Late gadolinium enhancement (LGE) imaging overestimates acute infarct size. The main aim of this study was to investigate whether acute extracellular volume (ECV) maps can reliably quantify myocardial area at risk (AAR) and final infarct size (IS).Methods and Results—
Fifty patients underwent cardiovascular magnetic resonance imaging acutely (24–72 hours) and at convalescence (3 months). The cardiovascular magnetic resonance protocol included cines, T2-weighted imaging, native T1 maps, 15-minute post-contrast T1 maps, and LGE. Optimal AAR and IS ECV thresholds were derived in a validation group of 10 cases (160 segments). Eight hundred segments (16 per patient) were analyzed to quantify AAR/IS by ECV maps (ECV thresholds for AAR is 33% and IS is 46%), T2-weighted imaging, T1 maps, and acute LGE. Follow-up LGE imaging was used as the reference standard for final IS and viability assessment. The AAR derived from ECV maps (threshold of >33) demonstrated good agreement with T2-weighted imaging–derived AAR (bias, 0.18; 95% confidence interval [CI], −1.6 to 1.3) and AAR derived from native T1 maps (bias=1; 95% CI, −0.37 to 2.4). ECV demonstrated the best linear correlation to final IS at a threshold of >46% (R=0.96; 95% CI, 0.92–0.98; P<0.0001). ECV maps demonstrated better agreement with final IS than acute IS on LGE (ECV maps: bias, 1.9; 95% CI, 0.4–3.4 versus LGE imaging: bias, 10; 95% CI, 7.7–12.4). On multiple variable regression analysis, the number of nonviable segments was independently associated with IS by ECV maps (β=0.86; P<0.0001).Conclusions—
ECV maps can reliably quantify AAR and final IS in reperfused acute myocardial infarction. Acute ECV maps were superior to acute LGE in terms of agreement with final IS. IS quantified by ECV maps are independently associated with viability at follow-up.