Defining "liver-up": does the volume of liver herniation predict outcome for fetuses with isolated left-sided congenital diaphragmatic hernia?

PURPOSE: The aim of the study was to evaluate the relationship between the degree of liver herniation and mortality or need for extracorporeal membrane oxygenation (ECMO) in isolated left-sided congenital diaphragmatic hernia (CDH).
METHODS: This is a retrospective review of all fetuses with isolated left-sided Bochdalek-type CDH evaluated at our center with ultrafast fetal magnetic resonance imaging between January 2004 and December 2010. Percentage of liver herniation was defined as the ratio of herniated liver volume to total fetal liver volume. The liver/thoracic volume ratio was defined as the ratio of herniated liver volume to total thoracic volume. Data were analyzed using receiver operating characteristic curves and Fisher's Exact and Mann-Whitney U tests.
RESULTS: Fifty-three fetuses with isolated left-sided CDH were evaluated. Fetuses with "liver-up" (n = 32) and "liver-down" (n = 21) anatomy had similar rates of mortality (25% vs 14%, P = .49) and ECMO use (41% vs 29%, P = .40). The accuracy of liver-up (a dichotomous variable) to predict mortality or need for ECMO was 49% and 53%, respectively. Percentage of liver herniation greater than 21% was associated with mortality (P < .001) or need of ECMO (P < .001), with an accuracy of 87% and 79%, respectively. Liver/thoracic volume ratio of greater than 14% was also associated with mortality or ECMO use (P < .001 and P = .01, respectively), with an accuracy of 85% and 72%, respectively.
CONCLUSION: Increased amounts of liver herniation in fetuses with isolated left-sided CDH are associated with higher rates of mortality or the need for ECMO support. Quantification of liver herniation (a continuous variable) is superior to overall liver herniation as a dichotomous variable (liver-up vs liver-down) in the prediction of perinatal morbidity and mortality.