Development of machine learning prognostic models for overall survival of prostate cancer patients with lymph node-positive.

Prostate cancer (PCa) patients with lymph node involvement (LNI) constitute a single-risk group with varied prognoses. Existing studies on this group have focused solely on those who underwent prostatectomy (RP), using statistical models to predict prognosis. This study aimed to develop an easily accessible individual survival prediction tool based on multiple machine learning (ML) algorithms to predict survival probability for PCa patients with LNI. A total of 3280 PCa patients with LNI were identified from the Surveillance, Epidemiology, and End Results (SEER) database, covering the years 2000-2019. The primary endpoint was overall survival (OS). Gradient Boosting Survival Analysis (GBSA), Random Survival Forest (RSF), and Extra Survival Trees (EST) were used to develop prognosis models, which were compared to Cox regression. Discrimination was evaluated using the time-dependent areas under the receiver operating characteristic curve (time-dependent AUC) and the concordance index (c-index). Calibration was assessed using the time-dependent Brier score (time-dependent BS) and the integrated Brier score (IBS). Moreover, the beeswarm summary plot in SHAP (SHapley Additive exPlanations) was used to display the contribution of variables to the results. The 3280 patients were randomly split into a training cohort (n = 2624) and a validation cohort (n = 656). Nine variables including age at diagnosis, race, marital status, clinical T stage, prostate-specific antigen (PSA) level at diagnosis, Gleason Score (GS), number of positive lymph nodes, radical prostatectomy (RP), and radiotherapy (RT) were used to develop models. The mean time-dependent AUC for GBSA, RSF, and EST was 0.782 (95% confidence interval [CI] 0.779-0.783), 0.779 (95% CI 0.776-0.780), and 0.781 (95% CI 0.778-0.782), respectively, which were higher than the Cox regression model of 0.770 (95% CI 0.769-0.773). Additionally, all models demonstrated almost similar calibration, with low IBS. A web-based prediction tool was developed using the best-performing GBSA, which is accessible at https://pengzihexjtu-pca-n1.streamlit.app/ . ML algorithms showed better performance compared with Cox regression and we developed a web-based tool, which may help to guide patient treatment and follow-up.

Corrigendum: A New Prognostic Risk Score: Based on the Analysis of Autophagy-Related Genes and Renal Cell Carcinoma.

[This corrects the article DOI: 10.3389/fgene.2021.820154.].

A preliminary study on the diagnostic value of PSADR, DPC and TSRP in the distinction of prostatitis and prostate cancer.

BACKGROUND: The purpose of this study was to investigate the ability of differential diagnosis of prostate specific antigen decline rate (PSADR) per week, degree of prostatic collapse (DPC) and tissue signal rate of prostate (TSRP) between prostatitis and prostate cancer. METHODS: The clinical data of 92 patients [prostate specific antigen (PSA) > 10 ng/mL] who underwent prostate biopsy in the Department of Urology, the Second Affiliated Hospital of Xi 'an Jiaotong University from May 2017 to April 2020 were reviewed retrospectively. They were divided into two groups, prostatitis group (n = 42) and prostate cancer (PCa) group (n = 50), according to pathological results. Parameters, like patient characteristics, PSADR, DPC, TSRP and infectious indicators, were compared and analyzed by t test or non-parametric test to identify if there were significant differences. The thresholds of parameters were determined by the receiver operating characteristic curve (ROC), and the data were analyzed to investigate the diagnostic value in distinguishing of prostatitis and prostate cancer. RESULTS: There were statistical differences in age, PSADR, DPC, TSRP, neutrophil percentage in serum, white blood cell (WBC) in urine and prostate volume between prostatitis group and PCa group (P < 0.001, < 0.001, = 0.001, 0.001, 0.024, 0.014, < 0.001 respectively). There was no statistical difference in serum WBC count, serum neutrophil count, monocyte percentage and urine bacterial count between two groups (P = 0.089, 0.087, 0.248, 0.119, respectively). Determined by ROC curve, when the thresholds of PSADR per week as 3.175 ng/mL/week, DPC as 1.113, TSRP as 2.708 were cutoffs of distinguishing prostatitis and prostate cancer. When combining these three indexes to diagnose, the accuracy rate of diagnosis of prostatitis was 78.85%, the accuracy rate of diagnosis of prostate cancer was 97.50%. Univariate analysis suggested that PSADR, DPC and TSRP played an important role in differentiating prostate cancer from prostatitis (P < 0.05), multivariate analysis suggested PSADR > 3.175 might be good indicators when distinguishing prostate disease with prostatitis (OR = 14.305, 95%CI = 3.779 ~ 54.147), while DPC > 1.113 and TSRP > 2.708 might be associated with a higher risk of prostate cancer (OR = 0.151, 95%CI = 0.039 ~ 0.588; OR = 0.012, 95%CI = 0.005 ~ 0.524, respectively). CONCLUSION: The combination of PSADR per week, DPC, and TSRP might be helpful to distinguish prostate cancer and prostatitis, and can reduce unnecessary invasive and histological procedure.

Abnormal expression and the significant prognostic value of aquaporins in clear cell renal cell carcinoma.

Aquaporins (AQPs) are a kind of transmembrane proteins that exist in various organs of the human body. AQPs play an important role in regulating water transport, lipid metabolism and glycolysis of cells. Clear cell renal cell carcinoma (ccRCC) is a common malignant tumor of the kidney, and the prognosis is worse than other types of renal cell cancer (RCC). The impact of AQPs on the prognosis of ccRCC and the potential relationship between AQPs and the occurrence and development of ccRCC are demanded to be investigated. In this study, we first explored the expression pattern of AQPs by using Oncomine, UALCAN, and HPA databases. Secondly, we constructed protein-protein interaction (PPI) network and performed function enrichment analysis through STRING, GeneMANIA, and Metascape. Then a comprehensive analysis of the genetic mutant frequency of AQPs in ccRCC was carried out using the cBioPortal database. In addition, we also analyzed the main enriched biological functions of AQPs and the correlation with seven main immune cells. Finally, we confirmed the prognostic value of AQPs throughGEPIA and Cox regression analysis. We found that the mRNA expression levels of AQP0/8/9/10 were up-regulated in patients with ccRCC, while those of AQP1/2/3/4/5/6/7/11 showed the opposite. Among them, the expression differences of AQP1/2/3/4/5/6/7/8/9/11 were statistically significant. The differences in protein expression levels of AQP1/2/3/4/5/6 in ccRCC and normal renal tissues were consistent with the change trends of mRNA. The biological functions of AQPs were mainly concentrated in water transport, homeostasis maintenance, glycerol transport, and intracellular movement of sugar transporters. The high mRNA expression levels of AQP0/8/9 were significantly correlated with worse overall survival (OS), while those of AQP1/4/7 were correlated with better OS. AQP0/1/4/9 were prognostic-related factors, and AQP1/9 were independent prognostic factors. In general, this research has investigated the values of AQPs in ccRCC, which could become new survival markers for ccRCC targeted therapy.

Prognostic value of circulating tumor cells and immune-inflammatory cells in patients with renal cell carcinoma.

PURPOSE: The relationships among circulating tumor cells (CTCs), inflammatory cells, and platelets in patients with renal cell carcinoma (RCC) are not transparent. We evaluated the correlations among CTCs, blood inflammatory cells, and platelets in patients with RCC and their prognostic value for metastasis-free survival. METHODS: CTC and typical tumor cell chip data were collected and analyzed by the GEO database. The baseline data, survival data, CTCs data, and blood test results were statistically analyzed. RESULTS: Bioinformatics analysis showed that the function of the differentially expressed genes between CTCs and normal tumor cells mainly involved platelets and immune inflammation. A total of 82 patients whose follow-up time was 3 to 68 months were included in the analysis. Clinical data of the patients confirmed that there is a correlation between platelets and mesenchymal CTCs. Simultaneously, there was a correlation between immune inflammatory cells and platelets. The univariate Cox proportional hazards model indicated that staging, mesenchymal CTCs, and the monocyte-to-neutrophil ratio (MNR) had prognostic value. The multivariate Cox proportional hazards model indicated that staging and the MNR had prognostic value and high accuracy. CONCLUSIONS: Bioinformatics analysis showed that CTCs were related to platelets and immune-inflammatory cells. Furthermore, the clinical data confirmed that platelets were correlated with mesenchymal CTCs and immune-inflammatory cells in the blood. By using mesenchymal CTCs, the MNR, or staging respectively, it is possible to predict the risk of postoperative metastasis in RCC patients. As a compound prognostic factor, staging, and the MNR can provide more convenient and accurate condition monitoring.

A New Prognostic Risk Score: Based on the Analysis of Autophagy-Related Genes and Renal Cell Carcinoma.

Introduction: Clear cell renal cell carcinoma (ccRCC) patients suffer from its high recurrence and metastasis rate, and a new prognostic risk score to predict individuals with high possibility of recurrence or metastasis is in urgent need. Autophagy has been found to have a dual influence on tumorigenesis. In this study we aim to analyze autophagy related genes (ATGs) and ccRCC patients and find a new prognostic risk score. Method: Analyzing differential expression genes (DEGs) in TCGA-KIRC dataset, and took intersection with ATGs. Through lasso, univariate, and multivariate cox regression, DEGs were chosen, and the coefficients and expression levels of them were components constructing the formula of risk score. We analyzed mRNA expression of DEGs in tumor and normal tissue in ONCOMINE database and TCGA-KIRC dataset. The Human Protein Atlas (HPA) was used to analyze protein levels of DEGs. The protein-protein interaction (PPI) network was examined in STRING and visualized in cytoscape. Functional enrichment analysis was performed in RStudio. To prove the ability and practicibility of risk score, we analyzed univariate and multivariate cox regression, Kaplan-Meier curve (K-M curve), risk factor association diagram, receiver operating characteristic curve (ROC curve) of survival and nomogram, and the performance of nomogram was evaluated by calibration curve. Then we further explored functional enrichment related to risk groups through Gene Set Enrichment Analysis (GSEA), weighted gene co-expression network analysis (WGCNA), and Metascape database. At last, we investigated immune cell infiltration of DEGs and two risk groups through TIMER database and "Cibersort" algorithm. Result: We identified 7 DEGs (BIRC5, CAPS, CLDN7, CLVS1, GMIP, IFI16, and TCIRG1) as components of construction of risk score. All 7 DEGs were differently expressed in ccRCC and normal tissue according to ONCOMINE database and TCGA-KIRC dataset. Functional enrichment analysis indicated DEGs, and their most associated genes were shown to be abundant in autophagy-related pathways and played roles in tumorigenesis and progression processes. A serious analysis proved that this risk score is independent from the risk signature of ccRCC patients. Conclusion: The risk score constructed by 7 DEGs had the ability of predicting prognosis of ccRCC patients and was conducive to the identification of novel prognostic molecular markers. However, further experiment is still needed to verify its ability and practicability.

Pathology of circulating tumor cells and the available capture tools (Review).

Circulating tumor cells (CTCs), are tumor cells that diffuse into the circulating blood and serve an important role in the progress of cancer. During the early stages of cancer, CTCs undergo an epithelial­mesenchymal transition and obtain a more invasive phenotype. Subsequently, the tumor cells enter the circulating blood with the aid of immune cells, and enter a dormant state upon reaching distal organs. As the tumor progresses, metastasis may occur under certain conditions. The capture technologies available for CTCs are based on antibody­based capture, or capture based on the physical properties of CTCs, as well as modern technologies that integrate both these methods. Emerging modern technologies have increased the accuracy and efficiency of tumor cell capture, and have thus improved our understanding of tumor cells, and the molecular mechanisms underlying their properties. CTCs serve an important role in disease progression, prediction of patient prognosis and individualized treatment.