Clinical utility of post-mortem micro-CT of the fetal heart: an exploratory study of diagnostic imaging versus macroscopic dissection

Abstract

BackgroundCongenital cardiac malformations are commonly seen at perinatal autopsy. Accurate diagnosis can be challenging in small, macerated fetuses. The diagnostic accuracy of post-mortem MRI is diminished below 18 gestational weeks; however, micro-CT can provide micrometre resolution images but has not been evaluated for human diagnostic accuracy. We hypothesised that micro-CT could provide useful diagnostic information in fetal congenital heart disease and compared it with the gold standard of autopsy. MethodsReferrals to Great Ormond Street Hospital were prospectively included. Ex-vivo fetal hearts underwent micro-CT examination. 21 indices normally assessed at autopsy were evaluated for each heart at both micro-CT and autopsy in a single-masked fashion. All parents provided written consent for both autopsy, including post-mortem imaging, and research. The study was part of a larger study given ethics approval (IRAS ID 13942, R&D 09MH01). FindingsWe examined six fetal hearts including five with complex congenital cardiac malformations (17–23 weeks' gestation). Heart weights ranged between 1·1 g and 5·3 g (median 4·6 g). All specimens demonstrated excellent internal contrast at micro-CT examination, and the correct overall diagnosis was made from micro-CT data in all cases. There was agreement for 114 of 126 indices assessed on micro-CT and autopsy dissection (overall concordance 95·8% (95% CI 90·5–98·2). Micro-CT was particularly useful in the assessment of ventricular morphology in macerated fetuses. Micro-CT examination for diagnosis of complex congenital heart disease had a sensitivity of 85·2% (95% CI 67·5–94·1) and specificity of 98·9% (94·1–99·8). InterpretationOur preliminary study shows that micro-CT can provide diagnostically accurate three-dimensional volumes of complex congenital fetal heart disease. This approach potentially represents a substantial advance in post-mortem imaging and confirms the potential of this technology for non-invasive examination of small fetuses and organs. Centres that do not have specialist paediatric cardiac pathology expertise might especially benefit. Limitations include the small study size. Future studies should also include in-vivo post-mortem evaluation. FundingNone.