Purpose: Our study aims to investigate the underlying molecular mechanism between Lymphatic vascular space invasion (LVSI) and parametrial invasion (PMI) patients, and we screen biomarkers for patients with LVSI+ and PMI+.Methods: The main molecular mechanism of the LVSI+ and PMI+ groups was observed by using differential expression analysis and GO enrichment. Based on the results of Go enrichment, the distribution of immune infiltration was compared between the LVSI+ group and the PMI+ group by using ssGSEA analysis. Then we identified immunological differentially expressed genes (IDGs) by taking the intersection of DEGs and immune-related genes. The prognostic IDGs were screened by univariate Cox regression analysis. The Cox model was constructed by multivariate Cox regression. The prognostic ability of the two subgroups’ models was evaluated by receiver operating characteristic (ROC) curves and the area under the curve (AUC) values. Based on the genes chosen for the LVSI and PMI models, the drug sensitivity was determined on the ImmPort website.Results: The immune-related pathway differentiate LVSI from PMI in cervical cancer. The ssGSEA result showed that adaptive immunity was suppressed in LVSI+ patients, whereas in PMI+ patients, innate immunity was suppressed. The Cox model was constructed using interaction genes EREG and IL-9R for LVSI+ patients, and NODAL and IL-12A for PMI+ patients, respectively. The LVSI model and the PMI model all had better prediction power in the TCGA and GEO cohorts. we found difference in drug sensitivity between the LVSI and the PMI group.Conclusion: We proposed that the distribution of immune infiltration was the fundamental distinction in the molecular mechanism between LVSI and PMI. This study identified four metastasis mode-specific genes related to the immune infiltration, these genes strongly influenced the prognosis of LVSI+ and PMI+ cervical cancer patients, respectively.

No abstract