COMBINING prenatal lung measurements with information collected after birth may help identify infants with congenital diaphragmatic hernia (CDH) who develop chronic lung disease (CLD), a retrospective study suggests.
The analysis found that predictive models distinguished infants with and without CLD more effectively as early postnatal information became available. The findings highlight the potential value of updating risk assessments beyond prenatal imaging alone.
Researchers at University Medical Center Mannheim, Germany, studied 202 infants treated between 2013 and 2021 who had available fetal magnetic resonance imaging data. Infants who died during their initial hospital stay were excluded. The team examined prenatal measurements, first-day findings and subsequent clinical-course factors.
CLD was defined by supplemental oxygen dependency on day 28. Severity was graded on day 56 or at discharge, whichever occurred first. Overall, 116 infants, or 57.4% of the study population, had CLD.
Prenatal imaging showed that infants who developed CLD had lower observed-to-expected fetal lung volumes. Their liver and stomach were also more frequently positioned within the chest.
Within the first day of life, the CLD group had lower birth weights, higher arterial carbon dioxide levels and higher oxygenation indices than infants without CLD.
Several features of the clinical course were also more frequent among infants with CLD. These included extracorporeal membrane oxygenation (ECMO) support, larger diaphragmatic defects, nitric oxide therapy and chylothorax.
The prenatal prediction model had an area under the receiver operating characteristic curve of 0.844. Adding first-day clinical data increased this to 0.909. An extended model incorporating fetal lung volume, the lowest arterial carbon dioxide level during the first day or before ECMO, and ECMO requirement achieved 0.946.
This measure describes how well a model separates infants with and without the outcome. It should not be read as a percentage probability that an individual infant will develop CLD, or as a direct measure of accuracy in another hospital.
The authors suggest that these data could support earlier risk stratification in CDH. Incorporating postnatal findings may provide information that prenatal measurements cannot capture on their own.
However, excluding infants who died introduces survivor bias, while requiring fetal MRI may also limit representativeness. There was no longitudinal respiratory follow-up after discharge. Independent validation is needed, particularly because treatment practices, including ECMO initiation, vary between centres.
The associations with ECMO and nitric oxide treatment also do not establish that these therapies cause CLD. Such treatments may identify infants with more severe underlying illness, underscoring the distinction between predicting an outcome and explaining its cause.
Reference
Hartmann J et al. Pre- and postnatal predictors of chronic lung disease in congenital diaphragmatic hernia. Pediatric Pulmonology. 2026;61(11):e71887. doi:10.1002/ppul.71887.
Featured image: Tomsickova on AdobeStock
Each article is made available under the terms of the Creative Commons Attribution-Non Commercial 4.0 License.
