Cancer-associated fibroblasts (CAFs), the central players in the tumor microenvironment (TME), can promote tumor progression and metastasis via various functions. However, the properties of CAFs in prostate cancer (PCa) have not been fully assessed. Therefore, we aimed to examine the CAF characteristics in PCa and construct a CAF-derived signature to predict PCa prognosis. CAFs were identified using single-cell RNA sequencing (scRNA-seq) data from 3 studies. We performed the FindAllMarkers function to extract CAF marker genes and constructed a signature to predict the biochemical relapse-free survival (bRFS) of PCa in the Cancer Genome Atlas (TCGA) cohort. Subsequently, different algorithms were applied to reveal the differences of the TME, immune infiltration, treatment responses in the high- and low-risk groups. Additionally, the CAF heterogeneity was assessed in PCa, which were confirmed by the functional enrichment analysis, gene set enrichment analysis (GSEA), and AUCell method. The scRNA-seq analysis identified a CAF cluster with 783 cells and determined 183 CAF marker genes. Cell-cell communication revealed extensive interactions between fibroblasts and immune cells. A CAF-related prognostic model, containing 7 genes (ASPN, AEBP1, ALDH1A1, BGN, COL1A1, PAGE4 and RASD1), was developed to predict bRFS and validated by 4 independent bulk RNA-seq cohorts. Moreover, the high-risk group of the signature score connected with an immunosuppressive TME, such as a higher level of M2 macrophages and lower levels of plasma cells and CD8⁺ T cells, and a reduced reaction rate for immunotherapy compared with low-risk group. After re-clustering CAFs via unsupervised clustering, we revealed 3 biologically distinct CAF subsets, namely myofibroblast-like CAFs (myCAFs), immune and inflammatory CAFs (iCAFs) and antigen-presenting CAFs (apCAFs). In conclusion, the CAF-derived signature, the first of its kind, can effectively predict PCa prognosis and serve as an indicator for immunotherapy. Furthermore, our study identified 3 CAF subpopulations with distinct functions in PCa.

Background: Early detection of prostate cancer (PCa) remains a clinical challenge. Plasma proteomics provides a non-invasive tool for identifying individuals at elevated risk prior to symptom onset or PSA elevation. Methods: We quantified 1463 plasma proteins in 23 825 PCa-free men from the UK Biobank (UKB). Participants were split into training and validation sets. Cox regression and Light Gradient Boosting Machine (LightGBM) with forward feature selection were used to identify and rank predictive proteins. Model performance was assessed by area under the receiver operating characteristic curve (AUC) in the validation set, and SHAP values were used to interpret feature contributions. Results: TSPAN1 and GP2 consistently ranked as top predictors across all analyses. In the training set, both proteins remained significantly associated with PCa risk after Bonferroni correction in multivariable Cox models. LightGBM with forward selection further prioritized TSPAN1 and GP2 as key contributors, and SHAP analysis confirmed their dominant importance. In the validation set, a model combining TSPAN1, GP2, and demographic variables achieved an AUC of 0.728 for overall PCa prediction and 0.760 for 5-year risk. Based on Youden Index–derived thresholds, high-expression groups of TSPAN1 and GP2 were associated with hazard ratios of 1.75 and 1.60, respectively. Longitudinal profiling showed that TSPAN1 levels began rising approximately 9 years before diagnosis, while GP2 increased from 6 years prior. Conclusions: TSPAN1 and GP2 are promising long-term predictive biomarkers for PCa. A streamlined proteomics-based model may enable individualized risk stratification and inform earlier, less invasive screening strategies.

Objective: Few studies have examined the relationship between systemic oxidative stress and prostate cancer (PCa) risk. This study aimed to explore potential correlations between PCa and oxidative balance score (OBS), which measures systemic oxidative stress. Design: A cross-sectional study. Setting: The National Health and Nutrition Examination Survey. Participants: A total of 8156 individuals were included in this study. Primary and secondary outcome measures: Weighted logistic regression with multivariable adjustment and restricted cubic splines (RCS) were used to assess the correlation between PCa and OBS. A sensitivity analysis was conducted specifically on patients with PCa to verify the results. Results: The prevalence of PCa was 2.55%. The multivariable logistic regression model revealed no correlation between OBS, dietary OBS, lifestyle OBS and PCa. Compared with the lowest quartile of OBS, the adjusted ORs for the highest quartile of OBS, dietary OBS and lifestyle OBS were 1.852 (95% CI 1.028–3.339), 1.565 (95% CI 0.841–2.913) and 1.575 (95% CI 0.915–2.710), respectively. Additionally, all p values for trend were greater than 0.05. Subgroup analysis revealed a consistent lack of association between OBS and PCa across various population settings. Furthermore, analysis using RCS confirmed this absence of association, indicating no significant relationship in either a linear or non-linear context. A sensitivity analysis focusing exclusively on patients with PCa showed a strong association (OR=2.737, p=0.008). Conclusion: This cross-sectional study reveals no significant association between systemic oxidative stress, measured by OBS, and PCa risk. Notably, a sensitivity analysis focusing solely on PCa patients suggested a potential link, warranting further investigation.