Transcriptomic Analysis of Epstein-Barr Virus and Mitochondrial Dynamics-Related Genes in Nasopharyngeal Carcinoma Prognosis

Transcriptomic Analysis of Epstein-Barr Virus and Mitochondrial Dynamics-Related Genes in Nasopharyngeal Carcinoma Prognosis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Transcriptomic Analysis of Epstein-Barr Virus and Mitochondrial Dynamics-Related Genes in Nasopharyngeal Carcinoma Prognosis Ping Li, Xin Xu, Xueyu Zhang, Huifen Xie, Cuirong Xiao, Yinggui Yang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8214331/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Nasopharyngeal carcinoma (NPC) is a kind of malignant tumor. The Epstein-Barr virus (EBV) and mitochondrial dynamics may be related to NPC. However, the mechanisms of mitochondrial dynamics and EBV in NPC need to be further explored. Methods This study obtained transcriptomic data from public databases, identified NPC-related prognostic genes through univariate Cox regression analysis and other methods, and subsequently constructed a risk model and a nomogram. Furthermore, based on the prognostic genes, gene set enrichment analysis (GSEA), immune infiltration analysis, and drug sensitivity analysis were performed, and the expression trends of the prognostic genes were verified.All experimental protocols were approved by the Ethics Committee of Affiliated Shenzhen Hospital of Southern Medical University. Results This study identified ARHGAP4, MEIS1, and XCR1 as prognostic genes for NPC, and the constructed risk model exhibited good predictive performance. Furthermore, through GSEA, it was found that the two risk groups were differentially enriched in pathways related to ribosomes and other pathways; meanwhile, immune cell infiltration analysis also showed significant differences and correlations. In addition, differences in the sensitivity to chemotherapeutic drugs such as docetaxel were detected, and the prognostic gene ARHGAP4 was up-regulated in tumor samples, while MEIS1 and XCR1 were down-regulated. Conclusion The research pinpointed three predictive genes (ARHGAP4, MEIS1 and XCR1) and utilized them in developing a risk model, providing new insights into potential therapeutic strategies for NPC. Nasopharyngeal carcinoma Prognostic genes Risk model Epstein-Barr virus Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1. Background Nasopharyngeal carcinoma (NPC) is a malignant tumor that arises from the mucosal epithelium of the nasopharynx. Its causes are complex, involving Epstein-Barr virus (EBV) infection, genetic factors, and environmental factors. Notable long-term consumption of pickled foods containing nitrosamine compounds, exposure to chemical carcinogens in the environment, and smoking may all increase the risk of developing the disease. Early-stage NPC typically presents with subtle, easily overlooked symptoms. However, as the disease progresses, patients may gradually exhibit a series of symptoms such as nasal congestion, epistaxis, tinnitus, hearing loss, headache, and cervical lymph node enlargement[ 1 , 2 ]. Currently, the accurate diagnosis of NPC relies on nasopharyngoscopy, pathological biopsy, CT and MRI imaging examinations, as well as EBV serological tests. In terms of treatment strategies, the treatment of NPC includes radiotherapy, chemotherapy, surgical operations, and immunotherapy[ 3 , 4 ]. Clinical research data indicate that NPC is particularly sensitive to radiotherapy in the early stage, with significant treatment effects; however, as the disease advances to the late stage, a single treatment method is often insufficient to achieve the desired results. At this point, a comprehensive treatment strategy including radiotherapy, chemotherapy, surgery, and immunotherapy is required to control the disease progression as much as possible and prolong the patient's survival time[ 5 , 6 ]. Given that early-stage NPC symptoms are easily overlooked, regular physical examinations and screenings are crucial for early detection and treatment. This not only helps to promptly identify early signs of NPC but, more importantly, it can secure valuable treatment windows for patients, thereby improving treatment outcomes and long-term prognosis. Therefore, the research and promotion of effective early screening methods and related prognostic approaches are of great importance for the prevention and intervention of NPC. Recent studies have shown that overexpression of EGFR, SSTR2, and WTAP is related to poor prognosis and cancer progression in NPC, and they are all promising prognostic biomarkers for NPC[ 7 – 9 ]. EBV is a ubiquitous virus associated with many different human malignancies and autoimmune diseases. As an extremely adaptable human herpesvirus, EBV infects almost everyone at least once in their lifetime[ 10 ]. EBV infection has been implicated in multiple malignancies, including gastric cancer and NPC[ 11 ]. Emerging evidence from diverse disciplines indicates EBV's pathogenic role in both oncogenesis and multiple sclerosis development[ 12 ]. Mechanistically, viral glycoproteins (TNC, FN1, GFBP3) mediate malignant transformation of epithelial cells through molecular interactions[ 13 ]. Research has found that EBV also plays a core role in mediating the tumor suppressor effect of MAOA, and the loss of MAOA may be an important step in the pathogenesis of NPC[ 14 ]. The above-mentioned research indicates that the association between EBV and NPC offers a new perspective for understanding pathological conditions such as NPC. In-depth exploration of the specific mechanisms and interactions between the two will facilitate the development of new treatment strategies. Mitochondria are important organelles that provide energy and maintain the metabolism of most eukaryotic cells. Mitochondrial dynamics—including fusion, fission, and mitophagy—are essential for maintaining optimal mitochondrial function in energy metabolism[ 15 ]. Increasing evidence indicates that mitochondrial fission is highly associated with the occurrence of diseases and cancers, and mitochondrial fission has been proven to be a key mechanism in carcinogenesis mediated by oncogenic viruses[ 16 ]. Additionally, studies have identified that mitochondrial Drp1 is crucial for the development and outcome of EBV latent membrane protein 1 (EBV-LMP1) positive NPC tumors[ 17 ]. The above studies indicate that dysregulation of genes involved in mitochondrial dynamics (particularly fission) may also be strongly associated with NPC initiation and progression, and a comprehensive analysis of the specific mechanisms and interactions between the two will also help us develop new therapeutic strategies. This study identified prognosis-related genes associated with EBV and mitochondrial dynamics through a series of bioinformatics approaches using NPC-related data from public databases. Additionally, the molecular mechanisms of the prognosis genes were further examined via enrichment, immune infiltration, and regulatory network analyses, providing new references for the clinical diagnosis and treatment of NPC. 2. Materials and methods 2.1 Data source The head and neck cancer dataset (TCGA-HNSC) was chosen from the Cancer Genome Atlas Program (TCGA) database, and the clinical information, progression-free survival (PFS) survival information and RNA sequencing (RNA-seq) data of samples from 6 sites (Larynx, Floor of mouth, Tonsil, Base of tongue, Oropharynx, Hypopharynx) were extracted as NPC analysis data, recorded as the training set (high-throughput sequencing), covering 243 NPC tumor tissue samples and 44 normal nasopharyngeal epithelial tissue samples (samples with no survival information were excluded)[ 18 , 19 ]. Specimens of tissue taken from 243 NPC patients, containing survival data, were split into two categories following a 7:3 ratio (162: 81). The group that accounted for 70% was used as the internal training set, while the other group that accounted for 30% was used as the internal testing set. The GSE102349 dataset (GPL11154) obtained from the Gene Expression Omnibus (GEO) database through sequencing mode on chip was used as the validation set and contained 88 NPC tissue samples with PFS[ 20 ]. A search by the molecular signatures database (MsigDB) and removal of overlapping genes resulted in 1,697 EVB-related genes (EBVRGs). The 23 mitochondrial dynamic-related genes (MDRGs) from the reported literature[ 21 ]. 2.2 Acquisition of differentially expressed genes (DEGs) Differential expression analysis was performed using DESeq2 (v 1.38.0)[ 22 ]to identify DEGs between NPC and normal specimens in the training cohort, with significance thresholds set at adjusted p 0.5. Visualization was achieved through ggplot2 (v 3.4.4) for generating volcano plots highlighting the top 10 most differentially expressed genes, supplemented by ComplexHeatmap (v 2.14.0)[ 23 ] for displaying expression patterns of the 40 most significantly up- and down-regulated genes. 2.3 Weighted gene co-expression network analysis (WGCNA) and acquisition of candidate genes The ssGSEA algorithm implemented in GSVA (v 1.50.0)[ 24 ] was employed to quantify enrichment scores for 23 MDRGs across 243 training set tumor specimens. Using an optimized cutoff value, specimens were stratified into high- and low-score cohorts, followed by Kaplan-Meier progression-free survival analysis performed with the survminer package (v 0.4.9)[ 25 ]. MDRG scores were used as traits, and WGCNA analysis was performed by “WGCNA” (v 1.71)[ 26 ]. To pinpoint and eliminate anomalies, a hierarchical grouping of all samples was employed, utilizing the Euclidean distance from the expression profiles of the samples. Following this, we adjusted R 2 to 0.85, mean connectivity to 0 for selecting the best soft threshold (β). The process of determining gene adjacency resulted in the computation of gene similarity, leading to the derivation of a gene dissimilarity coefficient, thereby forming a hierarchical gene clustering tree. By the dynamic tree cutting algorithm's criteria, the least number of genes per module was established at 100, and the gene module. To explore the gene modules highly correlated with the MDRGs score, the Pearson function of the WGCNA was utilized to analyze the correlation between each module and the MDRGs score. A heatmap of the relationship between each module and the MDRGs score was drawn to screen for key modules. The conditions for selecting the modules were: non-grey modules, a p -value 0.4. Finally, the key module genes were obtained. To obtain DEGs associated with EBVRGs and MDRGs in NPC patients, the “ggvenn” package (v 0.1.10)[ 27 ]was utilized to perform intersection operations on the DEGs, EBVRGs, and the key module genes, and the resulting genes were the candidate genes. 2.4 Pathway analysis and protein-protein interaction (PPI) network For investigate the functions performed by the candidate genes more deeply, this study conducted enrichment analysis by Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases with the help of the “clusterProfiler” (v 4.7.1.003)[ 28 ](p.adj < 0.05) and the org.Hs.eg.db human gene annotation package, and the results show the top 5 pathways. Next, the protein interactions between the candidate genes were further explored. In this study, the PPI network of the candidate genes was built using the STRING database and then visualized using Cytoscape[ 29 ]. 2.5 Screening process of prognostic genes and risk modelling To identify survival-related genes associated with the prognosis of NPC, univariate Cox regression analysis was implemented in the internal training set by the “survival” (v 3.5-7)[ 30 ]to screen candidate genes (cut-off value: p 0.05) for survival-related genes, and the genes that passed the PH assumption test were defined as candidate prognostic genes and visualized by drawing a forest plot. Subsequently, a multivariate Cox regression model was built for these candidate prognostic genes. Then, the multivariate regression model was adjusted by applying the stepwise regression function step, with the direction parameter configured as “both”. After that, an optimal combination of a multivariate Cox regression model was chosen, and the genes within this optimal model served as the prognostic genes. To determine whether prognostic gene expression could influence the survival of NPC patients, the NPC in the internal training set were categorized into groups with high/low prognostic gene expression by the median value of prognostic gene expression. KM curve analysis was performed by the “survminer” (v 0.4.9)[ 31 ]package (p < 0.05), and the survival discrepancy between the high and low prognostic gene expression groups was compared using the Log-rank test (p < 0.05). 2.6 Assessment and validation of risk models To assess the prognostic value, in this study, the risk value of each patient was computed in the internal training set, and the risk scores were computed applying the formula: $$\:\text{R}\text{i}\text{s}\text{k}\text{s}\text{c}\text{o}\text{r}\text{e}\:=\:\sum\:_{\text{i}\:=\:1}^{\text{n}}\text{c}\text{o}\text{e}\text{f}\left({\text{g}\text{e}\text{n}\text{e}}_{\text{i}}\right)\ast\:\text{e}\text{x}\text{p}\text{r}\left({\text{g}\text{e}\text{n}\text{e}}_{\text{i}}\right)$$ Cofe: coefficient, Expr: expression NPCs were classified into high and low-risk groups by the optimal threshold of the risk indicator. Immediately thereafter, KM survival was performed by the “survminer” (v 0.4.9)[ 32 ] for the two risk groups in the internal training set. Subsequently, using 1/2/3 years as the survival time node, receiver operating characteristic (ROC) curves were drawn, and the area under the curve (AUC) was calculated using the “survivalROC” (v 1.0.3)[ 33 ]. Heatmaps were plotted with the help of “pheatmap” (v 1.0.12)[ 34 ]to show the expression of prognostic genes. The NPC prognostic model was tested by the internal testing set as well as the validation set GSE102349. Again, patients were grouped with the median risk score as the threshold. Then, risk curves and KM curves were plotted. Subsequently, ROC curves were plotted, and AUC values were determined over periods of 1, 2, and 3 years. In addition, prognostic gene expression heat maps were mapped for the two risk groups. 2.7 Association of risk stratification with clinicopathological parameters The ggpubr package (v 0.6.0)[ 35 ]was employed to assess and graphically present variations in risk scores across demographic (age, gender) and pathological (tumor stage, T/N classification) subgroups within the training cohort. 2.8 Nomogram construction and evaluation The risk score and clinical characteristics of the internal training set were included in the univariate Cox regression analysis to obtain independent prognostic factors. First, a univariate Cox analysis was carried out. Secondly, the PH assumption test and independent prognostic factors were obtained by multivariate Cox analysis. To assess the reliability of the nomogram, by this nomogram prediction model, the nomogram's calibration curve was developed using the R package “rms” (v 6.7-1)[ 36 ], and similarly, the ROC curve for the nomogram was formulated with the assistance of the R package “timeROC” (v 0.4)[ 37 ]. In addition, this study utilized decision curve analysis (DCA) to plot decision curves for 1/2/3 years, which was used to evaluate the clinical utility of the nomogram prediction model. 2.9 Enrichment analysis and immune infiltration analysis To further clarify the functional enrichment in the two risk groups, the “DESeq2” (v 1.42.0)[ 38 ]was applied to analyze the differential expression, and log 2 FC was calculated and ranked. The c2.cp.kegg. The obtained by the MSigDB was applied as the background gene set, and the “clusterProfiler” (v 4.7.1.003)[ 39 ]was applied to perform gene set enrichment analysis (GSEA). The significance threshold for GSEA results was set at |NES| > 1 and p.adjust < 0.05, and the results were visualized through GSEA ridges generated with the “GseaVis” package.(v 0.0.5). To grasp the immune cell infiltration status in the two risk groups, the ssGSEA algorithm in the “GSVA” (v 1.50.0)[ 40 ]was applied to evaluate the infiltration status of 28 immune cells by the internal training set, and the histograms were drawn with the help of the “ggplot2” (v 0.1.10)[ 41 ]to show the proportion of various immune cells infiltrated. Next, the Wilcoxon test was applied, and the immune cells with remarkable differences (p < 0.05) were selected. Finally, Spearman correlation analysis was applied to prognostic genes and immune cells with significant differences, and the results were plotted using the R package “ggplot2” (v 0.1.10). 2.10 Immunotherapeutic response and immune checkpoints To compare the immune pathway and immune checkpoint differences in the two risk groups, the ssGSEA score analysis of immune-related pathways was first performed by the “GSVA” package (v 1.50.0) in the internal training set samples, and the pathway differences were compared (Wilcoxon, p < 0.05). Then, the differences in immune checkpoint-related genes in the two risk groups were compared in the internal training set (p < 0.05). 2.11 Drug sensitivity analysis and expression analysis of prognostic genes To investigate the difference in sensitivity to anti-tumor drugs among two risk groups of NPC patients, the IC 50 of 138 chemotherapy/targeted therapy drugs for each patient in the internal training set was assessed using the “pRRophetic” package (v 0.5)[ 42 ] based on the Genomics of Cancer Drug Sensitivity (GDSC) database. Next, the Wilcoxon test was applied to compare the difference in IC50 among drugs in the two risk groups (p < 0.05), and the 10 most significant drugs were plotted on a box plot with the help of the “ggplot2” package (v 3.4.4)[ 43 ]. Besides, to explore the expression status of prognostic genes in the training dataset, the Wilcoxon test was applied to dissect the expression variation in prognostic gene expression among NPC samples and normal samples, which was demonstrated by violin plots (p < 0.05). 2.12 Statistical analysis Statistical computations were implemented in R (v 4.2.2), employing the Wilcoxon rank-sum test for intergroup comparisons with a predefined significance threshold (p < 0.05). 3. Results 3.1 Acquisition of 4,636 DEGs and 2,952 MDRGs By comprehensively analyzing the DEGs and module genes of NPC in the training set, potential molecular mechanisms and biological processes were revealed, thus deepening our understanding of NPC pathogenesis. We analyzed the DEGs in the training set and identified 4,636 DEGs, of which 2,415 were up-regulated and 2,221 down-regulated for expression in disease samples (Fig. 1 A- 1 B). Analysis of MDRGs scores indicated a remarkable difference in patient survival among high/low scoring groups (p = 0.011) (Fig. 1 C). Hierarchical cluster analysis of NPC samples in the training set revealed no outliers (Fig. 1 D). The scale-free networks created by setting R 2 = 0.85 and determined the soft threshold β as 7 (Fig. 1 E). Then, the genes were divided into modules based on soft thresholds, and 12 co-expression modules were selected (Fig. 1 F). In addition, the correlation analysis among co-expression modules and MDRGs scores showed that the MEbrown (cor = -0.51, p = 4e-17, 2,082), MEmagenta (cor = -0.43, p = 1e-12, 231), and MEblack (cor = -0.43, p = 2e-12, 639) modules were the key modules, and finally 2,952 MDRGs were obtained as key model_genes (Fig. 1 G). 3.2 Detection and functional description of candidate genes The overlap of DEGs, EBVRGs, and key model genes was used to obtain 108 candidate genes (Fig. 2 A). Furthermore, GO enrichment results showed that candidate genes were overrepresented in biological processes, including regulation of immune effector process and humoral immune response. In cellular components, they were mainly involved in endoplasmic reticulum lumen, and in molecular functions, they were mainly represented by oxidoreductase activity (Fig. 2 B). KEGG enrichment analysis enriched a total of 22 pathways, such as endoplasmic reticulum lumen (Fig. 2 C). Additionally, the PPI network showed that there were 80 nodes and 168 edges in the network, indicating that the 80 candidate genes have some interactions (Fig. 2 D). Among them, the CCL2 and CXCL8 genes had the greatest interaction. 3.3 Three prognostic genes were identified Univariate Cox regression analysis and PH assumption test identified 35 candidate prognostic genes (Fig. 3 A, FigS1). Of these, 6 genes (CTTN, DHCR7, GALNS, TMEM158, UGP2, and HMGA2) were risk factors (Hazard Ratio (HR) > 1), and the remaining 29 genes were protective factors (HR < 1). Subsequently, multivariate Cox regression analysis selected three genes (MEIS1, XCR1, ARHGAP4), and these three genes were regarded as prognostic genes (Table 1 , Fig. 3 B). KM curve showed that the survival rate was higher in the high-expression group of all three prognostic genes (Fig. 3 C). The NPC were categorized by the optimal threshold of the risk score (riskscore = 1.2728), and the survival of the high-risk group was remarkably lower (Fig. 3 D). In the training set, there was a remarkable survival difference among the two risk groups (p < 0.0001) (Fig. 3 E). Besides, in the ROC analysis for the 1-, 2-, and 3-year risk models, the AUC values exceeded 0.7, reinforcing its effectiveness in forecasting NPC patient survival rates (Fig. 3 F). Heat map results indicated that the MEIS1, XCR1, and ARHGAP4 genes were lower in the high-risk group (Fig. 3 G). The prognostic value of prognostic models was verified in the internal training set (Fig. 4 A-D) as well as in the validation set (Fig. 4 E-H), and the results obtained were identical to the training set. Table 1 Results of multifactorial Cox analysis id coef HR HR.95L HR.95H pvalue MEIS1 0.22231542 0.80066278 0.59444329 1.07842226 0.14344205 XCR1 0.26684995 0.76578797 0.55833958 1.0503128 0.09783532 ARHGAP4 0.31963429 0.72641464 0.58177756 0.90701028 0.00477955 3.4 Prognostic nomogram for NPC based on independent prognostic factors Independent prognostic analyses were of vital importance for establishing robust clinical decision support systems. The risk score and N stage were significant (HR > 1), and PH assumption was satisfied in monovariate Cox regression analysis (Fig. 5 A). Polyvariate Cox regression analysis indicated that risk score was remarkable, so risk score was taken as an independent prognostic factor(Fig. 5 B). Nomogram founded on independent prognostic factor was constructed to display the possibility of survival of NPC at 1, 2 and 3 years (1, 2, and 3-year survival probability 28.2%, 44.7%, 49.6%, respectively). The scoring relationship showed that 3-year was greater than 2-year was greater than 1-year, indicating that NPC patients had the highest risk at 3-year (Fig. 5 C). Calibration curves and ROC curves further validated the effectiveness of the nomogram. The slope (approximating 1) of the calibration curve (Fig. 5 D) confirmed the nomogram's exceptional predictive accuracy. ROC analysis (Fig. 5 E) further validated its discriminative power (AUC > 0.7), with the nomogram exhibiting superior true positive detection compared to individual predictors. At the same time, the DCA curve showed a better clinical utility of the nomogram (Fig. 5 F). Correlation analysis among risk scores and clinical characteristics indicated that risk scores grouped by age and gender did not show significant differences between groups, and again did not differ significantly in stage; however, there were remarkable differences in risk scores among stages T2-T4, and between stages N0-N3, and N1-N3 (FigS2). FigS2 Risk score analysis (A-B) Correlation between risk score and age, gender, where the horizontal axis represents gender group (A) and gender (B), and the vertical axis represents risk score. (C) Correlation between risk score and Stage, where the horizontal axis represents stage. The vertical axis represents the risk score. (D-E) The correlation between the risk score and different clinical characteristics, where the horizontal axis represents the gender N stage (D) and T stage (E), and the vertical axis represents the risk score. 3.5 GSEA and comprehensive study of immune infiltration GSEA was conducted to delve deeper into the possible processes of prognostic gene differences among patients across two risk groups, with findings presented in Fig. 6 A. The outcomes of the study suggested that the two risk groups showed significant enrichment in pathways (such as ribosome and primary immunodeficiency). Furthermore, the quantity of 28 immune-infiltrating cells in the two risk groups was shown in Fig. 6 B. The results indicated that the infiltration degree of central memory CD4 T cells was more abundant in the high-risk group. Next, analysis of immune cell differences showed that 14 immune cells exhibited remarkable differences among the two risk groups (p < 0.05) (such as activated B cells). Among them, the infiltration level of the other 13 differential immune cells was greater in the low-risk group, and neutrophils was greater in the high-risk group (Fig. 6 C). The correlation analysis among prognostic genes and differential cells showed activated B cell was remarkably positively realted to ARHGAP4 (cor = 0.45), MEIS1 (cor = 0.40), and XCR1 (cor = 0.54) (p < 0.01) (Fig. 6 D). The differential distribution of these immune cells provided important clues for further research on risk assessment and treatment strategies for NPC. 3.6 Immune checkpoint and therapy response analysis in two risk groups The analysis of immune therapy response and immune checkpoints showed that there were 10 immune related pathways (such as Cytolytic-activity) with differences among the two risk groups (p < 0.05) (Fig. 7 A). In addition, among the 45 immune checkpoint genes, 38 (such as LAG3 and IDO2) indicated remarkable differences among the two risk groups (TIM3, VISTA, OX40 had no expression data in the training set) (Fig. 7 B). These findings offered valuable insights into NPC's immune responses and checkpoints, potentially aiding in developing targeted immunotherapies for different risk groups. 3.7 Drug sensitivity analysis and prognostic gene expression level validation To explore the difference in sensitivity to anti-tumor drugs of NPC, drug sensitivity analysis was applied in two risk groups. The IC50 of docetaxel, shikonin, FTI.277, and Imatinib was lower in the high-risk group (Fig. 8 A). By observing the efficacy of these drugs in patients belonging to different risk groups, we can have a better comprehension of the responsiveness of various patients to chemotherapeutic drugs, thereby offering a foundation for individualized treatment.Ultimately,three prognostic genes exhibited significant differences between tumor and normal samples. Amidst them, ARHGAP4 expression was remarkably up-regulated in tumor samples, while MEIS1 and XCR1 were remarkably down-regulated (p < 0.001) (Fig. 8 B). More importantly, Our experimental results also show the consistent trend of ARHGAP4, MEIS1 and XCR1 in NP and NPC tissues (Fig. 8 C). 4. Discussion Discussion on the risk model based on prognostic genes: By combining existing literature and our analysis results, we discuss three prognostic genes (ARHGAP4, MEIS1, and XCR1) and the risk model constructed based on them. ARHGAP4, also known as Rho GTPase activating protein 4, plays a role in various biological processes. Studies have found that ARHGAP4 is a significant Rho family GTPase-activating protein and is closely related to the occurrence and development of certain tumors. Studies have demonstrated elevated ARHGAP4 levels in AML cases, correlating with unfavorable clinical outcomes[ 44 ]. This molecular alteration potentially contributes to oncogenesis, progression, and metastatic dissemination in various malignancies. Its specific mechanism of action may involve regulating cytoskeletal reorganization, affecting cell movement and adhesion, etc. ARHGAP4 is highly expressed in colorectal cancer (CRC), and overexpression of ARHGAP4 is associated with a poor prognosis[ 45 ]. Thus, it is speculated that ARHGAP4 may be a promising indicator for the prognosis of malignant tumors. Myeloid ecotropic viral integration site 1 homolog(MEIS1), also known as bone marrow ecotropic viral integration site 1, is a transcription factor that plays an important role in normal development and disease processes. MEIS1 was first identified as upregulated in myeloid leukemia cell lines. Notably, as a transcriptional regulator, MEIS1 contributes to leukemogenesis and the progression of various malignancies, including solid tumors[ 46 ]. Emerging evidence indicates a significant reduction of MEIS1 transcriptional activity in colorectal carcinoma, which strongly predicts diminished overall survival[ 47 ]. The attenuated MEIS1 levels appear to facilitate neoplastic advancement through multiple mechanisms, including modulation of critical cellular processes (proliferation, differentiation, and programmed cell death) and crosstalk with various oncogenic pathways. XCR1, also known as XC motif chemokine receptor 1, is mainly expressed on specific types of immune cells, such as dendritic cells[ 48 ]. Moreover, XCR1 is a marker of terminally differentiated cDC1 and can mediate the antiviral effector function of human cDC1[ 49 ]. Thus, in tumors, downregulation of XCR1 expression may impair the antitumor immune response in the tumor microenvironment. The above results indicate that the three genes ARHGAP4, MEIS1, and XCR1 play critical tumor-suppressive roles in the initiation and progression of NPC. Our findings showed that high expression of these genes is closely associated with better overall survival in NPC patients, whereas their expression is lower in high-risk patients, further confirming their potential in tumor prognosis assessment. This result suggests that MEIS1, XCR1, and ARHGAP4 may affect the clinical progression of NPC by regulating tumor cell proliferation, metastasis, and immune escape. Although their specific mechanisms require further validation, our findings provide strong evidence for these genes as potential prognostic biomarkers for NPC and may provide new ideas for the formulation of individualized treatment strategies. The risk scoring model based on these genes is expected to assist clinicians in more accurately identifying high-risk NPC patients, thereby enabling them to formulate more personalized treatment regimens. Simultaneously, the expression levels of these genes also provide potential directions for the development of new targeted therapies. For instance, by regulating the expression of these genes or their downstream signaling pathways, it may be possible to effectively inhibit tumor progression and bring better treatment outcomes to patients. Furthermore, immune checkpoint inhibition (ICB) therapy has emerged as a promising treatment strategy and has demonstrated substantial efficacy for treating various cancers, including NPC[ 50 ]; Immune checkpoint inhibitors enhance anti-tumor immunity by reversing T cell exhaustion, thereby improving the targeting and efficacy of treatment[ 51 ]. Our data further show that 14 immune cell subsets exhibit significant differences between high- and low-risk groups—consistent with the expression patterns of ARHGAP4, MEIS1, and XCR1—providing a basis for early high-risk assessment of NPC and screening of high-risk patients through the detection of changes in these immune cell subpopulations. Investigating the association between these prognostic genes and drug sensitivity can, on one hand, explore the quantitative relationship between their expression levels and drug responses. On the other hand, it can analyze the impact of different prognostic gene mutation statuses on drug efficacy to clarify their relationship[ 52 , 53 ]. Our research results show that the IC50 values of docetaxel, shikonin, FTI-277, and imatinib in the high-risk group are lower, which is consistent with existing studies. Docetaxel and imatinib are commonly used clinical anti-tumor drugs, and more clinical studies are currently exploring the therapeutic effects of docetaxel and imatinib on NPC. The results indicate that both have certain advantages in the treatment of NPC[ 54 , 55 ]. Shikonin has also been proven to have anti-tumor effects. Studies have shown that shikonin can inhibit the growth of NPC cells by inactivating the phosphatidylinositol 3-kinase/AKT signaling pathway[ 56 ].FTI-277 is a CAAX peptidomimetic that has been shown to inhibit breast cancer cell invasion and migration by blocking H-Ras activation[ 57 , 58 ]. Notably, our study is the first to link FTI-277 to NPC, highlighting its potential therapeutic value for NPC. This study identified three prognostic genes (ARHGAP4, MEIS1, and XCR1) in nasopharyngeal carcinoma (NPC) and constructed a risk model using these genes via bioinformatics approaches, providing new insights into potential treatment strategies for NPC. Although the transcriptomic data analysis approaches used in this study yielded valuable insights, there are still several technical limitations. For example, the algorithms, software, and models we used may introduce biases or inaccuracies when processing complex data—particularly during data preprocessing, normalization, and result interpretation—all of which may compromise the analytical accuracy. Additionally, although our study identified potential prognostic biomarkers for NPC, the translation of these biomarkers into clinical practice is hindered by multiple challenges. On one hand, existing biomarker detection methods have limitations in sensitivity, specificity, and reproducibility, which limit their feasibility for NPC clinical practice. Although some biomarkers have shown potential in early diagnosis or treatment monitoring, their actual impact on clinical decision-making still needs further validation, especially considering the interaction of multiple factors and individual differences. In future research, we will strive to overcome these challenges. Firstly, we plan to adopt more sophisticated and efficient data analysis approaches—including deep learning and multi-omics data integration approaches—to improve the reliability and predictive performance of our risk model. Secondly, for the clinical translation of biomarkers, we will further optimize detection methods for our identified biomarkers and validate their clinical utility in real-world clinical settings through multi-center clinical trials to ensure their substantial role in clinical decision-making. Ultimately, in line with the concept of precision medicine, we will explore individualized diagnosis and treatment strategies to better serve patients with our research findings. Abbreviations Abbreviations Full name NPC Nasopharyngeal carcinoma EBV Epstein-Barr virus WGCNA weighted gene co-expression network analysis ROC receiver operating characteristic GSEA gene set enrichment analysis EBV-LMP1 EBV latent membrane protein 1 GEO Gene Expression Omnibus PFS progression-free survival MsigDB molecular signatures database EBVRGs EVB-related genes MDRGs mitochondrial dynamic-related genes DEGs differentially expressed genes WGCNA Weighted gene co-expression network analysis PPI protein-protein interaction GO Gene ontology KEGG Kyoto Encyclopedia of Genes and Genomes AUC area under the curve DCA decision curve analysis GDSC Genomics of Cancer Drug Sensitivity CRC colorectal cancer MEIS1 Myeloid ecotropic viral integration site 1 homolog ICB immune checkpoint inhibition Declarations Ethics approval and consent to participate All experimental protocols were approved by the Ethics Committee of Affiliated Shenzhen Hospital of Southern Medical University. Written informed consent was obtained from all the participants. All methods were carried out in accordance with Declaration of Helsinki. Clinical trial number not applicable. Consent for publication Not applicable. Competing interests All the authors declare that they have no conflicts of interest. Funding This work was supported by The Science and Technology Planning Project of Shenzhen, China (JCYJ20210324130801004), the Postdoctoral Research Foundation of Shenzhen (UN-KC-BHKY202205), and Research Foundation of Shenzhen Hospital of Southern Medical University (CNGZRJJPY202008, UN-KJ-KY200024-YYPT, PT2020GZR07, 22H3ATF05). Author Contribution Ping Li, and Xueyu Zhang contributed equally to this work. Yinggui Yang and Cuirong Xiao conceived and designed the research; Ping Li, and Xueyu Zhang developed the methodology and acquired the data; Xin Xu analysed and interpreted the data; Yinggui Yang and Cuirong Xiao revised and approved the manuscript. All authors approved the submission of the manuscript. Data Availability The datasets analyzed during this study are publicly available in The Cancer Genome Atlas (TCGA) database [http://xena.ucsc.edu/] (TCGA-HNSC) and the Gene Expression Omnibus(GEO) repository [https://www.ncbi.nlm.nih.govgeo/] under accession number GSE102349. All accession numbers and associated files have been fully released and are accessible for verification. References Guan S, Wei J, Huang L, Wu L. Chemotherapy and chemo-resistance in nasopharyngeal carcinoma. Eur J Med Chem. 2020;207:112758. Lee AWM, Ng WT, Chan JYW, Corry J, Mäkitie A, Mendenhall WM, Rinaldo A, Rodrigo JP, Saba NF, Strojan P, Suárez C, Vermorken JB, Yom SS, Ferlito A. Management of locally recurrent nasopharyngeal carcinoma. Cancer Treat Rev. 2019;79:101890. Li JY, Zhao Y, Gong S, Wang MM, Liu X, He QM, Li YQ, Huang SY, Qiao H, Tan XR, Ye ML, Zhu XH, He SW, Li Q, Liang YL, Chen KL, Huang SW, Li QJ, Ma J, Liu N. TRIM21 inhibits irradiation-induced mitochondrial DNA release and impairs antitumour immunity in nasopharyngeal carcinoma tumour models. Nat Commun. 2023;14(1):865. Tang LL, Chen YP, Chen CB, Chen MY, Chen NY, Chen XZ, Du XJ, Fang WF, Feng M, Gao J, Han F, He X, Hu CS, Hu DS, Hu GY, Jiang H, Jiang W, Jin F, Lang JY, Li JG, Lin SJ, Liu X, Liu QF, Ma L, Mai HQ, Qin JY, Shen LF, Sun Y, Wang PG, Wang RS, Wang RZ, Wang XS, Wang Y, Wu H, Xia YF, Xiao SW, Yang KY, Yi JL, Zhu XD, Ma J. The Chinese Society of Clinical Oncology (CSCO) clinical guidelines for the diagnosis and treatment of nasopharyngeal carcinoma. Cancer Commun (Lond). 2021;41(11):1195–227. Li T, Li F, Guo X, Hong C, Yu X, Wu B, Lian S, Song L, Tang J, Wen S, Gao K, Hao M, Cheng W, Su Y, Zhang S, Huang S, Fang M, Wang Y, Ng MH, Chen H, Luo W, Ge S, Zhang J, Xia N, Ji M. Anti-Epstein-Barr Virus BNLF2b for Mass Screening for Nasopharyngeal Cancer. N Engl J Med. 2023;389(9):808–19. Ge J, Wang J, Xiong F, Jiang X, Zhu K, Wang Y, Mo Y, Gong Z, Zhang S, He Y, Li X, Shi L, Guo C, Wang F, Zhou M, Xiang B, Li Y, Li G, Xiong W, Zeng Z. Epstein-Barr Virus-Encoded Circular RNA CircBART2.2 Promotes Immune Escape of Nasopharyngeal Carcinoma by Regulating PD-L1. Cancer Res. 2021;81(19):5074–88. Xu Y, Quan Z, Zhan Y, Wang H, Luo J, Wang W, Fan S. SSTR2 positively associates with EGFR and predicts poor prognosis in nasopharyngeal carcinoma. J Clin Pathol. 2024;77(12):829–34. Tao CJ, Zhang P, Chen YY, Chen M. High expression of WTAP is related to poor prognosis in nasopharyngeal carcinoma. Neoplasma. 2023;70(2):229–39. Damania B, Kenney SC, Raab-Traub N. Epstein-Barr virus: Biology and clinical disease. Cell. 2022;185(20):3652–70. Houen G, Trier NH. Epstein-Barr Virus and Systemic Autoimmune Diseases. Front Immunol. 2021;11:587380. Yang J, Liu Z, Zeng B, Hu G, Gan R. Epstein-Barr virus-associated gastric cancer: A distinct subtype. Cancer Lett. 2020;495:191–9. Aloisi F, Giovannoni G, Salvetti M. Epstein-Barr virus as a cause of multiple sclerosis: opportunities for prevention and therapy. Lancet Neurol. 2023;22(4):338–49. Zeng C, Qiao M, Chen Y, Xie H. EBV-positive glycoproteins associated with nasopharyngeal carcinoma. Pathol Res Pract. 2024;260:155427. Lee HM, Sia APE, Li L, Sathasivam HP, Chan MSA, Rajadurai P, Tsang CM, Tsao SW, Murray PG, Tao Q, Paterson IC, Yap LF. Monoamine oxidase A is down-regulated in EBV-associated nasopharyngeal carcinoma. Sci Rep. 2020;10(1):6115. Chan DC. Mitochondrial Dynamics and Its Involvement in Disease. Annu Rev Pathol. 2020;15:235–59. Song S, Gong S, Singh P, Lyu J, Bai Y. The interaction between mitochondria and oncoviruses. Biochim Biophys Acta Mol Basis Dis. 2018;1864(2):481–7. Xie L, Shi F, Li Y, Li W, Yu X, Zhao L, Zhou M, Hu J, Luo X, Tang M, Fan J, Zhou J, Gao Q, Wu W, Zhang X, Liao W, Bode AM, Cao Y. Drp1-dependent remodeling of mitochondrial morphology triggered by EBV-LMP1 increases cisplatin resistance. Signal Transduct Target Ther. 2020;5(1):56. Yu C, Xue B, Li J, Zhang Q. Tumor cell-derived exosome RNF126 affects the immune microenvironment and promotes nasopharyngeal carcinoma progression by regulating PTEN ubiquitination. Apoptosis. 2022;27(7–8):590–605. Dai Z, Zhong Y. Bioinformatics analysis based on ferroptosis-related lncRNAs: construction of a clinical prognostic model for nasopharyngeal carcinoma and correlation analysis. Transl Cancer Res. 2022;11(6):1665–77. Su J, Zhong G, Qin W, Zhou L, Ye J, Ye Y, Chen C, Liang P, Zhao W, Xiao X, Wen W, Luo W, Zhou X, Zhang Z, Cai Y, Li C. Integrating iron metabolism-related gene signature to evaluate prognosis and immune infiltration in nasopharyngeal carcinoma. Discov Oncol. 2024;15(1):112. Zhang L, Zhang X, Liu H, Yang C, Yu J, Zhao W, Guo J, Zhou B, Jiang N. MTFR2-dependent mitochondrial fission promotes HCC progression. J Transl Med. 2024;22(1):73. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. Gu Z, Hübschmann D. Make Interactive Complex Heatmaps in R. Bioinformatics. 2022;38(5):1460–2. Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7. Ramsay IS, Ma S, Fisher M, Loewy RL, Ragland JD, Niendam T, Carter CS, Vinogradov S. Model selection and prediction of outcomes in recent onset schizophrenia patients who undergo cognitive training. Schizophr Res Cogn. 2017;11:1–5. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. Zheng Y, Gao W, Zhang Q, Cheng X, Liu Y, Qi Z, Li T. Ferroptosis and Autophagy-Related Genes in the Pathogenesis of Ischemic Cardiomyopathy. Front Cardiovasc Med. 2022;9:906753. Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284–7. Liu P, Xu H, Shi Y, Deng L, Chen X. Potential Molecular Mechanisms of Plantain in the Treatment of Gout and Hyperuricemia Based on Network Pharmacology. Evid Based Complement Alternat Med. 2020;2020:3023127. Lei J, Qu T, Cha L, Tian L, Qiu F, Guo W, Cao J, Sun C, Zhou B. Clinicopathological characteristics of pheochromocytoma/paraganglioma and screening of prognostic markers. J Surg Oncol. 2023;128(4):510–8. Ramsay IS, Ma S, Fisher M, Loewy RL, Ragland JD, Niendam T, Carter CS, Vinogradov S. Model selection and prediction of outcomes in recent onset schizophrenia patients who undergo cognitive training. Schizophr Res Cogn. 2017;11:1–5. Ramsay IS, Ma S, Fisher M, Loewy RL, Ragland JD, Niendam T, Carter CS, Vinogradov S. Model selection and prediction of outcomes in recent onset schizophrenia patients who undergo cognitive training. Schizophr Res Cogn. 2017;11:1–5. Heagerty PJ, Lumley T, Pepe MS. Time-dependent ROC curves for censored survival data and a diagnostic marker. Biometrics. 2000;56(2):337–44. Gu Z, Hübschmann D. Make Interactive Complex Heatmaps in R. Bioinformatics. 2022;38(5):1460–2. Cheng Q, Chen X, Wu H, Du Y. Three hematologic/immune system-specific expressed genes are considered as the potential biomarkers for the diagnosis of early rheumatoid arthritis through bioinformatics analysis. J Transl Med. 2021;19(1):18. Sachs MC. plotROC: A Tool for Plotting ROC Curves. J Stat Softw. 2017;79:2. Blanche P, Dartigues JF, Jacqmin-Gadda H. Estimating and comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Stat Med. 2013;32(30):5381–97. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284–7. Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7. Charoentong P, Finotello F, Angelova M, Mayer C, Efremova M, Rieder D, Hackl H, Trajanoski Z. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell Rep. 2017;18(1):248–62. Geeleher P, Cox N, Huang RS. pRRophetic: an R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PLoS ONE. 2014;9(9):e107468. Gustavsson EK, Zhang D, Reynolds RH, Garcia-Ruiz S, Ryten M. ggtranscript: an R package for the visualization and interpretation of transcript isoforms using ggplot2. Bioinformatics. 2022;38(15):3844–6. Qi Y, Hu M, Han C, Wang J, Chen F, Guo H, She Y, Zhang M, Zhang J, Zhao Z, Xie H, Wang S, Chen M, Wang J, Zeng D. ARHGAP4 promotes leukemogenesis in acute myeloid leukemia by inhibiting DRAM1 signaling. Oncogene. 2023;42(34):2547–57. Fu MS, Pan SX, Cai XQ, Hu YX, Zhang WJ, Pan QC. Analysis of ARHGAP4 Expression With Colorectal Cancer Clinical Characteristics and Prognosis. Front Oncol. 2022;12:899837. Yao M, Gu Y, Yang Z, Zhong K, Chen Z. MEIS1 and its potential as a cancer therapeutic target (Review). Int J Mol Med. 2021;48(3):181. Li Y, Gan Y, Liu J, Li J, Zhou Z, Tian R, Sun R, Liu J, Xiao Q, Li Y, Lu P, Peng Y, Peng Y, Shu G, Yin G. Downregulation of MEIS1 mediated by ELFN1-AS1/EZH2/DNMT3a axis promotes tumorigenesis and oxaliplatin resistance in colorectal cancer. Signal Transduct Target Ther. 2022;7(1):87. Wu Z, Shih B, Macdonald J, Meunier D, Hogan K, Chintoan-Uta C, Gilhooley H, Hu T, Beltran M, Henderson NC, Sang HM, Stevens MP, McGrew MJ, Balic A. Development and function of chicken XCR1 + conventional dendritic cells. Front Immunol. 2023;14:1273661. Heger L, Hatscher L, Liang C, Lehmann CHK, Amon L, Lühr JJ, Kaszubowski T, Nzirorera R, Schaft N, Dörrie J, Irrgang P, Tenbusch M, Kunz M, Socher E, Autenrieth SE, Purbojo A, Sirbu H, Hartmann A, Alexiou C, Cesnjevar R, Dudziak D. XCR1 expression distinguishes human conventional dendritic cell type 1 with full effector functions from their immediate precursors. Proc Natl Acad Sci U S A. 2023;120(33):e2300343120. Sun Y, Liu Y, Chu H. Nasopharyngeal Carcinoma Subtype Discovery via Immune Cell Scores from Tumor Microenvironment. J Immunol Res. 2023;2023:2242577. Kalbasi A, Ribas A. Tumour-intrinsic resistance to immune checkpoint blockade. Nat Rev Immunol. 2020;20(1):25–39. Ma Z, Zhang W, Dong B, Xin Z, Ji Y, Su R, Shen K, Pan J, Wang Q, Xue W. Docetaxel remodels prostate cancer immune microenvironment and enhances checkpoint inhibitor-based immunotherapy. Theranostics. 2022;12(11):4965–79. Yadav S, Sharma A, Nayik GA, Cooper R, Bhardwaj G, Sohal HS, Mutreja V, Kaur R, Areche FO, AlOudat M, Shaikh AM, Kovács B, Mohamed Ahmed AE. Review of Shikonin and Derivatives: Isolation, Chemistry, Biosynthesis, Pharmacology and Toxicology. Front Pharmacol. 2022;13:905755. Liu T, Sun Q, Chen J, Wang F, Li B, Qin W, Ye Z, Hu F. A comparison of neoadjuvant chemotherapy with gemcitabine versus docetaxel plus cisplatin in locoregionally advanced nasopharyngeal carcinoma: a propensity score matching analysis. Cancer Manag Res. 2018;10:6237–45. Waldman D, Weintraub M, Freeman A, Neumann Y, Rechavi G, Toren A. Favorable early response of secondary chronic myeloid leukemia to imatinib. Am J Hematol. 2004;75(4):217–9. Zhang J, Zhou J, Xiao S. Shikonin inhibits growth, invasion and glycolysis of nasopharyngeal carcinoma cells through inactivating the phosphatidylinositol 3 kinase/AKT signal pathway. Anticancer Drugs. 2020;31(9):932–41. Ponnusamy A, Sinha S, Hyde GD, Borland SJ, Taylor RF, Pond E, Eyre HJ, Inkson CA, Gilmore A, Ashton N, Kalra PA, Canfield AE. FTI-277 inhibits smooth muscle cell calcification by up-regulating PI3K/Akt signaling and inhibiting apoptosis. PLoS ONE. 2018;13(4):e0196232. Lee KH, Koh M, Moon A. Farnesyl transferase inhibitor FTI-277 inhibits breast cell invasion and migration by blocking H-Ras activation. Oncol Lett. 2016;12(3):2222–6. Additional Declarations No competing interests reported. Supplementary Files 20251125SupplementaryInformation.zip Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8214331","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":564921973,"identity":"65e3caf7-f15a-47ba-89b7-47015257f6f9","order_by":0,"name":"Ping Li","email":"","orcid":"","institution":"Shenzhen University General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Ping","middleName":"","lastName":"Li","suffix":""},{"id":564921974,"identity":"595f5902-f8dd-4beb-9161-0e89d7dab852","order_by":1,"name":"Xin Xu","email":"","orcid":"","institution":"Southern Medical University Shenzhen Hospital","correspondingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"Xu","suffix":""},{"id":564921975,"identity":"37597fe1-cfa8-4404-8df3-9336781bfa31","order_by":2,"name":"Xueyu Zhang","email":"","orcid":"","institution":"Southern Medical University Shenzhen Hospital","correspondingAuthor":false,"prefix":"","firstName":"Xueyu","middleName":"","lastName":"Zhang","suffix":""},{"id":564921976,"identity":"f6193fc3-bfcd-4560-b668-b28444c34217","order_by":3,"name":"Huifen Xie","email":"","orcid":"","institution":"Shenzhen University General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Huifen","middleName":"","lastName":"Xie","suffix":""},{"id":564921977,"identity":"927715a2-a1a6-45ed-8f1b-07f2def4e2db","order_by":4,"name":"Cuirong Xiao","email":"","orcid":"","institution":"Southern Medical University Shenzhen Hospital","correspondingAuthor":false,"prefix":"","firstName":"Cuirong","middleName":"","lastName":"Xiao","suffix":""},{"id":564921978,"identity":"9cee68e1-b94b-45a3-a6b4-10ed1f01d731","order_by":5,"name":"Yinggui Yang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6klEQVRIiWNgGAWjYBADHjCZUCEhJ0+iljMWxoYNJNnF2FaRyHCAgCL5GTmGnwt+HZYx519+TeLhPIkExgbmh49u4NFicCPHWHpm32EeyxlvyiQSt0nksTOwGRvn4NMikWMgzdtzmMfgxpk0kJZixgYeNml8WoAOM/6N0DJHIrHhAAEtDDdyzKR5fgC1nG8/BlRPhBaDM8/KrHkb0oG28DBbJByTMDZsJuAX+fbkzbd5/ljbG5w//vDmj5o6OXn25oeP8TpMIMMAGB3NDAzAcICIMONTDgL8xx8wMPypgzJGwSgYBaNgFGABABKOTyzk5W8MAAAAAElFTkSuQmCC","orcid":"","institution":"Southern Medical University Shenzhen Hospital","correspondingAuthor":true,"prefix":"","firstName":"Yinggui","middleName":"","lastName":"Yang","suffix":""}],"badges":[],"createdAt":"2025-11-26 15:08:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8214331/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8214331/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99312687,"identity":"6f6ca6da-f8c6-4991-aed6-65821d4b9002","added_by":"auto","created_at":"2025-12-31 16:19:20","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2340872,"visible":true,"origin":"","legend":"","description":"","filename":"2025129TranscriptomicAnalysisofEpsteinBarrVirusandMitochondrialDynamicsRelatedGenesinNasopharyngealCarcinomaPrognosis.docx","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/f4bf8403d06b7790b18898db.docx"},{"id":98988361,"identity":"06565629-fa71-4fa5-84b8-aa56e353c9d3","added_by":"auto","created_at":"2025-12-25 09:31:00","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7760,"visible":true,"origin":"","legend":"","description":"","filename":"d06d036efc8e4ae9b799d81607d52247.json","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/e9a6505752a4d174000b3187.json"},{"id":98988364,"identity":"4f1d8ed9-df16-4252-a2b4-ec719db66eb1","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"zip","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5091990,"visible":true,"origin":"","legend":"","description":"","filename":"20251125SupplementaryInformation.zip","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/0c6a00026e368ec889cec719.zip"},{"id":98988366,"identity":"246a7f3d-c087-4450-82e3-ac53c1af949b","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":162324,"visible":true,"origin":"","legend":"","description":"","filename":"d06d036efc8e4ae9b799d81607d522471enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/4c15b394410a26cac741e815.xml"},{"id":98988371,"identity":"8441f158-98cb-4818-9f43-1810c9b3c3bc","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7191712,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/68197dc9df1a2687bf1e3048.jpeg"},{"id":98988368,"identity":"39994277-6ff9-4208-9d6f-d6328998911d","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6487348,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/50045100abb0a773c2ec19eb.jpeg"},{"id":98988386,"identity":"4aa7a7a0-dca6-4c00-9b73-44b0e5b69306","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6871856,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/6b476bda6ae4e1d152bff775.jpeg"},{"id":98988380,"identity":"b1cb1795-5c04-41c8-8b9d-34677836c5ac","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6885460,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/8ee34d474eb2e054aa488f55.jpeg"},{"id":98988376,"identity":"ae41d126-55dd-4817-980f-6a4eb3349207","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4770392,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/b84793d94edba486c1662dfd.jpeg"},{"id":99312117,"identity":"9b78b89e-fc42-45ad-b30a-6d34dc3f1b0f","added_by":"auto","created_at":"2025-12-31 16:18:09","extension":"jpeg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6734804,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/21ea1461711ff7e31996b11a.jpeg"},{"id":98988374,"identity":"6a23c0c5-5b7b-4cab-8054-bb11d33dc7bc","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"jpeg","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4530550,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/ae750348cf7d52229583a374.jpeg"},{"id":98988379,"identity":"4e951452-0844-484f-b5af-4de3e7bf4e38","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"jpeg","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7425628,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/0b0b221a7e6549afc0d82070.jpeg"},{"id":98988375,"identity":"998bba20-1089-4ea4-9b29-86a7da9fe816","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":102674,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/1f6832c613f7894e03875271.png"},{"id":98988378,"identity":"7158a89e-dbec-4ed5-a150-424d53b59b70","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":108040,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/2db9ac9f5fedaeaec351b331.png"},{"id":98988383,"identity":"0c6305ef-53eb-499f-b48e-9dc5e234a190","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":98077,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/bf7b7126e453ba3b7e994362.png"},{"id":98988370,"identity":"a4b65f45-877b-4630-820e-dedbb2181e4c","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":59069,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/e953f7f580677976b9a53ff7.png"},{"id":98988372,"identity":"eb08fb4c-60b6-41b5-9069-579d0e14525a","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":56116,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/14e311cc0693d47abe0b165c.png"},{"id":99312599,"identity":"a500c575-29df-4d66-b7be-9780774d16ac","added_by":"auto","created_at":"2025-12-31 16:19:10","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":113702,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/0a220c4f9a8237f530a88827.png"},{"id":98988384,"identity":"5a6587fe-66fe-428b-ae26-950a042aa5f0","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":48109,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/f464efed6850b6b83b3e3232.png"},{"id":98988382,"identity":"3e38ac44-80c8-4d8f-b62c-945f5296eab0","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":68259,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/677ae4c94fa5a85b7d580f8b.png"},{"id":98988385,"identity":"3e634ca5-7138-4e4f-9ef9-bbce1ba24bed","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"xml","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":161681,"visible":true,"origin":"","legend":"","description":"","filename":"d06d036efc8e4ae9b799d81607d522471structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/e41704031efaf29211fa2d92.xml"},{"id":99312569,"identity":"e511b79c-9b8f-42df-b792-f0ea57bdc2e6","added_by":"auto","created_at":"2025-12-31 16:19:07","extension":"html","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":174119,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/dd48cdc1c05407d86ec84d82.html"},{"id":98988357,"identity":"91c1d8e4-1a1a-4aad-89cc-d5675bd54e4f","added_by":"auto","created_at":"2025-12-25 09:31:00","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2401684,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eScreening and Sample Clustering of Differentially Expressed Genes \u003c/strong\u003e(A-B) Volcano plot and heatmap of differentially expressed genes. In Figure A, each point is a gene. Blue and red points are significantly differentially expressed genes: red for upregulated, blue for downregulated. Gray points are non-significant; top 10 genes with largest fold change (FC) are labeled. In Figure B, each small square is a gene; color indicates Z-score of gene expression after normalization. Higher expression is redder, lower is bluer. First row of color bars represents sample groups. Each row shows gene expression in different samples, each column shows expression of all differentially expressed genes in each sample. (C) K-M curve of MDRGs score. (D) Sample clustering and trait heatmap. Upper half shows sample clustering (branches = samples, y-axis = hierarchical clustering height). Lower half shows traits (corresponding to branches), colors indicate MDRGs score. (E) Scale-free soft threshold distribution. The x-axis of the graph represents the power value of the weight parameter. The y-axis of the left graph represents the scale-free fit index, that is, the signed R2. The y-axis of the right graph represents the mean of the adjacency function of all genes in the corresponding gene module. (F) Module clustering dendrogram. Genes are divided into various modules through hierarchical clustering, with different colors representing different modules. Gray is the default color for genes that cannot be classified into any module. (G) Heatmap of module and clinical trait correlation. The value in each square represents the correlation coefficient and significance P value between the module and the MDRGs score.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/75170c3ce363fd4d4f8a7a40.png"},{"id":99311974,"identity":"b936c6c7-a3f3-4a97-8c77-07a55cec8bcf","added_by":"auto","created_at":"2025-12-31 16:17:31","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2142703,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCandidate gene enrichment analysis and PPI network construction \u003c/strong\u003e(A) Venn diagram of candidate genes. (B) GO enrichment results of candidate genes. A represents the top 5 results in BP and are indicated by different colors. The genes in the outermost circle, indicated by different colors, are enriched in the corresponding BP entries of the same color. B represents the top 5 results in CC and are indicated by different colors. The genes in the outermost circle, indicated by different colors, are enriched in the corresponding CC entries of the same color. C represents the top 5 results in MF and are indicated by different colors. The genes in the outermost circle, indicated by different colors, are enriched in the corresponding MF entries of the same color. (C) KEGG pathway enrichment of candidate genes. (D) Protein-protein interaction network. The color and size of the nodes are labeled according to the Degree of the nodes.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/cb3d20ccb1b73afba0a01eb0.png"},{"id":98988359,"identity":"878f82ae-0f10-4a62-8aa7-9b6749d75414","added_by":"auto","created_at":"2025-12-25 09:31:00","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2031586,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCOX regression analysis and KM curve construction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Forest plot of univariate Cox regression analysis. The left side represents the genes and their corresponding P values and HR values. The orange squares on the right indicate HR values greater than 1, while the blue squares indicate HR values less than 1. The lines on both sides of the squares represent the 95% confidence interval of the HR value. (B) Forest plot of multivariate Cox regression results. The left side represents the genes and their corresponding P values and HR values. The orange squares on the right indicate HR values greater than 1, while the blue squares indicate HR values less than 1. The lines on both sides of the squares represent the 95% confidence interval of the HR value. (C) K-M curve of prognostic genes. (D) Risk curves of high-risk and low-risk groups in the training set. (E-G) K-M survival curves, ROC curves, and heat maps of prognostic genes in the training set. The vertical axis represents the survival rate, and the horizontal axis represents the overall survival time. The orange curve represents the high-risk group, and the cyan curve represents the low-risk group. The difference between the high-risk and low-risk groups is significant (p \u0026lt; 0.0001), showing a significant difference.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/f94360d11c22156f84bce26c.png"},{"id":99312657,"identity":"473440ac-1240-4593-8a77-996d9c844626","added_by":"auto","created_at":"2025-12-31 16:19:18","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1087902,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConstruction of predictive models for high-risk groups\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Risk curves of the high and low risk groups in TCGA-NB-INNER. (B-D) Survival curves, ROC curves and prognostic gene heat maps of the high and low risk groups in TCGA-NB-Inner. (E) Risk curves and scatter plots of the high and low risk groups in the GSE102349 dataset. (F-H)GSE102349 The survival curves, ROC curves and prognostic gene heat maps of the high and low-risk groups in the dataset. In the F map, the vertical axis represents survival rate and the horizontal axis represents overall survival time. Orange represents the high-risk group and green represents the low-risk group. The significant difference between the high-risk and low-risk groups was p = 0.03, showing a significant difference.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/5f44182b8912f1b298689fe3.png"},{"id":98988362,"identity":"72a512b6-c1c2-445c-bbfb-c683c91eb540","added_by":"auto","created_at":"2025-12-25 09:31:00","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":986792,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConstruction of independent post rain and Nomogram\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Independent prognosis - Forest plot of univariate cox results and PH test A-1 on the left represents the factor and the corresponding P value and HR value; The red square on the right indicates that the HR value is greater than 1, and the green square indicates that the HR value is less than 1. The line segments on both sides of the square represent the 95% confidence interval of the HR value.(B) Independent prognosis - Forest plot of multivariate cox results. (C) The histogram for predicting survival rate can be divided into three parts: the upper part is the contribution degree of the prognostic gene to the outcome variable (the size of the regression coefficient), and each value level of each factor is assigned a score. The Points in the figure represent the individual scores corresponding to each clinical factor under different values. The total score in the middle is the sum of each individual clinical factor. The lower part is to compare the total point obtained through calculation to determine the death probability of the patient in one or two or three years. (D-E) Nomogram correction curve and ROC curve.(F) The vertical axis represents the Net Benefit (NB) after subtracting the disadvantages from the advantages. The different curved and slanted lines in the figure represent different clinical diagnostic models respectively (see the legend for identification).\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/1f9bb07b0b7bb0c791b3125a.png"},{"id":98988377,"identity":"85fc92b8-01c3-469d-b851-fb9e830b0e92","added_by":"auto","created_at":"2025-12-25 09:31:01","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":2728284,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGSEA enrichment analysis and immune infiltration analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) GSEA Enrichment Ridge Chart This chart is divided into three parts. The first part is the line graph of gene Enrichment Score, with each gene under this gene on the horizontal axis and the corresponding Running ES on the vertical axis. There is a peak in the line graph. This peak is the Enrichment score of this gene set, and the genes before the peak are the core genes under this gene set. The second part is hit. The barcode-like lines in the middle represent the positions of the genes in the gene set in the background genes. Each vertical line represents the genes under that pathway, sorted from left to right according to the expression level. The third part is the rank value distribution map of all genes, a list of gene correlations calculated by the spearman method, corresponding to the title on the vertical axis. (B) The horizontal axis of the degree of immune cell infiltration in the high and low risk groups represents the sample, and the vertical axis represents the overall proportion of different immune cells. (C) Box plot of the difference in the proportion of immune cells in the high and low risk groups * represents P \u0026lt; 0.05, ** represents P \u0026lt; 0.01, *** represents P \u0026lt; 0.001, and **** represents P \u0026lt; 0.0001. (D) The scatter plot of the correlation between prognostic genes and differential immune cells has the expression value of prognostic genes on the vertical axis and the infiltration of immune cells on the horizontal axis. The absence of an asterisk (*) indicates insignificance.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/399d27c567185509562b0aca.png"},{"id":99312052,"identity":"1319ae1f-7003-43ff-a40a-8a45e619f6b3","added_by":"auto","created_at":"2025-12-31 16:17:48","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":771844,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAnalysis of differences between high-risk and low-risk populations \u003c/strong\u003e(A) Differences in immune pathways among patients with high and low risk groups. (B) Differences in the expression of ICI genes among patients with high and low risk groups.\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/bc5533adb0fe2d53cf92b6cc.png"},{"id":99311968,"identity":"c63b3eb3-9f0c-4975-a707-433573c84a00","added_by":"auto","created_at":"2025-12-31 16:17:31","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1333793,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDifferential analysis and RT qPCR expression level validation experiment of drugs in high and low-risk groups\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) IC50 differences of 10 drugs between high and low-risk groups. (B) Expression of prognostic genes in TCGA-NPC. (C) RT-qPCR check the expression of prognostic genes in nomal(NP) and tumor tissues (NPC).\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/8c51590ef60a888cd04d5937.png"},{"id":102745415,"identity":"2fdfd589-5fcb-426c-9062-df1f8a891f1c","added_by":"auto","created_at":"2026-02-16 08:49:29","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":14508219,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/4ccf5f8c-8b8e-49ac-b3c8-f0221dbeff2b.pdf"},{"id":99312065,"identity":"8ff9e781-af4c-42be-a5d9-a8a3d63b92bb","added_by":"auto","created_at":"2025-12-31 16:17:49","extension":"zip","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":5091990,"visible":true,"origin":"","legend":"","description":"","filename":"20251125SupplementaryInformation.zip","url":"https://assets-eu.researchsquare.com/files/rs-8214331/v1/ca4b668411d2a3b0cc563378.zip"}],"financialInterests":"No competing interests reported.","formattedTitle":"Transcriptomic Analysis of Epstein-Barr Virus and Mitochondrial Dynamics-Related Genes in Nasopharyngeal Carcinoma Prognosis","fulltext":[{"header":"1. Background","content":"\u003cp\u003eNasopharyngeal carcinoma (NPC) is a malignant tumor that arises from the mucosal epithelium of the nasopharynx. Its causes are complex, involving Epstein-Barr virus (EBV) infection, genetic factors, and environmental factors. Notable long-term consumption of pickled foods containing nitrosamine compounds, exposure to chemical carcinogens in the environment, and smoking may all increase the risk of developing the disease. Early-stage NPC typically presents with subtle, easily overlooked symptoms. However, as the disease progresses, patients may gradually exhibit a series of symptoms such as nasal congestion, epistaxis, tinnitus, hearing loss, headache, and cervical lymph node enlargement[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Currently, the accurate diagnosis of NPC relies on nasopharyngoscopy, pathological biopsy, CT and MRI imaging examinations, as well as EBV serological tests. In terms of treatment strategies, the treatment of NPC includes radiotherapy, chemotherapy, surgical operations, and immunotherapy[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Clinical research data indicate that NPC is particularly sensitive to radiotherapy in the early stage, with significant treatment effects; however, as the disease advances to the late stage, a single treatment method is often insufficient to achieve the desired results. At this point, a comprehensive treatment strategy including radiotherapy, chemotherapy, surgery, and immunotherapy is required to control the disease progression as much as possible and prolong the patient's survival time[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Given that early-stage NPC symptoms are easily overlooked, regular physical examinations and screenings are crucial for early detection and treatment. This not only helps to promptly identify early signs of NPC but, more importantly, it can secure valuable treatment windows for patients, thereby improving treatment outcomes and long-term prognosis. Therefore, the research and promotion of effective early screening methods and related prognostic approaches are of great importance for the prevention and intervention of NPC. Recent studies have shown that overexpression of EGFR, SSTR2, and WTAP is related to poor prognosis and cancer progression in NPC, and they are all promising prognostic biomarkers for NPC[\u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eEBV is a ubiquitous virus associated with many different human malignancies and autoimmune diseases. As an extremely adaptable human herpesvirus, EBV infects almost everyone at least once in their lifetime[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. EBV infection has been implicated in multiple malignancies, including gastric cancer and NPC[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Emerging evidence from diverse disciplines indicates EBV's pathogenic role in both oncogenesis and multiple sclerosis development[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Mechanistically, viral glycoproteins (TNC, FN1, GFBP3) mediate malignant transformation of epithelial cells through molecular interactions[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Research has found that EBV also plays a core role in mediating the tumor suppressor effect of MAOA, and the loss of MAOA may be an important step in the pathogenesis of NPC[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. The above-mentioned research indicates that the association between EBV and NPC offers a new perspective for understanding pathological conditions such as NPC. In-depth exploration of the specific mechanisms and interactions between the two will facilitate the development of new treatment strategies.\u003c/p\u003e \u003cp\u003eMitochondria are important organelles that provide energy and maintain the metabolism of most eukaryotic cells. Mitochondrial dynamics\u0026mdash;including fusion, fission, and mitophagy\u0026mdash;are essential for maintaining optimal mitochondrial function in energy metabolism[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Increasing evidence indicates that mitochondrial fission is highly associated with the occurrence of diseases and cancers, and mitochondrial fission has been proven to be a key mechanism in carcinogenesis mediated by oncogenic viruses[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Additionally, studies have identified that mitochondrial Drp1 is crucial for the development and outcome of EBV latent membrane protein 1 (EBV-LMP1) positive NPC tumors[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. The above studies indicate that dysregulation of genes involved in mitochondrial dynamics (particularly fission) may also be strongly associated with NPC initiation and progression, and a comprehensive analysis of the specific mechanisms and interactions between the two will also help us develop new therapeutic strategies.\u003c/p\u003e \u003cp\u003eThis study identified prognosis-related genes associated with EBV and mitochondrial dynamics through a series of bioinformatics approaches using NPC-related data from public databases. Additionally, the molecular mechanisms of the prognosis genes were further examined via enrichment, immune infiltration, and regulatory network analyses, providing new references for the clinical diagnosis and treatment of NPC.\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Data source\u003c/h2\u003e \u003cp\u003eThe head and neck cancer dataset (TCGA-HNSC) was chosen from the Cancer Genome Atlas Program (TCGA) database, and the clinical information, progression-free survival (PFS) survival information and RNA sequencing (RNA-seq) data of samples from 6 sites (Larynx, Floor of mouth, Tonsil, Base of tongue, Oropharynx, Hypopharynx) were extracted as NPC analysis data, recorded as the training set (high-throughput sequencing), covering 243 NPC tumor tissue samples and 44 normal nasopharyngeal epithelial tissue samples (samples with no survival information were excluded)[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Specimens of tissue taken from 243 NPC patients, containing survival data, were split into two categories following a 7:3 ratio (162: 81). The group that accounted for 70% was used as the internal training set, while the other group that accounted for 30% was used as the internal testing set. The GSE102349 dataset (GPL11154) obtained from the Gene Expression Omnibus (GEO) database through sequencing mode on chip was used as the validation set and contained 88 NPC tissue samples with PFS[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eA search by the molecular signatures database (MsigDB) and removal of overlapping genes resulted in 1,697 EVB-related genes (EBVRGs). The 23 mitochondrial dynamic-related genes (MDRGs) from the reported literature[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Acquisition of differentially expressed genes (DEGs)\u003c/h2\u003e \u003cp\u003eDifferential expression analysis was performed using DESeq2 (v 1.38.0)[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]to identify DEGs between NPC and normal specimens in the training cohort, with significance thresholds set at adjusted p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 and |log\u003csub\u003e2\u003c/sub\u003eFC | \u0026gt; 0.5. Visualization was achieved through ggplot2 (v 3.4.4) for generating volcano plots highlighting the top 10 most differentially expressed genes, supplemented by ComplexHeatmap (v 2.14.0)[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] for displaying expression patterns of the 40 most significantly up- and down-regulated genes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Weighted gene co-expression network analysis (WGCNA) and acquisition of candidate genes\u003c/h2\u003e \u003cp\u003eThe ssGSEA algorithm implemented in GSVA (v 1.50.0)[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e] was employed to quantify enrichment scores for 23 MDRGs across 243 training set tumor specimens. Using an optimized cutoff value, specimens were stratified into high- and low-score cohorts, followed by Kaplan-Meier progression-free survival analysis performed with the survminer package (v 0.4.9)[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eMDRG scores were used as traits, and WGCNA analysis was performed by \u0026ldquo;WGCNA\u0026rdquo; (v 1.71)[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. To pinpoint and eliminate anomalies, a hierarchical grouping of all samples was employed, utilizing the Euclidean distance from the expression profiles of the samples. Following this, we adjusted R\u003csup\u003e2\u003c/sup\u003e to 0.85, mean connectivity to 0 for selecting the best soft threshold (β). The process of determining gene adjacency resulted in the computation of gene similarity, leading to the derivation of a gene dissimilarity coefficient, thereby forming a hierarchical gene clustering tree. By the dynamic tree cutting algorithm's criteria, the least number of genes per module was established at 100, and the gene module. To explore the gene modules highly correlated with the MDRGs score, the Pearson function of the WGCNA was utilized to analyze the correlation between each module and the MDRGs score. A heatmap of the relationship between each module and the MDRGs score was drawn to screen for key modules. The conditions for selecting the modules were: non-grey modules, a \u003cem\u003ep\u003c/em\u003e-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05, and the absolute value of the correlation with the MDRGs score (|cor|)\u0026thinsp;\u0026gt;\u0026thinsp;0.4. Finally, the key module genes were obtained. To obtain DEGs associated with EBVRGs and MDRGs in NPC patients, the \u0026ldquo;ggvenn\u0026rdquo; package (v 0.1.10)[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]was utilized to perform intersection operations on the DEGs, EBVRGs, and the key module genes, and the resulting genes were the candidate genes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Pathway analysis and protein-protein interaction (PPI) network\u003c/h2\u003e \u003cp\u003eFor investigate the functions performed by the candidate genes more deeply, this study conducted enrichment analysis by Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases with the help of the \u0026ldquo;clusterProfiler\u0026rdquo; (v 4.7.1.003)[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e](p.adj\u0026thinsp;\u0026lt;\u0026thinsp;0.05) and the org.Hs.eg.db human gene annotation package, and the results show the top 5 pathways. Next, the protein interactions between the candidate genes were further explored. In this study, the PPI network of the candidate genes was built using the STRING database and then visualized using Cytoscape[\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Screening process of prognostic genes and risk modelling\u003c/h2\u003e \u003cp\u003eTo identify survival-related genes associated with the prognosis of NPC, univariate Cox regression analysis was implemented in the internal training set by the \u0026ldquo;survival\u0026rdquo; (v 3.5-7)[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]to screen candidate genes (cut-off value: p\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Then, the cox_zph function was used to perform the proportional hazards (PH) assumption test (p\u0026thinsp;\u0026gt;\u0026thinsp;0.05) for survival-related genes, and the genes that passed the PH assumption test were defined as candidate prognostic genes and visualized by drawing a forest plot. Subsequently, a multivariate Cox regression model was built for these candidate prognostic genes. Then, the multivariate regression model was adjusted by applying the stepwise regression function step, with the direction parameter configured as \u0026ldquo;both\u0026rdquo;. After that, an optimal combination of a multivariate Cox regression model was chosen, and the genes within this optimal model served as the prognostic genes.\u003c/p\u003e \u003cp\u003eTo determine whether prognostic gene expression could influence the survival of NPC patients, the NPC in the internal training set were categorized into groups with high/low prognostic gene expression by the median value of prognostic gene expression. KM curve analysis was performed by the \u0026ldquo;survminer\u0026rdquo; (v 0.4.9)[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]package (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), and the survival discrepancy between the high and low prognostic gene expression groups was compared using the Log-rank test (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Assessment and validation of risk models\u003c/h2\u003e \u003cp\u003eTo assess the prognostic value, in this study, the risk value of each patient was computed in the internal training set, and the risk scores were computed applying the formula:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\text{R}\\text{i}\\text{s}\\text{k}\\text{s}\\text{c}\\text{o}\\text{r}\\text{e}\\:=\\:\\sum\\:_{\\text{i}\\:=\\:1}^{\\text{n}}\\text{c}\\text{o}\\text{e}\\text{f}\\left({\\text{g}\\text{e}\\text{n}\\text{e}}_{\\text{i}}\\right)\\ast\\:\\text{e}\\text{x}\\text{p}\\text{r}\\left({\\text{g}\\text{e}\\text{n}\\text{e}}_{\\text{i}}\\right)$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eCofe: coefficient, Expr: expression\u003c/p\u003e \u003cp\u003eNPCs were classified into high and low-risk groups by the optimal threshold of the risk indicator. Immediately thereafter, KM survival was performed by the \u0026ldquo;survminer\u0026rdquo; (v 0.4.9)[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e] for the two risk groups in the internal training set. Subsequently, using 1/2/3 years as the survival time node, receiver operating characteristic (ROC) curves were drawn, and the area under the curve (AUC) was calculated using the \u0026ldquo;survivalROC\u0026rdquo; (v 1.0.3)[\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Heatmaps were plotted with the help of \u0026ldquo;pheatmap\u0026rdquo; (v 1.0.12)[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]to show the expression of prognostic genes.\u003c/p\u003e \u003cp\u003eThe NPC prognostic model was tested by the internal testing set as well as the validation set GSE102349. Again, patients were grouped with the median risk score as the threshold. Then, risk curves and KM curves were plotted. Subsequently, ROC curves were plotted, and AUC values were determined over periods of 1, 2, and 3 years. In addition, prognostic gene expression heat maps were mapped for the two risk groups.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7 Association of risk stratification with clinicopathological parameters\u003c/h2\u003e \u003cp\u003eThe ggpubr package (v 0.6.0)[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]was employed to assess and graphically present variations in risk scores across demographic (age, gender) and pathological (tumor stage, T/N classification) subgroups within the training cohort.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.8 Nomogram construction and evaluation\u003c/h2\u003e \u003cp\u003eThe risk score and clinical characteristics of the internal training set were included in the univariate Cox regression analysis to obtain independent prognostic factors. First, a univariate Cox analysis was carried out. Secondly, the PH assumption test and independent prognostic factors were obtained by multivariate Cox analysis. To assess the reliability of the nomogram, by this nomogram prediction model, the nomogram's calibration curve was developed using the R package \u0026ldquo;rms\u0026rdquo; (v 6.7-1)[\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e], and similarly, the ROC curve for the nomogram was formulated with the assistance of the R package \u0026ldquo;timeROC\u0026rdquo; (v 0.4)[\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. In addition, this study utilized decision curve analysis (DCA) to plot decision curves for 1/2/3 years, which was used to evaluate the clinical utility of the nomogram prediction model.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e2.9 Enrichment analysis and immune infiltration analysis\u003c/h2\u003e \u003cp\u003eTo further clarify the functional enrichment in the two risk groups, the \u0026ldquo;DESeq2\u0026rdquo; (v 1.42.0)[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]was applied to analyze the differential expression, and log\u003csub\u003e2\u003c/sub\u003eFC was calculated and ranked. The c2.cp.kegg. The obtained by the MSigDB was applied as the background gene set, and the \u0026ldquo;clusterProfiler\u0026rdquo; (v 4.7.1.003)[\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]was applied to perform gene set enrichment analysis (GSEA). The significance threshold for GSEA results was set at |NES| \u0026gt; 1 and p.adjust\u0026thinsp;\u0026lt;\u0026thinsp;0.05, and the results were visualized through GSEA ridges generated with the \u0026ldquo;GseaVis\u0026rdquo; package.(v 0.0.5).\u003c/p\u003e \u003cp\u003eTo grasp the immune cell infiltration status in the two risk groups, the ssGSEA algorithm in the \u0026ldquo;GSVA\u0026rdquo; (v 1.50.0)[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]was applied to evaluate the infiltration status of 28 immune cells by the internal training set, and the histograms were drawn with the help of the \u0026ldquo;ggplot2\u0026rdquo; (v 0.1.10)[\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]to show the proportion of various immune cells infiltrated. Next, the Wilcoxon test was applied, and the immune cells with remarkable differences (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) were selected. Finally, Spearman correlation analysis was applied to prognostic genes and immune cells with significant differences, and the results were plotted using the R package \u0026ldquo;ggplot2\u0026rdquo; (v 0.1.10).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e2.10 Immunotherapeutic response and immune checkpoints\u003c/h2\u003e \u003cp\u003eTo compare the immune pathway and immune checkpoint differences in the two risk groups, the ssGSEA score analysis of immune-related pathways was first performed by the \u0026ldquo;GSVA\u0026rdquo; package (v 1.50.0) in the internal training set samples, and the pathway differences were compared (Wilcoxon, p\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Then, the differences in immune checkpoint-related genes in the two risk groups were compared in the internal training set (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e2.11 Drug sensitivity analysis and expression analysis of prognostic genes\u003c/h2\u003e \u003cp\u003eTo investigate the difference in sensitivity to anti-tumor drugs among two risk groups of NPC patients, the IC\u003csub\u003e50\u003c/sub\u003e of 138 chemotherapy/targeted therapy drugs for each patient in the internal training set was assessed using the \u0026ldquo;pRRophetic\u0026rdquo; package (v 0.5)[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e] based on the Genomics of Cancer Drug Sensitivity (GDSC) database. Next, the Wilcoxon test was applied to compare the difference in IC50 among drugs in the two risk groups (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), and the 10 most significant drugs were plotted on a box plot with the help of the \u0026ldquo;ggplot2\u0026rdquo; package (v 3.4.4)[\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. Besides, to explore the expression status of prognostic genes in the training dataset, the Wilcoxon test was applied to dissect the expression variation in prognostic gene expression among NPC samples and normal samples, which was demonstrated by violin plots (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e2.12 Statistical analysis\u003c/h2\u003e \u003cp\u003eStatistical computations were implemented in R (v 4.2.2), employing the Wilcoxon rank-sum test for intergroup comparisons with a predefined significance threshold (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n\u003ch2\u003e\u003cstrong\u003e3.1 Acquisition of 4,636 DEGs and 2,952 MDRGs\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eBy comprehensively analyzing the DEGs and module genes of NPC in the training set, potential molecular mechanisms and biological processes were revealed, thus deepening our understanding of NPC pathogenesis. We analyzed the DEGs in the training set and identified 4,636 DEGs, of which 2,415 were up-regulated and 2,221 down-regulated for expression in disease samples (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eA-\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eB). Analysis of MDRGs scores indicated a remarkable difference in patient survival among high/low scoring groups (p\u0026thinsp;=\u0026thinsp;0.011) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eC). Hierarchical cluster analysis of NPC samples in the training set revealed no outliers (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eD). The scale-free networks created by setting R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.85 and determined the soft threshold \u0026beta; as 7 (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eE). Then, the genes were divided into modules based on soft thresholds, and 12 co-expression modules were selected (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eF). In addition, the correlation analysis among co-expression modules and MDRGs scores showed that the MEbrown (cor = -0.51, p\u0026thinsp;=\u0026thinsp;4e-17, 2,082), MEmagenta (cor = -0.43, p\u0026thinsp;=\u0026thinsp;1e-12, 231), and MEblack (cor = -0.43, p\u0026thinsp;=\u0026thinsp;2e-12, 639) modules were the key modules, and finally 2,952 MDRGs were obtained as key model_genes (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eG).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n\u003ch2\u003e3.2 Detection and functional description of candidate genes\u003c/h2\u003e\n\u003cp\u003eThe overlap of DEGs, EBVRGs, and key model genes was used to obtain 108 candidate genes (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eA). Furthermore, GO enrichment results showed that candidate genes were overrepresented in biological processes, including regulation of immune effector process and humoral immune response. In cellular components, they were mainly involved in endoplasmic reticulum lumen, and in molecular functions, they were mainly represented by oxidoreductase activity (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eB). KEGG enrichment analysis enriched a total of 22 pathways, such as endoplasmic reticulum lumen (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eC). Additionally, the PPI network showed that there were 80 nodes and 168 edges in the network, indicating that the 80 candidate genes have some interactions (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eD). Among them, the CCL2 and CXCL8 genes had the greatest interaction.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n\u003ch2\u003e3.3 Three prognostic genes were identified\u003c/h2\u003e\n\u003cp\u003eUnivariate Cox regression analysis and PH assumption test identified 35 candidate prognostic genes (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eA, FigS1). Of these, 6 genes (CTTN, DHCR7, GALNS, TMEM158, UGP2, and HMGA2) were risk factors (Hazard Ratio (HR)\u0026thinsp;\u0026gt;\u0026thinsp;1), and the remaining 29 genes were protective factors (HR\u0026thinsp;\u0026lt;\u0026thinsp;1). Subsequently, multivariate Cox regression analysis selected three genes (MEIS1, XCR1, ARHGAP4), and these three genes were regarded as prognostic genes (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eB). KM curve showed that the survival rate was higher in the high-expression group of all three prognostic genes (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eC). The NPC were categorized by the optimal threshold of the risk score (riskscore\u0026thinsp;=\u0026thinsp;1.2728), and the survival of the high-risk group was remarkably lower (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eD). In the training set, there was a remarkable survival difference among the two risk groups (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eE). Besides, in the ROC analysis for the 1-, 2-, and 3-year risk models, the AUC values exceeded 0.7, reinforcing its effectiveness in forecasting NPC patient survival rates (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eF). Heat map results indicated that the MEIS1, XCR1, and ARHGAP4 genes were lower in the high-risk group (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eG). The prognostic value of prognostic models was verified in the internal training set (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eA-D) as well as in the validation set (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eE-H), and the results obtained were identical to the training set.\u0026nbsp;\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003ctable id=\"Tab1\" border=\"1\"\u003e\u003ccaption\u003e\n\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n\u003cdiv class=\"CaptionContent\"\u003e\n\u003cp\u003e\u003cstrong\u003eResults of multifactorial Cox analysis\u003c/strong\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003c/caption\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eid\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003ecoef\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eHR\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eHR.95L\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003eHR.95H\u003c/p\u003e\n\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003epvalue\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eMEIS1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.22231542\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.80066278\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.59444329\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.07842226\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.14344205\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eXCR1\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.26684995\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.76578797\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.55833958\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e1.0503128\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.09783532\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003eARHGAP4\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.31963429\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.72641464\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.58177756\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.90701028\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"char\" char=\".\"\u003e\n\u003cp\u003e0.00477955\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\n\u003ch2\u003e3.4 Prognostic nomogram for NPC based on independent prognostic factors\u003c/h2\u003e\n\u003cp\u003eIndependent prognostic analyses were of vital importance for establishing robust clinical decision support systems. The risk score and N stage were significant (HR\u0026thinsp;\u0026gt;\u0026thinsp;1), and PH assumption was satisfied in monovariate Cox regression analysis (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eA). Polyvariate Cox regression analysis indicated that risk score was remarkable, so risk score was taken as an independent prognostic factor(Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eB). Nomogram founded on independent prognostic factor was constructed to display the possibility of survival of NPC at 1, 2 and 3 years (1, 2, and 3-year survival probability 28.2%, 44.7%, 49.6%, respectively). The scoring relationship showed that 3-year was greater than 2-year was greater than 1-year, indicating that NPC patients had the highest risk at 3-year (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eC). Calibration curves and ROC curves further validated the effectiveness of the nomogram. The slope (approximating 1) of the calibration curve (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eD) confirmed the nomogram's exceptional predictive accuracy. ROC analysis (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eE) further validated its discriminative power (AUC\u0026thinsp;\u0026gt;\u0026thinsp;0.7), with the nomogram exhibiting superior true positive detection compared to individual predictors. At the same time, the DCA curve showed a better clinical utility of the nomogram (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eF).\u003c/p\u003e\n\u003cp\u003eCorrelation analysis among risk scores and clinical characteristics indicated that risk scores grouped by age and gender did not show significant differences between groups, and again did not differ significantly in stage; however, there were remarkable differences in risk scores among stages T2-T4, and between stages N0-N3, and N1-N3 (FigS2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFigS2 Risk score analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A-B) Correlation between risk score and age, gender, where the horizontal axis represents gender group (A) and gender (B), and the vertical axis represents risk score. (C) Correlation between risk score and Stage, where the horizontal axis represents stage. The vertical axis represents the risk score. (D-E) The correlation between the risk score and different clinical characteristics, where the horizontal axis represents the gender N stage (D) and T stage (E), and the vertical axis represents the risk score.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\n\u003ch2\u003e3.5 GSEA and comprehensive study of immune infiltration\u003c/h2\u003e\n\u003cp\u003eGSEA was conducted to delve deeper into the possible processes of prognostic gene differences among patients across two risk groups, with findings presented in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eA. The outcomes of the study suggested that the two risk groups showed significant enrichment in pathways (such as ribosome and primary immunodeficiency). Furthermore, the quantity of 28 immune-infiltrating cells in the two risk groups was shown in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eB. The results indicated that the infiltration degree of central memory CD4 T cells was more abundant in the high-risk group. Next, analysis of immune cell differences showed that 14 immune cells exhibited remarkable differences among the two risk groups (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (such as activated B cells). Among them, the infiltration level of the other 13 differential immune cells was greater in the low-risk group, and neutrophils was greater in the high-risk group (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eC). The correlation analysis among prognostic genes and differential cells showed activated B cell was remarkably positively realted to ARHGAP4 (cor\u0026thinsp;=\u0026thinsp;0.45), MEIS1 (cor\u0026thinsp;=\u0026thinsp;0.40), and XCR1 (cor\u0026thinsp;=\u0026thinsp;0.54) (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eD). The differential distribution of these immune cells provided important clues for further research on risk assessment and treatment strategies for NPC.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\n\u003ch2\u003e3.6 Immune checkpoint and therapy response analysis in two risk groups\u003c/h2\u003e\n\u003cp\u003eThe analysis of immune therapy response and immune checkpoints showed that there were 10 immune related pathways (such as Cytolytic-activity) with differences among the two risk groups (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003eA). In addition, among the 45 immune checkpoint genes, 38 (such as LAG3 and IDO2) indicated remarkable differences among the two risk groups (TIM3, VISTA, OX40 had no expression data in the training set) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003eB). These findings offered valuable insights into NPC's immune responses and checkpoints, potentially aiding in developing targeted immunotherapies for different risk groups.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\n\u003ch2\u003e3.7 Drug sensitivity analysis and prognostic gene expression level validation\u003c/h2\u003e\n\u003cp\u003eTo explore the difference in sensitivity to anti-tumor drugs of NPC, drug sensitivity analysis was applied in two risk groups. The IC50 of docetaxel, shikonin, FTI.277, and Imatinib was lower in the high-risk group (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003eA). By observing the efficacy of these drugs in patients belonging to different risk groups, we can have a better comprehension of the responsiveness of various patients to chemotherapeutic drugs, thereby offering a foundation for individualized treatment.Ultimately,three prognostic genes exhibited significant differences between tumor and normal samples. Amidst them, ARHGAP4 expression was remarkably up-regulated in tumor samples, while MEIS1 and XCR1 were remarkably down-regulated (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003eB). More importantly, Our experimental results also show the consistent trend of ARHGAP4, MEIS1 and XCR1 in NP and NPC tissues (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003eC).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eDiscussion on the risk model based on prognostic genes: By combining existing literature and our analysis results, we discuss three prognostic genes (ARHGAP4, MEIS1, and XCR1) and the risk model constructed based on them.\u003c/p\u003e \u003cp\u003eARHGAP4, also known as Rho GTPase activating protein 4, plays a role in various biological processes. Studies have found that ARHGAP4 is a significant Rho family GTPase-activating protein and is closely related to the occurrence and development of certain tumors. Studies have demonstrated elevated ARHGAP4 levels in AML cases, correlating with unfavorable clinical outcomes[\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. This molecular alteration potentially contributes to oncogenesis, progression, and metastatic dissemination in various malignancies. Its specific mechanism of action may involve regulating cytoskeletal reorganization, affecting cell movement and adhesion, etc. ARHGAP4 is highly expressed in colorectal cancer (CRC), and overexpression of ARHGAP4 is associated with a poor prognosis[\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. Thus, it is speculated that ARHGAP4 may be a promising indicator for the prognosis of malignant tumors. Myeloid ecotropic viral integration site 1 homolog(MEIS1), also known as bone marrow ecotropic viral integration site 1, is a transcription factor that plays an important role in normal development and disease processes. MEIS1 was first identified as upregulated in myeloid leukemia cell lines. Notably, as a transcriptional regulator, MEIS1 contributes to leukemogenesis and the progression of various malignancies, including solid tumors[\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eEmerging evidence indicates a significant reduction of MEIS1 transcriptional activity in colorectal carcinoma, which strongly predicts diminished overall survival[\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. The attenuated MEIS1 levels appear to facilitate neoplastic advancement through multiple mechanisms, including modulation of critical cellular processes (proliferation, differentiation, and programmed cell death) and crosstalk with various oncogenic pathways.\u003c/p\u003e \u003cp\u003eXCR1, also known as XC motif chemokine receptor 1, is mainly expressed on specific types of immune cells, such as dendritic cells[\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. Moreover, XCR1 is a marker of terminally differentiated cDC1 and can mediate the antiviral effector function of human cDC1[\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. Thus, in tumors, downregulation of XCR1 expression may impair the antitumor immune response in the tumor microenvironment.\u003c/p\u003e \u003cp\u003eThe above results indicate that the three genes ARHGAP4, MEIS1, and XCR1 play critical tumor-suppressive roles in the initiation and progression of NPC. Our findings showed that high expression of these genes is closely associated with better overall survival in NPC patients, whereas their expression is lower in high-risk patients, further confirming their potential in tumor prognosis assessment. This result suggests that MEIS1, XCR1, and ARHGAP4 may affect the clinical progression of NPC by regulating tumor cell proliferation, metastasis, and immune escape. Although their specific mechanisms require further validation, our findings provide strong evidence for these genes as potential prognostic biomarkers for NPC and may provide new ideas for the formulation of individualized treatment strategies. The risk scoring model based on these genes is expected to assist clinicians in more accurately identifying high-risk NPC patients, thereby enabling them to formulate more personalized treatment regimens. Simultaneously, the expression levels of these genes also provide potential directions for the development of new targeted therapies. For instance, by regulating the expression of these genes or their downstream signaling pathways, it may be possible to effectively inhibit tumor progression and bring better treatment outcomes to patients.\u003c/p\u003e \u003cp\u003eFurthermore, immune checkpoint inhibition (ICB) therapy has emerged as a promising treatment strategy and has demonstrated substantial efficacy for treating various cancers, including NPC[\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]; Immune checkpoint inhibitors enhance anti-tumor immunity by reversing T cell exhaustion, thereby improving the targeting and efficacy of treatment[\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. Our data further show that 14 immune cell subsets exhibit significant differences between high- and low-risk groups\u0026mdash;consistent with the expression patterns of ARHGAP4, MEIS1, and XCR1\u0026mdash;providing a basis for early high-risk assessment of NPC and screening of high-risk patients through the detection of changes in these immune cell subpopulations.\u003c/p\u003e \u003cp\u003eInvestigating the association between these prognostic genes and drug sensitivity can, on one hand, explore the quantitative relationship between their expression levels and drug responses. On the other hand, it can analyze the impact of different prognostic gene mutation statuses on drug efficacy to clarify their relationship[\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]. Our research results show that the IC50 values of docetaxel, shikonin, FTI-277, and imatinib in the high-risk group are lower, which is consistent with existing studies. Docetaxel and imatinib are commonly used clinical anti-tumor drugs, and more clinical studies are currently exploring the therapeutic effects of docetaxel and imatinib on NPC. The results indicate that both have certain advantages in the treatment of NPC[\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e]. Shikonin has also been proven to have anti-tumor effects. Studies have shown that shikonin can inhibit the growth of NPC cells by inactivating the phosphatidylinositol 3-kinase/AKT signaling pathway[\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e].FTI-277 is a CAAX peptidomimetic that has been shown to inhibit breast cancer cell invasion and migration by blocking H-Ras activation[\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e]. Notably, our study is the first to link FTI-277 to NPC, highlighting its potential therapeutic value for NPC.\u003c/p\u003e \u003cp\u003eThis study identified three prognostic genes (ARHGAP4, MEIS1, and XCR1) in nasopharyngeal carcinoma (NPC) and constructed a risk model using these genes via bioinformatics approaches, providing new insights into potential treatment strategies for NPC. Although the transcriptomic data analysis approaches used in this study yielded valuable insights, there are still several technical limitations. For example, the algorithms, software, and models we used may introduce biases or inaccuracies when processing complex data\u0026mdash;particularly during data preprocessing, normalization, and result interpretation\u0026mdash;all of which may compromise the analytical accuracy. Additionally, although our study identified potential prognostic biomarkers for NPC, the translation of these biomarkers into clinical practice is hindered by multiple challenges. On one hand, existing biomarker detection methods have limitations in sensitivity, specificity, and reproducibility, which limit their feasibility for NPC clinical practice. Although some biomarkers have shown potential in early diagnosis or treatment monitoring, their actual impact on clinical decision-making still needs further validation, especially considering the interaction of multiple factors and individual differences.\u003c/p\u003e \u003cp\u003eIn future research, we will strive to overcome these challenges. Firstly, we plan to adopt more sophisticated and efficient data analysis approaches\u0026mdash;including deep learning and multi-omics data integration approaches\u0026mdash;to improve the reliability and predictive performance of our risk model. Secondly, for the clinical translation of biomarkers, we will further optimize detection methods for our identified biomarkers and validate their clinical utility in real-world clinical settings through multi-center clinical trials to ensure their substantial role in clinical decision-making. Ultimately, in line with the concept of precision medicine, we will explore individualized diagnosis and treatment strategies to better serve patients with our research findings.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAbbreviations\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eFull name\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eNPC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eNasopharyngeal carcinoma\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eEBV\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eEpstein-Barr virus\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eWGCNA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eweighted gene co-expression network analysis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eROC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ereceiver operating characteristic\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGSEA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003egene set enrichment analysis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eEBV-LMP1\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eEBV latent membrane protein 1\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGEO\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGene Expression Omnibus\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePFS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eprogression-free survival\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMsigDB\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003emolecular signatures database\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eEBVRGs\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eEVB-related genes\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMDRGs\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003emitochondrial dynamic-related genes\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eDEGs\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003edifferentially expressed genes\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eWGCNA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eWeighted gene co-expression network analysis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePPI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eprotein-protein interaction\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGO\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGene ontology\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eKEGG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eKyoto Encyclopedia of Genes and Genomes\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAUC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003earea under the curve\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eDCA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003edecision curve analysis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGDSC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGenomics of Cancer Drug Sensitivity\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCRC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ecolorectal cancer\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMEIS1\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eMyeloid ecotropic viral integration site 1 homolog\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eICB\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eimmune checkpoint inhibition\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e \u003cp\u003eAll experimental protocols were approved by the Ethics Committee of Affiliated Shenzhen Hospital of Southern Medical University. Written informed consent was obtained from all the participants. All methods were carried out in accordance with Declaration of Helsinki.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eClinical trial number\u003c/h2\u003e \u003cp\u003enot applicable.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eConsent for publication\u003c/h2\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eCompeting interests\u003c/h2\u003e \u003cp\u003eAll the authors declare that they have no conflicts of interest.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis work was supported by The Science and Technology Planning Project of Shenzhen, China (JCYJ20210324130801004), the Postdoctoral Research Foundation of Shenzhen (UN-KC-BHKY202205), and Research Foundation of Shenzhen Hospital of Southern Medical University (CNGZRJJPY202008, UN-KJ-KY200024-YYPT, PT2020GZR07, 22H3ATF05).\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003ePing Li, and Xueyu Zhang contributed equally to this work. Yinggui Yang and Cuirong Xiao conceived and designed the research; Ping Li, and Xueyu Zhang developed the methodology and acquired the data; Xin Xu analysed and interpreted the data; Yinggui Yang and Cuirong Xiao revised and approved the manuscript. All authors approved the submission of the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets analyzed during this study are publicly available in The Cancer Genome Atlas (TCGA) database [http://xena.ucsc.edu/] (TCGA-HNSC) and the Gene Expression Omnibus(GEO) repository [https://www.ncbi.nlm.nih.govgeo/] under accession number GSE102349. All accession numbers and associated files have been fully released and are accessible for verification.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eGuan S, Wei J, Huang L, Wu L. Chemotherapy and chemo-resistance in nasopharyngeal carcinoma. Eur J Med Chem. 2020;207:112758.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee AWM, Ng WT, Chan JYW, Corry J, M\u0026auml;kitie A, Mendenhall WM, Rinaldo A, Rodrigo JP, Saba NF, Strojan P, Su\u0026aacute;rez C, Vermorken JB, Yom SS, Ferlito A. Management of locally recurrent nasopharyngeal carcinoma. Cancer Treat Rev. 2019;79:101890.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi JY, Zhao Y, Gong S, Wang MM, Liu X, He QM, Li YQ, Huang SY, Qiao H, Tan XR, Ye ML, Zhu XH, He SW, Li Q, Liang YL, Chen KL, Huang SW, Li QJ, Ma J, Liu N. TRIM21 inhibits irradiation-induced mitochondrial DNA release and impairs antitumour immunity in nasopharyngeal carcinoma tumour models. Nat Commun. 2023;14(1):865.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang LL, Chen YP, Chen CB, Chen MY, Chen NY, Chen XZ, Du XJ, Fang WF, Feng M, Gao J, Han F, He X, Hu CS, Hu DS, Hu GY, Jiang H, Jiang W, Jin F, Lang JY, Li JG, Lin SJ, Liu X, Liu QF, Ma L, Mai HQ, Qin JY, Shen LF, Sun Y, Wang PG, Wang RS, Wang RZ, Wang XS, Wang Y, Wu H, Xia YF, Xiao SW, Yang KY, Yi JL, Zhu XD, Ma J. The Chinese Society of Clinical Oncology (CSCO) clinical guidelines for the diagnosis and treatment of nasopharyngeal carcinoma. Cancer Commun (Lond). 2021;41(11):1195\u0026ndash;227.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi T, Li F, Guo X, Hong C, Yu X, Wu B, Lian S, Song L, Tang J, Wen S, Gao K, Hao M, Cheng W, Su Y, Zhang S, Huang S, Fang M, Wang Y, Ng MH, Chen H, Luo W, Ge S, Zhang J, Xia N, Ji M. Anti-Epstein-Barr Virus BNLF2b for Mass Screening for Nasopharyngeal Cancer. N Engl J Med. 2023;389(9):808\u0026ndash;19.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGe J, Wang J, Xiong F, Jiang X, Zhu K, Wang Y, Mo Y, Gong Z, Zhang S, He Y, Li X, Shi L, Guo C, Wang F, Zhou M, Xiang B, Li Y, Li G, Xiong W, Zeng Z. Epstein-Barr Virus-Encoded Circular RNA CircBART2.2 Promotes Immune Escape of Nasopharyngeal Carcinoma by Regulating PD-L1. Cancer Res. 2021;81(19):5074\u0026ndash;88.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu Y, Quan Z, Zhan Y, Wang H, Luo J, Wang W, Fan S. SSTR2 positively associates with EGFR and predicts poor prognosis in nasopharyngeal carcinoma. J Clin Pathol. 2024;77(12):829\u0026ndash;34.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTao CJ, Zhang P, Chen YY, Chen M. High expression of WTAP is related to poor prognosis in nasopharyngeal carcinoma. Neoplasma. 2023;70(2):229\u0026ndash;39.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDamania B, Kenney SC, Raab-Traub N. Epstein-Barr virus: Biology and clinical disease. Cell. 2022;185(20):3652\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHouen G, Trier NH. Epstein-Barr Virus and Systemic Autoimmune Diseases. Front Immunol. 2021;11:587380.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang J, Liu Z, Zeng B, Hu G, Gan R. Epstein-Barr virus-associated gastric cancer: A distinct subtype. Cancer Lett. 2020;495:191\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAloisi F, Giovannoni G, Salvetti M. Epstein-Barr virus as a cause of multiple sclerosis: opportunities for prevention and therapy. Lancet Neurol. 2023;22(4):338\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeng C, Qiao M, Chen Y, Xie H. EBV-positive glycoproteins associated with nasopharyngeal carcinoma. Pathol Res Pract. 2024;260:155427.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee HM, Sia APE, Li L, Sathasivam HP, Chan MSA, Rajadurai P, Tsang CM, Tsao SW, Murray PG, Tao Q, Paterson IC, Yap LF. Monoamine oxidase A is down-regulated in EBV-associated nasopharyngeal carcinoma. Sci Rep. 2020;10(1):6115.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChan DC. Mitochondrial Dynamics and Its Involvement in Disease. Annu Rev Pathol. 2020;15:235\u0026ndash;59.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSong S, Gong S, Singh P, Lyu J, Bai Y. The interaction between mitochondria and oncoviruses. Biochim Biophys Acta Mol Basis Dis. 2018;1864(2):481\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXie L, Shi F, Li Y, Li W, Yu X, Zhao L, Zhou M, Hu J, Luo X, Tang M, Fan J, Zhou J, Gao Q, Wu W, Zhang X, Liao W, Bode AM, Cao Y. Drp1-dependent remodeling of mitochondrial morphology triggered by EBV-LMP1 increases cisplatin resistance. Signal Transduct Target Ther. 2020;5(1):56.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu C, Xue B, Li J, Zhang Q. Tumor cell-derived exosome RNF126 affects the immune microenvironment and promotes nasopharyngeal carcinoma progression by regulating PTEN ubiquitination. Apoptosis. 2022;27(7\u0026ndash;8):590\u0026ndash;605.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDai Z, Zhong Y. Bioinformatics analysis based on ferroptosis-related lncRNAs: construction of a clinical prognostic model for nasopharyngeal carcinoma and correlation analysis. Transl Cancer Res. 2022;11(6):1665\u0026ndash;77.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu J, Zhong G, Qin W, Zhou L, Ye J, Ye Y, Chen C, Liang P, Zhao W, Xiao X, Wen W, Luo W, Zhou X, Zhang Z, Cai Y, Li C. Integrating iron metabolism-related gene signature to evaluate prognosis and immune infiltration in nasopharyngeal carcinoma. Discov Oncol. 2024;15(1):112.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang L, Zhang X, Liu H, Yang C, Yu J, Zhao W, Guo J, Zhou B, Jiang N. MTFR2-dependent mitochondrial fission promotes HCC progression. J Transl Med. 2024;22(1):73.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLove MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu Z, H\u0026uuml;bschmann D. Make Interactive Complex Heatmaps in R. Bioinformatics. 2022;38(5):1460\u0026ndash;2.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eH\u0026auml;nzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamsay IS, Ma S, Fisher M, Loewy RL, Ragland JD, Niendam T, Carter CS, Vinogradov S. Model selection and prediction of outcomes in recent onset schizophrenia patients who undergo cognitive training. Schizophr Res Cogn. 2017;11:1\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLangfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng Y, Gao W, Zhang Q, Cheng X, Liu Y, Qi Z, Li T. Ferroptosis and Autophagy-Related Genes in the Pathogenesis of Ischemic Cardiomyopathy. Front Cardiovasc Med. 2022;9:906753.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu P, Xu H, Shi Y, Deng L, Chen X. Potential Molecular Mechanisms of Plantain in the Treatment of Gout and Hyperuricemia Based on Network Pharmacology. Evid Based Complement Alternat Med. 2020;2020:3023127.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLei J, Qu T, Cha L, Tian L, Qiu F, Guo W, Cao J, Sun C, Zhou B. Clinicopathological characteristics of pheochromocytoma/paraganglioma and screening of prognostic markers. J Surg Oncol. 2023;128(4):510\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamsay IS, Ma S, Fisher M, Loewy RL, Ragland JD, Niendam T, Carter CS, Vinogradov S. Model selection and prediction of outcomes in recent onset schizophrenia patients who undergo cognitive training. Schizophr Res Cogn. 2017;11:1\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamsay IS, Ma S, Fisher M, Loewy RL, Ragland JD, Niendam T, Carter CS, Vinogradov S. Model selection and prediction of outcomes in recent onset schizophrenia patients who undergo cognitive training. Schizophr Res Cogn. 2017;11:1\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeagerty PJ, Lumley T, Pepe MS. Time-dependent ROC curves for censored survival data and a diagnostic marker. Biometrics. 2000;56(2):337\u0026ndash;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu Z, H\u0026uuml;bschmann D. Make Interactive Complex Heatmaps in R. Bioinformatics. 2022;38(5):1460\u0026ndash;2.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng Q, Chen X, Wu H, Du Y. Three hematologic/immune system-specific expressed genes are considered as the potential biomarkers for the diagnosis of early rheumatoid arthritis through bioinformatics analysis. J Transl Med. 2021;19(1):18.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSachs MC. plotROC: A Tool for Plotting ROC Curves. J Stat Softw. 2017;79:2.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlanche P, Dartigues JF, Jacqmin-Gadda H. Estimating and comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Stat Med. 2013;32(30):5381\u0026ndash;97.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLove MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eH\u0026auml;nzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCharoentong P, Finotello F, Angelova M, Mayer C, Efremova M, Rieder D, Hackl H, Trajanoski Z. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell Rep. 2017;18(1):248\u0026ndash;62.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeeleher P, Cox N, Huang RS. pRRophetic: an R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PLoS ONE. 2014;9(9):e107468.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGustavsson EK, Zhang D, Reynolds RH, Garcia-Ruiz S, Ryten M. ggtranscript: an R package for the visualization and interpretation of transcript isoforms using ggplot2. Bioinformatics. 2022;38(15):3844\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQi Y, Hu M, Han C, Wang J, Chen F, Guo H, She Y, Zhang M, Zhang J, Zhao Z, Xie H, Wang S, Chen M, Wang J, Zeng D. ARHGAP4 promotes leukemogenesis in acute myeloid leukemia by inhibiting DRAM1 signaling. Oncogene. 2023;42(34):2547\u0026ndash;57.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFu MS, Pan SX, Cai XQ, Hu YX, Zhang WJ, Pan QC. Analysis of ARHGAP4 Expression With Colorectal Cancer Clinical Characteristics and Prognosis. Front Oncol. 2022;12:899837.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYao M, Gu Y, Yang Z, Zhong K, Chen Z. MEIS1 and its potential as a cancer therapeutic target (Review). Int J Mol Med. 2021;48(3):181.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Y, Gan Y, Liu J, Li J, Zhou Z, Tian R, Sun R, Liu J, Xiao Q, Li Y, Lu P, Peng Y, Peng Y, Shu G, Yin G. Downregulation of MEIS1 mediated by ELFN1-AS1/EZH2/DNMT3a axis promotes tumorigenesis and oxaliplatin resistance in colorectal cancer. Signal Transduct Target Ther. 2022;7(1):87.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu Z, Shih B, Macdonald J, Meunier D, Hogan K, Chintoan-Uta C, Gilhooley H, Hu T, Beltran M, Henderson NC, Sang HM, Stevens MP, McGrew MJ, Balic A. Development and function of chicken XCR1\u0026thinsp;+\u0026thinsp;conventional dendritic cells. Front Immunol. 2023;14:1273661.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeger L, Hatscher L, Liang C, Lehmann CHK, Amon L, L\u0026uuml;hr JJ, Kaszubowski T, Nzirorera R, Schaft N, D\u0026ouml;rrie J, Irrgang P, Tenbusch M, Kunz M, Socher E, Autenrieth SE, Purbojo A, Sirbu H, Hartmann A, Alexiou C, Cesnjevar R, Dudziak D. XCR1 expression distinguishes human conventional dendritic cell type 1 with full effector functions from their immediate precursors. Proc Natl Acad Sci U S A. 2023;120(33):e2300343120.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun Y, Liu Y, Chu H. Nasopharyngeal Carcinoma Subtype Discovery via Immune Cell Scores from Tumor Microenvironment. J Immunol Res. 2023;2023:2242577.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKalbasi A, Ribas A. Tumour-intrinsic resistance to immune checkpoint blockade. Nat Rev Immunol. 2020;20(1):25\u0026ndash;39.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMa Z, Zhang W, Dong B, Xin Z, Ji Y, Su R, Shen K, Pan J, Wang Q, Xue W. Docetaxel remodels prostate cancer immune microenvironment and enhances checkpoint inhibitor-based immunotherapy. Theranostics. 2022;12(11):4965\u0026ndash;79.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYadav S, Sharma A, Nayik GA, Cooper R, Bhardwaj G, Sohal HS, Mutreja V, Kaur R, Areche FO, AlOudat M, Shaikh AM, Kov\u0026aacute;cs B, Mohamed Ahmed AE. Review of Shikonin and Derivatives: Isolation, Chemistry, Biosynthesis, Pharmacology and Toxicology. Front Pharmacol. 2022;13:905755.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu T, Sun Q, Chen J, Wang F, Li B, Qin W, Ye Z, Hu F. A comparison of neoadjuvant chemotherapy with gemcitabine versus docetaxel plus cisplatin in locoregionally advanced nasopharyngeal carcinoma: a propensity score matching analysis. Cancer Manag Res. 2018;10:6237\u0026ndash;45.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWaldman D, Weintraub M, Freeman A, Neumann Y, Rechavi G, Toren A. Favorable early response of secondary chronic myeloid leukemia to imatinib. Am J Hematol. 2004;75(4):217\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang J, Zhou J, Xiao S. Shikonin inhibits growth, invasion and glycolysis of nasopharyngeal carcinoma cells through inactivating the phosphatidylinositol 3 kinase/AKT signal pathway. Anticancer Drugs. 2020;31(9):932\u0026ndash;41.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePonnusamy A, Sinha S, Hyde GD, Borland SJ, Taylor RF, Pond E, Eyre HJ, Inkson CA, Gilmore A, Ashton N, Kalra PA, Canfield AE. FTI-277 inhibits smooth muscle cell calcification by up-regulating PI3K/Akt signaling and inhibiting apoptosis. PLoS ONE. 2018;13(4):e0196232.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee KH, Koh M, Moon A. Farnesyl transferase inhibitor FTI-277 inhibits breast cell invasion and migration by blocking H-Ras activation. Oncol Lett. 2016;12(3):2222\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Nasopharyngeal carcinoma, Prognostic genes, Risk model, Epstein-Barr virus","lastPublishedDoi":"10.21203/rs.3.rs-8214331/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8214331/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eNasopharyngeal carcinoma (NPC) is a kind of malignant tumor. The Epstein-Barr virus (EBV) and mitochondrial dynamics may be related to NPC. However, the mechanisms of mitochondrial dynamics and EBV in NPC need to be further explored.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThis study obtained transcriptomic data from public databases, identified NPC-related prognostic genes through univariate Cox regression analysis and other methods, and subsequently constructed a risk model and a nomogram. Furthermore, based on the prognostic genes, gene set enrichment analysis (GSEA), immune infiltration analysis, and drug sensitivity analysis were performed, and the expression trends of the prognostic genes were verified.All experimental protocols were approved by the Ethics Committee of Affiliated Shenzhen Hospital of Southern Medical University.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eThis study identified ARHGAP4, MEIS1, and XCR1 as prognostic genes for NPC, and the constructed risk model exhibited good predictive performance. Furthermore, through GSEA, it was found that the two risk groups were differentially enriched in pathways related to ribosomes and other pathways; meanwhile, immune cell infiltration analysis also showed significant differences and correlations. In addition, differences in the sensitivity to chemotherapeutic drugs such as docetaxel were detected, and the prognostic gene ARHGAP4 was up-regulated in tumor samples, while MEIS1 and XCR1 were down-regulated.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eThe research pinpointed three predictive genes (ARHGAP4, MEIS1 and XCR1) and utilized them in developing a risk model, providing new insights into potential therapeutic strategies for NPC.\u003c/p\u003e","manuscriptTitle":"Transcriptomic Analysis of Epstein-Barr Virus and Mitochondrial Dynamics-Related Genes in Nasopharyngeal Carcinoma Prognosis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-25 09:30:56","doi":"10.21203/rs.3.rs-8214331/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"550b1d12-6cb8-4b56-8420-c5010f7bcfd8","owner":[],"postedDate":"December 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-11T12:42:58+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-25 09:30:56","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8214331","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8214331","identity":"rs-8214331","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}