The relationship between tumor infiltrating immune cells and the prognosis of patients with lung adenocarcinoma

Zhuang Tong; Xu Wang; Hongyu Liu; Jian Ding; Yinling Chu; Xin Zhou

doi:10.21037/jtd-22-1837

The relationship between tumor infiltrating immune cells and the prognosis of patients with lung adenocarcinoma

J Thorac Dis. 2023 Feb 28;15(2):600-610. doi: 10.21037/jtd-22-1837.

Authors

Zhuang Tong¹, Xu Wang², Hongyu Liu³, Jian Ding⁴, Yinling Chu⁴, Xin Zhou⁴

Affiliations

¹ Department of Thoracic Surgery, Cancer Hospital of Dalian University of Technology, Liaoning Cancer Hospital & Institute, Shenyang, China.
² Department of Gerontology, The Second Hospital of Shanxi Medical University, Taiyuan, China.
³ Department of Pathology, Qiqihar Hospital Affiliated to Southern Medical University, Qiqihar, China.
⁴ Department of Respiratory Medicine, First Hospital of Qiqihar, Qiqihar, China.

Abstract

Background: To depict the immune infiltration characteristics of tumor cells in patients with lung adenocarcinoma (LUAD) and evaluate the predictive value and significance of tumor immune cells on the prognosis of LUAD patients.

Methods: The clinical characteristics and transcriptome of LUAD patients were obtained from The Cancer Genome Atlas (TCGA), and the immune cell abundance in LUAD tissue was evaluated using the CIBERSORT algorithm. We created a simplified immune cell-based Cox regression model according to the survival status of patients and clarified the correlation between the survival status of patients and seven types of immune cells. An immune cell-based risk prediction model was created by Cox proportional hazards regression. Subsequently, the gene expression profile of LUAD patients was obtained from the Gene Expression Omnibus (GEO) database to validate the tumor immune infiltration and patient prognosis prediction model attained using the CIBERSORT algorithm.

Results: The abundance of 22 tumor-infiltrating immune cells in these patients was detected using the CIBERSORT algorithm. According to Pearson correlation analysis, the immune cells appeared to be closely related to each other. The immune cell composition was remarkably different between the LUAD tumor tissue and paracancerous tissue. The simplified COX model showed that seven kinds of immune cells have predictive value for the prognosis and survival status of LUAD. The receiver operating characteristic curve (ROC) curve confirmed that the prediction model performed well for 1-, 3-, and 5-year survival status. The calibration curve suggested that the prediction model was consistent with the clinical results. Correlation analysis revealed that the clinical features were significantly related to immune cell infiltration. A total of 246 LUAD specimens were from the GEO database, and the risk score model suggested that high risk scores were indicative of a poor prognosis. Finally, enzyme-linked immunosorbent assay (ELISA) revealed that the expressions of tumor necrosis factor-α (TNF-α), interleukin 8 (IL-8), IL-6, and interferon-γ (IFN-γ) in tumor tissues were remarkably higher compared with those in adjacent tissues.

Conclusions: There is a close correlation between the tumor-infiltrating immune cells and the prognosis and clinical characteristics of LUAD patients. The risk score model based on TCGA and GEO designed in this study can be applied in clinical practice.

Keywords: CIBERSORT; Lung adenocarcinoma (LUAD); bioinformatics analysis; tumor-infiltrating immune cells.