Management of the single liver nodule in a cirrhotic patient: a decision analysis model

The widespread use of liver imaging in patients with cirrhosis results in the discovery of small (<3 cm) nodules. Although the subsequent management of these patients is variable, it is generally focused on the diagnosis and treatment of hepatocellular carcinoma (HCC). We aimed to compare the 3-year survival associated with several competing strategies used in the management of patients with compensated liver cirrhosis in whom a single small liver lesion is detected during surveillance. We constructed a decision analysis model using a decision tree and Markov model. We assumed that all patients undergo an initial "diagnostic phase" consisting of an imaging study and serum alpha-fetoprotein (AFP). Patients with a "positive initial diagnostic phase" for HCC are referred for either imaging-guided biopsy (IGB) or surgical resection or orthotopic liver transplantation (OLT) without preceding IGB. IGB, if positive for HCC, was followed by OLT, surgical resection, or local therapy. Patients with a "negative initial diagnostic phase" undergo either repeat diagnostic testing (imaging, AFP) every 4 months or are referred for either OLT, surgical resection, or IGB followed by interventions. Probability assumptions were estimated from the published literature. The outcomes compared were 3-year overall survival and recurrence-free survival. When the initial diagnostic phase is positive for HCC, OLT it is associated with the longest survival. In the sensitivity analysis, when the 3-year overall survival for patients referred to OLT is <54%, surgical resection or IGB preceding therapy become more favorable strategies. This 3-year overall survival (<54%) associated with OLT is reached after a waiting time of 4 months on the transplant list, if a 4% monthly dropout rate is assumed. When the initial diagnostic phase is negative for HCC, then performing IGB, before proceeding to therapeutic intervention, is associated with the longest 3-year overall survival. If the IGB is positive, subsequent OLT is associated with the longest survival. The higher the predictive value of the initial diagnostic phase for HCC, the more favorable is OLT (for the "positive results" arm), and follow-up testing (for the "negative results" arm). In conclusion, given a high pretest likelihood of HCC in a single liver nodule detected during surveillance in patients with cirrhosis, IGB may not be required in the presence of a positive noninvasive diagnostic testing. The long waiting time prior to OLT limits its advantage over surgical resection in the treatment of patients with early HCC.