A Venous Thromboembolism-Associated Model for Predicting Tumor Microenvironment Features in Breast Cancer.

BACKGROUND

Breast cancer (BC) demonstrates a high worldwide prevalence and remains among the primary contributors to cancer mortality in females. Venous thromboembolism (VTE), a common oncological complication, strongly correlates with poorer survival in BC patients. However, the predictive value of VTE-related genes (VTERGs) in BC prognosis and their impact on clinical outcomes remain underexplored, lacking systematic investigation.

METHODS

This study developed a novel VTERG-driven predictive framework for BC. Using consensus clustering based on VTERG signatures, we analyzed transcriptomic profiles of BC specimens, identified 2 molecular subtypes, and constructed a prognostic risk model from subtype-specific differentially expressed genes (DEGs). The model was validated in independent datasets. We further compared immune cell infiltration, immunotherapeutic response, somatic mutation patterns, and drug sensitivity between high- and low-risk groups.

RESULTS

VTERG-based clustering distinguished 2 BC subtypes with distinct biological features. By screening SERPINA1 and IGHA1 as characteristic genes, a prognostic risk model was developed to achieve robust and independent prediction of BC clinical outcomes. Further integration of clinical covariates confirmed that risk score functioned as an independent prognostic determinant. Following the categorization of patients into high- and low-risk groups, marked disparities were observed between the 2 groups across multiple aspects, including survival prognosis, immune cell infiltration characteristics, human leukocyte antigen expression profiles, immune checkpoint molecular characteristics, and drug sensitivity landscape.

CONCLUSIONS

The study developed a BC prognostic risk signature that may help predict patient outcomes and offer preliminary insights for personalized immunotherapy strategies.