Relation of Magnetic Resonance Elastography to Fontan Failure and Portal Hypertension

Fontan associated liver disease is associated with morbidity and mortality in palliated single-ventricle congenital heart disease patients. Magnetic resonance elastography (MRE) provides a quantitative assessment of liver stiffness in Fontan patients. We hypothesized that MRE liver stiffness correlates with liver enzymes, hemodynamics, portal hypertension, and Fontan failure (FF). All adult Fontan patients who had MRE between 2011 and 2018 were included. Radiologic portal hypertension was defined as splenomegaly, ascites, and/or varices. FF was defined as death, transplantation, or heart failure symptoms requiring escalation of diuretics. Seventy patients with a median age of 24.7 years and a median follow-up from MRE of 3.9 years were included. The median liver stiffness was 4.3 kPa (interquartile range [IQR]: 3.8 to 5.0 kPa). There was a weak, positive correlation between liver stiffness and Fontan pathway pressure (r = 0.34, p = 0.03). There was a moderate negative correlation of liver stiffness with ventricular ejection fraction (r = -0.52, p = 0.03). Liver stiffness was weakly positively correlated with liver transaminases and gamma glutamyl transferase. Patients with portal hypertension had higher liver stiffness compared to patients without (5.2 ± 1.3 vs 4.2 ± 0.8 kPa, p = 0.03). At MRE or during follow-up, 13 patients (19%) met definition of FF and had significantly higher liver stiffness compared to patients without FF (5.1 [IQR: 4.3 to 6.3] vs 4.2 [IQR: 3.7 to 4.7] kPa, p = 0.01). Liver stiffness above 4.5 kPa differentiated FF with a sensitivity of 77% and specificity of 77%. In conclusion, elevated MRE-derived liver stiffness is associated with worse hemodynamics, liver enzymes and clinical outcomes in Fontan patients. This measure may serve as a global imaging biomarker of Fontan health.