Artificial intelligence for the diagnosis of clinically significant prostate cancer based on multimodal data: a multicenter study.

BACKGROUND: The introduction of multiparameter MRI and novel biomarkers has greatly improved the prediction of clinically significant prostate cancer (csPCa). However, decision-making regarding prostate biopsy and prebiopsy examinations is still difficult. We aimed to establish a quick and economic tool to improve the detection of csPCa based on routinely performed clinical examinations through an automated machine learning platform (AutoML). METHODS: This study included a multicenter retrospective cohort and two prospective cohorts with 4747 cases from 9 hospitals across China. The multimodal data, including demographics, clinical characteristics, laboratory tests, and ultrasound reports, of consecutive participants were retrieved using extract-transform-load tools. AutoML was applied to explore potential data processing patterns and the most suitable algorithm to build the Prostate Cancer Artificial Intelligence Diagnostic System (PCAIDS). The diagnostic performance was determined by the receiver operating characteristic curve (ROC) for discriminating csPCa from insignificant prostate cancer (PCa) and benign disease. The clinical utility was evaluated by decision curve analysis (DCA) and waterfall plots. RESULTS: The random forest algorithm was applied in the feature selection, and the AutoML algorithm was applied for model establishment. The area under the curve (AUC) value in identifying csPCa was 0.853 in the training cohort, 0.820 in the validation cohort, 0.807 in the Changhai prospective cohort, and 0.850 in the Zhongda prospective cohort. DCA showed that the PCAIDS was superior to PSA or fPSA/tPSA for diagnosing csPCa with a higher net benefit for all threshold probabilities in all cohorts. Setting a fixed sensitivity of 95%, a total of 32.2%, 17.6%, and 26.3% of unnecessary biopsies could be avoided with less than 5% of csPCa missed in the validation cohort, Changhai and Zhongda prospective cohorts, respectively. CONCLUSIONS: The PCAIDS was an effective tool to inform decision-making regarding the need for prostate biopsy and prebiopsy examinations such as mpMRI. Further prospective and international studies are warranted to validate the findings of this study. TRIAL REGISTRATION: Chinese Clinical Trial Registry ChiCTR2100048428. Registered on 06 July 2021.

Composition and changes in breast cancer patients' diagnosis and treatment expenses under the influence of medical insurance policy reform-A study on 3 950 patients in Guangxi Medical University Cancer Hospital. / åŒ»ä¿æ”¿ç­–æ”¹é©å½±å“ä¸‹ä¹³è ºç™Œæ‚£è€ è¯Šç–—è´¹ç”¨çš„æž„æˆåŠå˜åŒ–­­åŸºäºŽå¹¿è¥¿åŒ»ç§‘大学附属肿瘤医院3 950ä¾‹æ‚£è€ çš„ç ”ç©¶.

OBJECTIVES: To understand the influence of medical insurance policy reforms in Guangxi on the hospitalization expenses of breast cancer patients by analyzing the composition and changing trend in breast cancer diagnosis and treatment expenses in the Guangxi Medical University Cancer Hospital, and to provide the evidence for the improvement of medical insurance policy reform. METHODS: A total of 3 950 breast cancer patients were collected from 2014 to 2017 and analyzed. Kruskal-Wallis test and multiple linear regression model were used to discuss the breast cancer related epidemiology and analyze the composition of hospitalization expenses and its influential factors. RESULTS: The median hospitalization cost of breast cancer patients in our hospital from 2014 to 2017 was 29 266.94 Chinese Yuan. Single factor analysis showed that the impact of year, hospitalization days, age, payment method, tumor stage, and treatment method on hospitalization cost was significant (all P<0.01). Multiple linear regression analysis showed that the power of influential factors of hospitalization costs arranged descending from 2014 to 2017 as follows: hospitalization days, treatment methods, payment method, tumor staging, and age. CONCLUSIONS: Reasonably controlling hospitalization days and actively promoting the integration of urban and rural medical insurance can effectively reduce the economic burden for breast cancer patients.

Developing an advanced diagnostic model for hepatocellular carcinoma through multi-omics integration leveraging diverse cell-death patterns.

Introduction: Hepatocellular carcinoma (HCC), representing more than 80% of primary liver cancer cases, lacks satisfactory etiology and diagnostic methods. This study aimed to elucidate the role of programmed cell death-associated genes (CDRGs) in HCC by constructing a diagnostic model using single-cell RNA sequencing (scRNA-seq) and RNA sequencing (RNA-seq) data. Methods: Six categories of CDRGs, including apoptosis, necroptosis, autophagy, pyroptosis, ferroptosis, and cuproptosis, were collected. RNA-seq data from blood-derived exosomes were sourced from the exoRBase database, RNA-seq data from cancer tissues from the TCGA database, and scRNA-seq data from the GEO database. Subsequently, we intersected the differentially expressed genes (DEGs) of the HCC cohort from exoRBase and TCGA databases with CDRGs, as well as DEGs obtained from single-cell datasets. Candidate biomarker genes were then screened using clinical indicators and a machine learning approach, resulting in the construction of a seven-gene diagnostic model for HCC. Additionally, scRNA-seq and spatial transcriptome sequencing (stRNA-seq) data of HCC from the Mendeley data portal were used to investigate the underlying mechanisms of these seven key genes and their association with immune checkpoint blockade (ICB) therapy. Finally, we validated the expression of key molecules in tissues and blood-derived exosomes through quantitative Polymerase Chain Reaction (qPCR) and immunohistochemistry experiments. Results: Collectively, we obtained a total of 50 samples and 104,288 single cells. Following the meticulous screening, we established a seven-gene diagnostic model for HCC, demonstrating high diagnostic efficacy in both the exoRBase HCC cohort (training set: AUC = 1; testing set: AUC = 0.847) and TCGA HCC cohort (training set: AUC = 1; testing set: AUC = 0.976). Subsequent analysis revealed that HCC cluster 3 exhibited a higher stemness index and could serve as the starting point for the differentiation trajectory of HCC cells, also displaying more abundant interactions with other cell types in the microenvironment. Notably, key genes TRIB3 and NQO1 displayed elevated expression levels in HCC cells. Experimental validation further confirmed their elevated expression in both tumor tissues and blood-derived exosomes of cancer patients. Additionally, stRNA analysis not only substantiated these findings but also suggested that patients with high TRIB3 and NQO1 expression might respond more favorably to ICB therapy. Conclusions: The seven-gene diagnostic model demonstrated remarkable accuracy in HCC screening, with TRIB3 emerging as a promising diagnostic tool and therapeutic target for HCC.

Unfolding the mysteries of heterogeneity from a high-resolution perspective: integration analysis of single-cell multi-omics and spatial omics revealed functionally heterogeneous cancer cells in ccRCC.

The genomic landscape of clear cell renal cell carcinoma (ccRCC) has a considerable intra-tumor heterogeneity, which is a significant obstacle in the field of precision oncology and plays a pivotal role in metastasis, recurrence, and therapeutic resistance of cancer. The mechanisms of intra-tumor heterogeneity in ccRCC have yet to be fully established. We integrated single-cell RNA sequencing (scRNA-seq) and transposase-accessible chromatin sequencing (scATAC-seq) data from a single-cell multi-omics perspective. Based on consensus non-negative matrix factorization (cNMF) algorithm, functionally heterogeneous cancer cells were classified into metabolism, inflammatory, and EMT meta programs, with spatial transcriptomics sequencing (stRNA-seq) providing spatial information of such disparate meta programs of cancer cells. The bulk RNA sequencing (RNA-seq) data revealed high clinical prognostic values of functionally heterogeneous cancer cells of three meta programs, with transcription factor regulatory network and motif activities revealing the key transcription factors that regulate functionally heterogeneous ccRCC cells. The interactions between varying meta programs and other cell subpopulations in the microenvironment were investigated. Finally, we assessed the sensitivity of cancer cells of disparate meta programs to different anti-cancer agents. Our findings inform on the intra-tumor heterogeneity of ccRCC and its regulatory networks and offers new perspectives to facilitate the designs of rational therapeutic strategies.

Deciphering the molecular classification of pediatric sepsis: integrating WGCNA and machine learning-based classification with immune signatures for the development of an advanced diagnostic model.

Introduction: Pediatric sepsis (PS) is a life-threatening infection associated with high mortality rates, necessitating a deeper understanding of its underlying pathological mechanisms. Recently discovered programmed cell death induced by copper has been implicated in various medical conditions, but its potential involvement in PS remains largely unexplored. Methods: We first analyzed the expression patterns of cuproptosis-related genes (CRGs) and assessed the immune landscape of PS using the GSE66099 dataset. Subsequently, PS samples were isolated from the same dataset, and consensus clustering was performed based on differentially expressed CRGs. We applied weighted gene co-expression network analysis to identify hub genes associated with PS and cuproptosis. Results: We observed aberrant expression of 27 CRGs and a specific immune landscape in PS samples. Our findings revealed that patients in the GSE66099 dataset could be categorized into two cuproptosis clusters, each characterized by unique immune landscapes and varying functional classifications or enriched pathways. Among the machine learning approaches, Extreme Gradient Boosting demonstrated optimal performance as a diagnostic model for PS. Discussion: Our study provides valuable insights into the molecular mechanisms underlying PS, highlighting the involvement of cuproptosis-related genes and immune cell infiltration.

Investigating cellular similarities and differences between upper tract urothelial carcinoma and bladder urothelial carcinoma using single-cell sequencing.

Background: Upper tract urothelial carcinoma (UTUC) and bladder urothelial carcinoma (BLCA) both originate from uroepithelial tissue, sharing remarkably similar clinical manifestations and therapeutic modalities. However, emerging evidence suggests that identical treatment regimens may lead to less favorable outcomes in UTUC compared to BLCA. Therefore, it is imperative to explore molecular processes of UTUC and identify biological differences between UTUC and BLCA. Methods: In this study, we performed a comprehensive analysis using single-cell RNA sequencing (scRNA-seq) on three UTUC cases and four normal ureteral tissues. These data were combined with publicly available datasets from previous BLCA studies and RNA sequencing (RNA-seq) data for both cancer types. This pooled analysis allowed us to delineate the transcriptional differences among distinct cell subsets within the microenvironment, thus identifying critical factors contributing to UTUC progression and phenotypic differences between UTUC and BLCA. Results: scRNA-seq analysis revealed seemingly similar but transcriptionally distinct cellular identities within the UTUC and BLCA ecosystems. Notably, we observed striking differences in acquired immunological landscapes and varied cellular functional phenotypes between these two cancers. In addition, we uncovered the immunomodulatory functions of vein endothelial cells (ECs) in UTUC, and intercellular network analysis demonstrated that fibroblasts play important roles in the microenvironment. Further intersection analysis showed that MARCKS promote UTUC progression, and immunohistochemistry (IHC) staining revealed that the diverse expression patterns of MARCKS in UTUC, BLCA and normal ureter tissues. Conclusion: This study expands our multidimensional understanding of the similarities and distinctions between UTUC and BLCA. Our findings lay the foundation for further investigations to develop diagnostic and therapeutic targets for UTUC.

Disulfidptosis-related genes serve as potential prognostic biomarkers and indicate tumor microenvironment characteristics and immunotherapy response in prostate cancer.

Disulfidptosis, a newly identified programmed cell death pathway in prostate cancer (PCa), is closely associated with intracellular disulfide stress and glycolysis. This study aims to elucidate the roles of disulfidptosis-related genes (DRGs) in the pathogenesis and progression of PCa, with the goal of improving diagnostic and therapeutic approaches. We analyzed PCa datasets and normal tissue transcriptome data from TCGA, GEO, and MSKCC. Using consensus clustering analysis and LASSO regression, we developed a risk scoring model, which was validated in an independent cohort. The model's predictive accuracy was confirmed through Kaplan-Meier curves, receiver operating characteristic (ROC) curves, and nomograms. Additionally, we explored the relationship between the risk score and immune cell infiltration, and examined the tumor microenvironment and somatic mutations across different risk groups. We also investigated responses to immunotherapy and drug sensitivity. Our analysis identified two disulfidosis subtypes with significant differences in survival, immune environments, and treatment responses. According to our risk score, the high-risk group exhibited poorer progression-free survival (PFS) and higher tumor mutational burden (TMB), associated with increased immune suppression. Functional enrichment analysis linked high-risk features to key cancer pathways, including the IL-17 signaling pathway. Moreover, drug sensitivity analysis revealed varied responses to chemotherapy, suggesting the potential for disulfidosis-based personalized treatment strategies. Notably, we identified PROK1 as a crucial prognostic marker in PCa, with its reduced expression correlating with disease progression. In summary, our study comprehensively assessed the clinical implications of DRGs in PCa progression and prognosis, offering vital insights for tailored precision medicine approaches.

Explainable and visualizable machine learning models to predict biochemical recurrence of prostate cancer.

PURPOSE: Machine learning (ML) models presented an excellent performance in the prognosis prediction. However, the black box characteristic of ML models limited the clinical applications. Here, we aimed to establish explainable and visualizable ML models to predict biochemical recurrence (BCR) of prostate cancer (PCa). MATERIALS AND METHODS: A total of 647 PCa patients were retrospectively evaluated. Clinical parameters were identified using LASSO regression. Then, cohort was split into training and validation datasets with a ratio of 0.75:0.25 and BCR-related features were included in Cox regression and five ML algorithm to construct BCR prediction models. The clinical utility of each model was evaluated by concordance index (C-index) values and decision curve analyses (DCA). Besides, Shapley Additive Explanation (SHAP) values were used to explain the features in the models. RESULTS: We identified 11 BCR-related features using LASSO regression, then establishing five ML-based models, including random survival forest (RSF), survival support vector machine (SSVM), survival Tree (sTree), gradient boosting decision tree (GBDT), extreme gradient boosting (XGBoost), and a Cox regression model, C-index were 0.846 (95%CI 0.796-0.894), 0.774 (95%CI 0.712-0.834), 0.757 (95%CI 0.694-0.818), 0.820 (95%CI 0.765-0.869), 0.793 (95%CI 0.735-0.852), and 0.807 (95%CI 0.753-0.858), respectively. The DCA showed that RSF model had significant advantages over all models. In interpretability of ML models, the SHAP value demonstrated the tangible contribution of each feature in RSF model. CONCLUSIONS: Our score system provide reference for the identification for BCR, and the crafting of a framework for making therapeutic decisions for PCa on a personalized basis.

Dissecting order amidst chaos of programmed cell deaths: construction of a diagnostic model for KIRC using transcriptomic information in blood-derived exosomes and single-cell multi-omics data in tumor microenvironment.

Background: Kidney renal clear cell carcinoma (KIRC) is the most frequently diagnosed subtype of renal cell carcinoma (RCC); however, the pathogenesis and diagnostic approaches for KIRC remain elusive. Using single-cell transcriptomic information of KIRC, we constructed a diagnostic model depicting the landscape of programmed cell death (PCD)-associated genes, namely cell death-related genes (CDRGs). Methods: In this study, six CDRG categories, including apoptosis, necroptosis, autophagy, pyroptosis, ferroptosis, and cuproptosis, were collected. RNA sequencing (RNA-seq) data of blood-derived exosomes from the exoRBase database, RNA-seq data of tissues from The Cancer Genome Atlas (TCGA) combined with control samples from the GTEx databases, and single-cell RNA sequencing (scRNA-seq) data from the Gene Expression Omnibus (GEO) database were downloaded. Next, we intersected the differentially expressed genes (DEGs) of the KIRC cohort from exoRBase and the TCGA databases with CDRGs and DEGs obtained from single-cell datasets, further screening out the candidate biomarker genes using clinical indicators and machine learning methods and thus constructing a diagnostic model for KIRC. Finally, we investigated the underlying mechanisms of key genes and their roles in the tumor microenvironment using scRNA-seq, single-cell assays for transposase-accessible chromatin sequencing (scATAC-seq), and the spatial transcriptomics sequencing (stRNA-seq) data of KIRC provided by the GEO database. Result: We obtained 1,428 samples and 216,155 single cells. After the rational screening, we constructed a 13-gene diagnostic model for KIRC, which had high diagnostic efficacy in the exoRBase KIRC cohort (training set: AUC = 1; testing set: AUC = 0.965) and TCGA KIRC cohort (training set: AUC = 1; testing set: AUC = 0.982), with an additional validation cohort from GEO databases presenting an AUC value of 0.914. The results of a subsequent analysis revealed a specific tumor epithelial cell of TRIB3high subset. Moreover, the results of a mechanical analysis showed the relatively elevated chromatin accessibility of TRIB3 in tumor epithelial cells in the scATAC data, while stRNA-seq verified that TRIB3 was predominantly expressed in cancer tissues. Conclusions: The 13-gene diagnostic model yielded high accuracy in KIRC screening, and TRIB3high tumor epithelial cells could be a promising therapeutic target for KIRC.

Urine-derived exosomal PSMA is a promising diagnostic biomarker for the detection of prostate cancer on initial biopsy.

PURPOSE: It is well-established that the lack of accurate diagnostic modalities for prostate cancer (PCa) leads to overdiagnosis and overtreatments. Accordingly, this study aimed to assess the value of urine-derived exosomal prostate-specific membrane antigen (PSMA) as a biomarker for the diagnosis of PCa and clinically significant prostate cancer (csPCa). METHODS: A total of 284 urine samples were collected from patients after the digital rectal examination (DRE). Urinary exosomes were extracted using commercial kits, and urine-derived exosomal PSMA was determined via enzyme-linked immunosorbent assay (ELISA). Evaluation of diagnostic accuracy of PSMA was performed via receiver operating characteristic (ROC) analysis, decision curve analysis (DCA), and waterfall plots. RESULTS: We found that urine-derived exosomal PSMA was significantly higher in PCa and csPCa than in benign prostatic hyperplasia (BPH) and BPH + non-aggressive prostate cancer (naPCa) groups (P < 0.001). Furthermore, the urine-derived exosome PSMA yielded area under the ROC curve (AUC) values of 0.876 and 0.826 for detecting PCa and csPCa, respectively, suggesting better performance than traditional clinical biomarkers. Besides, when the cutoff value used corresponded to a sensitivity of 95%, urine-derived exosomal PSMA could avoid unnecessary biopsies in 41.2% of cases and missed only 0.7% of csPCa cases. CONCLUSIONS: Urine-derived exosomal PSMA exhibits a good diagnostic yield for detecting PCa and csPCa. Findings of the present study provide the foothold for future studies on cancer management and research in this patient population.

A Prognostic Model for Predicting Tumor Mutation Burden and Tumor-Infiltrating Immune Cells in Bladder Urothelial Carcinoma.

Tremendous progress has been made in development of immunotherapeutic approaches for treatment of bladder urothelial carcinoma (BLCA). However, efficacy and safety of these approaches remain unsatisfactory, necessitating further investigations for identification of indicators for predicting prognosis and efficacy. In this study, we downloaded transcriptomic and clinical data of BLCA patients from The Cancer Genome Atlas (TCGA) database, and identified differentially expressed genes (DEGs) between tumor and normal tissues. We incorporated these DEGs in an intersection analysis with immune-related genes (IRGs) obtained from the Immunology Database and Analysis Portal (ImmPort) database, and identified immune-related DEGs. These genes were subjected to Cox and least absolute shrinkage and selection operator (LASSO) regression analyses, then a prognostic model containing AHNAK, OAS1, NGF, PPY and SCG2 genes was constructed, for prediction of prognosis of BLCA and efficacy of immunotherapy. Finally, we explored the relationship between the prognostic model and tumor mutational burden (TMB), abundance of tumor-infiltrating immune cells (TICs) and immunotherapeutic targets, and found that patients with higher risk score (RS) had poorer prognosis and significantly lower levels of TMB. Patients in the low-RS group exhibited higher numbers of lymphoid cells, whereas those in the high-RS group exhibited higher proportions of myeloid cells. However, patients with high-RS tended to respond better to immunotherapy relative to those in the low-RS group. The constructed prognostic model provides a new tool for predicting prognosis of BLCA patients and efficacy of immunotherapy, offering a feasible option for management of the disease.

KCNN4 is a potential prognostic marker and critical factor affecting the immune status of the tumor microenvironment in kidney renal clear cell carcinoma.

BACKGROUND: The tumor microenvironment (TME) has emerged as a crucial factor in cancer development and progression. Recent findings have indicated that tumor-infiltrating immune cells (TICs) in the TME may predict cancer prognosis and response to treatment. Herein, we sought to identify critical modulators of the kidney renal clear cell carcinoma (KIRC) TME. METHODS: KIRC datasets from The Cancer Genome Atlas (TCGA) were analyzed using the ESTIMATE algorithm to determine the ImmuneScore and StromalScore. By profiling the differentially expressed genes (DEGs) in the ImmuneScore and StromalScore, we finally identified the immune- and stromal-related DEGs of the cases, through which we then performed intersection analysis to determine the immune-related genes (IRGs). Cox regression analysis and least absolute shrinkage and selection operator (LASSO) regression analysis were used to identify critical IRGs and construct a prognostic model. The CIBERSORT algorithm was used to calculate the relative content of 22 immune cell types. Finally, the datasets from the Gene Expression Omnibus (GEO) database were analyzed to validate results from the above analyses. Experimental validation was used on KIRC tissues by quantitative polymerase chain reaction (qPCR) and western blot. RESULTS: We found that the ImmuneScore was negatively correlated with patients' prognosis. Intersection analysis of the ImmuneScore and StromalScore identified 118 IRGs that were enriched in immune-related functions. Following IRGs screening by Cox and LASSO regression analyses, six genes were identified and used to construct a KIRC prognostic model. Intersection analysis of these six genes and protein-protein interaction (PPI) were performed and obtained the most critical gene: Potassium Calcium-Activated Channel Subfamily N Member 4 (KCNN4). Further analysis showed that KCNN4 expression was higher in tumor samples relative to normal controls, and was negatively correlated with prognosis. CIBERSORT analysis revealed significant correlation between KCNN4 expression and multiple types of TICs, demonstrating that KCNN4 may affect KIRC prognosis by influencing the TME immune status. Ultimately, the GEO datasets and validation experiments confirmed that KCNN4 was highly expressed in tumor tissues compared to the corresponding normal tissues. CONCLUSIONS: Our study demonstrated that KCNN4 might be a potential prognostic marker in KIRC, offering a novel therapeutic avenue.

Key Genes And Pathways Controlled By E2F1 In Human Castration-Resistant Prostate Cancer Cells.

BACKGROUND: Treatment of castration-resistant prostate cancer (CRPC) is an enormous challenge. As E2F transcription factor 1 (E2F1) is an essential factor in CRPC, this study investigated the genes and pathways controlled by E2F1 and their effects on cellular behavior in CRPC. METHODS: In vitro assays were used to evaluate cellular proliferation, apoptosis, and behavior. Cellular expression was quantified by RNA sequencing (RNA-seq). Gene co-expression was assessed using the GeneMANIA database, and correlations were analyzed with the GEPIA server. Altered pathways of differentially expressed genes (DEGs) were revealed by functional annotation. Module analysis was performed using the STRING database and hub genes were filtered with the Cytoscape software. Some DEGs were validated by real-time quantitative PCR (RT-qPCR). RESULTS: Knockdown of E2F1 significantly inhibited proliferation and accelerated apoptosis in PC3 cells but not in DU145 cells. Invasion and migration were reduced for both cell lines. A total of 1811 DEGs were identified in PC3 cells and 27 DEGs in DU145 cells exhibiting E2F1 knockdown. Ten overlapping DEGs, including TMOD2 and AIF1L, were identified in both knockdown cell lines and were significantly enriched for association with actin filament organization pathways. TMOD2 and KREMEN2 were genes co-expressed with E2F1; six overlapping DEGs were positively correlated with transcription factor E2F1. DEGs of the PC3 and DU145 groups were associated with multiple pathways. Five DEGs that overlapped between the two cell lines and three hub DEGs from PC3 cells were validated by RT-qPCR. CONCLUSION: The results of this study suggest that E2F1 has a critical role in regulating actin filaments, as indicated by the change in expression level of several genes, including TMOD2 and AIF1L, in CRPC. This extends our understanding of the cellular responses affected by E2F1 in CRPC.