Page 96 - Read Online

Page 2 of 26 Li et al. Cancer Drug Resist. 2025;8:31

Results: Two immune subtypes were identified: high-risk subgroups displayed TP53 mutations, increased tumor
mutation burden (TMB), and enriched energy metabolism pathways. ScRNA-seq delineated three PCa cell clusters,
with high-risk subtypes being sensitive to bendamustine/dacomitinib and resistant to apalutamide/neratinib. A
10-gene prognostic model (e.g., MUC5B, TREM1) categorized patients into high/low-risk groups with distinct
survival outcomes (log-rank P <​ 0.0001). Validation in external datasets confirmed the robust predictive accuracy
(AUC: 0.854-0.889). Experimental assays verified subtype-specific drug responses and dysregulation of key model
genes.

Discussion: This study establishes a TME-driven prognostic framework that connects immune heterogeneity,
genomic instability, and therapeutic resistance in PCa. By pinpointing metabolic dependencies and subtype-specific
vulnerabilities, our findings provide actionable strategies to circumvent treatment failure, such as targeting energy
metabolism or tailoring therapies based on resistance signatures.

INTRODUCTION
Prostate cancer (PCa) is a common solid malignancy among men worldwide . Despite considerable
[1]
advancements in the diagnosis and treatment, the incidence and mortality rates of metastatic PCa remain
high, and accurate prognostication remains a challenge.

The tumor microenvironment (TME) includes tumor stromal cells such as immune cells, fibroblasts, and
endothelial cells, in addition to numerous signaling molecules such as cytokines and chemokines . In PCa,
[2]
the TME is exceptionally complex, and the interactions between these different cellular components
profoundly influence disease progression and treatment outcomes . In recent years, the introduction of
[3]
high-throughput -omics technologies and bioinformatics methods has advanced our understanding of the
interplay between PCa and the immune system . These tools enable a comprehensive characterization of the
[4]
immune landscape within tumors and the identification of immune-related biomarkers with prognostic
significance .
[5,6]
Elucidating the immune environment in PCa is not only crucial for accurate prognosis prediction but also
for the development of therapeutic interventions targeting the immune system. Immune system-based
therapies have shown great promise in treating PCa, given their successful application in other
malignancies . Nonetheless, owing to the heterogeneity of PCa and the intricate nature of the immune
[7]
microenvironment, the clinical application of immunotherapy in PCa remains challenging, creating a need
to establish robust immune environment-related prognostic models that can identify patients most likely to
benefit from immunotherapeutic approaches. Such a model may enhance prognostic accuracy and provide
new avenues for personalized treatment interventions.

In this study, we aimed to use bioinformatics techniques to examine the immune landscape of PCa,
leveraging The Cancer Genome Atlas (TCGA) dataset to identify TME subtypes. We aimed to identify the
pivotal immune-related genes and pathways linked to prognosis within the high-risk subtype and to
construct a novel immune-related prognostic prediction model. By integrating multiomics data and clinical
information, we aimed to document the TME heterogeneity and its clinically relevant characteristics. In
addition, we aimed to use this model to identify immune-related treatment targets. This study may advance
our understanding of the TME in PCa and may support clinical practice and treatment selection.

89