J Cardiovasc Comput Tomogr. 2026 Oct 7:S1934-5925(26)00491-0. doi: 10.1016/j.jcct.2026.09.008. Online ahead of print.
ABSTRACT
OBJECTIVES: Although myocardial CT late enhancement (CT-LE) has been reported in various studies, its diagnostic performance using shuttle-mode dual-source CT has not been fully validated for detecting myocardial infarction (MI) and assessing its transmural extent against late gadolinium enhancement cardiovascular MR (LGE-CMR). This study aimed to evaluate the diagnostic performance of shuttle-mode myocardial CT-LE in detecting MI with LGE-CMR as a reference standard.
MATERIALS & METHODS: This study included 56 patients with suspected or known coronary artery disease who underwent both comprehensive cardiac CT, including myocardial CT-LE, and LGE-CMR from a prospective, multicenter trial (AMPLIFiED). Sixteen myocardial segments per patient (excluding the apical cap based on AHA classification) were assessed. Segments were classified as having or not having MI; segments with MI were further categorized by transmural extent: <50%, 50-99%, or 100%. Two observers independently assessed myocardial CT-LE images. LGE-CMR findings were determined by consensus between two observers.
RESULTS: In the segment-based analysis (n = 896), myocardial CT-LE showed a sensitivity of 97.6% and a positive predictive value (PPV) of 93.0%-94.1% for MI detection (specificity, 99.3%-99.4%; negative predictive value [NPV], 99.8%). Agreement for transmural extent was high (weighted kappa, 0.82-0.83). Inter-reader reproducibility for binary MI detection was excellent (agreement, 99.4%; kappa, 0.97). In an exploratory patient-level analysis based on 19 LGE-CMR-positive patients, sensitivity was 94.7% (95% confidence interval [CI], 74.0%-99.9%) for Observer A and 100% (95% CI, 82.4%-100%) for Observer B; the corresponding PPVs were 94.7% and 90.5%.
CONCLUSION: Shuttle-mode myocardial CT-LE demonstrated excellent agreement with LGE-CMR for the detection and transmural assessment of MI. Its high sensitivity for subendocardial MI supports its value as a reliable alternative to LGE-CMR.
PMID:42838816 | DOI:10.1016/j.jcct.2026.09.008