Abstract
BACKGROUND: Hepatocellular carcinoma (HCC) is the second leading cause of cancer-related mortality worldwide. Non-coding RNAs play an important role in HCC. This study aims to identify a senescence-related non-coding RNA network-based prognostic model for individualized therapies for HCC. METHODS: HCC subtypes with senescence status were identified on the basis of the senescence-related genes. Immune status of the subtypes was analyzed by CIBERSORT and ESTIMATE algorithm. The differentially expressed mRNAs, microRNAs (miRNAs), and long non-coding RNAs (lncRNAs) were identified between the two HCC subtypes. A senescence-based competing endogenous RNA (ceRNA) co-expression network in HCC was constructed. On the basis of the ceRNA network, Lasso Cox regression was used to construct the senescence-related prognostic model (S score). The prognosis potential of the S score was evaluated in the training dataset and four external validation datasets. Finally, the potential of the prognostic model in predicting immune features and response to immunotherapy was evaluated. RESULTS: The HCC samples were classified into senescence active and inactivate subtypes. The senescence active group showed an immune suppressive microenvironment compared to the senescence inactive group. A total of 2,902 mRNAs, 19 miRNAs, and 308 lncRNAs were identified between the two subtypes. A ceRNA network was constructed using these differentially expressed genes. On the basis of the ceRNA network, S score was constructed to predict the prognosis of patients with HCC. The S score was correlated with immune features and can predict response to immunotherapy of cancer. CONCLUSION: The present study analyzed the biological heterogeneity across senescence-related subtypes and constructed a senescence-related ceRNA-network-based prognostic model for predicting prognosis and immunotherapy responsiveness.