Characterization of Tumor Microenvironment and Prognosis of Regulatory T cells-Related Subtypes.

Abstract

Introduction: Regulatory T cells (Tregs) play an important role in the tumor microenvironment (TME). Currently, there have been no studies of Treg-related genes (TRGs) in lung adenocarcinoma (LUAD).

Methods: We integrated the Cancer Genome Atlas (TCGA) dataset with the Gene Expression Omnibus (GEO) dataset and divided the TCGA-GEO dataset patient samples into different cohorts by unsupervised clustering analysis based on the expression of TRGs in LUAD. By analyzing the TME characteristics of different cohorts, we assessed immune cell infiltration and function. In addition, we constructed Cox risk proportional regression models based on TRGs to predict patient prognosis.

Results: The results of unsupervised cluster analysis classified the TCGA-GEO dataset as "immune desert", "immune evasion" and "immune inflammation". Moreover, there was a significant survival differential among the three cohorts (p-value < 0.05). Based on the expression of 61 TRGs in LUAD, we screened TFRC, CTLA4, IL1R2, NPTN NPTN and METTL7A to construct a Cox risk proportional regression model to divide the TCGA-GEO dataset into a training cohort and a test cohort. Survival was significantly worse in the high-risk group than in the low-risk group in both the training and test cohorts (p-value < 0.05). Finally, the nomogram scoring system constructed by integrating the model risk scores with clinical parameters can well predict the 1, 3 and 5 year survival of patients.

Conclusion: In conclusion, based on our analysis of the TRGs of LUAD patients, we can classify the patient TME into different immune statuses, which provides insights into adopting appropriate treatment regimens for different patients.

.