Construction of Lymph Node Metastasis-Related Prognostic Model and Analysis of Immune Infiltration Mode in Lung Adenocarcinoma
Related Experiment Video
Updated: Sep 4, 2025
Author Spotlight: Investigating Immune Cell Dynamics in the Tumor Microenvironment — Challenges and Innovations in Cancer Prognosis
Published on: April 12, 2024
Wujin Li1, Debin Ou1, Jiguang Zhang1
1Thoracic Surgery Department, Fujian Provincial Hospital, Fuzhou 350001, China.
Background:
Lung adenocarcinoma (LUAD) is a major cause for global cancer-related deaths. Research reports demonstrate that lymph node metastasis (LNM) is pertinent to the survival rate of LUAD patients, and crux lies in the lack of biomarkers that could distinguish patients with LNM. We aimed to verify the LNM-related prognostic biomarkers in LUAD.
Methods:
We firstly accessed the expression data of mRNA from The Cancer Genome Atlas (TCGA) database and then obtained samples with LNM (N+) and without LNM (N-). Differential expression analysis was conducted to acquire differentially expressed genes (DEGs). Univariate-LASSO-multivariate Cox regression analyses were performed on DEGs to build a risk model and obtain optimal genes. Afterwards, effectiveness and independence of risk model were assessed based on TCGA-LUAD and GSE31210 datasets. Moreover, a nomogram was established combining clinical factors and riskscores. Nomogram performance was measured by calibration curves. The infiltration abundance of immune cells was scored with CIBERSORT to explore the differences between high- and low-risk groups. Lastly, gene set enrichment analysis (GSEA) was used to investigate differences in immune features between the two risk groups.
Results:
Nine optimal feature genes closely related to LNM in LUAD were identified to construct a risk model. Prognostic ability of the risk model was verified in independent databases. Patients were classified into high- and low-risk groups in accordance with their median riskscores. CIBERSORT score displayed differences in immune cell infiltration like T cells CD4 memory resting between high/low-risk groups. LNM-related genes may also be closely relevant to immune features. Additionally, GSEA indicated that differential genes in the two risk groups were enriched in genes related to immune cells.
Conclusion:
This research built a risk model including nine optimal feature genes, which may be potential biomarkers for LUAD.
