A bioinformatics-based immune-related prognostic index for lung adenocarcinoma that predicts patient response to immunotherapy and common treatments

Chenghao Wang; Tong Lu; Ran Xu; Xiaoyan Chang; Shan Luo; Bo Peng; Jun Wang; Lingqi Yao; Kaiyu Wang; Zhiping Shen; Jiaying Zhao; Linyou Zhang

doi:10.21037/jtd-22-494

A bioinformatics-based immune-related prognostic index for lung adenocarcinoma that predicts patient response to immunotherapy and common treatments

J Thorac Dis. 2022 Jun;14(6):2131-2146. doi: 10.21037/jtd-22-494.

Authors

Chenghao Wang^#¹, Tong Lu^#¹, Ran Xu¹, Xiaoyan Chang¹, Shan Luo², Bo Peng¹, Jun Wang¹, Lingqi Yao¹, Kaiyu Wang¹, Zhiping Shen¹, Jiaying Zhao¹, Linyou Zhang¹

Affiliations

¹ Department of Thoracic Surgery, The Second Affiliated Hospital of Harbin Medical University, Harbin, China.
² Second Clinical College of Medicine, Harbin Medical University, Harbin, China.

^# Contributed equally.

Abstract

Background: There is increasing evidence of the effectiveness of immune checkpoint blockade (ICB) therapy for the treatment of lung adenocarcinoma (LUAD). However, the benefits of ICB therapy vary among LUAD patients. Due to the research dimension, existing biomarkers, such as programmed death-ligand 1 (PD-L1) expression and tumor mutation burden (TMB), could not reflect the complex tumor environment, and had low prediction accuracy of ICB. Therefore, we aimed to uncover a prognostic biomarker that could also predict whether a patient would benefit from ICB therapy and other common treatments from multiple dimensions, so as to improve the prediction accuracy of pre-treatment patients.

Methods: Based on the LUAD dataset retrieved from The Cancer Genome Atlas (TCGA) database, 50 immune-related hub genes were identified using weighted gene co-expression network analysis and univariate Cox regression analyses. An immune-related gene prognostic index (IRGPI) was constructed using a Cox proportional-hazards model based on 15 genes and validated using GSE72094 dataset. We tested its prognostic accuracy by Kaplan-Meier (K-M) survival curves of the two datasets and assessed its predictive power by comparing area under curve (AUC) of IRGPI with existing biomarkers. Subsequently, we analyzed the molecular and immune characteristics, and evaluated the benefits of ICB by PD-L1 expression and Tumor Immune Dysfunction and Exclusion (TIDE) analysis, predicted the inhibitory concentration 50 of common treatments drugs for two IRGPI score-related subgroups.

Results: Patients in the IRGPI-high subgroup had lower overall survival (OS) than patients in the IRGPI-low subgroup in K-M survival curve in two cohorts. And IRGPI has AUC values of 0.715, 0.724, and 0.743 in 1, 2, and 3 years, respectively. A higher tumor mutation burden and PD-L1 expression and the tumor microenvironment (TME) landscape demonstrated that IRGPI-high subgroup patients may respond better to ICB therapy. Genomics of Drug Sensitivity in Cancer (GDSC) analysis indicated that the IRGPI-high subgroup showed greater sensitivity to chemotherapy.

Conclusions: IRGPI is a prospective biomarker for evaluating whether a patient will benefit from ICB therapy and other treatments, and distinguishing patients with different molecular and immune characteristics.

Keywords: Immune-related genes; immune cell infiltration; immune checkpoint blockade therapy (ICB therapy); lung adenocarcinoma (LUAD); prognosis biomarker.