Huang X, Feng Y, Li Y, Ding H, Huang X, Chen C, Yu Z, Zhang J, Xu X, Ma D, Yu S, Chen C. A novel transcriptomic signature associated with lymphovascular invasion predicts clinical outcomes, tumor microenvironment, and therapeutic response in lung adenocarcinoma.
Int Immunopharmacol 2024;
127:111286. [PMID:
38064818 DOI:
10.1016/j.intimp.2023.111286]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/03/2023] [Revised: 10/31/2023] [Accepted: 11/22/2023] [Indexed: 01/18/2024]
Abstract
PURPOSE
Since TNM staging has limitations for predicting post-operative outcomes and relapse, more effective prediction tools need to be researched and developed. Lymphovascular invasion, LVI, as a histopathological feature, has been widely shown to have a correlation with poor prognosis and early recurrence of lung adenocarcinoma (LUAD). However, LVI assessment is limited by subjective bias, and therefore its efficacy in practical clinical application needs further clarification. The aim of this study was to formulate a new signature based on LVI-related genes to predict prognosis and recurrence in patients with lung adenocarcinoma.
METHODS
Clinicopathological information, gene sequencing data and whole slide images (WSIs) of LUAD patients were downloaded from the Cancer Genome Atlas (TCGA) databases. LVI statue were evaluated by professional pathologists, and then the differentially expressed genes (LVI DEGs) associated with LVI were screened. The least absolute shrinkage and selection operator (LASSO) and Step Cox regression models were used to construct LVI-associated risk scores (LVRS), including PAQR4, ARGHEF6, CKS1B, CFTR and SEC14L4. The validity of the LVRS score was evaluated on multiple external datasets and our JSSZL cohort dataset. Using LVRS scores and clinical information, nomogram were constructed for use by clinicians. In addition, we further explored the relationship between LVRS score and clinicopathological features, immune infiltration, tumor mutational load, and immunotherapy response, and confirmed the expression of key genes in LVRS score in lung adenocarcinoma tissues using qRT-PCR and immunohistochemistry (IHC) techniques.
RESULTS
Based on the LVRS, patients could be classified into high-LVRS and low-LVRS groups. It was found that OS and PFS were significantly worse in the high-LVRS group than in the low-LVRS group (p < 0.001). By ROC curve analysis, it could be found that the nomogram combining LVRS and clinical information could accurately predict the prognosis of LUAD patients with the area under the curve of 1,3,5-year survival rate could reach 0.754, 0.741 and 0.735. The results of univariate and multivariate analysis showed that LVRS was an independent prognostic factor. At the same time, there were significant differences in the mutation profiles and immune microenvironment between the high-LVRS and low-LVRS groups, with the high-LVRS group having a significantly higher mutation rate than the low-LVRS group and exhibiting immunological "cold" features. By the experimental results, higher expression levels of PAQR4 and CKS1B were found in LUAD tissues, while lower expression levels of ARGHEF6, CFTR and SEC14L4 were observed.
CONCLUSIONS
The LVRS established in this study serves as a valid tool to predict the prognosis and recurrence status of lung adenocarcinoma patients and has a predictive effect on the response to postoperative treatment. The establishment of LVRS may offer some theoretical support to clinical treatment strategies for patients with lung adenocarcinoma following surgical intervention.
Collapse