Multi-omics analyses construct an inflammatory response based prognostic gene signature for cervical cancer and suggest tumor infiltrating monocytes subgroups as key players

Abstract

Background: Inflammatory response in the tumor micro-environment contributes to the progression and treatment response of various types of cancers. However, for cervical cancer, a type of cancer initiated by the infection of HPV, the clinical relevance of the inflammatory response and the underlying mechanisms remain to be elucidated. Methods: The RNA-seq and clinical data of cervical cancer patients in several public datasets was used to construct and validate a prognostic gene signature based on the inflammatory response related genes. Risk stratification of patients was carried out according to this gene signature, and bioinformatic analyses were conducted to depict the immune landscape, identify the enriched biological pathways and predict patients’ treatment response. Single-cell and bulk RNA-seq data was jointly analysed to explore the underlying cellular and molecular mechanisms of the gene signature. The RNA-seq data of our own cohort and additional public datasets was used to further validate the findings made in this study. Results: A prognostic gene signature consisting of 16 inflammatory response related genes was constructed and successfully validated on multiple testing datasets. Patients in the low-risk group defined by this gene signature had significantly better survival (hazard ratio [HR]=0.48, 95% Confidence Interval [CI]: 0.275-0.85; Multivariate analysis on the CGCI testing dataset). The two risk groups had different immune landscapes, enriched biological pathways and predicted sensitivity to chemo-, radio- and immune-therapy. Two subgroups of tumor infiltrating monocytes with possibly opposite functions might be actively involved in the inflammatory response. SERPINE1 and ITGA5 expressed on endothelial cells might have synergic effects and regulate the infiltration of monocytes and macrophages. Findings were successfully validated with our own RNA-seq data and on additional public datasets. Conclusion: The inflammatory response in the tumor micro-environment of cervical cancer, possibly jointly regulated by multiple TIM subgroups, is associated with the prognosis and treatment response of cervical cancer patients and may be potential treatment targets.