We read the article published by Martín-Garre1 with interest. However, in the light of that reading, we would like to make a few comments that we believe to be important.

The Fontan procedure has been of particular benefit to infants with functional single-ventricle complexes but an inevitable consequence is systemic venous hypertension. Hepatic histology of patients with Fontan circulation usually begins with sinusoidal dilatation, parenchymal atrophy, and progressive fibrosis secondary to repetitive mechanical stretch due to persistent chronic passive venous congestion and limited cardiac output, which favors tissue hypoxia. Hepatocarcinogenesis forms part of the continuum of dedifferentiation that includes hypervascular nodules, regenerative nodules, dysplastic nodules, and hepatocellular carcinoma (HCC). Although ultrasound (US) remains inexpensive and is recommended as the first choice for the screening and surveillance of HCC by the guidelines of almost all international societies, Fontan patients have some peculiarities that must be taken into account.

First, US imaging findings in long-standing Fontan patients may be characterized by hepatomegaly, hepatic vein and suprahepatic inferior vena cava dilation, surface nodularity, increased parenchymal echogenicity, and HCC, which is usually a nodule greater than 1cm in diameter. The classic US findings of HCC include hypoechoic nodules or mixed echogenic nodules due to tumor necrosis or fatty metamorphosis or a surrounding thin hypoechoic band indicating a capsule that is characteristic of these tumors.2 In addition, as mentioned by Martín-Garre, the form of presentation of HCC may vary (multiplicity of nodules, small sized nodules, and “nodules within nodules”).

Second, standard US can assess nodularity with variable accuracy (the sensitivity and specificity for HCC diagnosis are 60% and 93%, respectively, and are even poorer for HCC less than 1cm). Doppler US may be used to assess portal vein flow and the presence of collateral vessels suggesting portal hypertension. In addition, color Doppler flow imaging may show hypervascularity and tumor vascular shunting. Nonetheless, both nodularity and portal flow changes are late findings and are not therefore helpful in detecting signs of early hepatic compromise,3 which is of particular importance due to the significant impact of even mild liver disease on the outcome of cardiac surgery. Similarly, contrast-enhanced US may improve the detection of cirrhosis and may reflect the real-time dynamics of blood supply of the lesion, which is helpful in both the detection and characterization of HCCs, but again does not accurately distinguish earlier stages of fibrosis.

Third, US may be adequate for screening cirrhosis in general but is not the preferred option in Fontan patients due to the high incidence of nonmalignant vascular lesions. In fact, the presence of arterialized nodules in Fontan patients is relatively frequent and, although these nodules are benign and pathologically identifiable as focal nodular hyperplasia, they can be confused with HCC, which is increasingly reported even in the absence of frank cirrhosis.4

Finally, although there are no data on the precise incidence of HCC, the fact that most Fontan patients have structural hepatic derangements at the initiation of screening confirms that we are initiating imaging too late. Therefore, periodic imaging should be started in childhood or early adolescence.5 The ideal screening test is gadolinium-enhanced magnetic resonance imaging (with a sensitivity and a specificity of 91% and 95%, respectively, to detect HCC)2 at intervals of 3 to 4 years. In patients with contraindications to magnetic resonance, a computed tomography scan may be performed, although the risk of radiation exposure must be considered. Meanwhile, liver stiffness estimated by US and magnetic resonance elastography techniques may be used as a quantitative imaging biomarker for the detection, staging, characterization, and monitoring of liver fibrosis. However, the use of elastography in Fontan-associated liver disease is problematic because any altered hepatic stiffness beyond fibrosis, particularly the vascular congestion universally present in Fontan patients, may have an impact on the results.

For all these reasons, some authors recommend that patients with a Fontan procedure performed more than 10 years previously should undergo cardiac assessment, liver imaging, and even liver biopsy4 to stay ahead of neoplastic transformation. Even after heart transplant, patients who have undergone the Fontan procedure will require vigilant screening for HCC.