Abstract
BACKGROUND: Fontan patients frequently develop liver cirrhosis (LC); however, the diagnostic accuracy of ultrasound (US) for detecting LC and the clinical implications of such diagnoses have not been clearly established. OBJECTIVES: This study aims to evaluate the diagnostic performance of US for detecting LC in an adult population with Fontan circulation and to determine the correlation between LC and mortality/transplantation. METHODS: This was a retrospective study. Data on cross-sectional imaging, liver USs, and clinical visits that occurred within 12 months of the cross-sectional imaging were collected. Liver US diagnostic accuracy was evaluated against cross-sectional imaging. Kappa agreement between methods was assessed. Univariate Cox proportional hazards regression analysis was employed to compare mortality and transplant outcomes. RESULTS: Overall, 131 patients were included. Liver US and cross-sectional imaging (computed tomography 74, magnetic resonance imaging 57) was performed in all patients. Liver US reported heterogeneous parenchyma, lobar redistribution, and surface nodularity in 85.4%, 72.5%, and 65.6% of cases. Cross-sectional imaging reported these features in 60.3%, 87.0%, and 84.9% of cases, respectively. US sensitivity was greater than 0.75 for all variables, while specificity was 0.21, 0.58, and 0.85, respectively. LC was diagnosed in 78% of cases by US and in 90% by cross-sectional imaging, with a kappa agreement of 0.21 between techniques. There was no significant correlation between the presence of hepatic parenchymal changes or cirrhosis and mortality/transplantation. CONCLUSIONS: Liver US is effective for screening and monitoring liver cirrhotic features in the adult Fontan population. In a univariate analysis, there was no association between LC and mortality or transplantation.