Wei Wang

SRSF9 Forms Phase-Separated Condensates to Promote Ovarian Cancer Progression by Inducing RNA Alternative Splicing That Is Inhibited by m6A Modification

Abstract Deregulation of RNA alternative splicing and modification can play an important role in tumor initiation and progression. Elucidation of the interplay between alternative splicing and modifications of RNA could provide important insights into cancer biology. In this study, we showed that serine/arginine-rich splicing factor 9 (SRSF9) recognized non-N6-methyladenosine (m6A)–modified NUMB mRNA and induced an oncogenic isoform switch in ovarian cancer. NUMB mRNA m6A modification antagonized SRSF9-mediated alternative splicing. Notably, SRSF9 formed phase-separated condensates within the nucleus, which was indispensable for its splicing function as well as its tumor-promoting effect in ovarian cancer. Furthermore, SRSF9 was aberrantly upregulated in ovarian cancer, correlating with poor patient prognosis. Loss of SRSF9 or antisense oligonucleotide–mediated isoform switch of NUMB mRNA inhibited ovarian cancer growth in vitro and in vivo. In conclusion, this study reveals that SRSF9 condensation promotes ovarian cancer progression through modulation of alternative splicing, in competition with m6A modification. Significance: Phase separation increases activity of the splicing factor SRSF9 to support progression of ovarian cancer by generating an oncogenic isoform of NUMB mRNA competitively with m6A modification, which provides promising therapeutic targets.

hnRNPL phase separation activates PIK3CB transcription and promotes glycolysis in ovarian cancer

Ovarian cancer has the highest mortality rate among gynecologic tumors worldwide, with unclear underlying mechanisms of pathogenesis. RNA-binding proteins (RBPs) primarily direct post-transcriptional regulation through modulating RNA metabolism. Recent evidence demonstrates that RBPs are also implicated in transcriptional control. However, the role and mechanism of RBP-mediated transcriptional regulation in tumorigenesis remain largely unexplored. Here, we show that the RBP heterogeneous ribonucleoprotein L (hnRNPL) interacts with chromatin and regulates gene transcription by forming phase-separated condensates in ovarian cancer. hnRNPL phase separation activates PIK3CB transcription and glycolysis, thus promoting ovarian cancer progression. Notably, we observe that the PIK3CB promoter is transcribed to produce a non-coding RNA which interacts with hnRNPL and promotes hnRNPL condensation. Furthermore, hnRNPL is significantly amplified in ovarian cancer, and its high expression predicts poor prognosis for ovarian cancer patients. By using cell-derived xenograft and patient-derived organoid models, we show that hnRNPL knockdown suppresses ovarian tumorigenesis. Together, our study reveals that phase separation of the chromatin-associated RBP hnRNPL promotes PIK3CB transcription and glycolysis to facilitate tumorigenesis in ovarian cancer. The formed hnRNPL-PIK3CB-AKT axis depending on phase separation can serve as a potential therapeutic target for ovarian cancer.

ALYREF condensation stabilizes m5C-modified PARP10 mRNA and promotes PI3K-AKT signaling in ovarian cancer

Abstract The role of epigenetic regulation of RNAs in the tumorigenesis remains incompletely understood. This study uncovers a critical function of the 5-methylcytosine (m 5 C) RNA modification reader protein ALYREF (also termed, ALY; BEF) in ovarian cancer. ALYREF is elevated in ovarian cancer patient samples, and its depletion reduces ovarian tumorigenesis and metastasis in mice in a m 5 C-dependent manner. Mechanistically, ALYREF binds to the m 5 C-modified mRNA of ADP-ribosyltransferase PARP10, competing with exosome complex component MTR4, and enhancing the stability and nuclear export of PARP10 mRNA. Further, ALYREF forms condensates in the nucleus of ovarian cancer cells, and depletion or mutation of ALYREF’s intrinsically disordered regions rescues its control on PARP10 mRNA nucleoplasmic distribution and stability, reduces tumor growth and is required for promotion of ovarian cancer aggressiveness and proliferation. Finally, ALYREF and PARP10 expression correlate with poor prognosis in ovarian cancer patients. Together, these findings suggest that ALYREF phase separation facilitates the malignant progression of ovarian cancer by promoting PARP10 expression and thereby enhancing PARP10-dependent proliferative pathways in a m 5 C-dependent manner.

Integrating homologous recombination deficiency subtyping with TCGA molecular classification for enhanced prognostic stratification and personalised therapy in endometrial cancer

Homologous recombination deficiency (HRD) has emerged as a functional biomarker reflecting genome-wide DNA repair defects and genomic instability. While the Cancer Genome Atlas (TCGA) molecular classification provides valuable prognostic guidance in endometrial cancer (EC), it lacks resolution for DNA repair competency and therapeutic responsiveness. This study aimed to investigate whether HRD subtyping could complement TCGA classification for improved prognostic stratification and therapeutic decision-making. A total of 142 EC patients were analysed using a next-generation sequencing panel and genomic scar-based HRD scoring (loss of heterozygosity, telomeric allelic imbalance, large-scale state transitions). Unsupervised clustering stratified patients into HRD-High, -Middle, and -Low groups. Maximally selected rank statistics were used to identify prognostic thresholds for HRD scores; the tumour-immune microenvironment was characterised by RNA-based immune gene expression profiling and multiplex immunohistochemistry. A support vector machine (SVM) model was developed for recurrence prediction. HRD subtyping identified distinct genomic, pathological, and immunological features. HRD-High tumours were associated with advanced FIGO stages, TP53 mutations, higher chromosomal instability, and elevated CD8⁺PD-1⁺ T-cell infiltration. HRD subtyping independently predicted disease-free survival and showed superior prognostic accuracy (C-index = 0.857) compared to TCGA subtyping (C-index = 0.751). Integrating HRD and TCGA classifiers further improved predictive performance (C-index = 0.903). An SVM model incorporating HRD score and immune features achieved an AUC of 0.733 for recurrence prediction. HRD subtyping refines risk stratification beyond traditional TCGA classification and identifies patients potentially responsive to immune checkpoint or DNA damage-targeted therapies. Integrating HRD-based genomic instability metrics with molecular and immune profiling supports precision oncology in endometrial cancer.

Comprehensive Analysis of DNA Methylation and Transcriptome to Identify PD-1-Negative Prognostic Methylated Signature in Endometrial Carcinoma

Background. Epigenetic mechanism plays an important role in endometrial carcinoma (EC). This study was designed to analyze the epigenetic mechanism between DNA methylation-driven genes (DEDGs) and drugs targeting DEDGs and to develop a DEDG score model for predicting the prognosis of EC. Methods. Expression profile and methylation profile data of PD-1-negative EC samples were obtained from TCGA. To obtain intersected DEDGs, differentially expressed genes (DEGs) and differentially methylated genes from tumor tissues and normal tissues were analyzed by limma. A linear discriminant classification model was constructed using the gene expression profile of DMDGs, methylation profile of TSS1500, TSS200, and gene body regions. Principal component analysis (PCA) and ROC analysis were conducted. The protein-drug interactions analysis of DMDGs was performed using Network Analyst 3.0 tool. Lasso Cox regression analysis was used to screen prognostic DNA methylation driving gene and to build a risk score model. The ROC curve and Kaplan-Meier survival curve were plotted to evaluate the model prediction capability. Results. A total of 96 DMDGs were screened from the three regions, distributed on 22 chromosomes, with consistent methylation patterns in different gene regions. Both the expression profile and methylation profile of the three regions can neatly distinguish tumor samples from normal ones, with a high classifying performance. A gene signature, which consisted of ELFN1-AS1 and ZNF132, could classify EC patients into a high-risk group and low-risk group. Prognosis of the high-risk group was significantly worse than that of the low-risk group. The risk model showed a high performance in predicting the prognosis of EC. Conclusion. We successfully established a risk score system with two DMDGs, which showed a high prediction accuracy of EC prognosis.

Immediate risk and 5-year cumulative risk of high-grade cervical lesions in high-risk HPV-positive patients with minor cytological abnormalities: a retrospective single-center study in China

Although minor cytological abnormalities predict a low risk of high-grade lesions, their high reporting rates lead to a considerable number of high-grade lesion cases. We carried out this study to analyze the immediate risk and 5-year cumulative risk of high-grade cervical lesions in high-risk human papillomavirus (Hr-HPV)-positive patients with minor cytological abnormalities and to investigate the clinical significance of minor cytological abnormalities during follow-up in our single-center. A total of 1892 patients with positive Hr-HPV, cytology result of atypical squamous cells of undetermined significance (ASC-US) or low-grade squamous intraepithelial lesion (LSIL) and also underwent colposcopy and biopsy were selected to analyze the immediate risk of high-grade cervical lesions. Besides, a total of 832 patients with baseline histological results of CIN1 or below and 5-year follow-up data available were further used to analyze the 5-year cumulative risk of high-grade cervical lesions. The immediate incidence rates of CIN3 + in the ASC-US and LSIL groups were 6.27% (63/1005) and 5.64% (50/887), respectively. When CIN3 + was used as the study endpoint, the multivariate logistic regression analysis indicated that there was no significant difference in either the immediate risk or the 5-year cumulative risk of CIN3 + between the ASC-US and LSIL groups. In summary, since both the immediate and 5-year follow-up risks for CIN3 + were similar in patients with ASC-US and LSIL, routine follow-up should be performed in minor cytological abnormalities, regardless of whether the cytology result is ASC-US or LSIL. Through the risk assessment of Hr-HPV and cytology combined screening, the 2019 ASCCP guidelines were suitable for cervical cancer screening at our single center.

Unveiling Gamma‐Interferon‐Inducible Lysosomal Thiol Reductase (IFI30) as a Regulator of Macrophage Polarization and Prognostic Biomarker by Multi‐Transcriptome Analysis in Cervical Cancer

ABSTRACTCervical cancer remains a significant challenge to global health, necessitating the development of reliable clinical prognostic models to predict patient survival outcomes with accuracy. This study aims to develop an mRNA signature model based on tumor immune infiltration characteristics of cervical cancer. By employing RNA sequencing technologies at both tissue and single‐cell resolutions, a survival predictive gene signature was constructed for cervical cancer through the application of machine learning methods. To further validate the key prognostic genes identified in the prognostic signature, we performed additional experiments, including tissue microarray (TMA) analysis and in vitro assays. Our developed signature model comprised nine genes, which ranks at the top tier when compared to previously published mRNA signature models. Gamma‐interferon‐inducible lysosomal thiol reductase (IFI30) emerged as a critical prognostic marker, validated externally through immunohistochemistry (IHC) and multiplex immunohistochemistry staining (mIHC) on cervical cancer TMAs. Notably, IFI30 exhibited pronounced expression in macrophages compared to other cell types within the tumor microenvironment (TME). We further investigated the potential role of IFI30 in regulating macrophage polarization. Specifically, a reduced expression of IFI30 in macrophages co‐cultured with HeLa cells induced a polarization transition from the M2 to the M1 phenotype. In conclusion, we have successfully established a prognostic model on the basis of tumor immune infiltration characteristic of cervical cancer, highlighting IFI30 as a pivotal prognostic marker potentially involved in macrophage polarization. Future investigation is required to explore the underlying mechanisms for the advancement of therapeutic strategies in cervical cancer.

Inhibitory effect of human interleukin-24 on the proliferation, migration, and invasion of cervical cancer cells

Objective This study aimed to identify significantly differentially expressed genes (DEGs) related to cervical cancer by exploring extensive gene expression datasets to unveil new therapeutic targets. Methods Gene expression profiles were extracted from the Gene Expression Omnibus, The Cancer Genome Atlas, and the Genotype-Tissue Expression platforms. A differential expression analysis identified DEGs in cervical cancer cases. Weighted gene co-expression network analysis (WGCNA) was implemented to locate genes closely linked to the clinical traits of diseases. Machine learning algorithms, including LASSO regression and the random forest algorithm, were applied to pinpoint key genes. Results The investigation successfully isolated DEGs pertinent to cervical cancer. Interleukin-24 was recognized as a pivotal gene via WGCNA and machine learning techniques. Experimental validations demonstrated that human interleukin (hIL)-24 inhibited proliferation, migration, and invasion, while promoting apoptosis, in SiHa and HeLa cervical cancer cells, affirming its role as a therapeutic target. Conclusion The multi-database analysis strategy employed herein emphasized hIL-24 as a principal gene in cervical cancer pathogenesis. The findings suggest hIL-24 as a promising candidate for targeted therapy, offering a potential avenue for innovative treatment modalities. This study enhances the understanding of molecular mechanisms of cervical cancer and aids in the pursuit of novel oncological therapies.

High neuropilin and tolloid‐like 1 expression associated with metastasis and poor survival in epithelial ovarian cancer via regulation of actin cytoskeleton

AbstractAbnormal expression of neuropilin and tolloid‐like 1 (NETO1) has been detected in some human carcinomas. However, the expression of NETO1 and the underlying mechanism in epithelial ovarian cancer (EOC) remain unknown. In this study, we found that a higher NETO1 expression in EOC tissue samples compared to normal ovarian tissue samples was significantly correlated with worse overall survival. Additionally, Cox regression analysis suggested that NETO 1 was independently associated with overall survival. NETO1 overexpression enhanced the EOC cells’ migration and invasion capability in vitro via regulation of actin cytoskeleton. Mechanistically, silencing NETO1 reduced the expression of β‐tubulin, F‐actin and KIF2A. In conclusion, our results demonstrated the critical role of NETO1 in EOC invasion, and therapies aimed at inhibiting its expression or activity might significantly control EOC growth, invasion and metastatic dissemination.

Plasma cells shape the mesenchymal identity of ovarian cancers through transfer of exosome-derived microRNAs

Plasma cells shape the mesenchymal identity of ovarian cancers through exosome-mediated miR-330-3p/JAM2 axis.

Periostin+cancer‐associated fibroblasts promote lymph node metastasis by impairing the lymphatic endothelial barriers in cervical squamous cell carcinoma

Lymph node metastasis (LNM), a critical prognostic determinant in cancer patients, is critically influenced by the presence of numerous heterogeneous cancer‐associated fibroblasts (CAFs) in the tumor microenvironment. However, the phenotypes and characteristics of the various pro‐metastatic CAF subsets in cervical squamous cell carcinoma (CSCC) remain unknown. Here, we describe a CAF subpopulation with elevated periostin expression (periostin+CAFs), located in the primary tumor sites and metastatic lymph nodes, that positively correlated with LNM and poor survival in CSCC patients. Mechanistically, periostin+CAFs impaired lymphatic endothelial barriers by activating the integrin‐FAK/Src‐VE‐cadherin signaling pathway in lymphatic endothelial cells and consequently enhanced metastatic dissemination. In contrast, inhibition of the FAK/Src signaling pathway alleviated periostin‐induced lymphatic endothelial barrier dysfunction and its related effects. Notably, periostin‐CAFs were incapable of impairing endothelial barrier integrity, which may explain the occurrence of CAF‐enriched cases without LNM. In conclusion, we identified a specific periostin+CAF subset that promotes LNM in CSCC, mainly by impairing the lymphatic endothelial barriers, thus providing the basis for potential stromal fibroblast‐targeted interventions that block CAF‐dependent metastasis.

Inhibiting ubiquitin conjugating enzyme E2 N by microRNA‐590‐3p reduced cell growth of cervical carcinoma

AbstractThe ubiquitin conjugating enzyme E2 N (UBE2N) has been reported to be involved in the tumorigenesis of several tumors, but its function in cervical carcinoma has not been investigated yet. In the present study, UBE2N was found elevated in cervical carcinoma, and patients with high UBE2N had a shorter overall survival than patients with low expression. Additionally, knockdown of UBE2N decreased the activation of MEK1/2 and p38 in cervical carcinoma cells, and UBE2N knockdown also markedly inhibited cervical carcinoma cell growth. Our further studies found that microRNA‐590‐3p (miR‐590‐3p) was significantly decreased in cervical carcinoma, and patients with high miR‐590‐3p had a longer overall survival than patients with low expression. Moreover, miR‐590‐3p expression was found negatively correlated with UBE2N expression in cervical carcinoma, and our further studies showed that miR‐590‐3p targeted UBE2N and inhibited its expression in cervical carcinoma. Overexpression of miR‐590‐3p could inhibit cervical carcinoma cell growth, but enhanced UBE2N could rescue miR‐590‐3p‐induced cell growth inhibition in cervical carcinoma. This study indicated that targeting miR‐590‐3p/UBE2N axis could be a potential strategy for the treatment of cervical carcinoma.

Prevalence, genotype distribution and risk factors of cervical HPV infection in Yangqu, China: a population-based survey of 10086 women

Human papillomavirus(HPV) infection is a necessary factor for the development of cervical cancer. The HPV vaccine is currently available, but there is still a lack of large-scale research on the distribution and risk factors of HPV. The aim of this study is to investigate the genotype distribution and risk factors of HPV infection in Yangqu which is located in North China. This study enrolled 10086 women aged <65 years from Yangqu County. HPV genotypes were identified via standard HPV DNA testing. The overall prevalence of HPV infection was 8.92%. The prevalence of high-risk HPV types was 8.80%, and it was 0.38% for low-risk HPV types. Single genotype infection accounted for 67.91% in HPV-positive cases. The most common HPV genotypes were HPV-16, -52, and -58. HPV-18 was only the 11th most common type in HPV-positive cases. Women ≥50 years of age had the highest prevalence rate of HPV, and women <30 years had the lowest prevalence rate. The distribution of HPV genotypes also varied among the three age groups: <30, 30-49, and ≥50 years. The risk factors that contributed to the rate of HPV infection included low educational level, low income, smoking, age at first sexual encounter <23 years old, and number of births ≥3 times. This large routine clinical practice report of HPV prevalence and genotype distribution revealed the characteristics of HPV infection-type distributions in Shanxi Province, which should be considered in formulating comprehensive prevention strategies including vaccination for cervical cancer in China.

Multiparametric MRI-based radiomics analysis: differentiation of subtypes of cervical cancer in the early stage

Background There are significant differences in outcomes for different histological subtypes of cervical cancer (CC). Yet, it is difficult to distinguish CC subtypes using non-invasive methods. Purpose To investigate whether multiparametric magnetic resonance imaging (MRI)-based radiomics analysis can differentiate CC subtypes and explore tumor heterogeneity. Material and Methods This study retrospectively analyzed 96 patients with CC (squamous cell carcinoma [SCC] = 50, adenocarcinoma [AC] = 46) who underwent pelvic MRI before surgery. Radiomics features were extracted from the tumor volumes on five sequences (sagittal T2-weighted imaging [T2SAG], transverse T2-weighted imaging [T2TRA], sagittal contrast-enhanced T1-weighted imaging [CESAG], transverse contrast-enhanced T1-weighted imaging [CETRA], and apparent diffusion coefficient [ADC]). Clustering and logistic regression were used to examine the distinguishing capabilities of radiomics features extracted from five different MR sequences. Results Among the 105 extracted radiomics features, there were 51, 38, 37, and 2 features that showed intergroup differences for T2SAG, T2TRA, ADC, and CESAG, respectively (all P &lt; 0.05). AC had greater textural heterogeneity than SCC ( P &lt; 0.05). Upon unsupervised clustering of significantly different features, T2SAG achieved the highest accuracy (0.844; sensitivity = 0.920; specificity = 0.761). The largest area under the curve (AUC) for classification ability was 0.86 for T2SAG. Hence, the radiomics model from five combined MR sequences (AUC = 0.89; accuracy = 0.81; sensitivity = 0.67; specificity = 0.94) exhibited better differentiation ability than any MR sequence alone. Conclusion Multiparametric MRI-based radiomics models may be a promising method to differentiate AC and SCC. AC showed more heterogeneous features than SCC.

Analysis of the agreement between colposcopic impression and histopathological diagnosis of cervical biopsy in a single tertiary center of Chengdu

The aim of this retrospective study was to analyze the agreement between colposcopic impression and histopathological diagnosis of cervical biopsy. The medical records of patients underwent a colposcopy-guided cervical biopsy at Chengdu Women's and Children's Central Hospital between January 2017 and January 2019 were collected, including age, menopausal status, cervical cytology and human papillomavirus (HPV) test results, type of transformation zone, colposcopic diagnosis and histopathological outcomes of cervical biopsy. Colposcopy was carried out using 2011 colposcopic terminology of International Federation for Cervical Pathology and Colposcopy (IFCPC). Related variables were analyzed. A total of 495 patients were collected in this study. The perfect agreement between colposcopic impression and histopathological diagnosis was 46.9%, and the strength of agreement with kappa value was 0.283 (P < 0.001), and the agreement within 1 grade was 93.5%. Positive predictive value (PPV), negative predictive value (NPV), sensitivity, specificity, false-positive rate and false-negative rate of detecting HSIL or more (HSIL +) were 93.1%, 57.8%, 80.9%, 93.9%, 6.1% and 45.3%, respectively. Colposcopic diagnosis more often underestimated (43.2%) [especially in HSIL (59.3%) and carcinoma (70.7%) patients] than overestimated (9.9%) in cervical lesions. The results of cytology, HPV status, patients' age and different experiences of practitioners were the factors for under-diagnosis of HSIL + by colposcopy. Colposcopy is an excellent tool to estimate cervical high-grade lesion but is imprecise. Many factors can bias the diagnosis of colposcopy, especially the known results of cervical cytology and HPV. Precise diagnosis of cervical lesion should rely on the colposcopy-directed biopsy.

PreliminaryMRIStudy of Extracellular Volume Fraction for Identification of Lymphovascular Space Invasion of Cervical Cancer

BackgroundLymphovascular space invasion (LVSI) is a risk factor for poor prognosis of cervical cancer. Preoperative identification of LVSI is very difficult.PurposeTo evaluate the potential of extracellular volume (ECV) fraction based on T1 mapping in preoperative identification of LVSI in cervical cancer compared with dynamic contrast‐enhanced MRI (DCE‐MRI).Study TypeRetrospective.SubjectsA total of 79 patients (median age 54 years) with cervical cancer were classified into LVSI group (n = 29) and without LVSI group (n = 50) according to postoperative pathology.Field Strength/SequenceA 3‐T, noncontrast and contrast‐enhanced T1 mapping performed with volume interpolated breath hold examination (VIBE) sequence, DCE‐MRI applied with 3D T1‐weighted VIBE sequence.AssessmentRegions of interest along the medial edge of the lesion were drawn on slices depicting the maximum cross‐section of the tumor. The noncontrast and contrast‐enhanced T1 value of the tumor and arteries in the same slice were measured, and ECV was calculated from T1 values. The parametric maps (Ktrans,kep, andve) derived from DCE‐MRI standard Toft's model were evaluated.Statistical TestsECV,Ktrans,kep, andvebetween groups with and without LVSI were compared using Student'st‐test. The receiver operating characteristic (ROC) curve and DeLong test were used to evaluate and compare the diagnostic performance of ECV,Ktrans,kep, andvefor differentiating LVSI.P &lt; 0.05 was considered statistically significant.ResultsThe ECV andKtransof the LVSI group were significantly higher than that of non‐LVSI group (52.86% vs. 36.77%, 0.239 vs. 0.176, respectively), and no significant differences inKeporvevalues were observed (P = 0.071 andP = 0.168, respectively). The ECV fraction showed significantly higher area under ROC curve thanKtransfor differentiating LVSI (0.874 vs. 0.655, respectively).Data ConclusionECV measurements based on T1 mapping might improve the discrimination between patients with and without LVSI in cervical cancer, showing better performance for this purpose than DCE‐MRI.Evidence Level2Technical EfficacyStage 2

Downregulation of LNMAS orchestrates partial EMT and immune escape from macrophage phagocytosis to promote lymph node metastasis of cervical cancer

AbstractEpithelial-mesenchymal transition (EMT) is an essential step to drive the metastatic cascade to lymph nodes (LNs) in cervical cancer cells. However, few of them metastasize successfully partially due to increased susceptibility to immunosurveillance conferred by EMT. The precise mechanisms of cancer cells orchestrate EMT and immune evasion remain largely unexplored. In this study, we identified a lncRNA termed lymph node metastasis associated suppressor (LNMAS), which was downregulated in LN-positive cervical cancer patients and correlated with LN metastasis and prognosis. Functionally, LNMAS suppressed cervical cancer cells metastasis in vitro and in vivo. Mechanistically, LNMAS exerts its metastasis suppressive activity by competitively interacting with HMGB1 and abrogating the chromatin accessibility of TWIST1 and STC1, inhibiting TWIST1-mediated partial EMT and STC1-dependent immune escape from macrophage phagocytosis. We further demonstrated that the CpG sites in the promoter region of LNMAS was hypermethylated and contributed to the downregulation of LNMAS. Taken together, our results reveal the essential role of LNMAS in the LN metastasis of cervical cancer and provide mechanistic insights into the regulation of LNMAS in EMT and immune evasion.

Identification of lymph node metastasis in pre‐operation cervical cancer patients by weakly supervised deep learning from histopathological whole‐slide biopsy images

AbstractBackgroundLymph node metastasis (LNM) significantly impacts the prognosis of individuals diagnosed with cervical cancer, as it is closely linked to disease recurrence and mortality, thereby impacting therapeutic schedule choices for patients. However, accurately predicting LNM prior to treatment remains challenging. Consequently, this study seeks to utilize digital pathological features extracted from histopathological slides of primary cervical cancer patients to preoperatively predict the presence of LNM.MethodsA deep learning (DL) model was trained using the Vision transformer (ViT) and recurrent neural network (RNN) frameworks to predict LNM. This prediction was based on the analysis of 554 histopathological whole‐slide images (WSIs) obtained from Qilu Hospital of Shandong University. To validate the model's performance, an external test was conducted using 336 WSIs from four other hospitals. Additionally, the efficiency of the DL model was evaluated using 190 cervical biopsies WSIs in a prospective set.ResultsIn the internal test set, our DL model achieved an area under the curve (AUC) of 0.919, with sensitivity and specificity values of 0.923 and 0.905, respectively, and an accuracy (ACC) of 0.909. The performance of the DL model remained strong in the external test set. In the prospective cohort, the AUC was 0.91, and the ACC was 0.895. Additionally, the DL model exhibited higher accuracy compared to imaging examination in the evaluation of LNM. By utilizing the transformer visualization method, we generated a heatmap that illustrates the local pathological features in primary lesions relevant to LNM.ConclusionDL‐based image analysis has demonstrated efficiency in predicting LNM in early operable cervical cancer through the utilization of biopsies WSI. This approach has the potential to enhance therapeutic decision‐making for patients diagnosed with cervical cancer.

Targeted sequencing and functional interrogation identified novel variant at 12q14.2 associated with risk of ovarian cancer in Han Chinese women

Abstract Chromosome 12q14.2 has been reported as a potential risk locus for epithelial ovarian cancer (EOC) in genome-wide association study (GWAS). We performed targeted sequencing around the rs11175194 at chromosome 12q14.2 and identified five potential risk variants. The association between these five variants and EOC risk was evaluated in 893 EOC cases and 1292 controls. We identified that rs11175195 (P = 1.94 × 10−6, OR = 1.36, 95% CI = 1.20–1.54) was significantly associated with EOC risk in validation study and after meta-analysis with previous GWAS data, rs11175195 reached genome-wide significant level (P &amp;lt; 5 × 10−8). Functional annotation and expression quantitative trait loci analysis prioritized rs11175194 as a causal variant at this locus. The presence of G-rs11175194 risk allele increased binding affinity of the transcription factor NR1H4 and upregulate SRGAP1 gene expression. Overexpression of SRGAP1 promotes the proliferation and invasion in ovarian cancer cell lines. In conclusion, we identified a novel susceptibility locus of ovarian cancer and revealed a potential molecular mechanism for ovarian cancer carcinogenesis. These results may provide a potential biomarker and therapeutic target for ovarian cancer.

Early prediction and risk stratification of ovarian cancer based on clinical data using machine learning approaches

Our study was aimed to construct a predictive model to advance ovarian cancer diagnosis by machine learning. A retrospective analysis of patients with pelvic/adnexal/ovarian mass was performed. Potential features related to ovarian cancer were obtained as many as possible. The optimal machine learning algorithm was selected among six candidates through 5-fold cross validation. Top 20 features having the most powerful predictive significance were ranked by Shapley Additive Interpretation (Shap) method. Clinical validation was further performed to confirm whether our model could advance diagnosis of ovarian cancer. A total of 9,799 patients were collected. The inclusion criteria included age >18 years old, the first diagnosis being pelvic/adnexal/ovarian mass of undetermined significance, and pathological report indispensable. Four hundred and thirty-eight dimensional features were obtained after filtration. LightGBM showed the best performance with accuracy 88%. Among the top 20 features, 55% belonged to laboratory test report, 35% came from imaging examination report, and 10% were attributed to basic demographics and main symptom. Age, CA125, and risk of ovarian malignancy algorithm were the top three. Our predictive model performed stably in testing and clinical validation datasets, and was found to advance the diagnosis of ovarian cancer about 17 days before clinical pathological examination. LightGBM was the optimal algorithm for our predictive model with accuracy of 88%. Laboratory test and imaging examination played essential roles in diagnosing ovarian cancer. Our model could advance the diagnosis of ovarian cancer before clinical pathological examination.

Repression of EGFR by new biguanide 4C potentiated ovarian cancer to PARP inhibitors through down-regulation of BRCA2 and Rad51

Abstract EGFR, one of the most successful therapeutic targets, has recently been found to exert a novel function for regulating homologous recombination (HR). Activation of HR is the critical event of treatment failure of PARPi in BRCA1/2 wild-type ovarian cancer (OC). Besides, the antitumor effects of biguanides have also been a focus of attention. Here, we discovered that the new biguanide 4C inhibited HR and sensitized BRCA1/2 wild-type OC cells to PARPi by targeting EGFR. Mechanistically, EGFR promoted nuclear accumulation of both BRCA2 and Rad51, and HR activation by competitively inhibiting the binding of BRCA2 and Rad51 to E3 ubiquitin ligase c-Cbl, thereby reducing cancer cell sensitivity to PARPi following ATM-mediated DNA damage signal transmission from the nucleus to the cytoplasm. Interestingly, EGFR was downregulated by 4C , which in turn enhanced the interaction of BRCA2 and Rad51 with c-Cbl. Consequently, BRCA2 and Rad51 were then ubiquitinated and degraded to inhibit HR and increase the sensitivity of OC to PARPi. Thus, these findings reveal that the combination of 4C with PARPi leading to “synthetic lethality” is an effective strategy for treating BRCA1/2 wild-type OC.