Child-Pugh, MELD, MELD-Na, and ALBI scores: which liver function models best predicts prognosis for HCC patient with ascites?

Objective It remains controversial whether certain treatments should apply to HCC patients with ascites due to concerns about worsening liver function. The objective of the present study is to compare the prognostic performance of 4 liver function models currently in use for HCC patients with ascites. Methods A total of 437 treatment-naive, newly diagnosed HCC patients were analyzed. The predictive performance of Child-Pugh, MELD, MELD-Na, and ALBI scores were examined using ROC curve analysis. Results MELD-Na score showed good performance in predicting 1-, 2-, and 3-year mortality, particularly 1-year mortality. MELD-Na score significantly increased at 30 days after treatment in cases initially receiving best supportive care (14-17,p < .001), TACE (9-11,p < .001), and other treatment (radiotherapy, sorafenib, or systemic chemotherapy) (9-11,p = .021). For patients with advanced tumor stage and MELD-Na score >= 12, HCC-specific treatment did not offer significantly better prognosis compared with only the best supportive care (median survival: 2.2 vs. 1.8 months for HCC-specific treatment vs. best supportive care,p = .15). Conclusion MELD-Na can effectively identify liver functional reserve and prognosis in HCC patients with ascites. MELD-Na, together with the tumor stage, may help establish a therapeutic strategy for them.