A Prognostic Risk Model for Breast Cancer Integrating Migrasome and Tumor Microenvironment Features to Predict Immunological Characteristics.
Breast cancer (BRCA) is a heterogeneous disease with a complex etiology. The prognostic value of genes related to migrasomes and the tumor microenvironment (MTMERGs) in BRCA is unclear.
A prognostic risk model was constructed using six core signature genes identified from MTMERGs via differential expression analysis and Cox regression. Its reliability was validated in an independent cohort using Kaplan-Meier and time-dependent ROC curves. A nomogram was developed and assessed via Decision Curve Analysis (DCA). Biological functions and immune infiltration were evaluated with GSEA, CIBERSORT, and ssGSEA. Immunotherapy sensitivity was predicted using TIDE/IPS scores and the IMvigor210 cohort. Tumor Mutation Burden (TMB) analysis and the pRRophetic algorithm were used for further clinical correlation and drug sensitivity prediction.
The six-MTMERG model effectively stratified patients into high- and low-risk groups with distinct survival outcomes. The high-risk group was associated with a predicted immunosuppressive microenvironment (estimated enrichment of M0/M2 macrophages), higher TMB, and poorer prognosis. In contrast, the low-risk group was estimated to possess an immunologically active profile and showed a better predicted response to immune checkpoint inhibitors. Predicted differential sensitivities to conventional chemotherapy were also computationally evaluated between the subgroups.
We developed a computationally derived and externally validated prognostic model for BRCA based on migrasome and tumor microenvironment features. It successfully stratifies patients into groups with divergent clinical outcomes, immune profiles, and therapeutic responses, providing insights into BRCA heterogeneity and prognosis. Further prospective and experimental validation is warranted before clinical application.
A prognostic risk model was constructed using six core signature genes identified from MTMERGs via differential expression analysis and Cox regression. Its reliability was validated in an independent cohort using Kaplan-Meier and time-dependent ROC curves. A nomogram was developed and assessed via Decision Curve Analysis (DCA). Biological functions and immune infiltration were evaluated with GSEA, CIBERSORT, and ssGSEA. Immunotherapy sensitivity was predicted using TIDE/IPS scores and the IMvigor210 cohort. Tumor Mutation Burden (TMB) analysis and the pRRophetic algorithm were used for further clinical correlation and drug sensitivity prediction.
The six-MTMERG model effectively stratified patients into high- and low-risk groups with distinct survival outcomes. The high-risk group was associated with a predicted immunosuppressive microenvironment (estimated enrichment of M0/M2 macrophages), higher TMB, and poorer prognosis. In contrast, the low-risk group was estimated to possess an immunologically active profile and showed a better predicted response to immune checkpoint inhibitors. Predicted differential sensitivities to conventional chemotherapy were also computationally evaluated between the subgroups.
We developed a computationally derived and externally validated prognostic model for BRCA based on migrasome and tumor microenvironment features. It successfully stratifies patients into groups with divergent clinical outcomes, immune profiles, and therapeutic responses, providing insights into BRCA heterogeneity and prognosis. Further prospective and experimental validation is warranted before clinical application.