Chemoresistance-associated alternative splicing signatures in serous ovarian cancer

Predictive Value of Machine Learning for Platinum Chemotherapy Responses in Ovarian Cancer: Systematic Review and Meta-Analysis

BACKGROUND

Machine learning is a potentially effective method for predicting the response to platinum-based treatment for ovarian cancer. However, the predictive performance of various machine learning methods and variables is still a matter of controversy and debate.

OBJECTIVE

This study aims to systematically review relevant literature on the predictive value of machine learning for platinum-based chemotherapy responses in patients with ovarian cancer.

METHODS

Following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, we systematically searched the PubMed, Embase, Web of Science, and Cochrane databases for relevant studies on predictive models for platinum-based therapies for the treatment of ovarian cancer published before April 26, 2023. The Prediction Model Risk of Bias Assessment tool was used to evaluate the risk of bias in the included articles. Concordance index (C-index), sensitivity, and specificity were used to evaluate the performance of the prediction models to investigate the predictive value of machine learning for platinum chemotherapy responses in patients with ovarian cancer.

RESULTS

A total of 1749 articles were examined, and 19 of them involving 39 models were eligible for this study. The most commonly used modeling methods were logistic regression (16/39, 41%), Extreme Gradient Boosting (4/39, 10%), and support vector machine (4/39, 10%). The training cohort reported C-index in 39 predictive models, with a pooled value of 0.806; the validation cohort reported C-index in 12 predictive models, with a pooled value of 0.831. Support vector machine performed well in both the training and validation cohorts, with a C-index of 0.942 and 0.879, respectively. The pooled sensitivity was 0.890, and the pooled specificity was 0.790 in the training cohort.

CONCLUSIONS

Machine learning can effectively predict how patients with ovarian cancer respond to platinum-based chemotherapy and may provide a reference for the development or updating of subsequent scoring systems.

Multi-omics analysis of the Indian ovarian cancer cohort revealed histotype-specific mutation and gene expression patterns

Introduction: In India, OVCa is women’s third most common and lethal cancer type, accounting for 6.7% of observed cancer incidences. The contribution of somatic mutations, aberrant expression of gene and splice forms in determining the cell fate, gene networks, tumour-specific variants, and the role of immune fraction infiltration have been proven essential in understanding tumorigenesis. However, their interplay in OVCa in a histotype-specific manner remains unclear in the Indian context. In the present study, we aimed to unravel the Indian population histotype-specific exome variants, differentially expressed gene modules, splice events and immune profiles of OVCa samples.

Methods: We analysed 10 tumour samples across 4 ovarian cancer histotypes along with 2 normal patient samples. This included BCFtool utilities and CNVkit for exome, WGCNA and DESeq2 for obtaining differential module hub genes and dysregulated miRNA targets, CIBERSORTx for individual immune profiles and rMATS for tumour specific splice variants.

Result: We identified population-specific novel mutations in Cancer Gene Census Tier1 and Tier2 genes. MUC16, MUC4, CIITA, and NCOR2 were among the most mutated genes, along with TP53. Transcriptome analysis showed significant overexpression of mutated genes MUC16, MUC4, and CIITA, whereas NCOR2 was downregulated. WGCNA revealed histotype-specific gene hubs and networks. Among the significant pathways, alteration in the immune system was one of the pathways, and immune profiling using CIBERSORTx revealed histotype-specific immune cell fraction. miRNA analysis revealed miR-200 family, miR-200a and miR-429 were upregulated in HGSOCs.Splice factor abrasion caused splicing perturbations, with the most abundant alternative splice event being exon skipping and the most spliced gene, SNHG17. Pathway analysis of spliced genes revealed translational elongation and Base excision repair as the pathways altered in OVCa.

Conclusion: Integrated exome, transcriptome, and splicing patterns revealed different population-specific molecular signatures of ovarian cancer histotypes in the Indian Cohort.

Comprehensive characterization of the alternative splicing landscape in ovarian cancer reveals novel events associated with tumor-immune microenvironment

Background: Ovarian cancer (OV) is a serious threat to women’s health. Immunotherapy is a new approach. Alternative splicing (AS) of messenger RNA (mRNA) and its regulation are highly relevant for understanding every cancer hallmark and may offer a broadened target space.

Methods: We downloaded the clinical information and mRNA expression profiles of 587 tumor tissues from The Cancer Genome Atlas (TCGA) database. We constructed a risk score model to predict the prognosis of OV patients. The association between AS-based clusters and tumor-immune microenvironment features was further explored. The ESTIMATE algorithm was also carried out on each OV sample depending on the risk score groups. A total of three immune checkpoint genes that have a significant correlation with risk scores were screened.

Results: The AS-events were a reliable and stable independent risk predictor in the OV cohort. Patients in the high-risk score group had a poor prognosis (P<0.001). Mast cells activated, NK cells resting, and Neutrophils positively correlated with the risk score. The number of Macrophages M1 was also more numerous in the low-risk score group (P<0.05). Checkpoint genes CD274, CTLA-4, and PDCD1LG2, showed a negative correlation with the risk score of AS in OV.

Conclusions: The proposed AS signature is a promising biomarker for estimating overall survival (OS) in OV. The AS-events signature combined with tumor-immune microenvironment enabled a deeper understanding of the immune status of OV patients, and also provided new insights for exploring novel prognostic predictors and precise therapy methods.

Prognostic Signatures of Alternative Splicing Events in Esophageal Carcinoma Based on TCGA Splice-Seq Data

An alternative splicing (AS) event is a highly complex process that plays an essential role in post-transcriptional gene expression. Several studies have suggested that abnormal AS events were the primary element in the pathological process of cancer. However, few works are dedicated to the study of AS events in esophageal carcinoma (EC). In the present study, clinical information and RNA-seq data of EC patients were downloaded from The Cancer Genome Atlas (TCGA) database. The percent spliced in (PSI) values of AS events were acquired from the TCGA Splice-seq. A total of 183 EC patients were enrolled in this study, and 2,212 AS events were found significantly associated with the overall survival of these patients by univariate Cox regression analysis. The prognostic signatures based on AS events were built by multivariate Cox analysis. Receiver operating characteristic (ROC) curves displayed that the area under the curve (AUC) of the following prognostic signatures, including exon skip (ES), alternate terminator (AT), alternate acceptor site (AA), alternate promoter (AP), alternate donor site (AD), retained intron (RI), and total events, was greater than 0.8, suggesting that these seven signatures had valuable prognosis prediction capacity. Finally, the risk score of prognostic signatures was indicated as an independent risk factor of survival. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed to explore the function of splicing factors (SFs) that were associated with AS events. Also, the interactive network between AS events and SFs identified several hub genes and AS events which need further study. This was a comprehensive study that explored prognosis-related AS events and established valuable prognosis signatures in EC patients. The network of interactions between AS events and SFs might offer novel insights into the fundamental mechanisms of tumorigenesis and progression of EC.

Identification of Phospho-Tyrosine Targets as a Strategy for the Treatment of Esophageal Adenocarcinoma Cells

Introduction

Esophageal cancer (EC) is an aggressive cancer type that is increasing at a high rate in the US and worldwide. Extensive sequencing of EC specimens has shown that there are no consistent driver mutations that can impact treatment strategies. The goal of this study was to identify activated tyrosine kinase receptors (TKRs) in EC samples as potential targets in the treatment of EC.

Methods

Activated tyrosine kinase receptors were detected using a dot-blot array for human TK receptors. Human esophageal cancer cell lines were transplanted into immunocompromised mice, and tumor xenografts were subjected to tyrosine kinase inhibitors based on the dot-blot array data.

Results

Using the OE33 esophageal cancer cell line, we identified activated EGF receptor (EGFR), as well as ErbB2 and ErbB3. Treatment of this cell line with erlotinib, a specific inhibitor of EGFR, did not impact the growth of this tumor cell line. Treating the OE33 cell line with afatinib, a pan-EGFR family inhibitor resulted in the growth inhibition of OE33, indicating that the ErbB2 and ErbB3 receptors were contributing to tumor cell proliferation. Afatinib treatment of mice growing OE33 tumors inhibited growth of the OE33 tumor cells.

Discussion

Activated tyrosine kinase receptors were readily detected in both cancer cell lines and human esophageal cancer samples. By identifying the activated receptors and then using the appropriate tyrosine kinase inhibitors, we can block tumor growth in vitro and in animal xenografts. We propose that identifying and targeting activated TKRs can be used as a personalized EC tumor treatment strategy.