Abstract
Objective: Given the crucial role of mitochondria in the prognosis and treatment of hepatocellular carcinoma (HCC), we aim to develop two independent mitochondrial scoring systems to separately predict patient prognosis and the likelihood of transarterial chemoembolization non-response (TACE NR). Methods: Mitochondria-related candidate genes were selected and analyzed using univariate Cox and LASSO Cox regression analyses to create a risk prognosis score (RPS). Univariate and LASSO logistic regression analyses were used to establish the risk diagnosis score (RDS). Alternative therapies for patients with TACE NR were explored using TIDE and oncoPredict algorithms. The Seurat package was used to study the involvement of the RDS genes in HCC differentiation. Results: The RPS accurately predicts the 1-5 year survival rates of patients with HCC, where higher RPS values were associated with poorer survival outcomes. The RDS model demonstrated a commendable performance in diagnosing TACE NR, as patients with a higher RDS exhibited a greater likelihood of TACE NR. RDS was associated with the infiltration of various immune cells, and patients with lower RDS tended to have higher response rates to immunotherapy and increased sensitivity to JAK1, rapamycin, and AZD2014. By contrast, patients with higher RDS values and a higher probability of TACE NR had more responsive to paclitaxel, dasatinib, and vincristine, suggesting that these drugs are potential alternative therapies. Single-cell sequencing studies have identified ACSM2A as a key player in HCC differentiation and a potential target for therapeutic intervention. Conclusion: The RPS and RDS are important reference points for predicting outcomes and guiding treatment decisions in patients with HCC. Additionally, ACSM2A shows promise as a potential therapeutic target for HCC.