A multi-model based on radiogenomics and deep learning techniques associated with histological grade and survival in clear cell renal cell carcinoma

OBJECTIVES

This study aims to evaluate the efficacy of multi-model incorporated by radiomics, deep learning, and transcriptomics features for predicting pathological grade and survival in patients with clear cell renal cell carcinoma (ccRCC).

METHODS

In this study, data were collected from 177 ccRCC patients, including radiomics features, deep learning (DL) features, and RNA sequencing data. Diagnostic models were then created using these data through least absolute shrinkage and selection operator (LASSO) analysis. Additionally, a multi-model was developed by combining radiomics, DL, and transcriptomics features. The prognostic performance of the multi-model was evaluated based on progression-free survival (PFS) and overall survival (OS) outcomes, assessed using Harrell's concordance index (C-index). Furthermore, we conducted an analysis to investigate the relationship between the multi-model and immune cell infiltration.

RESULTS

The multi-model demonstrated favorable performance in discriminating pathological grade, with area under the ROC curve (AUC) values of 0.946 (95% CI: 0.912-0.980) and 0.864 (95% CI: 0.734-0.994) in the training and testing cohorts, respectively. Additionally, it exhibited statistically significant prognostic performance for predicting PFS and OS. Furthermore, the high-grade group displayed a higher abundance of immune cells compared to the low-grade group.

CONCLUSIONS

The multi-model incorporated radiomics, DL, and transcriptomics features demonstrated promising performance in predicting pathological grade and prognosis in patients with ccRCC.

CRITICAL RELEVANCE STATEMENT

We developed a multi-model to predict the grade and survival in clear cell renal cell carcinoma and explored the molecular biological significance of the multi-model of different histological grades.

KEY POINTS

1. The multi-model achieved an AUC of 0.864 for assessing pathological grade. 2. The multi-model exhibited an association with survival in ccRCC patients. 3. The high-grade group demonstrated a greater abundance of immune cells.

Radiogenomics study to predict the nuclear grade of renal clear cell carcinoma

Purpose

To develop models based on radiomics and genomics for predicting the histopathologic nuclear grade with localized clear cell renal cell carcinoma (ccRCC) and to assess whether macro-radiomics models can predict the microscopic pathological changes.

Method

In this multi-institutional retrospective study, a computerized tomography (CT) radiomic model for nuclear grade prediction was developed. Utilizing a genomics analysis cohort, nuclear grade-associated gene modules were identified, and a gene model was constructed based on top 30 hub mRNA to predict the nuclear grade. Using a radiogenomic development cohort, biological pathways were enriched by hub genes and a radiogenomic map was created.

Results

The four-features-based SVM model predicted nuclear grade with an area under the curve (AUC) score of 0.94 in validation sets, while a five-gene-based model predicted nuclear grade with an AUC of 0.73 in the genomics analysis cohort. A total of five gene modules were identified to be associated with the nuclear grade. Radiomic features were only associated with 271 out of 603 genes in five gene modules and eight top 30 hub genes. Differences existed in the enrichment pathway between associated and un-associated with radiomic features, which were associated with two genes of five-gene signatures in the mRNA model.

Conclusion

The CT radiomics models exhibited higher predictive performance than mRNA models. The association between radiomic features and mRNA related to nuclear grade is not universal.

Diagnostic and Prognostic Biomarkers in Renal Clear Cell Carcinoma

Renal clear cell carcinoma (ccRCC) comprises over 75% of all renal tumors and arises in the epithelial cells of the proximal convoluted tubule. Molecularly ccRCC is characterized by copy number alterations (CNAs) such as the loss of chromosome 3p and VHL inactivation. Additional driver mutations (SETD2, PBRM1, BAP1, and others) promote genomic instability and tumor cell metastasis through the dysregulation of various metabolic and immune-response pathways. Many researchers identified mutation, gene expression, and proteomic signatures for early diagnosis and prognostics for ccRCC. Despite a tremendous influx of data regarding DNA alterations, gene expression, and protein expression, the incorporation of these analyses for diagnosis and prognosis of RCC into the clinical application has not been implemented yet. In this review, we focused on the molecular changes associated with ccRCC development, along with gene expression and protein signatures, to emphasize the utilization of these molecular profiles in clinical practice. These findings, in the context of machine learning and precision medicine, may help to overcome some of the barriers encountered for implementing molecular profiles of tumors into the diagnosis and treatment of ccRCC.

A Novel Machine Learning 13-Gene Signature: Improving Risk Analysis and Survival Prediction for Clear Cell Renal Cell Carcinoma Patients

Simple Summary

Clear cell renal cell carcinoma is a type of kidney cancer which comprises the majority of all renal cell carcinomas. Many efforts have been made to identify biomarkers which could help healthcare professionals better treat this kind of cancer. With extensive public data available, we conducted a machine learning study to determine a gene signature that could indicate patient survival with high accuracy. Through the min-Redundancy and Max-Relevance algorithm we generated a signature of 13 genes highly correlated with patient outcomes. These findings reveal potential strategies for personalized medicine in the clinical practice.

Abstract

Patients with clear cell renal cell carcinoma (ccRCC) have poor survival outcomes, especially if it has metastasized. It is of paramount importance to identify biomarkers in genomic data that could help predict the aggressiveness of ccRCC and its resistance to drugs. Thus, we conducted a study with the aims of evaluating gene signatures and proposing a novel one with higher predictive power and generalization in comparison to the former signatures. Using ccRCC cohorts of the Cancer Genome Atlas (TCGA-KIRC) and International Cancer Genome Consortium (ICGC-RECA), we evaluated linear survival models of Cox regression with 14 signatures and six methods of feature selection, and performed functional analysis and differential gene expression approaches. In this study, we established a 13-gene signature (AR, AL353637.1, DPP6, FOXJ1, GNB3, HHLA2, IL4, LIMCH1, LINC01732, OTX1, SAA1, SEMA3G, ZIC2) whose expression levels are able to predict distinct outcomes of patients with ccRCC. Moreover, we performed a comparison between our signature and others from the literature. The best-performing gene signature was achieved using the ensemble method Min-Redundancy and Max-Relevance (mRMR). This signature comprises unique features in comparison to the others, such as generalization through different cohorts and being functionally enriched in significant pathways: Urothelial Carcinoma, Chronic Kidney disease, and Transitional cell carcinoma, Nephrolithiasis. From the 13 genes in our signature, eight are known to be correlated with ccRCC patient survival and four are immune-related. Our model showed a performance of 0.82 using the Receiver Operator Characteristic (ROC) Area Under Curve (AUC) metric and it generalized well between the cohorts. Our findings revealed two clusters of genes with high expression (SAA1, OTX1, ZIC2, LINC01732, GNB3 and IL4) and low expression (AL353637.1, AR, HHLA2, LIMCH1, SEMA3G, DPP6, and FOXJ1) which are both correlated with poor prognosis. This signature can potentially be used in clinical practice to support patient treatment care and follow-up.

Atoh8 in Development and Disease

Atoh8 belongs to a large superfamily of transcriptional regulators called basic helix-loop-helix (bHLH) proteins. bHLH proteins have been identified in a wide range of organisms from yeast to humans. The members of this special group of transcription factors were found to be involved not only in embryonic development but also in disease initiation and its progression. Given their importance in several fundamental processes, the translation, subcellular location and turnover of bHLH proteins is tightly regulated. Alterations in the expression of bHLH proteins have been associated with multiple diseases also in context with Atoh8 which seems to unfold its functions as both transcriptional activator and repressor. Like many other bHLH transcription factors, so far, Atoh8 has also been observed to be involved in both embryonic development and carcinogenesis where it mainly acts as tumor suppressor. This review summarizes our current understanding of Atoh8 structure, function and regulation and its complex and partially controversial involvement in development and disease.

Detection of Hepatitis B Virus-Host Junction Sequences in Urine of Infected Patients

Integrated hepatitis B virus (HBV) DNA, found in more than 85% of HBV-associated hepatocellular carcinomas (HBV-HCCs), can play a significant role in HBV-related liver disease progression. HBV-host junction sequences (HBV-JSs), created through integration events, have been used to determine HBV-HCC clonality. Here, we investigate the feasibility of analyzing HBV integration in a noninvasive urine liquid biopsy. Using an HBV-targeted next-generation sequencing (NGS) assay, we first identified HBV-JSs in eight HBV-HCC tissues and designed short-amplicon junction-specific polymerase chain reaction assays to detect HBV-JSs in matched urine. We detected and validated tissue-derived junctions in five of eight matched urine samples. Next, we screened 32 urine samples collected from 25 patients infected with HBV (5 with hepatitis, 10 with cirrhosis, 4 with HCC, and 6 post-HCC). Encouragingly, all 32 urine samples contained HBV-JSs detectable by HBV-targeted NGS. Of the 712 total HBV-JSs detected in urine, 351 were in gene-coding regions, 11 of which, including TERT (telomerase reverse transcriptase), had previously been reported as recurrent integration sites in HCC tissue and were found only in the urine patients with cirrhosis or HCC. The integration breakpoints of HBV DNA detected in urine were found predominantly (~70%) at a previously identified integration hotspot, HBV DR1-2 (down-regulator of transcription 1-2). Conclusion: HBV viral-host junction DNA can be detected in urine of patients infected with HBV. This study demonstrates the potential for a noninvasive urine liquid biopsy of integrated HBV DNA to monitor patients infected with HBV for HBV-associated liver diseases and the efficacy of antiviral therapy.

An Exploration of the Tumor Microenvironment Identified a Novel Five-Gene Model for Predicting Outcomes in Bladder Cancer

Bladder cancer (BC) is one of the top ten most common cancer types globally, accounting for approximately 7% of all male malignancies. In the last few decades, cancer research has focused on identifying oncogenes and tumor suppressors. Recent studies have revealed that the interplay between tumor cells and the tumor microenvironment (TME) plays an important role in the initiation and development of cancer. However, the current knowledge regarding its effect on BC is scarce. This study aims to explore how the TME influences the development of BC. We focused on immune and stromal components, which represent the major components of TME. We found that the proportion of immune and stromal components within the TME was associated with the prognosis of BC. Furthermore, based on the scores of immune and stromal components, 811 TME-related differentially expressed genes were identified. Three subclasses with distinct biological features were divided based on these TME-genes. Finally, five prognostic genes were identified and used to develop a prognostic prediction model for BC patients based on TME-related genes. Additionally, we validated the prognostic value of the five-gene model using three independent cohorts. By further analyzing features based on the five-gene signature, higher CD8+ T cells, higher tumor mutational burden, and higher chemosensitivity were found in the low-risk group, which presented a better prognosis. In conclusion, our exploration comprehensively analyzed the TME and identified TME-related prognostic genes for BC, providing new insights into potential therapeutic targets.

ACAT1 and Metabolism-Related Pathways Are Essential for the Progression of Clear Cell Renal Cell Carcinoma (ccRCC), as Determined by Co-expression Network Analysis

Kidney cancer ranks as one of the top 10 causes of cancer death; this cancer is difficult to detect, difficult to treat, and poorly understood. The most common subtype of kidney cancer is clear cell renal cell carcinoma (ccRCC) and its progression is influenced by complex gene interactions. Few clinical studies have investigated the molecular markers associated with the progression of ccRCC. In this study, we collected microarray profiles of 72 ccRCCs and matched normal samples to identify differentially expressed genes (DEGs). Then a weighted gene co-expression network analysis (WGCNA) was conducted to identify co-expressed gene modules. By relating all co-expressed modules to clinical features, we found that the brown module and Fuhrman grade had the highest correlation (r = -0.8, p = 1e-09). Thus, the brown module was regarded as a clinically significant module and subsequently analyzed. Functional annotation showed that the brown module focused on metabolism-related biological processes and pathways, such as fatty acid oxidation and amino acid metabolism. We then performed a protein-protein interaction (PPI) network to identify the hub nodes in the brown module. It is worth noting that only one candidate, acetyl-CoA acetyltransferase (ACAT1), was considered to be the final target most relevant to the Fuhrman grade of ccRCC, by applying the intersection of hub genes in the co-expressed network and the PPI network. ACAT1 was subsequently validated using another two external microarray datasets and the TCGA dataset. Intriguingly, validation results indicated that ACAT1 was negatively correlated with four grades of ccRCC, which was also consistent with our results from qRT-PCR analysis and immunohistochemistry staining of clinical samples. Overexpression of ACAT1 inhibited the proliferation and migration of human ccRCC cells in vitro. In addition, the Kaplan-Meier survival curve showed that patients with a lower expression of ACAT1 showed a significantly lower overall survival rate and disease-free survival rate, indicating that ACAT1 could act as a prognostic and recurrence/progression biomarker of ccRCC. In summary, we found and confirmed that ACAT1 might help to identify the progression of ccRCC, which might have important clinical implications for enhancing risk stratification, therapeutic decision, and prognosis prediction in ccRCC patients.

Development and External Validation of a Novel 12-Gene Signature for Prediction of Overall Survival in Muscle-Invasive Bladder Cancer

Purpose: We aimed to develop and validate a novel gene signature from published data and improve the prediction of survival in muscle-invasive bladder cancer (MIBC).

Methods: We searched the published gene signatures associated with the overall survival (OS) of MIBC and compiled all 274 genes to develop a novel gene signature. RNAseq data of TCGA (the Cancer Genome Atlas) bladder cohort were downloaded. All genes were included in a univariate Cox hazard ratio model. We then used a reduced multivariate Cox regression model, which included only genes achieving P < 0.05 in the univariate model. A total of 172 patients at Fudan University Shanghai Cancer Center (FUSCC) and 61 patients from GEO datasets were used as an external validation set.

Results: A total of 327 patients in the TCGA cohort were enrolled. We identified 274 genes from eight published papers on the OS of MIBC. Using the TCGA database, we identified 12 genes that correlated with OS (P < 0.05 in both univariate and multivariate analyses). By integrating these genes with the RT-qPCR data in our validation dataset and GEO datasets, we confirmed that the power for predicting OS of the 12-gene panel (AUC of 0.741 and 0.727, respectively) was higher than just clinical data (including gender, age, T stage, grade, and N stage) alone in the TCGA and FUSCC cohort (AUC of 0.667 and 0.631, respectively). Additionally, upon combining the clinical data and 12-gene panel together, the AUC increased to 0.768, 0.757, and 0.88 in the TCGA, FUSCC and GSE13507 cohorts, respectively.

Conclusions: Applying published gene signatures and TCGA data, we successfully built and externally validated a novel 12-gene signature for the survival of MIBC.

bHLH Transcription Factor Math6 Antagonizes TGF-β Signalling in Reprogramming, Pluripotency and Early Cell Fate Decisions

The basic helix-loop-helix (bHLH) transcription factor Math6 (Atonal homolog 8; Atoh8) plays a crucial role in a number of cellular processes during embryonic development, iron metabolism and tumorigenesis. We report here on its involvement in cellular reprogramming from fibroblasts to induced pluripotent stem cells, in the maintenance of pluripotency and in early fate decisions during murine development. Loss of Math6 disrupts mesenchymal-to-epithelial transition during reprogramming and primes pluripotent stem cells towards the mesendodermal fate. Math6 can thus be considered a regulator of reprogramming and pluripotent stem cell fate. Additionally, our results demonstrate the involvement of Math6 in SMAD-dependent TGF beta signalling. We furthermore monitor the presence of the Math6 protein during these developmental processes using a newly generated Math6Flag-tag mouse. Taken together, our results suggest that Math6 counteracts TGF beta signalling and, by this, affects the initiating step of cellular reprogramming, as well as the maintenance of pluripotency and early differentiation.

An Integrated Score and Nomogram Combining Clinical and Immunohistochemistry Factors to Predict High ISUP Grade Clear Cell Renal Cell Carcinoma

Objective: The International Society of Urological Pathology (ISUP) has proposed a grading system to classify renal cell carcinoma (RCC). However, classification using biopsy specimens remains problematic and, consequently, the accuracy of a biopsy-based diagnosis is relatively poor. This study aims to combine clinical and immunohistochemical (IHC) factors for the prediction of high ISUP grade clear cell RCC (ccRCC) in an attempt to complement and improve the accuracy of a biopsy-based diagnosis. Methods: A total of 362 ccRCC patients were enrolled in this study and used for the training set. We performed IHC analysis of 18 protein markers on standard tissue sections using an automated stainer. Multivariate logistic regression models were developed to evaluate independent predictors for high ISUP grade. We evaluated different prediction models using receiver operating characteristic (ROC) curves and area under the ROC curve (AUC) analysis. A nomogram for the derivation of an integrated score for predicting high ISUP grade ccRCC and a calibration curve were also plotted. Finally, an internal validation cohort was examined to evaluate the performance of our integrated scoring system and nomogram. Results: Multivariate logistic analyses revealed seven credible candidates for predicting high grade ISUP. These were age, tumor diameter, surgery, and CK7, Ki-67, PTEN, and MTOR protein expression. The ROC curves for the clinical, IHC and integrated models were compared in the training set, and the AUC for each was 0.731, 0.744, and 0.801, respectively. DeLong's test showed that the integrated model was significantly better at predicting high ISUP grade, when compared with the other models. Internal validation confirmed the good performance of the integrated score in predicting ISUP grade. Conclusion: We have developed a nomogram integrating clinical and immunohistochemical parameters to predict high ISUP grade for M0 ccRCC patients. This nomogram may offer potentially useful information during preoperative individualized patient risk assessment, and consequently may help urologists when planning personalized management regimens.

Fifteen hub genes associated with progression and prognosis of clear cell renal cell carcinoma identified by coexpression analysis

Renal cell carcinoma (RCC) is the most common type of renal tumor, and the clear cell renal cell carcinoma (ccRCC) is the most frequent subtype. In this study, our aim is to identify potential biomarkers that could effectively predict the prognosis and progression of ccRCC. First, we used The Cancer Genome Atlas (TCGA) RNA-sequencing (RNA-seq) data of ccRCC to identify 2370 differentially expressed genes (DEGs). Second, the DEGs were used to construct a coexpression network by weighted gene coexpression network analysis (WGCNA). Moreover, we identified the yellow module, which was strongly related to the histologic grade and pathological stage of ccRCC. Then, the functional annotation of the yellow module and single-samples gene-set enrichment analysis of DEGs were performed and mainly enriched in cell cycle. Subsequently, 18 candidate hub genes were screened through WGCNA and protein-protein interaction (PPI) network analysis. After verification of TCGA's ccRCC data set, Gene Expression Omnibus (GEO) data set (GSE73731) and tissue validation, we finally identified 15 hub genes that can actually predict the progression of ccRCC. In addition, by using survival analysis, we found that patients of ccRCC with high expression of each hub gene were more likely to have poor prognosis than those with low expression. The receiver operating characteristic curve showed that each hub gene could effectively distinguish between localized and advanced ccRCC. In summary, our study indicates that 15 hub genes have great predictive value for the prognosis and progression of ccRCC, and may contribute to the exploration of the pathogenesis of ccRCC.

Identification of Biomarkers Associated With Pathological Stage and Prognosis of Clear Cell Renal Cell Carcinoma by Co-expression Network Analysis

Clear cell renal cell carcinoma (ccRCC) is the most common subtype among renal cancer whose prognostic is affected by the tumor progression associated with complex gene interactions. However, there is currently no available molecular markers associated with ccRCC progression and used or clinical application. In our study, microarray data of 101 ccRCC samples and 95 normal kidney samples were analyzed and 2,425 differentially expressed genes (DEGs) were screened. Weighted gene co-expression network analysis (WGCNA) was then conducted and 11 co-expressed gene modules were identified. Module preservation analysis revealed that two modules (red and black) were found to be most stable. In addition, Pearson's correlation analysis identified the module most relevant to pathological stage(patho-module) (r = 0.44, p = 3e-07). Functional enrichment analysis showed that biological processes of the patho-module focused on cell cycle and cell division related biological process and pathway. In addition, 29 network hub genes highly related to ccRCC progression were identified from the stage module. These 29 hub genes were subsequently validated using 2 other independent datasets including GSE53757 (n = 72) and TCGA (n = 530), and the results indicated that all hub genes were significantly positive correlated with the 4 stages of ccRCC progression. Kaplan-Meier survival curve showed that patients with higher expression of each hub gene had significantly lower overall survival rate and disease-free survival rate, indicating that all hub genes could act as prognosis and recurrence/progression biomarkers of ccRCC. In summary, we identified 29 molecular markers correlated with different pathological stages of ccRCC. They may have important clinical implications for improving risk stratification, therapeutic decision and prognosis prediction in ccRCC patients.

Notch 1 is a valuable therapeutic target against cell survival and proliferation in clear cell renal cell carcinoma

Notch 1 is a key component of the Notch pathway, which performs a crucial role in clear cell renal cell carcinoma (CCRCC) development. The present study aimed to investigate whether Notch 1 could serve as a potential target for CCRCC treatment. Firstly, an association analysis was performed using 52 CCRCC cases and 30 normal controls. The results indicated that Notch 1 protein expression in renal tissues was closely associated with the incidence of CCRCC. In addition, higher Notch 1 expression in CCRCC tissues was positively associated with higher tumor-node-metastasis stage and Fuhrman grade, in addition to larger tumor size. Subsequently, an in vitro study was conducted to examine the biological functions of Notch 1 in CCRCC 786-O cells through inhibiting the Notch 1 expression with Notch 1-specific small interfering RNA (siRNA). As a result, the inhibition of Notch 1 expression by increasing concentrations of Notch 1-specific siRNA dose-dependently decreased cell proliferation and increased cell apoptosis in 786-O cells. Furthermore, B-cell lymphoma-2 and procaspase-3 expression exhibited a dose-dependent decrease accompanied with a dose-dependent inactivation of the Akt/mammalian target of rapamycin (mTOR) signaling pathway in Notch 1 siRNA-treated 786-O cells. These findings demonstrated that Notch 1 was associated with CCRCC carcinogenesis and progression, the underlying mechanism of which was that Notch 1 acted as an activator for cell proliferation and a suppressor for cell apoptosis through the Akt/mTOR signaling-dependent pathway in CCRCC. In conclusion, the present study confirmed that Notch 1 is a valuable target against cell survival and proliferation in CCRCC treatment.