Observed-to-expected lung-area-to-head-circumference ratio on ultrasound examination vs total fetal lung volume on magnetic resonance imaging in prediction of survival in fetuses with left-sided diaphragmatic hernia.

Objective: To assess and compare the value of antenatally determined observed-to-expected (O/E) lung-area-to-head-circumference ratio (LHR) on ultrasound examination vs O/E total fetal lung volume (TFLV) on magnetic resonance imaging (MRI) examination to predict postnatal survival of fetuses with isolated, expectantly managed left-sided congenital diaphragmatic hernia (CDH).
Methods: This was a multicenter retrospective study including all consecutive fetuses with isolated CDH that were managed expectantly in Mannheim, Germany, and in five other European centers, that underwent at least one ultrasound examination for measurement of O/E-LHR and one MRI scan for measurement of O/E-TFLV during pregnancy. All MRI data were centralized, and lung volumes were measured by two experienced operators blinded to the pre- and postnatal data. Multiple logistic regression analyses were performed to examine the effect on survival at hospital discharge of various perinatal variables, including the center of management. In left-sided CDH with intrathoracic herniation of the liver, receiver-operating-characteristics (ROC) curves were constructed separately for cases from Mannheim and the other five European centers and were used to compare O/E-TFLV and O/E-LHR in the prediction of postnatal survival.
Results: From Mannheim, 309 patients were included with a median gestational age (GA) at ultrasound examination of 29.6 (range, 19.7-39.1) weeks and median GA at MRI examination of 31.1 (range, 18.0-39.9) weeks. From the other five European centers, 116 patients were included with a median GA at ultrasound examination of 26.7 (range, 20.6-37.6) weeks and median GA at MRI examination of 27.7 (range, 21.3-37.9) weeks. Regression analysis demonstrated that the survival rates at discharge were lower in left-sided CDH (odds ratio (OR), 0.349 (95% CI, 0.133-0.918), P = 0.033) and those with intrathoracic liver (OR, 0.297 (95% CI, 0.141-0.628), P = 0.001), and higher with increasing O/E-TFLV (OR, 1.123 (95% CI, 1.079-1.170), P < 0.001), advanced GA at birth (OR, 1.294 (95% CI, 1.055-1.588), P = 0.013) and when birth occurred in Mannheim (OR, 7.560 (95% CI, 3.368-16.967), P < 0.001). Given the difference in survival rate between Mannheim and the five other European centers, ROC curve comparisons between the two imaging modalities were presented separately. For cases of left-sided CDH with intrathoracic herniation of the liver, pairwise comparison showed no significant difference between the area under the ROC curves for the prediction of postnatal survival between O/E-TFLV and O/E-LHR in Mannheim (mean difference = 0.025, P = 0.610, standard error = 0.050), whereas there was a significant difference in the other European centers studied (mean difference = 0.056, P = 0.033, standard error = 0.056).
Conclusions: In fetuses with left-sided CDH and intrathoracic herniation of the liver, the predictive value for postnatal survival of O/E-TFLV on MRI examination and O/E-LHR on ultrasound examination was similar in one center (Mannheim), but O/E-TFLV had better predictive value compared to O/E-LHR in the five other European centers. Hence, in these five European centers, MRI should be included in the diagnostic process for left-sided CDH. © 2024 International Society of Ultrasound in Obstetrics and Gynecology.
(© 2024 International Society of Ultrasound in Obstetrics and Gynecology.)