This study sought to investigate whether early post-infarction cardiac magnetic resonance (CMR) parameters provide additional long-term prognostic value beyond traditional outcome predictors in ST-segment elevation myocardial infarction (STEMI) patients.
Long-term prognostic significance of CMR in STEMI patients has not been assessed yet.
This was a longitudinal study from a multicenter registry that prospectively included STEMI patients undergoing CMR afterinfarction. Between May 2003 and August 2015, 810 revascularized STEMI patients were included. CMR was performed at a median of 4 days after STEMI. Infarct size, microvascular obstruction (MVO), and left ventricular (LV) volumes and function were measured. Primary endpoint was a composite of all death and decompensated heart failure (HF).
During median follow-up of 5.5 years (range 1.0 to 13.1 years), primary endpoint occurred in 99 patients (39 deaths and 60 HF hospitalization). MVO was a strong predictor of the composite endpoint after correction for important clinical, CMR, and angiographic parameters, including age, LV systolic function, and infarct size. The independent prognostic value of MVO was confirmed in all multivariate models irrespective of whether it was included as a dichotomous (presence of MVO, hazard ratio [HR]: 1.985 to 1.995), continuous (MVO extent as % LV, HR: 1.095 to 1.097), or optimal cutoff value (MVO extent ?2.6% of LV; HR: 3.185 to 3.199; p < 0.05 for all). MVO extent ?2.6% of LV was a strong independent predictor of all death (HR: 2.055; 95% confidence interval: 1.076 to 3.925; p = 0.029) and HF hospitalization (HR: 5.999; 95% confidence interval: 3.251 to 11.069; p < 0.001). Finally, MVO extent ?2.6% of LV provided incrementalprognostic value over traditional outcome predictors (net reclassification improvement index: 0.16 to 0.30; p < 0.05 for all models).
Early post-infarction CMR-based MVO is a strong independent prognosticator in revascularized STEMI patients. Remarkably, MVO extent ?2.6% of LV improved long-term risk stratification over traditional outcome predictors.