Prenatal Diagnosis of Aortic Coarctation: Prediction Algorithm according to Gestational Age
<b><i>Objectives:</i></b> The aim of the study was first to quantify the diagnostic accuracy of predictive anatomical factors of aortic coarctation (CoA) and second to design a postnatal CoA probability algorithm according to gestational age (GA) in prenatal period. <b><i>Methods:</i></b> Global and according to GA diagnostic performance of cardiac anatomical variables using the ROC curve were evaluated in a retrospective cohort of fetuses with suspicion of CoA (2004–2020). A serial testing strategy to predict postnatal CoA by fetal echocardiography was designed. <b><i>Results:</i></b> 114 fetuses were included. Isthmus-to-ductal (I/D) ratio provided the best discrimination between healthy fetuses and those with CoA (AUC 0.91, 95% CI: 0.86–0.96, I/D < 0.74 sensitivity 96.3%, I/D < 0.6, specificity 92.5%) with good classification capacity in both the second and third trimesters of gestation. Isthmus <i>z</i>-score and pulmonary/aortic valve ratio increased accuracy in fetuses >28 and tricuspid/mitral valve ratio (TV/MV) in fetuses ≤28 weeks. Study of I/D plus TV/MV ratio in fetuses ≤28 and I/D ratio plus isthmus <i>z</i>-scores in fetuses >28 weeks allowed to correctly classify 91.8% of fetuses as high or low probability of postnatal CoA. <b><i>Conclusions:</i></b> Diagnostic discrimination of anatomic predictive factors for CoA varies according to GA. Specific algorithms according to GA increase accuracy in CoA’s prenatal prediction.