The investigators will investigate the usefulness of biannual ultrasonography versus annual non-contrast magnetic resonance imaging for surveillance of hepatocellular carcinoma in single arm patients.

Hepatocellular carcinoma (HCC) is one of the major complications in patients with chronic liver disease. The prognosis of HCC relies on the extent of disease at the time of diagnosis. Hence to detect cancer at an earlier stage, a regular surveillance test is important for the subjects with a high risk of developing cancer. Current guidelines recommend a regular surveillance using ultrasonography (US) at a 6 month-interval. Ultrasonography is a non-invasive and safe procedure, yet it is limited by the skills of the operator and it is often difficult to differentiate cancer from regenerating nodules especially in atrophied cirrhotic liver. According to a recent meta-analysis, the overall sensitivity and specificity of detecting liver cancer using ultrasonography were both over 90%, however the sensitivity was decreased to 60% in detecting early lesions in which surgery or liver transplantation is indicated. Of note, addition of alpha feto protein(AFP) to ultrasonography also failed to increase the sensitivity in detecting small cancers. Taken together, there is a need to develop a new surveillance test with an improved sensitivity and specificity.

Recently, a retrospective study reported that CT or MRI showed a better sensitivity than ultrasonography in detecting early liver cancer. However, performing CT as a surveillance test is limited by frequent exposure to radiation and contrast-dye agent. Although MRI does not have the risk of radiation-exposure, it is limited by the high cost and limited availability. On the contrast, non-contrast MRI offers a cost that is comparable to US and an absence of exposure to radiation or contrast-dye agent, which suggests non-contrast MRI as a good alternative surveillance tool for early detection of HCC. Therefore, in this prospective trial, the investigators will investigate the usefulness of biannual ultrasonography versus annual non-contrast magnetic resonance imaging for surveillance of hepatocellular carcinoma.

Inclusion Criteria:

• Male or female subject older than 40 years of age at the enrollment
• Fulfill below conditions A and B

A. The evidence of cirrhosis of any etiology within 12 months prior to screening Definition of cirrhosis by any of following methods

1. Histologically by liver biopsy;

2. Non-histologically by evidence of portal hypertension in the presence of chronic liver disease;

   • Evidence of portal hypertension, including any of followings;

     • The identification of splenomegaly on ultrasonography, CT, or MRI examinations with typical features of cirrhosis
     • The identification of esophageal or gastric varices on endoscopic examination

B. Risk Index larger than 2.33; Risk Index = 1.65 (if the prothrombin activity is <=75%) + 1.41 (if the age is 50 years or older) + 0.92 (if the platelet count is <=100,000/mm3) + 0.74 (if the presence of HCV is positive).

Exclusion Criteria:

• Presence of liver cancer or other intrahepatic malignancy
• Has a history of malignancy within previous 5 years
• Is pregnant or breast-feeding

Patients with the risk of annual HCC occurrence larger than 5%