A ferroptosis-related gene signature and immune infiltration for predicting the lymph node metastasis of patients with lung adenocarcinoma

Abstract

Abstract Background The lymph node metastasis of LUAD is a pivotal factor leading to late TNM staging and poor prognosis. Ferroptosis plays a key role in promoting cancer cell death and immunotherapy. However, the roles of FRGs in lymph node metastasis and immunity of LUAD remain unclear. Methods LUAD patients obtained from TCGA database were divided into lymph node metastasis group and non-lymph node metastasis group, and differential analysis was performed to screen lymph node metastasis-related FRGs. Univariate and multivariate Cox regression analyses were performed to construct a prediction model of FGRs. Kaplan-Meier survival curves and ROC curves were performed to verify the validity of model. The CIBERSOFT method was used to study the degree and prognostic value of immune cells in different groups. Results The gene expression profiles of 301 LUAD samples without lymph node metastasis and 153 LUAD samples with lymph node metastasis obtained from the TCGA database were analyzed, 90 FRGs were obtained. Univariate analysis showed that 15 FRGs were significantly associated with OS in LUAD. Subsequently, we used multivariate Cox regression analysis to build a 9-FRGs model associated with LUAD survival, including CISD1, DDIT4, DECR1, IL33, PEBP1, PHKG2, PPP1R13L, SLC7A5 and VDAC2. The samples were divided into low-risk and high-risk subgroups. Kaplan-Meier survival curves showed better OS in the low-risk group. The ROC curve showed that this signature performed well in predicting OS. Finally, we systematically analyzed differences in immune infiltration profiles between high-risk and low-risk samples. We found that resting mast cells and resting memory CD4 T cells showed higher infiltration in low-risk group than in high-risk group, but M0 macrophages, activated mast cells and follicular helper T cells tended to infiltrate in high-risk group, and there were certain associations between above 5 TIICs with the risk scores and above 9 FGRs, and the high infiltration of activated mast cells was an adverse prognostic factor of LUAD. Conclusion We constructed a novel 9-FRGs model that could serve as a potential therapeutic target for lymph node metastasis in LUAD. Targeting FRGs seems to be an alternative to clinical therapy for lymph node metastasis of LUAD.