Identification of a 3-Gene Model as Prognostic Biomarker in Patients With Gastric Cancer

Siming Xue; Tianjiao Zheng; Juan Yan; Jinmin Ma; Cong Lin; Shichen Dong; Chen Wei; Tong Li; Xiaoyin Zhang; Guibo Li

doi:10.3389/fonc.2022.930586

Identification of a 3-Gene Model as Prognostic Biomarker in Patients With Gastric Cancer

Front Oncol. 2022 Jul 14:12:930586. doi: 10.3389/fonc.2022.930586. eCollection 2022.

Authors

Siming Xue^{1

2}, Tianjiao Zheng³, Juan Yan³, Jinmin Ma¹, Cong Lin¹, Shichen Dong^{1

2}, Chen Wei^{1

2}, Tong Li^{1

2}, Xiaoyin Zhang⁴, Guibo Li^{1

2}

Affiliations

¹ Beijing Genomics Institute (BGI)-Shenzhen, Shenzhen, China.
² Beijing Genomics Institute (BGI)-Henan, BGI-Shenzhen, Xinxiang, China.
³ Beijing Genomics Institute (BGl) College & Henan Institute of Medical and Pharmaceutical Sciences, Zhengzhou University, Zhengzhou, China.
⁴ Department of Gastroenterology, National Clinical Research Center of Infectious Disease, The Third People's Hospital of Shenzhen, The Second Affiliated Hospital of Southern University of Science and Technology, Shenzhen, China.

Abstract

Objective: Although the incidence of gastric cancer (GC) is decreasing, GC remains one of the leading cancers in the world. Surgical resection, radiotherapy, chemotherapy, and neoadjuvant therapy have advanced, but patients still face the risk of recurrence and poor prognosis. This study provides new insights for assessment of prognosis and postoperative recurrence of GC patients.

Methods: We collected paired cancer and adjacent tissues of 17 patients with early primary GC for bulk transcriptome sequencing. By comparing the transcriptome information of cancer and adjacent cancer, 321 differentially expressed genes (DEGs) were identified. These DEGs were further screened and analyzed with the GC cohort of TCGA to establish a 3-gene prognostic model (PLCL1, PLOD2 and ABCA6). At the same time, the predictive ability of this risk model is validated in multiple public data sets. Besides, the differences in immune cells proportion between the high- and low-risk groups were analyzed by the CIBERSORT algorithm with the Leukocyte signature matrix (LM22) gene signature to reveal the role of the immune microenvironment in the occurrence and development of GC.

Results: The model could divide GC samples from TCGA cohorts into two groups with significant differences in overall and disease-free survival. The excellent predictive ability of this model was also validated in multiple other public data sets. The proportion of these immune cells such as resting mast cells, T cells CD4+ memory activated and Macrophages M2 are significantly different between high and low risk group.

Conclusion: These three genes used to build the models were validated as biomarkers for predicting tumor recurrence and survival. They may have potential significance for the treatment and diagnosis of patients in the future, and may also promote the development of targeted drugs.

Keywords: RNA-seq; cox regression; gastric cancer; immune microenvironment; survival.