Decoding the Homeobox (HOX) Gene Expression Profile: Implications for Diagnosis, Prognosis and Survival in Prostate Cancer

Abstract

Objectives: This study provides a comprehensive analysis of the expression profiles and clinical significance of homeobox (HOX) genes in prostate cancer (PCa).

Methods: We utilized large-scale datasets including TCGA-PRAD, DKFZ-RNAseq, and metastatic castration-resistant prostate cancer (mCRPC) cohorts. We screened 228 HOX-related genes for gene expression related to disease progression, diagnosis, and prognosis. Gene expression was compared in patient subgroups stratified by clinical parameters. Survival prognosis was analyzed using the Kaplan-Meier survival curve approach. The AUC value of the altered gene was determined using the Receiver Operating Characteristic (ROC) curve analysis. The effect of castration on HOX-related gene expression was analyzed using the LuCaP35 xenograft dataset.

Results: We identified 42 up-regulated and 56 down-regulated genes with significant differential expression between benign and malignant tissues. Further investigation into 22 key genes revealed their profound impact on diagnosis and prognosis. HOXC6 and NKX2-3 emerged as highly effective diagnostic biomarkers, demonstrating area under the curve (AUC) values of 0.917 and 0.936, respectively. Prognostically, NKX6-1 expression showed the strongest predictive potential for 5-year disease-specific survival (AUC = 0.881), while HOXB7 also served as a critical survival indicator (AUC = 0.909). Conversely, the down-regulation of EVX2 was uniquely associated with all clinicopathological and survival parameters, suggesting a significant tumor-suppressive role. In early-onset PCa, HOXC5, HOXC6, and MEIS2 correlated strongly with tumor mutation burden and biochemical recurrence. In the context of advanced disease, NKX2-3 was associated with Androgen Receptor (AR) activation scores, whereas MEIS2 showed a strong negative correlation. Notably, this study identified HOXB8/HOXD13 genes, like LHX2, as novel and consistent markers, significantly increased in treatment-induced neuroendocrine prostate cancer (t-NEPC).

Conclusion: These findings highlight a distinct subset of HOX genes that govern prostate cancer progression and provide a framework for developing novel diagnostic and prognostic tools tailored to disease stage and subtype.