• Andrew M Taylor, Neil J Sebire, Michael T Ashworth, Silvia Schievano, Rosemary J Scott, Angie Wade, Lyn S Chitty, Nikki Robertson, Sudhin Thayyil, and Magnetic Resonance Imaging Autopsy Study Collaborative Group.
• Centre for Cardiovascular Imaging, University College London Institute of Cardiovascular Science, London, United Kingdom (A.M.T., S.S., S.T.); Cardiorespiratory Services, Great Ormond Street Hospital for Children (A.M.T.), Department of Histopathology, University College London Institute of Child Health and Great Ormond Street Hospital for Children (N.J.S., M.T.A.), and Department of Histopathology, University College London Hospitals (R.J.S.), National Health Service Foundation Trust, London, United Kingdom; Paediatric Epidemiology and Biostatistics, Institute of Child Health, University College London, London, United Kingdom (A.W.); Clinical Molecular Genetics Unit, University College London Institute of Child Health and University College London Hospitals National Health Service Foundation Trust, London, United Kingdom (L.S.C.); and Academic Neonatology, University College London Institute for Women's Health, London, United Kingdom (N.R., S.T.).
• Circulation. 2014 May 13; 129 (19): 1937-44.

BackgroundPerinatal and pediatric autopsies have declined worldwide in the past decade. We compared the diagnostic accuracy of postmortem, cardiovascular magnetic resonance (CMR) imaging with conventional autopsy and histopathology assessment in fetuses and children.Methods And ResultsWe performed postmortem magnetic resonance imaging in 400 fetuses and children, using a 1.5-T Siemens Avanto magnetic resonance scanner before conventional autopsy. A pediatric CMR imager reported the CMR images, masked to autopsy information. The pathologists were masked to the information from CMR images. The institutional research ethics committee approved the study, and parental consent was obtained. Assuming a diagnostic accuracy of 50%, 400 cases were required for a 5% precision of estimate. Three cases were excluded from analysis, 2 with no conventional autopsy performed and 1 with insufficient CMR sequences performed. Thirty-eight CMR data sets were nondiagnostic (37 in fetuses ≤24 weeks; 1 in a fetus >24 weeks). In the remaining 359 cases, 44 cardiac abnormalities were noted at autopsy. Overall sensitivity and specificity (95% confidence interval) of CMR was 72.7% (58.2-83.7%) and 96.2% (93.5-97.8%) for detecting any cardiac pathology, with positive and negative predictive values of 72.7% (58.2-83.7%) and 96.2% (93.5-97.8%), respectively. Higher sensitivity of 92.6% (76.6-97.9%), specificity of 99.1% (97.4-99.7%), positive predictive value of 89.3% (72.8-96.3%), and negative predictive value of 99.4% (97.8-99.8%) were seen for major structural heart disease.ConclusionsPostmortem CMR imaging may be a useful alternative to conventional cardiac autopsy in fetuses and children for detecting cardiac abnormalities.Clinical Trial Registration Urlhttp://www.clinicaltrials.gov. Unique identifier: NCT01417962.

22 keywords...

hide…