Objectives
In fetuses with isolated left‐sided congenital diaphragmatic hernia (LCDH), prenatal detection of severe pulmonary hypoplasia is important, as fetal therapy can improve survival. Cases with mild or moderate lung hypoplasia still carry a considerable risk of mortality and morbidity, but there has been less interest in the accurate prediction of outcome in these cases. In this study of fetuses with mild or moderate isolated LCDH, we aimed to investigate: (1) the association between intrapulmonary artery (IPA) Doppler findings and mortality at discharge; (2) whether adding IPA Doppler findings improves the prediction of mortality based on lung size and liver herniation; and (3) the association between IPA Doppler findings and early neonatal morbidity.
Methods
This was a retrospective study of all consecutive fetuses assessed at the BCNatal and UZ Leuven hospitals between 2008 and 2020 with a prenatal diagnosis of isolated, non‐severe LCDH, defined as observed‐to‐expected lung‐to‐head ratio (o/e‐LHR) > 25%, that were managed expectantly during pregnancy followed by standardized neonatal management. An additional inclusion criterion was the availability of IPA Doppler measurements. The primary outcome was the association between IPA Doppler findings and mortality at discharge. Other predictors included o/e‐LHR, liver herniation and gestational age at birth. Secondary outcomes were the association between IPA Doppler findings and the presence of pulmonary hypertension (PHT), need for supplemental oxygen at discharge and need for extracorporeal membrane oxygenation. IPA pulsatility index (PI) values were converted into Z‐scores. Logistic regression analysis was performed to investigate the associations between predictor variables and outcome, and the best model was chosen based on the Nagelkerke R2.
Results
Observations for 70 non‐severe LCDH cases were available. Fifty‐four (77%) fetuses survived until discharge. On logistic regression analysis, higher IPA‐PI was associated with an increased risk of mortality (odds ratio (OR), 3.96 (95% CI, 1.62–9.70)), independently of o/e‐LHR (OR, 0.87 (95% CI, 0.79–0.97)). An IPA‐PI Z‐score cut‐off of 1.8 predicted mortality with a detection rate of 69% and specificity of 93%. Adding IPA‐PI to o/e‐LHR improved significantly the model's performance (Nagelkerke R2, 46% for o/e‐LHR + IPA‐PI vs 28% for o/e‐LHR (P < 0.002)), with a detection rate of 81% at a 10% false‐positive rate. IPA‐PI was associated with PHT (OR, 2.20 (95% CI, 1.01–4.59)) and need for oxygen supplementation at discharge (OR, 1.90 (95% CI, 1.10–3.40)), independently of lung size.
Conclusions
In fetuses with mild or moderate LCDH, IPA‐PI was associated with mortality and morbidity, independently of lung size. A model combining o/e‐LHR with IPA‐PI identified up to four in five cases that eventually died, despite being considered to have non‐severe pulmonary hypoplasia. © 2021 International Society of Ultrasound in Obstetrics and Gynecology.