Abstract
BACKGROUND: The prognostic importance of periprocedural myocardial infarction (pMI) and its inclusion in the composite outcomes of coronary revascularization trials are controversial. We assessed whether pMI is a surrogate for all-cause or cardiac mortality and quality of life (QoL) outcomes in coronary revascularization trials. METHODS: All randomized trials comparing percutaneous coronary intervention vs coronary artery bypass grafting (MEDLINE, EMBASE, Cochrane Library) were identified. Trials were included if they reported data for pMI and mortality. Trial-level associations between pMI and all-cause or cardiac mortality and QoL were assessed using the coefficient of determination (R (2) ). The criterion for surrogacy was set at 0.7. Subgroup analyses based on pMI definition and on key clinical/procedural variables were performed. RESULTS: Twelve trials were included (11,549 patients; weighted mean follow-up: 5.6 years). There was a positive correlation between pMI and all-cause mortality (slope, 1.81; 95% CI, 1.00-2.63; R (2) = 0.72). In the trials that defined pMI as a rise in cardiac biomarkers >5 times the upper reference limit, pMI positively correlated with both all-cause (slope, 2.07; 95% CI, 1.00-3.14; R (2) = 0.93) and cardiac mortality (slope, 0.70; 95% CI, 0.20-1.19; R (2) = 0.87); no such relationships were present in trials that used a lower biomarker threshold. An inverse correlation was found between pMI and long-term changes in the Short Form Health Survey Physical Component score (slope, -4.66; 95% CI, -5.75 to -3.57; R (2) =0.99). CONCLUSIONS: In the published coronary revascularization trials, pMI defined by larger biomarker elevations was associated with subsequent mortality and reduced QoL. These findings suggest that large pMI should be included as an outcome measure in coronary revascularization trials.