Hongxi Chen1, Jinliang Xie2, Peng Jin2. 1. Department of Gastrointestinal Surgery, Hunan Provincial People's Hospital (The First Affiliated Hospital of Hunan Normal University) Changsha 410005, Hunan Province, China. 2. Organ Transplant Center, Xiangya Hospital, Central South University Changsha 410008, Hunan Province, China.
Abstract
BACKGROUND: The present study aimed to explore and validate a prognostic immune signature, to formulate a prognosis for ccRCC patients combined with immune-infiltration analysis. METHODS: Public datasets were used as our source of multi-omics data. Differential analysis was performed via the edgeR package. A prognostic immune signature was identified by univariate Cox analysis, and we constructed an integrative tumor-associated immune genes (TAIG) model from the multivariate Cox results. In order to interrogate and identify the related crosstalk, functional analysis was deployed. Significantly, we implemented the CIBERSORT algorithm to estimate the immune cell fractions in the ccRCC samples, and analyzed the differential abundance of tumor-infiltrating immune cells in two TAIG groups, using a Wilcoxon rank-sum test. The prognostic role of differential immune cells was further assessed via a Kaplan-Meier analysis. In addition, we investigated the associations of a single immune signature with specific immune cells. RESULTS: A total of 628 ccRCC patients were comprised in our integrative analysis, including 537 ccRCC patients in the discovery group and 91 patients in the validation group. Fourteen key immune signatures were subsequently identified. A figure of 0.802 was registered for AUC, and worse prognosis was evinced for those patients with a higher TAIG. Correlation analysis indicated that TAIG correlated closely with both clinical variables and TMB. Moreover, functional analysis implicated the immune-related GO items or crosstalk. Hence, we were able to identify the relationships obtaining between tumor-infiltrating immune cells and TAIG. The differential abundance of immune cells showed a significant prognostic difference and consisted of memory-activated CD4+ T cells, T follicular helper cells, T regulatory cells, and so on. Moreover, we also characterized the associations between identified signatures and specific immune cells. Finally, the five-year AUC in the ICGC cohort was 0.72, suggesting the robustness of the TAIG that we constructed. CONCLUSIONS: Overall, our team characterized the tumor-associated immune signature in ccRCC, and further identified the prognostic tumor-infiltrating immune cells related to TAIG. This in turn provided a solid foundation for investigating individualized immunotherapy, as well as other relevant mechanisms. AJTR
BACKGROUND: The present study aimed to explore and validate a prognostic immune signature, to formulate a prognosis for ccRCC patients combined with immune-infiltration analysis. METHODS: Public datasets were used as our source of multi-omics data. Differential analysis was performed via the edgeR package. A prognostic immune signature was identified by univariate Cox analysis, and we constructed an integrative tumor-associated immune genes (TAIG) model from the multivariate Cox results. In order to interrogate and identify the related crosstalk, functional analysis was deployed. Significantly, we implemented the CIBERSORT algorithm to estimate the immune cell fractions in the ccRCC samples, and analyzed the differential abundance of tumor-infiltrating immune cells in two TAIG groups, using a Wilcoxon rank-sum test. The prognostic role of differential immune cells was further assessed via a Kaplan-Meier analysis. In addition, we investigated the associations of a single immune signature with specific immune cells. RESULTS: A total of 628 ccRCC patients were comprised in our integrative analysis, including 537 ccRCC patients in the discovery group and 91 patients in the validation group. Fourteen key immune signatures were subsequently identified. A figure of 0.802 was registered for AUC, and worse prognosis was evinced for those patients with a higher TAIG. Correlation analysis indicated that TAIG correlated closely with both clinical variables and TMB. Moreover, functional analysis implicated the immune-related GO items or crosstalk. Hence, we were able to identify the relationships obtaining between tumor-infiltrating immune cells and TAIG. The differential abundance of immune cells showed a significant prognostic difference and consisted of memory-activated CD4+ T cells, T follicular helper cells, T regulatory cells, and so on. Moreover, we also characterized the associations between identified signatures and specific immune cells. Finally, the five-year AUC in the ICGC cohort was 0.72, suggesting the robustness of the TAIG that we constructed. CONCLUSIONS: Overall, our team characterized the tumor-associated immune signature in ccRCC, and further identified the prognostic tumor-infiltrating immune cells related to TAIG. This in turn provided a solid foundation for investigating individualized immunotherapy, as well as other relevant mechanisms. AJTR