Background: With the introduction of less invasive modalities to treat hepatocellular carcinoma (HCC), the debate regarding whether hepatectomy is required has emerged. Recent studies suggest that radiofrequency ablation (RFA) is equivalent to resection for small solitary tumors. Decision analysis modeling is a unique method of evaluating medical decisions and highlighting features that might persuade us that one method is, in fact, superior. The objective of this study was to use decision analysis to predict which of these two treatments would result in the longest average survival for patients with early hepatocellular carcinoma.
Methods: An English language literature review was performed in PubMed and Medline to locate articles reporting outcomes for patients with hepatocellular carcinoma after hepatectomy and/or RFA. A decision tree was constructed in TreeAge Pro (Williamstown, MA). Weighted averages of values from the literature were used to determine the probabilities entered in the decision tree. Sensitivity analyses vary individual variables over a range of values and were performed to determine if varying individual probabilities impacted the outcome of the model.
Results: The following weighted averages were used in the mathematical model: 3.5% peri-operative mortality after hepatectomy, 69.2% 5-year tumor recurrence rate, 22-month median disease free interval, and 80.5% of recurrences amenable to salvage therapies. For RFA, we used a 0.2% peri-procedural mortality, 72.4% 5-year tumor recurrence rate, 16-month median recurrence free interval, and 77.4% of post RFA recurrences amenable to additional treatment. A 46-month median survival for patients who recur and are amenable to additional treatment was used for both groups, as was a 7.5-month median survival for patients unable to received additional therapy. The decision model predicted a 68.9-month survival after hepatectomy and a 64.7-month survival after RFA for a 4.2-month survival benefit to for hepatectomy over RFA. This preferred treatment strategy was insensitive to changes within the published range for peri-procedural/operative mortality, 5-yr. tumor recurrence rate after RFA, median time to recurrence, percentage amenable to additional therapy, or the median survival of patients ineligible to additional treatment.
Conclusion: The baseline analysis predicts that hepatectomy results in a 4.2 month longer life expectancy compared to the less invasive RFA treatment. However, RFA does result in an equivalent or potentially superior survival benefit over hepatectomy for certain variables within the range reported in the literature.