Abstract
PURPOSE: Hepatocellular carcinoma (HCC) responds poorly to immunotherapy, and the durable response rate is 10-20%. Here, we aim to characterize HCC classifications based on lactate genes to identify patients who may benefit from immunotherapy. METHODS: Lactate-related genes were applied for HCC classification in the current study, and lactate Cluster 1 (LC1) and lactate Cluster 2 (LC2) were defined. Differential genes from LC1 and LC2 helped define the following lactate phenotype clusters: lactate phenotype Cluster 1 (LPC1), lactate phenotype Cluster 2 (LPC2) and lactate phenotype Cluster 3 (LPC3). Based on the cluster annotation, the lactate score was defined and analyzed to evaluate the immunotherapy response. RESULTS: All the classified clusters were analyzed, and they showed different immune signatures. The survival rate of LPC3 was higher than that of LPC2 (LPC3 vs. LPC2, P = 0.027) and LPC1 (LPC3 vs. LPC1, P = 0.027). Then, the lactate score was annotated and confirmed to be effective in predicting responses to immune checkpoint blockade therapy. CONCLUSION: In the current study, we developed a classification system for HCC and defined the lactate score, which was validated to be partially effective in estimating responses among tumor patients.