recommendations for non-culprit lesion PCI in STEMI based on limited evidence of reducing repeat revascularizations. How should the COMPLETE trial data influence the Class of Recommendation and Level of Evidence for routine staged PCI?

Key Response

Because COMPLETE demonstrated a definitive reduction in the hard clinical endpoints of cardiovascular death and recurrent MI in a large, well-powered, multi-center randomized controlled trial, it provides Level of Evidence A data. The committee should upgrade routine staged PCI of angiographically significant non-culprit lesions in hemodynamically stable STEMI patients to a Class I recommendation, establishing it as the standard of care to improve long-term prognosis.

Clinical Landscape

Noteworthy Related Trials

2013

PRAMI Trial

n = 465 · NEJM

Tested

Preventive PCI of non-infarct arteries

Population

STEMI patients with multivessel coronary artery disease

Comparator

Culprit-lesion-only PCI

Endpoint

Composite of CV death, nonfatal MI, or refractory angina

Key result: Preventive PCI reduced the risk of adverse cardiovascular events compared to culprit-only PCI.
2015

CvLPRIT Trial

n = 296 · JACC

Tested

Complete revascularization during index admission

Population

STEMI patients with multivessel disease

Comparator

Infarct-artery-only PCI

Endpoint

Composite of all-cause death, recurrent MI, heart failure, or ischemia-driven revascularization

Key result: Complete revascularization significantly reduced the primary composite endpoint at 12 months compared to infarct-artery-only PCI.
2015

DANAMI-3-PRIMULTI Trial

n = 627 · Lancet

Tested

FFR-guided complete revascularization prior to discharge

Population

STEMI patients with multivessel disease

Comparator

Culprit-only PCI

Endpoint

Composite of all-cause mortality, non-fatal reinfarction, or ischemia-driven revascularization

Key result: FFR-guided complete revascularization reduced the primary endpoint, driven primarily by a reduction in repeat revascularizations.

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