基于MRI的新型参数DHER预测左侧先天性膈疝结局

原文

BACKGROUND: In congenital diaphragmatic hernia (CDH), the size of the diaphragmatic defect is a key determinant of lung hypoplasia, pulmonary hypertension, survival and complexity of postnatal surgical repair. Current prenatal prognostic assessment relies on indirect surrogates of disease severity-such as lung volume measurements and liver herniation-that do not directly quantify the anatomical defect. Direct prenatal assessment of diaphragmatic defect size remains largely unexplored, with limited evidence available. OBJECTIVE: To introduce the Defect-to-HEmithorax Ratio (DHER), a novel prenatal magnetic resonance imaging (MRI)-derived parameter designed to quantify diaphragmatic defect size, and to assess its predictive value for postnatal outcomes and surgical repair complexity in fetuses with isolated left CDH. STUDY DESIGN: This retrospective study included all fetuses with isolated left-sided CDH consecutively evaluated with prenatal MRI at a single tertiary center between 2014 and 2024. Right-sided CDH, associated structural or genetic anomalies, and pregnancy termination were excluded. DHER was calculated as the ratio between the diaphragmatic defect area and the corresponding hemithorax area on standard MRI sequences. Associations between DHER and postnatal outcomes (survival at discharge, extracorporeal membrane oxygenation [ECMO] requirement, persistent pulmonary hypertension of the newborn [PPHN], duration of mechanical ventilation [MV], length of hospitalization [LOS], and oxygen dependence at 28 days of life) as well as indicators of surgical complexity (defect type according to CDH Study Group classification and need for patch repair) were evaluated. Correlations with established prenatal prognostic markers (observed-to-expected lung-to-head ratio [O/E LHR], liver herniation, stomach position, mediastinal shift angle [MSA], and total fetal lung volumes [TFLV]) and interobserver reproducibility were also assessed. RESULTS: Seventy-five fetuses met inclusion criteria. Higher DHER values were significantly associated with lower survival at discharge, increased ECMO requirement, longer MV duration and LOS, and increased PPHN (all P<.001), while no significant association was found with oxygen dependence at 28 days (P=.27). ROC analysis showed excellent predictive performance of DHER for survival (AUC=0.91, P<.001), with higher discriminative ability in our cohort than O/E LHR (AUC=0.74, P=.002) and O/E TFLV (AUC=0.74, P=.01). In multivariable analysis, DHER remained independently associated with survival and significantly improved model performance (LR test P<.001). It was also significantly associated with larger surgical defect types (C-D) and need for patch repair (both P<.001), strongly correlated with all existing prenatal severity markers (P≤.002), and demonstrated high interobserver reproducibility (ICC=0.86). CONCLUSIONS: DHER is a prenatal MRI-derived parameter that, through a geometrical simplification, provides an accurate and reproducible estimate of diaphragmatic defect size and is strongly associated with postnatal outcomes and surgical complexity in left-sided CDH. These findings support its potential role as a complementary tool for individualized prenatal risk stratification. Prospective multicenter studies are required to confirm its clinical value and to assess its applicability in right-sided CDH and cases undergoing FETO.