A novel diagnostic model for predicting immune microenvironment subclass based on costimulatory molecules in lung squamous carcinoma

There is still no ideal predictive biomarker for immunotherapy response among patients with non-small cell lung cancer. Costimulatory molecules play a role in anti-tumor immune response. Hence, they can be a potential biomarker for immunotherapy response. The current study comprehensively investigat...

Full description

Saved in:
Bibliographic Details
Published in:Frontiers in genetics 2022-12, Vol.13, p.1078790-1078790
Main Authors: Duan, Fangfang, Wang, Weisen, Zhai, Wenyu, Wang, Junye, Zhao, Zerui, Zheng, Lie, Rao, Bingyu, Zhou, Yuheng, Long, Hao, Lin, Yaobin
Format: Article
Language:English
Subjects:
costimulatory molecules
diagnostic biomarker
Genetics
lung squamous carcinoma
machine learning algorithm
tumor immune microenvironment
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:There is still no ideal predictive biomarker for immunotherapy response among patients with non-small cell lung cancer. Costimulatory molecules play a role in anti-tumor immune response. Hence, they can be a potential biomarker for immunotherapy response. The current study comprehensively investigated the expression of costimulatory molecules in lung squamous carcinoma (LUSC) and identified diagnostic biomarkers for immunotherapy response. The costimulatory molecule gene expression profiles of 627 patients were obtained from the The Cancer Genome Atlas, GSE73403, and GSE37745 datasets. Patients were divided into different clusters using the k-means clustering method and were further classified into two discrepant tumor microenvironment (TIME) subclasses (hot and cold tumors) according to the immune score of the ESTIMATE algorithm. A high proportion of activated immune cells, including activated memory CD4 T cells, CD8 T cells, and M1 macrophages. Five CMGs (FAS, TNFRSF14, TNFRSF17, TNFRSF1B, and TNFSF13B) were considered as diagnostic markers using the Least Absolute Shrinkage and Selection Operator and the Support Vector Machine-Recursive Feature Elimination machine learning algorithms. Based on the five CMGs, a diagnostic nomogram for predicting individual tumor immune microenvironment subclasses in the TCGA dataset was developed, and its predictive performance was validated using GSE73403 and GSE37745 datasets. The predictive accuracy of the diagnostic nomogram was satisfactory in all three datasets. Therefore, it can be used to identify patients who may benefit more from immunotherapy.
ISSN:1664-8021
1664-8021
DOI:10.3389/fgene.2022.1078790