Background: Bladder cancer (BLCA) is highly heterogeneous with distinct molecular subtypes. This research aimed to investigate the heterogeneity of different molecular subtypes from a tumor microenvironment perspective and develop a molecular-subtype-associated immune prognostic signature that can be recognized by MRI radiomics features. Methods: Individuals with BLCA in The Cancer Genome Atlas (TCGA) and IMvigor210 were classified into luminal and basal subtypes according to the UNC classification. The proportions of tumor-infiltrating immune cells (TIICs) were examined using The Cell Type Identification by Estimating Relative Subsets of RNA Transcripts algorithm. Immune-linked genes that were expressed differentially between luminal and basal subtypes and associated with prognosis were selected to develop the immune prognostic signature (IPS) and utilized for the classification of the selected individuals into low- and high-risk groups. Functional enrichment analysis (GSEA) was performed on the IPS. The data from RNA-sequencing and MRI images of 111 BLCA samples in our center were utilized to construct a least absolute shrinkage and selection operator (LASSO) model for the prediction of patients’ IPSs. Results: Half of the TIICs showed differential distributions between the luminal and basal subtypes. IPS was highly associated with molecular subtypes, critical immune checkpoint gene expression, prognoses, and immunotherapy response. The prognostic value of the IPS was further verified through several validation data sets (GSE32894, GSE31684, GSE13507, and GSE48277) and meta-analysis. GSEA revealed that some oncogenic pathways were co-enriched in the group at high risk. A novel performance of a LASSO model developed as per ten radiomics features was achieved in terms of IPS prediction in both the validation (area under the curve (AUC): 0.810) and the training (AUC: 0.839) sets. Conclusions: Dysregulation of TIICs contributed to the heterogeneity between the luminal and basal subtypes. The IPS can facilitate molecular subtyping, prognostic evaluation, and personalized immunotherapy. A LASSO model developed as per the MRI radiomics features can predict the IPSs of affected individuals.

Background: The field of urological tumor histopathology has long relied on subjective pathologist expertise, leading to diagnostic variability. Recent advances in digital pathology and artificial intelligence (AI) offer transformative potential by standardizing diagnoses, improving accuracy, and bridging healthcare disparities. This study conducted a 20-year bibliometric analysis to map global research trends and innovations in AI-driven urological pathology. Methods: For this bibliometric analysis, literature from 2004 to 2024 was retrieved from the Web of Science Core Collection. CiteSpace, VOSviewer, and Microsoft Excel were used to visualize coauthorship, cocitation, and co-occurrence analyses of countries/regions, institutions, authors, references, and keywords in the field of AI for urological tumor histopathology. Results: A total of 199 papers were included. Research on AI-driven urological tumor pathology has steadily increased since 2005, with a significant surge between 2020 and 2023. The United States made the largest contribution in terms of publications (131), citations (4725), and collaborations. The most productive institution was the University of Southern California, while Patel et al. and Epstein et al. were identified as the most active and most cocited authors, respectively. European Urology led in both publication volume and impact. Keyword analysis identified “machine learning,” “prostate cancer,” “deep learning,” and “diagnosis” as major research foci. Conclusions: The integration of AI into urological tumor pathology demonstrates transformative potential, significantly enhancing diagnostic accuracy and efficiency through automated analysis of whole-slide imaging and Gleason grading, comparable to pathologist-level performance. However, clinical translation encounters critical challenges, including data bias, model interpretability (“black-box” limitations), and regulatory-ethical complexities. Future advancements hinge on developing explainable AI frameworks, multimodal systems integrating histopathology, radiomics, and genomics and establishing global collaborative networks to address resource disparities. Prioritizing standardized data protocols, fairness-aware algorithms, and dynamic regulatory guidelines will be essential to ensure equitable, reliable, and clinically actionable AI solutions, ultimately advancing precision oncology in urological malignancies.

Background. Lymph node status is important for treatment decision making in prostate cancer (PCa). We aimed to develop a genomic-clinicopathologic nomogram for the prediction of lymph node invasion (LNI) in PCa. Methods. Differentially expressed genes between LNI and non-LNI PCa samples were identified in the Cancer Genome Atlas database. Univariate Cox regression analysis and minimum redundancy maximum relevance were performed for gene selection. The synthetic minority oversampling technique (SMOTE) was conducted to balance the minority group (LNI group). Machine learning models were constructed in the training set and assessed in the validation set. Univariable logistic regression and multivariable logistic regression were applied to build a nomogram. Furthermore, the RNA-sequence data from our center were used to validate the expression levels of hub genes between five matched primary PCa and the corresponding LNI samples. Results. The 37-gene-based support vector machine (SVM) model had the optimal synthesized performance in the SMOTE-balanced training (area under the curve (AUC): 0.947) and validation (AUC: 0.901) sets. Incorporating the SVM-based risk score and the Gleason grade, the genomic-clinicopathologic nomogram demonstrated good prediction and calibration both in the SMOTE-balanced training (AUC: 0.946) and validation (AUC: 0.910) sets. The dysregulated expression of hub genes between PCa and LNI samples was also validated. Conclusion. The proposed nomogram combining the 37-gene-based SVM model with the Gleason grade had the potential to preoperatively predict LNI in PCa. Some of the hub genes should be prioritized for functional studies and mechanistic analyses.

Background: Quantifying tumor–stroma architecture on routine hematoxylin and eosin slides may refine risk stratification in bladder cancer (BCa). We developed a convolutional neural network to segment whole-slide images, compute the mixed tumor–stroma ratio (MTSR), evaluate its prognostic value across multicenter cohorts, explore underlying molecular programs through multi-omics analysis, and construct a preoperative multiparametric MRI (mpMRI) radiomics model to estimate MTSR noninvasively. Methods: The ResNet50 convolutional network was customized using The Cancer Genome Atlas BCa slides labeled into 9 histological classes and background, followed by internal validation and multicenter external testing. Whole-slide-image-level segmentation yielded quantitative tissue ratios. The prognostic value was evaluated using Cox regression, Kaplan–Meier analysis, and meta-analysis, with a nomogram constructed by incorporating independent predictors. Prognostic significance was assessed by Cox regression, Kaplan–Meier analysis, and meta-analysis, and a nomogram was developed by integrating independent predictors. Bulk RNA sequencing underwent gene set variation analysis/gene set enrichment analysis, immune deconvolution, and ESTIMATE analyses, while single-cell RNA sequencing of high- vs. low-MTSR tumors profiled cellular heterogeneity, pseudotime trajectories, and regulon activity using SCENIC. An mpMRI-based random forest radiomics model was trained to predict high vs. low MTSR. Results: The convolutional neural network achieved >90% classification accuracy with Cohen’s kappa >0.95 in all cohorts. A nomogram combining MTSR and N stage outperformed clinicopathological predictors. Molecular analyses revealed that high-MTSR tumors displayed increased macrophage infiltration and enrichment of pathways related to extracellular matrix remodeling, cell adhesion, and transforming growth factor-β/WNT signaling. Single-cell analysis identified an integrin subunit beta 8 (ITGB8)-high urothelial subtype (cluster 8) with terminal differentiation, enhanced WNT activity, and sender-dominant communication networks. The mpMRI radiomics model achieved accuracies of 0.701 and 0.710 for predicting MTSR status in the training and validation sets, respectively. Conclusions: The deep learning-generated MTSR showed consistent reproducibility and prognostic independence across cohorts, mechanistically connected with an ITGB8-enriched stromal–oncogenic pathway. Its estimation via mpMRI radiomics enables integrative, noninvasive risk stratification for precision management of BCa.

Background: The integration of multi-omics approaches provides a powerful strategy for understanding cancer. By combining these methods, researchers gain insights into tumor diversity, gene activity, and the tumor microenvironment, which are essential for advancing cancer biology, improving early detection, refining prognostic tools, and developing targeted treatments. This study aims to explore key biomarkers in muscle-invasive bladder cancer (MIBC) and to develop a predictive model for better understanding disease progression and therapeutic responses Methods: Single-cell analysis of MIBC samples from public datasets identified basal-related genes. Using Cox analysis of The Cancer Genome Atlas (TCGA)-bladder urothelial carcinoma (BLCA) clinical data and the expression matrix, and combining it with weighted gene co-expression network analysis (WGCNA) of another MIBC dataset, a disintegrin and metalloprotease protein 17 (ADAM17) was identified as a key target gene. We collected BLCA patient samples from TCGA, Shanghai Tenth People’s Hospital (STPH), and Guangdong Second Provincial General Hospital (GD2H) to develop the pathomics model for predicting ADAM17 expression. Single-cell validation of ADAM17 expression was performed using GD2H MIBC samples. Results: ADAM17 was highly expressed in MIBC and associated with poor prognosis. Its expression correlated with clinical features such as non-papillary subtype, advanced pathological stage, and higher tissue grade. ADAM17 overexpression was linked to immune reprogramming and drug resistance. The pathomics model effectively predicted ADAM17 expression in BLCA samples. Single-cell analysis of GD2H MIBC samples confirmed ADAM17 overexpression in epithelial cells and identified key pathways in cell communication, matrix remodeling, tumor invasion, and immune regulation. Conclusions: Multi-omics approaches effectively identify biomarkers for MIBC, with ADAM17 emerging as a key biomarker. Further research is needed to clarify its role in MIBC progression.