Impact of large periprocedural myocardial infarction on mortality after PCI and coronary artery bypass grafting for left main disease

Selected in the European Heart Journal by E. Asher , S. Brugaletta

An analysis from the EXCEL trial

Reviewers

Elad Asher

@AsherElad

Interventional cardiologist / Cardiologist

Shaarey Zedek Medical Center - Jerusalem, Israel

Salvatore Brugaletta

@sbrugaletta

Interventional cardiologist / Cardiologist

Barcelona, Spain

Our Comment

This joint review is part of the PCRonline GLOBAL Journal Club Initiative by selected members of the EAPCI/PCR Journal Club and PCR NextGen, and is based on the underlying idea of “Bringing peers together, exchanging ideas, towards a common standard of care”.

Why this study – the rationale/objective?

The prognostic implications of periprocedural myocardial infarction (PMI) after percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) remain controversial. 
Periprocedural myocardial infarction can be detected by post-procedure elevation in cardiac biomarkers; however, cardiac biomarker increase is frequently observed after procedures that are deemed successful and otherwise uncomplicated.

The Third Universal Definition of Myocardial Infarction defines PMI by an elevation of troponin to ≥ 5X or ≥ 10X the 99th percentile upper reference limit (URL) within 48h after PCI or CABG, respectively in patients with normal baseline values, the Society for Cardiovascular Angiography and Interventions (SCAI) definition of a clinically relevant PMI uses a creatinine kinase-MB (CK-MB) cut-off of ≥ 10X URL for both PCI and CABG (or troponin ≥ 70X URL). Hence, the aim of the current study was to examine the 3-year rates of mortality among patients with and without PMI undergoing left main (LM) coronary artery intervention randomized to PCI vs. CABG.

How was it executed – the methodology?

A total of 1858 patients undergoing left main coronary artery intervention were 1:1 randomized to PCI with everolimus-eluting stents vs. CABG in the large-scale, multicentre, prospective, randomized EXCEL trial. PMI was defined using an identical threshold for PCI and CABG creatinine kinase-MB (CK-MB) elevation ≥ 10X the URL within 72h post-procedure, or ≥5X URL with new Q-waves, angiographic vessel occlusion, or loss of myocardium on imaging. CK-MB was routinely collected at baseline, 12± 2 and 24± 2h post-procedure. 
Cox proportional hazards modelling was performed controlling for age, sex, hypertension, diabetes mellitus, left ventricular ejection fraction, SYNTAX score, and chronic obstructive pulmonary disease (COPD). Rate of all-cause death and cardiac death were analyzed at 3-year.

What is the main result?

• Protocol criteria for PMI were met in 34/935(3.6%) patients treated by PCI vs. 56/923 (6.1%) patients treated by CABG [odds ratio 0.61, 95% confidence interval (CI) 0.40–0.93; P= 0.02].
• Periprocedural MI was associated with SYNTAX score, COPD, cross-clamp duration and total procedure duration, and not using antegrade cardioplegia. 
• By multivariable analysis, PMI was associated with cardiovascular death and all-cause death at 3 years [adjusted hazard ratio (HR) 2.63, 95% CI 1.19–5.81; P=0.02 and adjusted HR 2.28, 95% CI 1.22–4.29; P=0.01, respectively]. 
• The effect of PMI was consistent for PCI and CABG for cardiovascular death (P interaction = 0.56) and all-cause death (P interaction = 0.59). 
• Peak post-procedure CK-MB >_10 x URL strongly predicted mortality, whereas lesser degrees of myonecrosis were not associated with prognosis. 

Critical reading and the relevance for clinical practice

In the EXCEL trial, PMI was more common after CABG than PCI and it was strongly associated with increased 3-year mortality, after adjusting for confounding factors. Of note only extensive PMI was prognostically important.

One strong point of the EXCEL trial is the definition of PMI, which was classically difficult to point out in previous randomized trial, because of use of different threshold or requirements for symptoms, angiographic, electrocardiographic or imaging evidence of myocardial infarction. The EXCEL trial overcomes all these limitations of previous reports by using a common PMI definition defined by agreement between cardiac surgeons and interventional cardiologists.

Interestingly, large PMI (CK-MB ≥10X URL) were associated with a more than two-fold adjusted 3-year risk of cardiovascular death and all-cause mortality, whereas small PMI did not change prognosis. This may reinforce the justification for routine post-procedure cardiac biomarker assessment after both PCI and CABG, and expand the use of the SCAI criteria of a clinically relevant biomarker based PMI definition common to both revascularization procedures. 

The current study had several limitations, which should be taken in mind while analyzing its results. Only 90 patients met the protocol definition of PMI, so the present analyses are underpowered to identify all predictors of PMI. Secondly, PMI was defined using peak CK-MB, but troponin levels, a more sensitive marker for myocardial tissue injury, were not routinely collected. 

What is your approach regarding LM revascularization and routine biomarker assessment after LM revascularization in daily practice?