Abstract
BACKGROUND: We sought to evaluate the diagnostic accuracy of post-mortem cardiac magnetic resonance (PMCMR) of explanted hearts to detect the cardiac causes of sudden death. METHODS: PMCMR was performed in formalin-fixed explanted hearts of 115 cases of sudden death. Histological sampling of myocardium was performed using two different approaches: (1) guideline-based sampling; (2) guideline-based plus PMCMR-driven sampling. RESULTS: Forensic diagnosis of cardiac cause of death was ascertained in 72 (63%) patients. When the guideline-driven histological sampling was used, the PMCMR interpretation matched with final forensic diagnosis in 93 out of 115 cases (81%) with sensitivity of 88% (79-95%), specificity of 65% (47-80%), PPV of 84% (78-90%), NPV of 73% (58-84%), accuracy of 81% (72-88%), and AUC of 0.77 (0.68-0.84). When a PMCMR-driven approach was added to the guideline-based one, the matching increased to 102 (89%) cases with a PMCMR sensitivity of 89% (80-94%), a specificity of 86% (67-96%), PPV of 95% (89-98%), NPV of 73% (59-83%), accuracy of 89% (81-93%), and AUC of 0.88 (0.80-0.93). CONCLUSIONS: PMCMR has high accuracy to identify the cardiac cause of sudden death and may be considered a valid auxilium for forensic diagnosis. PMCMR could improve histological diagnosis in conditions with focal myocardial involvement or demonstrating signs of myocardial ischemia.