Hepatic adenoma masquerading as a hepatocellular carcinoma in a patient with a Fontan procedure on the oral contraceptive pill

Abstract

A 21-year-old woman with an extra-cardiac conduit Fontan for tricuspid atresia and pulmonary stenosis presented for routine review. Medications included warfarin, the oral contraceptive pill (OCP) and sulfasalazine (Crohn disease). She had good exercise tolerance. Biochemistry demonstrated: bilirubin 18, alanine aminotransferase (ALT) 65, aspartate aminostransferase (AST) 31, gamma-glutamyl transferase (GGT) 66, alkaline phosphatase (ALP) 118, albumin 42, alpha-fetoprotein 2.6 and brain-natriuretic-peptide <10. Transthoracic echocardiogram revealed an unobstructed Fontan pathway. Liver ultrasound showed a 12 mm nodule suggestive of a focal nodular hyperplasia (FNH)-like lesion. Quadruple-phase computed tomography showed arterial hyperenhancement and washout consistent with hepatocellular carcinoma (HCC). After multidisciplinary review, she proceeded to percutaneous biopsy and microwave ablation. Histology described a well differentiated hepatocellular lesion with sinusoidal ectasia, inflammation and solitary arterioles with ‘pseudoportal tracts’ (Fig. 1). The favoured differential was well differentiated HCC. A dysplastic nodule or nodular hyperplasia was considered. Further immunohistochemistry showed diffuse lesional C-reactive protein and serum amyloid-A positivity, consistent with an inflammatory-subtype hepatic adenoma (HA). β-catenin and glutamine synthetase staining were normal, indicating no underlying β-catenin activation. Background liver showed features of Fontan-associated liver disease (FALD). The OCP was ceased. Magnetic resonance imaging (MRI) 6 months later showed no new lesions. A Fontan circulation remains the operative approach for patients with single ventricle physiology. Unfortunately, numerous complications including FALD are recognised. Mechanistically, FALD is thought to result from increased systemic venous pressure, reduced cardiac output, hepatic hypoxia and sinusoidal dilation causing fibrosis and cirrhosis. Recent guidelines recommend screening with 2-yearly blood tests, liver ultrasound and non-invasive hepatic fibrosis assessment. Development of cirrhosis requires enrolment into HCC screening. Benign hepatic nodules are seen in 20–30%, mainly FNH-like lesions hypothesised to result from arterialisation secondary to reduced portal blood flow from central venous hypertension. HA are benign tumours, rarely associated with cardiac hepatopathy or malignant transformation, but at risk of haemorrhage particularly with concomitant anticoagulation. OCP use conveys a 30-fold-risk of developing HA and additionally associates with larger tumour size and greater bleeding risk. Thus, OCP cessation is recommended. It is usually possible to diagnose a hepatic mass with contrast-enhanced computed tomography or MRI. HA are typically well circumscribed isointense lesions with early arterial enhancement and delayed-phase isodensity. HCC are mosaic with arterial hyperenhancement and delayed-phase washout. Confident diagnosis is usually made without histology. However, imaging in FALD is challenging and biopsy with immunohistochemistry is an invaluable adjunct. OCP use in patients with a Fontan circulation is controversial due to thromboembolic risks of oestrogen in a stasis-prone circulation. Long-acting progesterone-only intrauterine devices (IUD) and subcutaneous rods are highly efficacious compared with oral progesterone-only contraception. However, IUD insertion can precipitate vasovagal reactions, which can be precarious in this population. Additionally, there is a small risk of infection and endocarditis. Some centres have a permissive approach to OCP use if accompanied by concomitant anticoagulation. This patient’s HA was thought likely to reflect OCP use. Diagnosis was complicated by the associated FALD changes. The present case exemplifies the importance of both FALD surveillance and multidisciplinary care for these medically complex individuals.