Procedural myocardial infarction (MI) is a far weaker predictor of 1-year mortality than spontaneous MI, according to an analysis published online October 31, 2012, ahead of print in the Journal of the American College of Cardiology. This disparity in prognostic value carries implications for the design and interpretation of clinical trials using MI as an endpoint.

Investigators led by Sergio Leonardi, MD, MHS, of the Duke Clinical Research Institute (Durham, NC), looked at mortality risk associated with procedural vs. spontaneous MI in NSTE ACS patients who underwent PCI in 2 randomized trials enrolling nearly 20,000 patients:

• EARLY-ACS (n = 5,559), which evaluated early use of the glycoprotein IIb/IIIa inhibitor eptifibatide
• SYNERGY (n = 4,700), which compared anticoagulation with enoxaparin vs. unfractionated heparin

In all, 9,087 patients with complete data were included in the analysis. Procedural MI was defined as a CK-MB level at least 3 times the upper limit of normal (ULN) and at least 50% higher than preprocedural levels if they exceeded the ULN. Spontaneous MI was defined as a CK-MB elevation more than twice the ULN.

Overall, 1-year mortality was 4.6%. The dataset included 773 procedural MIs and 239 spontaneous MIs that met study criteria; 28 patients had both types. The vast majority (96.5%) of procedural MIs occurred within 24 hours of PCI, while the median time from intervention to spontaneous MI was 2 days (interquartile range, 1-5 days).

Mortality Risk Higher with Spontaneous MI

After adjustment for baseline characteristics, the risk of the primary endpoint of 1-year mortality (excluding the first 24 hours after PCI) was markedly higher for spontaneous MI than for procedural MI in both the SYNERGY and EARLY-ACS cohorts as well as in the combined analysis (table 1).

Table 1. Association Between MI Definition and 1-Year Mortality Risk

^a Limited to CK-MB twice the ULN.

By incrementally increasing the ratio of peak CK-MB to the ULN for procedural MI, the investigators determined that the threshold where 1-year mortality risk for procedural MI would first match that for spontaneous MI was 27.7 times the ULN (95% CI 13.9 -58.4) in the combined dataset. The level required to achieve equipoise was higher in the SYNERGY cohort (57.9 times; 95% CI 17.9-63.6 times) than the EARLY-ACS cohort (17.7 times; 95% CI 4.1-23.9 times using the spontaneous MI definition of CK-MB twice the ULN that was common to the 2 trials).

Only 49 (6.3%) of the 773 procedural MI events met the threshold of 27.7 times the ULN.

Balancing Trial Needs

A trial endpoint that includes both procedural and spontaneous MI combines heterogeneous events with very different prognoses, the authors observe.

“This is of particular concern for interventions that are expected to have unequal effects on spontaneous MI and procedural MI, such as early invasive strategy compared with conservative management in patients with NSTE ACS,” they write. “In this situation, a positive effect on spontaneous MI associated with [an] early invasive strategy could be obscured by more common but less prognostically relevant procedural MI. This suggests a need for separate analysis by differentially weighting effects of experimental treatments on these MI endpoints, or alternatively, considering the use of higher degrees of CK-MB elevation for procedural MI.”

In clinical trial design, the main advantage of composite outcomes is the increase in statistical power provided by a greater number of endpoint events, Dr. Leonardi and colleagues note. Although maximizing prognostic homogeneity is desirable, the substantial reduction in procedural MI caused by raising the enzymatic bar leads to logistical challenges, such as increased sample size and/or study duration.

Thus, they conclude, the “ideal threshold [for procedural MI] should balance the competing objectives of prognostic homogeneity, clinical relevance, and maximizing endpoints.”

Leonardi S, Thomas L, Neely ML, et al. Comparison of the prognosis of spontaneous and percutaneous coronary intervention-related myocardial infarction. J Am Coll Cardiol. 2012;Epub ahead of print.

Related Stories:

• CK-MB More Accurate Than Troponin in Identifying Post-PCI MI
• Periprocedural MI Does Not Predict Long-Term Mortality
• Study Urges Higher Cutoff for Troponin Elevation in Diagnosing MI