We performed a single-centre retrospective study comparing the accuracy of non-invasive elastography with liver biopsy in accurate assessment of Fontan-associated liver disease.

Fontan patients who underwent combined assessment with a percutaneous liver biopsy and non-invasive elastography between January 2015 and December 2023 at our Children’s hospital were included. Liver biopsies were classified using the Congestive Hepatic Fibrosis Score as early Fontan-associated liver disease (scores 1, 2) and advanced Fontan-associated liver disease (score 3/bridging fibrosis and score 4/cirrhosis). Elastography values were categorised as advanced Fontan-associated liver disease for liver elasticity >2.1 m/s by ultrasound and liver stiffness >5 KPa on magnetic resonance elastography.

We included 130 patients (116 children, 89%, mean age at biopsy: 14.6 years ± 3.6) who underwent liver biopsy at a mean duration of 11.1 years (±0.3) following Fontan surgery. Advanced Fontan-associated liver disease was noted in 41 (31.5%) patients with 13 (10%) showing frank cirrhosis. Pre-biopsy ultrasound showed advanced liver fibrosis in 18/125 (14%), with low sensitivity (23%), high specificity (90%), and low accuracy (68%, k = 0.1) in diagnosing advanced Fontan-associated liver disease. Similarly, pre-biopsy magnetic resonance elastography showed advanced fibrosis in 23/86 (27%) of patients, with low sensitivity (30%), fair specificity (75%), and low accuracy (63%, k = 0.1). Interestingly, advanced Fontan-associated liver disease was missed by ultrasound in 29% and by magnetic resonance elastography in 25% of patients. Advanced Fontan-associated liver disease was associated with lower platelet count (p = 0.02) and higher Gamma-glutamyl Transferase levels (p = 0.02).

Advanced hepatic fibrosis is common among paediatric Fontan patients. Non-invasive elastography may overestimate and underestimate the degree of liver fibrosis, and therefore, liver biopsy may be required for confirming disease severity.