Predictive value of MRI findings for the identification of a hernia Sac in fetuses with congenital diaphragmatic hernia

OBJECTIVE. The purpose of this study was to identify MRI features of diaphragmatic hernia sac, as well as to assess the accuracy of diagnosing a sac prenatally. MATERIALS AND METHODS. All fetal MRI examinations performed for intrapleural congenital diaphragmatic hernia (CDH) from 2004 to 2013 were retrospectively reviewed by two pediatric radiologists blinded to the hernia sac status (defined intraoperatively or at autopsy). Reviewers noted whether a sac was present on the basis of identification of the following four MRI findings: 1, meniscus of lung posterior or apical to the hernia contents; 2, encapsulated appearance of hernia contents, exerting less than expected mass effect on the heart and mediastinum; 3, presence of pleural fluid outlining a sac from above; and 4, presence of ascites outlining a sac from below. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated for each finding and for various combinations. Contingency tables, chi-square testing, and logistic regression were applied to calculate the probability of a sac. RESULTS. Ninety patients were included: 21 with and 69 without a sac. The first three MRI findings correlated with the presence of a sac. Logistic regression yielded high predicted probability of a sac when one finding was identified (finding 1, 94.4%; finding 2, 96.2%). Adding a second and a third finding improved the probability to 99.7% and 99.9%, respectively. Sensitivity and specificity for the presence of a sac were 0.43 and 0.97, respectively. PPV and NPV were 83.8% and 80%, respectively. CONCLUSION. On fetal MRI, presence of a hernia sac in CDH can be diagnosed with high specificity when indicative findings are present.