Fusing WGCNA and Machine Learning for Immune-Related Gene Prognostic Index in Lung Adenocarcinoma: Precision Prognosis, Tumor Microenvironment Profiling, and Biomarker Discovery.

Background: The objective is to create an IRGPI (Immune-related genes prognostic index), which could predict the survival and effectiveness of immune checkpoint inhibitor (ICI) treatment for lung adenocarcinoma (LUAD).

Methods: By applying weighted gene co-expression network analysis (WGCNA), we ascertained 13 genes associated with immune functions. An IRGPI was constructed using four genes through multicox regression, and its validity was assessed in the GEO dataset. Next, we explored the immunological and molecular attributes and advantages of ICI treatment in subcategories delineated by IRGPI. The model genes were also validated by the random forest tree, and functional experiments were conducted to validate it.

Results: The IRGPI relied on the genes CD79A, IL11, CTLA-4, and CD27. Individuals categorized as low-risk exhibited significantly improved overall survival in comparison to those classified as high-risk. Extensive findings indicated that the low-risk category exhibited associations with immune pathways, significant infiltration of CD8 T cells, M1 macrophages, and CD4 T cells, a reduced rate of gene mutations, and improved sensitivity to ICI therapy. Conversely, the higher-risk group displayed metabolic signals, elevated frequencies of TP53, KRAS, and KEAP1 mutations, escalated levels of NK cells, M0, and M2 macrophage infiltration, and a diminished response to ICI therapy. Additionally, our study unveiled that the downregulation of IL11 effectively impedes the proliferation and migration of lung carcinoma cells, while also inducing cell cycle arrest.

Conclusion: IRGPI is a biomarker with significant potential for predicting the effectiveness of ICI treatment in LUAD patients and is closely related to the microenvironment and clinicopathological characteristics.