What can we learn from systematic segmental analysis of fetal heart by postmortem micro-CT: Is it time to change approach?

Background: Postmortem analysis of fetal heart is crucial but challenging. Alternative techniques to autopsy have been studied. We investigate micro-CT diagnostic accuracy in the analysis of fetal heart coming from early and late termination of pregnancy. Methods: Micro-CT and autopsy analyzed human fetal heart using twenty-six indices of cardiac anatomy derived from segmental approach. Sub-analysis of smaller specimens was performed. Primary outcome is the agreement between techniques. Secondary outcomes are sensitivity, specificity, positive and negative predictive value. Indices are defined "visible" or "non-visible" (primary end point) and "normal", "abnormal" or "non-diagnostic" (secondary end point). Results: Fifty-five cases were included. Agreement was 74.5% and 65.2% for primary end point and 93.3% and 91.8% for secondary end point in general population and in the group of smaller specimens, respectively. Sensitivity, specificity, positive and negative predictive value were 92.8%, 51.8%, 67.1%, 87.2% in the general population and 92.8%, 47.8%, 54.3%, 90.9% in the subgroup of smaller specimens for primary end point. They were 97.3%, 68.3%, 95.1%, 80.0% in the general population and 95.7%, 75.6%, 94.2%, 81.0% in the subgroup of smaller specimens for secondary end point. 86.1% and 91.5% of indices defined as "non-diagnostic" at autopsy would be visible by micro-CT in the two studied populations. Conclusion: Micro-CT represents a valid alternative to autopsy for postmortem evaluation of human fetal heart. We proposed to use micro-CT for initial analysis of fetal heart and to perform autopsy only in unsolved cases or if histological analysis is needed.