A three-gene expression signature predicts lymph node metastasis in cervical squamous cell carcinoma: development and validation using TCGA and clinical cohorts

ObjectiveTo investigate the clinicopathological features and outcomes between node-negative, early-stage cervical squamous cell carcinoma (SCC) and adenocarcinoma (AC) after hysterectomy.MethodsPatients diagnosed with International Federation of Gynecology and Obstetrics (FIGO) stages I–IIA cervical SCC and AC between 1988 and 2013 were retrospectively reviewed using the Surveillance, Epidemiology, and End Results database. We used propensity score-matching to balance patient baseline characteristics. Univariate and multivariate Cox regression analyses were used for prognostic analyses of cause-specific survival (CSS) and overall survival (OS).ResultsA total of 9,858 patients were identified, comprising 6,117 patients (62.1%) and 3,741 (37.9%) patients with cervical SCC and AC, respectively. Compared with cervical SCC, cervical AC cases were more likely to be younger, diagnosed after 2000, white, and have well-differentiated and FIGO stage IB1 disease. For SCC and AC, the 10-year CSS rates were 93.4% and 94.7%, respectively (p=0.011), and the 10-year OS rates were 89.6% and 92.2%, respectively (p<0.001). Multivariate analysis revealed that age, ethnicity, tumor grade, and FIGO stage were independent prognostic factors of CSS and OS, but that histologic subtype was not associated with CSS and OS. In the propensity score-matched patient population, univariate and multivariate analyses also showed that histologic subtype was not associated with survival outcomes.ConclusionCervical AC has equivalent survival to cervical SCC in node-negative, early-stage disease after hysterectomy and lymphadenectomy.

BackgroundSurgery combined with radiotherapy substantially escalates the likelihood of encountering complications in early-stage cervical squamous cell carcinoma(ESCSCC). We aimed to investigate the feasibility of Deep-learning-based radiomics of intratumoral and peritumoral MRI images to predict the pathological features of adjuvant radiotherapy in ESCSCC and minimize the occurrence of adverse events associated with the treatment.MethodsA dataset comprising MR images was obtained from 289 patients who underwent radical hysterectomy and pelvic lymph node dissection between January 2019 and April 2022. The dataset was randomly divided into two cohorts in a 4:1 ratio.The postoperative radiotherapy options were evaluated according to the Peter/Sedlis standard. We extracted clinical features, as well as intratumoral and peritumoral radiomic features, using the least absolute shrinkage and selection operator (LASSO) regression. We constructed the Clinical Signature (Clinic_Sig), Radiomics Signature (Rad_Sig) and the Deep Transformer Learning Signature (DTL_Sig). Additionally, we fused the Rad_Sig with the DTL_Sig to create the Deep Learning Radiomic Signature (DLR_Sig). We evaluated the prediction performance of the models using the Area Under the Curve (AUC), calibration curve, and Decision Curve Analysis (DCA).ResultsThe DLR_Sig showed a high level of accuracy and predictive capability, as demonstrated by the area under the curve (AUC) of 0.98(95% CI: 0.97–0.99) for the training cohort and 0.79(95% CI: 0.67–0.90) for the test cohort. In addition, the Hosmer-Lemeshow test, which provided p-values of 0.87 for the training cohort and 0.15 for the test cohort, respectively, indicated a good fit. DeLong test showed that the predictive effectiveness of DLR_Sig was significantly better than that of the Clinic_Sig(P < 0.05 both the training and test cohorts). The calibration plot of DLR_Sig indicated excellent consistency between the actual and predicted probabilities, while the DCA curve demonstrating greater clinical utility for predicting the pathological features for adjuvant radiotherapy.ConclusionDLR_Sig based on intratumoral and peritumoral MRI images has the potential to preoperatively predict the pathological features of adjuvant radiotherapy in early-stage cervical squamous cell carcinoma (ESCSCC).

Predicting lymph node status based on deep learning on histopathological images from primary tumor of early-stage cervical squamous cell carcinoma (080)

Squamous cell carcinoma (SCC) is histologically the most prominent type of cervical cancer. There is accumulating evidence suggesting that microRNAs (miRNAs) play important regulatory roles in the biological processes of cervical squamous cell carcinoma (CSCC). Deciphering the miRNA regulatory network in CSCC could deepen our understanding at the molecular level of CSCC initiation and progression. In the present study, we performed next‑generation sequencing (NGS) to profile miRNA expression in 3pairs of early-stage CSCC samples. Quantitative real-time polymerase chain reaction (qRT-PCR) was used to verify primary findings in another 20pairs of CSCC samples. We identified 37known miRNAs that exhibited significant alterations in expression (2-fold change or greater), among which 8miRNAs were upregulated and 29miRNAs were downregulated. Nine of these miRNAs were selected for further qRT-PCR validation. A novel miRNA candidate was also reported for the first time in the present study to be upregulated. The Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed that its target genes were involved in MAPK, calcium and adherent junction signaling pathways. The present study systematically characterized the miRNA expression variation in early-stage CSCC and provides novel biomarkers for diagnosis and treatment as well as an opportunity for further investigation of the molecular mechanisms underlying the pathogenesis and development of CSCC.

Background: Colorectal cancer (CRC) has a high morbidity and mortality. Ferroptosis is a phenomenon in which metabolism and cell death are closely related. The role of ferroptosis-related genes in the progression of CRC is still not clear. Therefore, we screened and validated the ferroptosis-related genes which could determine the prevalence, risk and prognosis of patients with CRC. Methods: We firstly screened differentially expressed ferroptosis-related genes by The Cancer Genome Atlas (TCGA) database. Then, these genes were used to construct a risk-score model using the least absolute shrinkage and selection operator (LASSO) regression algorithm. The function and prognosis of the ferroptosis-related genes were confirmed using multi-omics analysis. The gene expression results were validated using publicly available databases and qPCR. We also used publicly available data and ferroptosis-related genes to construct a prognostic prediction nomogram. Results: A total of 24 differential expressed genes associated with ferroptosis were screened in this study. A three-gene risk score model was then established based on these 24 genes and GPX3, CDKN2A and SLC7A11 were selected. The significant prognostic value of this novel three-gene signature was also assessed. Furthermore, we conducted RT-qPCR analysis on cell lines and tissues, and validated the high expression of CDKN2A, GPX3 and low expression of SLC7A11 in CRC cells. The observed mRNA expression of GPX3, CDKN2A and SLC7A11 was consistent with the predicted outcomes. Besides, eight variables including selected ferroptosis related genes were included to establish the prognostic prediction nomogram for patients with CRC. The calibration plots showed favorable consistency between the prediction of the nomogram and actual observations. Also, the time-dependent AUC (>0.7) indicated satisfactory discriminative ability of the nomogram. Conclusions: The present study constructed and validated a novel ferroptosis-related three-gene risk score signature and a prognostic prediction nomogram for patients with CRC. Also, we screened and validated the ferroptosis-related genes GPX3, CDKN2A, and SLC7A11 which could serve as novel biomarkers for patients with CRC.

Earlier research has reported that transcription factors play a crucial role in the anti-tumorigenic immune response of lung cancer patients. The aim of this study is to determine the relationship between post-translational modifications of transcription factors and histological fate and patient prognosis. Based on the information of 293 lung cancer patients in the Gene Expression Omnibus (GEO) database, differentially expressed genes (DEGs) related to the interferon regulatory factor (IRF) and signal transducer and activator of transcription (STAT) families between patients experiencing early death and those with long-term survival were identified and characterized. A survival prediction model was established by incorporating 7 STAT genes and 9 IRF genes into the least absolute shrinkage and selection operator (LASSO) algorithm. Gene Ontology (GO) enrichment analysis indicated that these two transcription factor families can govern lung cancer tissue differentiation and predict patient prognosis. Moreover, the Cox proportional hazards regression model was applied to select the genes with the highest predictive capability to construct a gene-based signature. Lastly, the data of 1,803 and 784 lung cancer patients from the Kaplan-Meier plotter (KMPLOT) and The Cancer Genome Atlas (TCGA) databases were used to evaluate the accuracy and sensitivity of the model. Based on the minimum criterion, TRIM28, IRF3, and STAT3 were employed to generate the prognostic model. The 1-, 3-, and 5-year area under the curve (AUC) values of the three-gene-based signature showed consistent results, signifying that the model had excellent accuracy and sensitivity in predicting overall survival (OS) for patients with lung cancer. Finally, the three-gene signature and tumor-node-metastasis (TNM) staging system were combined to construct a nomogram for evaluating the OS of lung cancer patients. TRIM28 may affect the stability of IRF3. Encouragingly, the predicted OS was highly consistent with the observed OS in multiple cohorts. Taken together, these findings implied that the predictive model based on the three-gene signature showed robust discriminatory performance.

Circulating microRNA expression levels can serve as diagnostic/prognostic biomarkers in several types of malignant tumors; however, to our knowledge, there have been reports describing their value in cervical squamous cell carcinoma (SCC). In this study, we used hybridization arrays to compare the microRNA expression profiles in cervical squamous cell carcinomas (SCC) samples among patients with lymph node metastasis (LNM) or without LNM; 89 microRNAs were found to fit our inclusion criteria. Using quantitative PCR (qPCR), we examined the expression levels of these microRNAs in cervical cancer tissue, as well as in serum from patients and healthy women. We compared the expression levels between patients with LNM (n=40) and those without LNM (n=40) and healthy controls (n=20). Using regression analysis, we generated a comprehensive set of marker microRNAs and drew the fitted binormal receiver operating characteristic (ROC) curves to access the predictive value. We identified 6 serum microRNAs that can predict LNM in cervical SCC patients; these microRNAs were miR-1246, miR-20a, miR-2392, miR-3147, miR-3162-5p and miR-4484. The area under the curve (AUC) of the comprehensive set of serum microRNAs predicting LNM was 0.932 (sensitivity, 0.856; specificity, 0.850). The predictive value of the serum microRNAs was inferior to that in tissue (AUC 0.992; sensitivity, 0.967; specificity, 0.950; P=0.018). We compared the LNM predictive value of serum microRNAs and SCC antigen (SCC-Ag) by drawing fitted binormal ROC curves However, serum microRNA analysis is by far superior to serum SCC-Ag analysis (AUC 0.713; sensitivity, 0.612; specificity, 0.700; P<0.0001). Serum microRNAs are a good predictor of LNM with clinical value in early-stage cervical SCC.

Background Cervical squamous cell carcinoma (CESC) is the gynecologic malignancy with high incidence rate and high mortality rate. Oxidative stress participates in gene regulation and malignant tumor progression, including CESC. Methods RNA-seq, clinical information, and genomic mutation were from The Cancer Genome Atlas- (TCGA-) CESC and GSE44001 datasets. Oxidative stress-related genes were obtained from the gene set enrichment analysis (GSEA) website. ConsensusClusterPlus was used for clustering, which was assessed by the Kaplan-Meier (KM) survival curve analysis, mutation analysis, immunocharacteristic analysis, and therapy. Prognostic signatures were built by combining weighted correlation network analysis (WGCNA), least absolute shrinkage and selection operator (LASSO) algorithm, and stepAIC. The prognostic power of this model was evaluated using the KM survival curve analysis, receiver operating characteristic (ROC) curve analysis, nomogram, and decision curve analysis (DCA). Results 218 of the 291 CESC cases (74.91%) presented oxidative stress-related gene mutation, especially FBXW7. Three clusters were determined based on oxidative stress-related genes, among which cluster 3 (C3) presented low-frequency mutation and hyperimmune state and was sensitive to immunotherapy. This research developed a 5-gene oxidative stress-related prognostic signature and a RiskScore model. As shown by ROC analysis, in the TCGA and GSE44001 datasets, the RiskScore model showed a high prediction accuracy for 1-, 3-, and 5-year CESC overall survival. High RiskScore was associated with enhanced immune status. The nomogram model was greatly predictive of the overall survival of CESC patients. Conclusion Our prognostic model was based on oxidative stress-related genes in CESC, potentially aids in CESC prognosis, and provides potential targets against CESC.

In this study, we evaluated the diagnostic value of key genes in myocardial infarction (MI) based on data from the Gene Expression Omnibus (GEO) database. We used data from GSE66360 to identify a set of significant differentially expressed genes (DEGs) between MI and healthy controls. Logistic regression, least absolute shrinkage and selection operator (LASSO) regression, support vector machine recursive feature elimination (SVM-RFE), and SignalP 3.0 server were used to identify the potential role of genes in predicting diagnosis in patients with MI. Principal component analysis (PCA), receiver operating characteristic (ROC) curve analyses, area under the curve (AUC) analyses, and C-index were used to estimate the diagnostic value of genes in patients with MI. The association was validated using six other independent data sets. Subsequently, bioinformatics analysis was conducted based on the aforementioned potential genes. A meta-analysis was performed to evaluate the diagnostic value of the genes in MI. Forty-four DEGs were selected from the GSE66360 dataset. A three-gene signature consisting of CCL20, IL1R2, and ITLN1 could effectively distinguish patients with MI. The three-gene signature was validated in seven independent cohorts. Functional enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to reveal the involvement of the three-gene signature in inflammation-related biological processes and pathways. Moreover, diagnostic meta-analysis results of the three-gene signature showed that the pooled sensitivity, specificity, and AUC for MI were 0.80, 0.90, and 0.93, respectively. These results suggest that the three-gene signature is a novel candidate biomarker for distinguishing MI from healthy controls.

BackgroundLung adenocarcinoma (LUAD) is the major cause of cancer mortality. Traditional prognostic factors have limited importance after including other parameters. Thus, developing a more credible prognostic model combined with genes and clinical parameters is necessary.MethodsThe messenger RNA (mRNA) expression and clinical information from The Cancer Genome Atlas (TCGA)-LUAD datasets and microarray data from three Gene Expression Omnibus (GEO) databases were obtained. We identified differentially-expressed genes (DEGs) between lung tumor and normal tissues through integrated analysis of the three GEO datasets. Univariate and multivariate Cox regression analyses were conducted to select survival-associated DEGs and to establish a prognostic gene signature which was associated with overall survival (OS). The expression of gene proteins was assessed in 180 LUAD tissue microarrays (TMAs) by immunohistochemistry (IHC). We verified its predictive performance with a Kaplan-Meier (KM) curve, receiver operating characteristic (ROC) curve, and Harrell’s concordance index (C-index) and validated it in external GEO databases. Multivariate Cox regression analysis was performed to identify the significant prognostic indicators in LUAD. Furthermore, we established a prognostic nomogram based on TCGA-LUAD dataset.ResultsA three-gene signature was constructed to predict the OS of LUAD patients. The KM analysis, ROC curve, and C-index present a good predictive ability of the gene signature in TCGA dataset [P<0.0001; C-index 0.6375; 95% confidence interval (CI): 0.5632–0.7118; area under the ROC curve (AUC) 0.674] and the external GEO datasets (P=0.05, 0.004, and 0.04, respectively). Univariate and multivariate Cox regression analyses also verified that LUAD patients with low-risk scores had a decreased risk of death compared to those with a high-risk score in TCGA database [hazard ratio (HR) =0.3898; 95% CI: 0.1938–0.7842; P<0.05]. Finally, we constructed a nomogram integrating the gene signature and clinicopathological parameters (P<0.0001; C-index 0.762; 95% CI: 0.714–0.845; AUC 0.8136). Compared with conventional staging, a nomogram can effectively improve prognosis prediction.ConclusionsThe nomogram is closely associated to the OS of LUAD patients. This consequence may be beneficial to individualized treatment and clinical decision-making.

As heterogeneity of cervical squamous cell carcinoma (CSCC), prognosis assessment for CSCC patients remain challenging. To develop novel prognostic strategies for CSCC patients, associated biomarkers are urgently needed. This study aimed to cluster CSCC samples from a molecular perspective. CSCC expression data sets were obtained from The Cancer Genome Atlas and based on the accessed expression profile, a co-expression network was constructed with weighted gene co-expression network analysis to form different gene modules. Tumor microenvironment was evaluated using ESTIMATE algorithm, observing that the brown module was highly associated with tumor immunity. CSCC samples were clustered into three subtypes by consensus clustering based on gene expression profiles in the module. Gene set variation analysis showed differences in immune-related pathways among the three subtypes. CIBERSORT and single-sample gene set enrichment analysis analyses showed the difference in immune cell infiltration among subtype groups. Also, Human leukocyte antigen protein expression varied considerably among subtypes. Subsequently, univariate, Lasso and multivariate Cox regression analyses were performed on the genes in the brown module and an 8-gene prognostic model was constructed. Kaplan–Meier analysis illuminated that the low-risk group manifested a favorable prognosis, and receiver operating characteristic curve showed that the model has good predictive performance. qRT-PCR was used to examine the expression status of the prognosis-associated genes. In conclusion, this study identified three types of CSCC from a molecular perspective and established an effective prognostic model for CSCC, which will provide guidance for clinical subtype identification of CSCC and treatment of patients.

Transient receptor potential vanilloid 6 (TRPV6) has been shown to promote caner proliferation in several solid tumors, leading to unfavorable clinical outcomes. Our study aimed to elucidate the clinical significance of TRPV6 in patients with early-stage cervical squamous cell carcinoma (CSCC). The mRNA expression of TRPV6 was measured in 12 paired early-stage CSCC specimens and six cervical carcinoma cell lines using quantitative real-time PCR (qRT-PCR). Western blotting and immunohistochemistry (IHC) were employed to examine the protein expression level of TRPV6 in four paired specimens, 175 paraffin-embedded early-stage CSCC specimens, and 50 normal cervical tissues (NCTs), respectively. Statistical analyses were performed to evaluate the clinical significance of TRPV6 expression. The expressions of TRPV6 mRNA and protein were both significantly downregulated in early-stage CSCC tissues and cervical cancer cell lines. IHC analyses revealed that TRPV6 was downregulated in 136 (77.7%) of 175 early-stage CSCC specimens. Moreover, TRPV6 expression in early-stage CSCC was significantly correlated with the tumor stage (P<0.001), tumor growth type (P<0.001), tumor size (P=0.008), and differentiation grade (P=0.003). The early-stage CSCC patients with a low TRPV6 expression level had a short progress-free survival (PFS) and overall survival (OS) duration. Univariate and multivariate analyses identified TRPV6 as an independent prognostic factor for early-stage CSCC patients' survival. We demonstrated that TRPV6 was downregulated in CSCC, which was correlated with unfavorable survival outcomes of early-stage CSCC patients. TRPV6 may be used as a novel prognostic marker for early-stage CSCC.

Triple-negative breast cancer (TNBC), a type of breast cancer, lacks immune-related markers that can be used for prognosis or prediction. Therefore, we created a predictive framework for TNBC using a risk assessment. Our previous study group consisted of 360 individuals who were diagnosed with TNBC through pathology using RNA sequencing and had clinical data from Fudan University Shanghai Cancer Center (FUSCC). A risk scoring model was constructed using the Cox regression method with the least absolute shrinkage and selection operator (LASSO). A multivariate Cox regression analysis was utilized to develop the prediction model, which was then assessed using the consistency index and calibration plots. The validation cohort of The Cancer Genome Atlas (TCGA) TNBC confirmed the strength of the signatures' predictive value. The prognostic risk score model included 12 genes: TDO2, CHIT1, CARML2, HLA-C, ADIRF, C19orf33, CA8, AHNAK2, RHOV, OPLAH, THEM6, and NEBL. The receiver operator characteristic (ROC) curves for survivability values at 1, 3, and 5 years in the FUSCC TNBC cohort demonstrated area under the curve (AUC) values of 0.78, 0.83, and 0.75, respectively. These results indicated a high level of accuracy in predicting outcomes, which was further confirmed through validation using TCGA database. The patients in the high-risk group showed worse prognoses and lower levels of immune cell infiltration, specifically CD8+ T cells, than those in the low-risk group. Furthermore, the low-risk group exhibited a significant upregulation of genes that encode immune checkpoints, including CD274 and CTLA4, suggesting that immunotherapy may yield enhanced efficacy within this particular group. In conclusion, the prognostic signature consisting of 12 genes can assist in the choice of immunotherapy for TNBC.

Programmed cell death (PCD) has been widely investigated in various human diseases. The present study aimed to identify a novel PCD-related genetic signature in cervical squamous cell carcinoma (CESC) to provide clues for survival, immunotherapy and drug sensitization prediction. Single-sample gene set enrichment analysis (ssGSEA) was used to quantify the PCD score and assess the distribution of PCD in clinicopathological characteristics in The Cancer Genome Atlas (TCGA)-CESC samples. Then, the ConsensusClusterPlus method was used to identify molecular subtypes in the TCGA-CESC database. Genomic mutation analysis, Gene Ontology and Kyoto Encyclopedia of Genes and Genomes functional enrichment, as well as tumor microenvironment (TME) infiltration analysis, were performed for each molecular subtype group. Finally, a prognostic model by Uni-Cox and least absolute shrinkage and selection operator-Cox analysis was established based on differentially expressed genes from molecular subtypes. ESTIMATE (i.e. Estimation of STromal and Immune cells in MAlignantTumours using Expression data) and ssGSEA were performed to assess the correlation between the model and TME. Drug sensitization prediction was carried out with the oncoPredict package. Preliminary analysis indicated that PCD had a potential association clinical characteristics of the TCGA-CESC cohort, and PCD-related genes mutated in 289 (70.59%) CESC patients. Next, four groups of CESC molecular typing were clustered based on 63 significantly prognostic PCD-related genes. Among four subtypes, C1 group displayed the worst prognosis combined with over expressed PCD genes and enriched cell cycle-related pathways. C4 group exhibited the best prognosis accompanied with high degree of immune infiltration. Finally, a five-gene (SERPINE1, TNF, CA9, CX3CL1 and JAK3) prognostic model was constructed. Patients in the high-risk group displayed unfavorable survival. Immune infiltration analysis found that the low-risk group had significantly higher levels of immune cell infiltration such as T cells, Macrophages_M1, relative to the high-risk group, and were significantly enriched in apoptosis-associated pathways, which predicted a higher level of immunity. Drug sensitivity correlation analysis revealed that the high-risk group was resistant to conventional chemotherapeutic drugs and sensitive to the Food and Drug Administration-approved drugs BI.2536_1086 and SCH772984_1564. In the present study, we first found that PCD-related gene expression patterns were correlated with clinical features of CESC patients, which predicts the feasibility of subsequent mining of prognostic features based on these genes. The five-PCD-associated-gene prognostic model showed good assessment ability in predicting patient prognosis, immune response and drug-sensitive response, and provided guidance for the elucidation of the mechanism by which PCD affects CESC, as well as for the clinical targeting of drugs.

BackgroundIt is widely acknowledged that hypoxia and m6A/m5C/m1A RNA modifications promote the occurrence and development of tumors by regulating the tumor microenvironment. This study aimed to establish a novel liver cancer risk signature based on hypoxia and m6A/m5C/m1A modifications.MethodsWe collected data from The Cancer Genome Atlas (TCGA-LIHC), the National Omics Data Encyclopedia (NODE-HCC), the International Cancer Genome Consortium (ICGC), and the Gene Expression Omnibus (GEO) databases for our study (GSE59729, GSE41666). Using Cox regression and least absolute shrinkage and selection operator (LASSO) method, we developed a risk signature for liver cancer based on differentially expressed genes related to hypoxia and genes regulated by m6A/m5C/m1A modifications. We stratified patients into high- and low-risk groups and assessed differences between these groups in terms of gene mutations, copy number variations, pathway enrichment, stemness scores, immune infiltration, and predictive capabilities of the model for immunotherapy and chemotherapy efficacy.ResultsOur analysis revealed a significantly correlated between hypoxia and methylation as well as m6A/m5C/m1A RNA methylation. The three-gene prognosis signature (CEP55, DPH2, SMS) combining hypoxia and m6A/m5C/m1A regulated genes exhibited strong predictive performance in TCGA-LIHC, NODE-HCC, and ICGC-LIHC-JP cohorts. The low-risk group demonstrated a significantly better overall survival compared to the high-risk group (p < 0.0001 in TCGA, p = 0.0043 in NODE, p = 0.0015 in ICGC). The area under the curve (AUC) values for survival at 1, 2, and 3 years are all greater than 0.65 in the three cohorts. Univariate and Multivariate Cox regression analyses of the three datasets indicated that the signature could serve as an independent prognostic predictor (p < 0.001 in the three cohorts). The high-risk group exhibited more genome changes and higher homologous recombination deficiency scores and stemness scores. Analysis of immune infiltration and immune activation confirmed that the signature was associated with various immune microenvironment characteristics. Finally, patients in the high-risk group experienced a more favorable response to immunotherapy, and various common chemotherapy drugs.ConclusionOur prognostic signature which integrates hypoxia and m6A/m5C/m1A-regulated genes, provides valuable insights for clinical prediction and treatment guidance for liver cancer patients.