Procedural MI Definition Makes a Big Difference in PCI vs CABG Showdown

Certain procedural heart attacks were more clinically significant than others after percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) surgery, reports suggested.

Just how clinically relevant procedural myocardial infarctions (MIs) are after PCI has long been a point of controversy, especially when comparing the procedure to CABG, where some degree of myocardial injury is inherently expected.

Just how sensitive the results of the EXCEL and SYNTAX trials were to how peri-procedural MI had been defined was detailed in two post hoc analyses published in the Oct. 6 issue of the Journal of the American College of Cardiology.

EXCEL vs Third Universal Definition

EXCEL investigators defended their pre-specified protocol definition of peri-procedural MI and the conclusions they drew from the trial despite conflicting findings when such heart attacks were counted according to the Third Universal Definition.

The incidence of procedural MI reached 3.6% of the PCI group and 6.1% of CABG recipients (P=0.015) using the pre-specified protocol definition, which required the same large elevation of creatine kinase-MB (CK-MB) after either procedure, according to Stuart Pocock, PhD, of London School of Hygiene and Tropical Medicine, and colleagues.

However, procedural MIs rose to 4.0% for PCI and dropped to 2.2% for CABG (P=0.025) when defined by the Third Universal Definition of MI, which required less in terms of biomarker elevation (troponins preferred, CK-MB otherwise) but more in that there had to be accompanying evidence of myocardial ischemia.

Both types of procedural MIs were associated with 5-year mortality. Protocol-defined ones were tied to cardiovascular mortality to a similar extent after PCI and CABG (adjusted HR 2.18, 95% CI 1.13-4.23), whereas Universal Definition MIs were only associated with such deaths after CABG and not PCI (adjusted HR 2.87, 95% CI 1.44-5.73; P=0.004 for interaction).

“The absolute and relative rates of PMI [procedural MI] after PCI and CABG and thus the primary composite outcome of the EXCEL trial varied greatly depending on the specific PMI definition utilized,” according to study authors.

“The consistent hazard after PCI and CABG and minimization of ascertainment bias supports the use of the pre-specified protocol definition of PMI in the EXCEL primary composite outcome and the principal conclusions that PCI and CABG provide similar 3- and 5-year rates of major adverse cardiovascular events for revascularization of selected patients with LMCAD [left main coronary artery disease],” they concluded.

Yet Donald Cutlip, MD, of Beth Israel Deaconess Medical Center and Harvard Medical School, raised some concerns with this interpretation.

In an accompanying editorial, he noted that none of the six cardiac deaths in the protocol-defined MI group at 5 years occurred in people who had had CABG, and that one out of the four deaths after protocol-defined MI in PCI had been moved to “no MI” classification using the Third Universal Definition due to lack of supporting criteria.

Notably, the NOBLE trial that found surgery to be better than PCI for left main revascularization had not included procedural MIs in the primary endpoint.

Included in the present EXCEL analysis were 1,858 patients with LMCAD randomized to PCI or CABG surgery.

Limitations of the study included the lack of routine troponin collection (which could have introduced various biases) and the lack of a biomarker core laboratory for the trial.

SYNTAX, EXCEL, ISCHEMIA, and More

Rates of procedural MI were similarly inconsistent and of varying clinical significance depending on how such heart attacks were defined in post hoc analysis of the SYNTAX Extended Survival study:

• CABG looked better, at least numerically, by the SYNTAX definition (2.7% PCI vs 2.4% CABG) and Fourth Universal Definition of MI (3.0% vs 2.1%), two definitions that combine elevation of CK-MB with other evidence of myocardial damage
• PCI was favored by the EXCEL and SCAI definitions (5.7% vs 16.5% for both), which define procedural MI by CK-MB elevation ≥10x upper limit of normal (ULN) or ≥5x ULN with additional criteria
• PCI also came out tops by the ISCHEMIA definition (6.0% vs 8.8%), wherein procedural MIs are defined by CK-MB ≥10x ULN for PCI and ≥15x ULN for CABG (or lower cutoffs with additional criteria)

Pooling the PCI and CABG patients, procedural MIs according to the SYNTAX or Fourth Universal Definition of MI were significantly associated with all-cause mortality “and are therefore ‘clinically relevant,'” reported Patrick Serruys, MD, PhD, of National University of Ireland, Galway, and colleagues.

But the association at 10 years wasn’t significant with the SCAI and EXCEL definitions relying solely on enzyme elevation. “This ‘enzymatic PMI event’ more selectively affects the time-to-event curve and the composite endpoint of the surgical cohort,” according to the group.

“PMIs after PCI were associated with 10-year mortality regardless of definition, whereas their impact on mortality after CABG was limited to 1 year,” they added.

SYNTAX enrolled 1,800 patients, of whom 1,652 had CK-MB recorded post-procedure were therefore included in the post hoc analysis.

The main finding of the trial was that certain patients with complex coronary lesions get different benefits from surgery vs stenting.

Because troponin was not collected routinely, CK-MB ratios were used as a substitute when applying the Fourth Universal Definition, Serruys and colleagues cautioned.

The Debate Continues

The EXCEL and SYNTAX reports had their limitations but shared common and important themes, Cutlip suggested.

“In both studies, the importance of supporting criteria (new Q-waves, documented vessel occlusion, or imaging evidence of loss of myocardium) had a substantial impact on MI diagnosis after CABG. Biomarker elevation alone, even at high levels, without these supporting criteria increased the frequency of procedural MI but appeared to add little, if any, to the value of MI as a correlate with subsequent mortality,” he observed.

“The SYNTAX and EXCEL analyses support a requirement for the supporting criteria proposed by UDMI [Universal Definition of MI] after CABG. However, the supporting criteria as proposed by the UDMI after PCI are not useful and should be removed in favor of a standalone biomarker threshold that is meaningful,” according to the editorialist.

That threshold remains a point of debate, he said.

When it comes to the next comparative effectiveness trial of PCI versus CABG, “[t]here might need to be acceptance that the same procedural MI definition cannot be fit to both procedures,” Cutlip suggested. “If we cannot find definitions that fit these purposes, then perhaps it is time to remove procedural MI from primary composite endpoints.”