Abstract
Histone deacetylases comprise a family of 18 genes, and classical HDACs are a promising class of novel anticancer drug targets. However, to date, no systematic study has been comprehensive to reveal the potential significance of these 18 genes in lung adenocarcinoma (LUAD). Here, we used a systematic bioinformatics approach to comprehensively describe the biological characteristics of the HDACs in LUAD. Unsupervised consensus clustering was performed to identify LUAD molecular subtypes. The ssGSEA, CIBERSORT, MCP counter, and ESTIMATE algorithms were used to depict the tumor microenvironment (TME) landscape. The Cox proportional hazards model and LASSO regression analyses were used to construct the HDAC scoring system for evaluating the prognosis of individual tumors. In this study, three distinct HDAC-mediated molecular subtypes were determined, which were also related to different clinical outcomes and biological pathways. HDACsCluster-C subtype had lowest PD-L1/PD-1/CTLA4 expression and immune score. The constructed HDAC scoring system (HDACsScore) could be used as an independent predictor to assess patient prognosis and effectively identify patients with different prognosis. High- and low-HDACsScore groups presented distinct genetic features, immune infiltration, and biological processes. The high-HDACsScore group was more likely to benefit from immunotherapy, as well as from the application of common chemotherapeutic agents (cyclopamine, docetaxel, doxorubicin, gemcitabine, paclitaxel, and pyrimethamine). Overall, HDAC family genes play important roles in LUAD, and the three LUAD subtypes and the HDAC scoring system identified in this study would help enhance our perception of LUAD prognostic differences and provide important insights into the efficacy of immunotherapy and chemotherapy.