Machine learning developed a macrophage signature for predicting prognosis, immune infiltration and immunotherapy features in head and neck squamous cell carcinoma

Machine learning developed a macrophage signature for predicting prognosis, immune infiltration and immunotherapy features in head and neck squamous cell carcinoma | 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 Article Machine learning developed a macrophage signature for predicting prognosis, immune infiltration and immunotherapy features in head and neck squamous cell carcinoma Yao Wang, Ya‐Kui Mou, Wan-Chen Liu, Han‐Rui Wang, Xiao-Yu Song, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4219358/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 22 Aug, 2024 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Macrophages played an important role in the progression and treatment of cancer. Nevertheless, there is a limited amount of research that has comprehensively elucidated the characteristics of macrophages associated genes in head and neck squamous cell carcinoma (HNSCC). We employed weighted gene co-expression network analysis (WGCNA) to identify macrophage-related genes (MRGs) and classify patients with HNSCC into two distinct subtypes. A macrophage-related risk signature (MRS) model, comprising nine genes: IGF2BP2, PPP1R14C, SLC7A5, KRT9, RAC2, NTN4, CTLA4, APOC1 , and CYP27A1 , was formulated by integrating 101 machine learning algorithm combinations. We observed lower overall survival (OS) in the high-risk group and the high-risk group showed elevated expression levels in most of the differentially expressed immune checkpoint and human leukocyte antigen (HLA) genes, suggesting a strong immune evasion capacity in these tumors. Correspondingly, TIDE score positively correlated with risk score, implying that high-risk tumors may resist immunotherapy more effectively. At the single-cell level, we noted macrophages in the TME predominantly stalled in the G2/M phase, potentially hindering epithelial-mesenchymal transition and playing a crucial role in the inhibition of tumor progression. Additionally, we validated MRS gene expression levels using RT-qPCR and immunohistochemistry (IHC). The current study constructed a novel MRS for HNSCC, which could serve as an indicator for predicting the prognosis, immune infiltration and immunotherapy benefits for HNSCC patients. Biological sciences/Computational biology and bioinformatics/Machine learning Biological sciences/Biochemistry Biological sciences/Biological techniques Biological sciences/Cancer Biological sciences/Computational biology and bioinformatics Health sciences/Biomarkers HNSCC Macrophage Prognostic model Risk score Immunotherapy Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 1. Introduction Head and neck squamous cell carcinoma (HNSCC) is the sixth most prevalent cancer worldwide, accounting for over 95% of all head and neck tumors [ 1 ]. While the TNM classification system is the standard for prognosis and treatment determination, patient outcomes vary significantly even within the same TNM stages [ 2 ]. This variability underscores the need for new biomarkers for an accurate survival prognosis and novel treatment targets to improve patient outcomes. The tumor microenvironment (TME) plays a crucial role in HNSCC progression by influencing recurrence, metastasis, and drug resistance [ 3 ]. Tumor-associated macrophages (TAMs), predominantly in the TME, either promote or hinder tumor growth depending on their M1 or M2 polarization [ 4 ]. The dynamic balance between these macrophage types is critical for cancer progression, with a higher M2/M1 ratio often indicating advanced tumor stages [ 5 ]. In this study, we explored macrophage-related genes (MRGs) by clustering HNSCC subtypes based on macrophage infiltration. We developed a macrophage-related risk-signature (MRS) model to predict patient outcomes and make diagnostic and therapeutic decisions. This study extended the examination of gene mutations, immune infiltration, and drug sensitivity, offering insights into the role of macrophages in HNSCC at the single-cell level. This research aimed to enhance targeted diagnosis and treatment strategies, fostering personalized patient care for HNSCC. 2. Materials and methods 2.1 Data collection and collation We accessed The Cancer Genome Atlas (TCGA) and GSE65858 datasets in the Gene Expression Omnibus (GEO) database to gather clinicopathological information, gene expression, and genomic mutation data of HNSCC patients [ 6 ]. Single-cell RNA sequencing (scRNA-seq) profiles from nine HNSCC samples, comprising of 1,4087 cells, were retrieved from the GSE182227 database [ 7 ]. The IMvigor210 dataset helped assess immunotherapy efficacy predictions by our risk model [ 8 ]. A total of 484 and 270 patients with HNSCC were collected from TCGA and GEO databases, respectively. Figure 1 shows a flowchart of the design of this study. 2.2 Macrophage co-expression network construction To estimate immune cell infiltration, HNSCC samples from the TCGA database were analyzed using the TIMER algorithm, while the CIBERSORT method was applied to patients in the GSE65858 dataset [ 9 , 10 ]. We employed the weighted gene co-expression network analysis (WGCNA) algorithm to create a scale-free network by grouping highly correlated genes into modules [ 11 ]. Pearson’s test was used to ascertain gene significance (GS) and module membership (MM), linking module eigengenes (MEs) with macrophages. 2.3 Acquisition of DEGs related to macrophage We employed the “DESeq2” R package to detect differentially expressed genes (DEGs) by contrasting HNSCC samples with paired healthy samples in the TCGA dataset [ 12 ]. DEGs were selected based on the criteria |log2FC| > 1 and FDRs < 0.05. We then cross-referenced these DEGs with genes related to macrophage modules, identifying 194 MRGs. Subsequent analyses focused on exploring the enriched cellular pathways of these MRGs, utilizing resources such as the Kyoto Encyclopedia of Genes and Genomes (KEGG), Molecular Signatures Database (MSigDB), and Reactome database. 2.4 Consensus Clustering Analysis Using the “ConsensusClusterPlus” R package, we clustered HNSCC subtypes based on intersected MRGs [ 13 ]. Cluster numbers were determined by analyzing the consensus cumulative distribution function (CDF), with a focus on significant delta area reductions. The subtypes were visualized using 3D principal component analysis (PCA). 2.5 Functional Enrichment Analysis We assessed the biological functions of prognostic genes using the “clusterProfiler” R package, analyzing Gene Ontology (GO) and KEGG pathways. Variations in gene enrichment were examined via gene set variation analysis (GSVA) using “GSVA” packages [ 14 ]. The Metascape database provided additional insights into GO/KEGG enrichment [ 15 ]. 2.6 Immune Landscape Construction The “estimate” R package calculated stromal and immune scores, estimation scores, and tumor purity. Immune pathway activities were analyzed, and human leukocyte antigen (HLA) and immune checkpoint gene expression were compared between the groups. 2.7 Tumor Mutation Burden (TMB) Calculation TMB was calculated from somatic mutation data on the GDC website ( https://portal.gdc.cancer.gov ), identifying genes with high mutation frequencies and visualizing them using “maftools” package [ 16 ]. TMB levels in MRG-related subgroups and risk groups were compared using the Wilcoxon test. 2.8 Immunotherapy Efficacy Prediction Immunophenoscore (IPS) and Tumor Immune Dysfunction and Exclusion (TIDE) scores have been used to predict the efficacy, which were obtained from TCIA ( https://tcia.at/home ) and TIDE ( http://tide.dfci.harvard.edu ), respectively [ 17 , 18 ]. The TISMO platform ( http://tismo.cistrome.org ) was used to evaluate the predictive capacity of hub genes for immunotherapy responses in syngeneic mouse immunotherapy cohorts and tumor cell lines. 2.9 MRS Model Construction A consensus MRS model was developed using ten machine learning techniques, integrating 101 algorithm permutations for optimal accuracy and stability [ 19 ]. The TCGA and GEO patients were divided into training and testing cohorts, and Harrell’s concordance index (C-index) was used to evaluate the accuracy of the model. 2.10 MRS Performance Assessment The “survminer” and “timeROC” R packages compared survival outcomes and assessed the MRS’s predictive ability in HNSCC, determining the optimal “cut-off” value for risk groups. 2.11 Clinical Prognostic Model Construction and Validation Cox regression analyses evaluated the risk score as an independent factor in the TCGA and GEO cohorts. The “regplot” package developed a nomogram incorporating patient characteristics for survival assessment. 2.12 PPI Network and Hub Gene Identification The “Cytoscape” software, aided by the “CytoHubba” plugin, constructed a PPI network for DEGs and identified hub genes [ 20 ]. 2.13 Chemotherapeutic Drug Sensitivity Assessment The “pRRophetic” and “ggplot2” packages analyzed chemotherapeutic drug sensitivity in HNSCC, comparing IC50 values between risk groups [ 21 ]. 2.14 scRNA-seq Analysis of HNSCC The “Seruat” package processed scRNA-seq data, identifying distinct cell clusters visualized using UMAP. The “TriCycle” algorithm analyzed cell cycle correlations, and the “irGSEA” method explored functional pathways [ 22 , 23 ]. 2.15 RT-qPCR and Immunohistochemistry Staining analysis for MRS-related genes We used VeZol Reagent (Vazyme, China) to process the collected 14 pairs of HNSCC tumor tissues and their adjacent normal tissues to extract total RNA, which was then reverse-transcribed into complementary DNA using HiScript QRT SuperMix for qPCR (Vazyme, China). RT-qPCR was performed on an FTC-3000P system (Funglyn Biotech, Canada) using qPCR SYBR Green Master Mix (Vazyme, China) as a fluorescent dye. GAPDH was used as an internal reference in the current research. The experiments were conducted separately and in duplicates. Immunohistochemistry (IHC) data were retrieved from the Human Protein Atlas (HPA, http://www.proteinatlas.org ) to evaluate the protein expression levels of MRS-related genes. 2.16 Statistical Analysis Statistical analyses were performed using R software (version 4.1.0, https://www.r-project.org/ ). Continuous variables were compared using the Student’s t-test and Wilcoxon rank-sum test. Survival curves were constructed using the Kaplan-Meier algorithm, with the Pearson and Spearman methods for correlation analysis. Statistical significance was set at a P-value = 0.05. 3 Results 3.1 Macrophage infiltration into HNSCC: hub module identification and enrichment analysis In the TCGA-HNSCC cohort, the blue module was significantly correlated with macrophages ( R 2 = 0.59, P = 5e-57) (Fig. 2 A). In the GSE146771 cohort, the M0, M1, and M2 macrophages had the strongest correlation with the red, tan, and turquoise modules, respectively (M0: R 2 = 0.4, P = 8e-12, M1: R 2 = 0.59, P = 4e-26, M2: R 2 = 0.35, P = 6e-09) (Fig. 2 B). And the correlation values between macrophage module membership and gene significance were shown in Fig. 2 C-F. Subsequently, we conducted a paired difference analysis of cancerous and adjacent tissues from patients with HNSCC in the TCGA database (Fig. 2 G). A total of 194 MRGs were identified by intersecting the hub modules and DEGs in the TCGA dataset ( Fig. 2 H ) . A functional enrichment analysis of the 194 MRGs revealed that their biological activities predominantly involved cellular immunity. Different databases emphasized different aspects of the immune system, such as “cytokine-cytokine receptor interaction” at the center of the KEGG database, while “inflammatory response” and “allograft rejection” were the focus of the MisgDB database, and maintenance of immune function and cytokine signaling transmission were the primary concerns of the Reactome database (Fig. 2 I-K). Meaning that the function of MRGs is mainly related to the immune inflammatory response of cells. 3.2 Construction and prognostic analysis of molecular subtypes of MRGs in HNSCC patients To delve deeper into the heterogeneity of macrophages among HNSCC tumors, unsupervised consensus analysis was used to classify the patients in TCGA-HNSCC into two distinct subtypes according to the 194 MRGs (Fig. 3 A-D). Patients in cluster 2 had considerably superior overall survival (OS) than that in cluster 1 (Fig. 3 E). We presented distribution differences between clusters 1 and 2, in terms of clinicopathological features, and simultaneously visualized the top ten genes with upregulated and downregulated levels between the two clusters (Fig. 3 F). GSVA revealed notable disparities between the two clusters in multiple functional pathways, and almost all immune and inflammatory response pathways were enriched in cluster 2 (Fig. 3 G). Finally, we performed KEGG and GO enrichment analyses of the DEGs between clusters 1 and 2 (Fig. 3 H-I). KEGG analysis suggested that the main pathways were involved in viral protein interactions with cytokine and cytokine receptors. GO analysis revealed that the DEGs were mostly enriched in the BP functional set, namely, leukocyte cell-cell adhesion and regulation of leukocyte proliferation, and were associated with the MF functional set. These findings indicated that MRGs may influence the interaction between cytokines and tumor cells. 3.3 Evaluation of TME and biological characteristics of each macrophage-related cluster Using the ESTIMATE algorithm, we computed the immune score, estimated score, stromal score, and tumor purity to compare the TME and activities of immune-related pathways between the two clusters. Cluster 1 exhibited a significantly lower immune score, stromal score, and estimated score than cluster 2 (Fig. 4 A); conversely, cluster 1 demonstrated a substantially higher tumor purity than cluster 2 (Fig. 4 B). Based on these findings, it appeared the two clusters had totally distinct TME infiltration patterns. There were significant differences in the immune cell functions involved between the two clusters, with almost all immune pathways enriched in cluster 2 (Fig. 4 C). Cluster 2 showed higher expression of HLA-related genes and immune checkpoint genes ( NRP1, IDO1, LGALS9, CD40, TNFRSF14, CD274 , etc.) and lower expression of CD44, TNFRSF18, TNFSF9 , and TNFSF18 (Fig. 4 D-E). To assess the predictive power of the clusters associated with macrophages for ICI response, IPS analysis was performed on patients with HNSCC to ascertain their immunotherapeutic sensitivity. As shown in Fig. 4 F-I, cluster 2 showed a higher IPS score in CTLA4 − _PD1 − , CTLA4 − _PD1 + , and CTLA4 + _PD1 + , which suggested that individuals in cluster 2 may potentially get more advantages from immunotherapy. TMB, or non-synonymous variation, is strongly correlated with the infiltration of immune cells and activation of immunological responses [ 24 ]. The somatic cell mutation frequency in patients is 94.18%, with the highest frequencies observed for TP53, TTN, CDKN2A , and FAT1 mutations (Fig. 4 J). Comparison of the TMB scores between clusters 1 and 2 revealed that cluster 1 was significantly higher than cluster 2 (Fig. 4 K). Furthermore, according to the TMB score, patients were divided into high and low TMB groups; patients with high TMB had a poorer prognosis than those with low TMB (Fig. 4 L). Additionally, we conducted a stratification study and found that combining the cluster with the TMB risk group could more accurately predict the prognosis of patients with HNSCC (Fig. 4 M). 3.4 Integrative machine learning algorithms constructed an optimal prognostic MRS We evaluated the MRGs and created a prognostic MRS, both accurately and stably, using an integrative approach combined with 10 machine-learning-based algorithms. Consequently, a total of 101 distinct prediction models were obtained, as shown in Fig. 5 A. The StepCox[both] + CoxBoost method yielded the optimum model, which consisted of APOC1, CTLA4, IGF2BP2, CYP27A1, NTN4, SLC7A5, PPP1R14C, KRT9 , and RAC2 , as evidenced by an average C-index of 0.6469 (Fig. 5 A). We subsequently classified HNSCC cases into high- and low-risk categories based on their risk scores. Differences in age, sex, pathological grade, and clinical stage distribution among patients in the various risk groups are shown in Supplementary Figure S1 A-B . As anticipated, HNSCC patients with high-risk scores had a significantly poor OS rate in TCGA training ( p < 0.001, Fig. 5 B), TCGA testing ( p = 0.002, Fig. 5 C), and GSE65858 ( p = 0.037, Fig. 5 D) cohorts. Analysis of the risk curve revealed that patients in the high-risk group had a greater likelihood of death and a shorter duration of survival. The population of patients in the high-risk group expanded and mortality rates increased with increasing risk scores (Fig. 5 E-G). Additionally, the AUCs for 1-, 3-, and 5-year OS were shown in Fig. 5 H-J. As shown in supplementary Figure S2 , age, pathological grade, T status, and N status were significantly associated with risk score in the TCGA training cohort. Supplementary Figure S3 shows the ROC curves for forecasting patient survival probability. Subsequently, COX analysis suggested that the risk score was an independent prognostic factor for predicting survival and nomogram was constructed to predict the survival duration of patients (Fig. 6 A-I). Furthermore, we conducted a paired difference analysis to confirm the expression of nine genes comprising MRS, utilizing data from patients with HNSCC in TCGA (Fig. 6 J-R). Additionally, we performed functional enrichment analysis and gene set variation analysis (GSVA) on the basis of MRS model, which was shown in Fig. 7 . 3.5 Identification of hub genes and variation characteristics in high and low risk groups A PPI network for the DEGs between high and low risk groups was constructed based on the “cytoscape” software (Fig. 8 A). We applied three different algorithms to identify the top 10 hub genes (Fig. 8 B-D). Further intersections were performed, resulting in the identification of four hub genes: CD80, CTLA4, CCL2 , and IFNG (Fig. 8 E). We employed TISMO to conduct a comparative analysis of the alterations in the expression of four hub genes in MOC22 tumor mouse models following the administration of anti-PD1 therapy. The findings indicated that the levels of CCL2 and IFNG were significantly increased in the group that responded positively to anti-PD1 therapy compared to the group that did not respond (Fig. 8 F-I), which suggest that CCL2 and IFNG can be used as biomarkers to predict the responsiveness of HNSCC patients to anti-PD1 treatment. A copy number variant (CNV) is a DNA segment and is associated with the heterogeneity of the genome and molecular phenotype, which contributes to the development of tumors [ 25 ]. We conducted a mutation frequency analysis and the results showed that IFNG had a greater occurrence of gain-of-function mutations in HNSCC, whereas CTLA4 displayed a higher frequency of loss-of-function mutations (Fig. 8 J-K). The TMB scores of the high- and low-risk groups were calculated and compared, and no significant differences were observed (Fig. 8 L). Subsequently, we conducted a correlation analysis, which revealed that the expression levels of CD80, CTLA4 , and CCL2 were negatively correlated with the TMB score, whereas the expression levels of IFNG and CD80 were positively correlated with the risk score. According to prior research, there is no notable link between risk and TMB scores. There was a possibility of a “chain effect” in the regulation of gene expression among the four hub genes, where their expression levels are positively associated with one another (Fig. 8 M). As shown in Fig. 8 N, the higher the TMB and risk score, the lower the probability of patient survival. In accordance with the TMB score, individuals classified as high-risk had a greater prevalence of mutations. TP53 had the highest mutation rate in the high-risk group, with a mutation frequency of 80%, whereas the mutation frequency was 60% in the low-risk group. Notably, TTN had a higher mutation rate in the low-risk group (40%) than in the high-risk group (36%). The main types of TP53 and TTN mutations were missense mutations and multiple hits (Fig. 8 O-P). 3.6 The estimate of immunotherapy for MRS model We conducted a comparative analysis of the proportions of the immunological subtypes of HNSCC across various risk groups. While the link between them may not be statistically significant, it was evident that clusters 1 and 2 exhibited a substantial association in terms of survival outcomes. Specifically, patients in cluster 1 predominantly exhibited survival results, whereas those in cluster 2 mainly died (Fig. 9 A-B). We observed substantial disparities in the expression of immune checkpoint genes between high- and low-risk groups. CD44, CD276, NRP1, CD274, LAIR1, HAVCR2, CD86, PDCD1LG2 , and CD80 were predominantly overexpressed in the high-risk group. Conversely, TNFRSF18, TNFRSF25 , and CD200 were highly expressed in the low-risk group (Fig. 9 C). In terms of HLA gene expression, except for HLA-DOB , which was significantly overexpressed in the low-risk group, all other differentially expressed genes were enriched in the high-risk group (Fig. 9 D). Using the TIDE methodology, we found that the risk score was positively correlated with the TIDE score (Fig. 9 E). The high-risk group exhibited a considerably lower response rate than the low-risk group. Conversely, the risk score of the immune response group was much lower than that of the non-response group (Fig. 9 F-G). Patients with metastatic urothelial carcinoma in the IMvigor210 cohort were divided into high- and low-risk groups. In both groups, the proportions of inflammatory immune subtypes and tumor-infiltrating immune cells expressing PD-L1 differed significantly (Fig. 9 H-I). Nevertheless, the treatment outcomes following immunotherapy did not differ considerably between two groups (Fig. 9 J, Supplementary Figure S1 C-D ). It is worth mentioning that the survival probability of patients in the high-risk group was significantly lower than that in the low-risk group, which is consistent with the conclusion obtained from the HNSCC dataset (Fig. 9 K-L). Additionally, the chemotherapy drugs sensitivity analysis was showed in Fig. 10 . 3.7 MRS gene expression patterns at single cell resolution. To further investigate the patterns of gene expression in the MRS model at the single-cell level, we analyzed the single-cell dataset GSE1822271. After quality control and dimensional reduction, a grand total of 30 cell clusters were obtained (Fig. 11 A-B). The top five differentially expressed marker genes in each cell type are shown as a heatmap (Fig. 11 C). We further visualized the distribution of nine genes in MRS across different cell types, among which SLC7A5, RAC2 , and APOC1 were enriched in macrophages, with APOC1 being the most highly expressed (Fig. 11 D-E). Every cell type plays a role in the onset and progression of the disease, and follows a distinct cell cycle. Data analysis revealed that macrophages were mostly in the transitional phase from G2 to M, with the lowest cell count observed in the S phase, suggesting that macrophage division and proliferation were more active. Simultaneously, fibroblasts mostly resided in the G2 and S phases and were primarily tasked with duplicating genetic material and preparing for cellular division (Fig. 11 F-G). Functional enrichment analysis showed that epithelial mesenchymal transition was significantly upregulated in fibroblasts, but downregulated in macrophages, mast cells, plasma cells, and T cells. We also found that epithelial cells, endothelial cells, fibroblasts, and macrophages were involved in the activation of tumor-related pathways, whereas mast cells, plasma cells, and T cells were mainly involved in the inhibition of related pathways (Fig. 11 H-I). It can be seen that macrophages, epithelial cells, and fibroblasts were more abundant in the HPV − group, while endothelial cells and T cells were more abundant in the HPV + group (Fig. 11 K). In addition, the nine genes in MRS were mainly enriched in macrophages, T cells, epithelial cells, and mast cells, with the highest proportion of high expression in macrophages and T cells (Fig. 11 L-M). To some extent, this confirms the effectiveness of our MRS prognostic model. 3.8 Authentication of MRS-related genes via qRT-PCR and HPA platform We conducted RT-qPCR to quantify mRNA expression levels of the nine genes in the MRS model to confirm our bioinformatic findings. Figure 12 A showing our experimental findings, consistent with earlier findings, all genes except CYP27A1 and NTN4 showed elevated expression levels in cancer. Additionally, we used the HPA database to obtain immunohistochemical images to evaluate the protein expression of MRS-related genes. As shown in Fig. 12 B, IGF2BP2, PPP1R14C, KRT9, RAC2, SLC7A5, and APOC1 protein expression was substantially higher in tumor tissues than in normal tissues, whereas CYP27A1 and NTN4 showed lower expression levels. Thus, the experimental outcomes of our study effectively confirmed the bioinformatics-based conclusions, thereby strengthening the importance of our work. Validating our bioinformatics predictions is crucial, as it enhances the dependability and trustworthiness of our results. The congruity between experimental and bioinformatics data offers compelling evidence to substantiate MRS-related genes and their regulatory networks. 4. Discussion HNSCC, the most prevalent malignant tumor of the head and neck, has increased in incidence owing to risk factors such as smoking, alcohol consumption, environmental pollutants, and pathogenic viruses such as HPV and EBV [ 26 ]. Immunotherapy has emerged as a promising treatment for HNSCC as it reactivates the ability of the immune system to recognize cancer cells [ 27 ]. TAMs, as a crucial role in immune infiltration in HNSCC can activate pathways that increase tumor stemness and chemotherapy resistance, reduce chemotherapy sensitivity, and elevate the risk of recurrence in some HNSCC patients [ 28 ]. Single-cell studies have also revealed that TAMs promote angiogenesis, lymph node metastasis, and tumor invasion by secreting CCL8 and CXCL18, thus promoting HNSCC progression [ 29 ]. We quantitatively analyzed the infiltration abundance of macrophages in the two cohorts, identified key modules related to MRGs in HNSCC, ultimately obtaining 194 key MRGs. We classified HNSCC into two molecular subtypes and found that the OS of patients in cluster 1 was lower than that in cluster 2, which may be linked to the differences in enriched functional pathways: cluster 1 was enriched in metabolic pathways, while cluster 2 was mainly related to inflammatory response, which may be due to the strong “immune clearance effect” of the body’s immune system on tumor cells. In addition, cluster 2 had higher matrix, immune, and infiltration scores than cluster 1, and the functional pathways of immune cells and the expression of HLA molecules were enriched in cluster 2. Owing to immune editing, the absence or downregulation of HLA molecules is a common early event in cancer occurrence, which can lead to impaired recognition of tumor cells by cytotoxic T cells, resulting in immune escape and is closely associated with tumor recurrence and metastasis [ 30 ]. The ICI gene was predominantly expressed in cluster 2, which was closely linked to malignant characteristics such as epithelial-mesenchymal transition, angiogenesis, and the dissemination and invasion of malignancies. We hypothesized that highly expressed ICI genes may not play a prominent role in cluster 2 [ 31 ]. The tumor purity of cluster 2 was significantly lower than that of cluster1, and the treatment sensitivity against CTLA4 and PD1 was higher, suggesting that attention should be paid to the role of immunotherapy in the treatment of HNSCC patients in cluster 2, which can be combined with conventional methods to further improve patient survival. To better apply macrophage-related typing to the diagnosis and treatment of HNSCC, we constructed an MRS using a combination of 101 machine learning algorithms to further identify molecular targets related to the prognosis of HNSCC. A total of nine genes were found to comprise the MRS, which was utilized to stratify patients with HNSCC into high- and low-risk groups. The high-risk group exhibited a considerably poorer survival rate, and the risk score may serve as an independent risk factor for predicting patient survival outcomes. Among the nine genes in MRS, APOC1 has been shown that inhibition of its expression can induce ferroptosis to reverse M2 macrophages to M1 macrophages, leading to immune activation and increased sensitivity to PD1 therapy [ 32 ]. More detailed studies have indicated that APOC1 inhibits KEAP1, promotes NRF2 nuclear translocation, and increases the expression of cystathionine-beta-synthase (CBS) to inhibit ferroptosis, thereby promoting tumor proliferation [ 33 ]. CYP27A1 mainly regulates cholesterol homeostasis using the CYP27A1/27-hydroxycholesterol (27-HC) axis as a mediator to regulate cell apoptosis and the cell cycle and is associated with good clinicopathological characteristics and prognosis [ 34 ]. Similarly, NTN4 is also associated with better survival rates facilitated by Wnt/β- Catenin signal transduction, and is positively correlated with the degree of infiltration of immune cells, including macrophages and neutrophils, as well as the immunosuppressive condition of tumors [ 35 ]. Previous studies have demonstrated that IGF2BP2 promotes the clearance, synthesis, metabolism, and growth of HNSCC cells and that its overexpression is a risk factor for poor prognosis in patients. Furthermore, by increasing the stability of CDK6, IGF2BP2 can upregulate its expression, which in turn promotes malignant characteristics [ 36 ]. The CTLA4 molecule, which is mainly located on the surface of T cells, can effectively bind to the B7 protein to induce T cell dysfunction and immune suppression [ 37 ]. SLC7A5 maintains the levels of essential amino acids in cancer cells through transcription and metabolic recording, which are crucial for the proliferation of cancer cells. Specific knockout of SLC7A5 significantly downregulates the mTORC1 signaling pathway in cancer cells, mobilizing the general amino acid control (GAAC) pathway to inhibit the synthesis of specific proteins, thereby hindering tumor progression [ 38 ]. Other genes in the MRS, including KRT9, PPP1R14C , and RAC2 , can influence to varying degrees the proliferation, invasion, metastasis, and angiogenesis of tumor cells, as well as the functions of immune cells in the TME. The specific mechanisms by which these genes regulate HNSCC warrant further investigation. CNV is a driving force in cancer development and plays a critical role in oncogene activation and tumor suppressor gene inactivation [ 39 ]. Data analysis revealed that CTLA4 is prone to copy number loss, whereas IFNG is prone to copy number amplification. TMB can reflect the number of gene mutations in tumor cells and predict the sensitivity and resistance of patients to ICI treatment; the more mutations, the more new antigens, and the higher the probability of triggering a T cell immune response [ 40 ]. Our analysis revealed a significant correlation between the expression levels of hub genes CD80 , CTLA4 , and CCL2 and TMB. Previous studies have demonstrated that T cells in the TME, activated in a CTLA4 dependent manner, exhibit strong immunogenicity in high-TMB tumor cells, leading to a series of immune responses [ 41 ]. In addition, TP53 showed a higher mutation frequency in the high-risk group, whereas TTN showed a higher mutation frequency in the low-risk group. TP53 , as the gene with the highest mutation frequency in human cancer, plays an important role in inducing cell apoptosis, aging, cell cycle arrest, and DNA damage repair, and is associated with increased chromosomal instability [ 42 , 43 ]. Previous studies have shown that TP53 mutations can increase the expression of relevant immune checkpoint molecules, and inhibit the infiltration of cytotoxic CD8 + T cells, leading to immunosuppressive states that induce macrophage polarization towards the M2 phenotype of TAMs, resulting in sustained tumor progression and a higher mutation frequency in high-risk groups [ 44 ]. TTN mutations can cause varied gene expression in cancers, boost tumor immune response, and increase susceptibility to immunotherapy [ 45 ]. Mutations in TTN are strongly associated with the conversion of macrophages into M1 type TAM, which could explain the increased occurrence of TTN mutations in the low-risk group [ 46 ]. Cancer immunotherapy is an innovative strategy that seeks to augment the capacity of the immune system to combat cancer by performing radical cancer treatment and preventing recurrence [ 47 ]. We observed that the vast majority of immune checkpoint genes and HLA genes were significantly upregulated in the high-risk group, which could potentially affect the level of immune infiltration and exacerbate tumor growth by recruiting various immune cells or cytokines. The TIDE score of tumors was positively correlated with the risk score of the MRS model, indicating that the tumor cells in the high-risk group have a stronger immune escape ability, whereas patients in the low-risk group may be more sensitive to ICI therapy and have a higher positive immune response rate. To confirm the reliability of the MRS model in accurately forecasting the effectiveness of immunotherapy and patient survival, we used clinical data from individuals with urothelial carcinoma in the IMvigor210 cohort to assess the predictive capacity of immunotherapy. We found the high-risk group had lower survival rates and the predictive power of the risk score for patient survival was higher. It is worth noting that the sensitivity of patients with HNSCC to chemotherapy drugs varied among different risk groups, which was related to tumor heterogeneity caused by mutations in drug resistance-related genes and crosstalk between different immune cells [ 48 ]. Hence, the MRS model score can be utilized to determine specific chemotherapeutic medications for patients with HNSCC, thereby attaining precision in medicine. Finally, our single-cell analysis indicated that SLC7A5, RAC2, and APOC1 were highly expressed in the macrophage subpopulation in the MRS. Macrophages in the TME mainly stagnated in G2.M phase, which is related to the downregulation of multiple functional pathways, with the inhibition of epithelial-mesenchymal transition being the most prominent and the highest abundance in the HPV − patient population, all of which confirm that macrophages themselves may play a favorable role in the progression of HNSCC. This phenomenon could be attributed to the scavenging functions of macrophages, which also regulate the immune response to tumor cells and preserve tissue homeostasis. By regulating the differentiation of TAM into M2 macrophages with cytokines or metabolites, the TME may induce anti-tumor immunity [ 49 , 50 ]. This study aimed to categorize HNSCC patients into subtypes based on MRGs, identify DEGs among these clusters, and develop an MRS model for patient prognosis assessment to aid immunotherapy selection. Through comprehensive validation across various perspectives and databases, we developed an MRS model that effectively predicts patient survival and provides information on treatment strategies. However, this study has some limitations. HNSCC patient data, sourced solely from public databases, often lack critical details, such as HPV infection status and microvascular infiltration. However, the effects of these factors remain unclear. To authenticate and assess the diagnostic and therapeutic efficacy of the MRS model, future studies should include broader multicenter clinical cohorts. Additionally, our validation of MRS-related gene expression was restricted to the RNA and protein levels. The specific molecular mechanisms and pathways in HNSCC have been inferred from the literature and warrant further exploration through in vivo and in vitro experiments. In summary, this study identified characteristic genes associated with macrophages in HNSCC and established an MRS model that fully validated its diagnostic efficacy in predicting patient survival prognosis, clarified its potential relationship with tumor cell genome mutations, and provided a theoretical basis for immunotherapy and chemotherapy drug selection. These results will help us further explore the characteristics and related molecular mechanisms of macrophage immune infiltration in HNSCC to identify new molecular targets for personalized diagnosis and treatment of HNSCC patients in the future. Declarations Acknowledgements : Not applicable. Ethics approval and consent to participate : The studies involving human participants were reviewed and approved by Ethics Committee of Yantai Yuhuangding Hospital. The patients/participants provided their written informed consent to participate in this study. Consent for publication : Not applicable. Availability of data and materials : The data underlying this article are available in the Gene Expression Omnibus (GEO) database at https://www.ncbi.nlm.nih.gov/geo/ and The Cancer Genome Atlas (TCGA) at https://portal.gdc.cancer.gov/, and can be accessed with GSE65858 and GSE182227. Competing interests : The authors declare that they have no competing interests. Funding: This work was supported by Taishan Scholar Project (No.ts20190991), the Key R&D Project of Shandong Province (2022CXPT023) and the Natural Science Foundation of Shandong Province (No. ZR2019PH113). Authors’ contributions : X-CS and CR: conception, design and administrative support. YW, Y-KM and W-CL: study conception, data analysis and visualization. YW, Y-KM and H-RW, X-YS and TY : data collection and manuscript revision. YW, Y-KM and W-CL, CR and X-CS: data analysis and interpretation. All authors: manuscript writing and final approval of manuscript. References Global Burden of Disease Cancer Collaboration; Fitzmaurice C, Allen C, et al. Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With Disability, and Disability-Adjusted Life-years for 32 Cancer Groups, 1990 to 2015: A Systematic Analysis for the Global Burden of Disease Study. JAMA Oncol. 2017;3(4):524–548. doi: 10.1001/jamaoncol.2016.5688. Erratum in: JAMA Oncol. 2017;3(3):418. PMID: 27918777. Takes RP, Rinaldo A, Silver CE, et al. Future of the TNM classification and staging system in head and neck cancer. Head Neck. 2010;32(12):1693 – 711. doi: 10.1002/hed.21361 . PMID: 20191627. Qin Y, Zheng X, Gao W, Wang B, Wu Y. Tumor microenvironment and immune-related therapies of head and neck squamous cell carcinoma. Mol Ther Oncolytics. 2021;20:342–351. doi: 10.1016/j.omto.2021.01.011 . PMID: 33614915. Mantovani A, Marchesi F, Malesci A, Laghi L, Allavena P. Tumour-associated macrophages as treatment targets in oncology. Nat Rev Clin Oncol. 2017;14(7):399–416. doi: 10.1038/nrclinonc.2016.217 . Epub 2017 Jan 24. PMID: 28117416. Yunna C, Mengru H, Lei W, Weidong C. Macrophage M1/M2 polarization. Eur J Pharmacol. 2020;877:173090. doi: 10.1016/j.ejphar.2020.173090 . Epub 2020 Mar 29. PMID: 32234529. Wichmann G, Rosolowski M, Krohn K, et al. The role of HPV RNA transcription, immune response-related gene expression and disruptive TP53 mutations in diagnostic and prognostic profiling of head and neck cancer. Int J Cancer. 2015;137(12):2846–57. doi: 10.1002/ijc.29649 . Epub 2015 Jul 6. PMID: 26095926. Puram SV, Mints M, Pal A, et al. Cellular states are coupled to genomic and viral heterogeneity in HPV-related oropharyngeal carcinoma. Nat Genet. 2023;55(4):640–650. doi: 10.1038/s41588-023-01357-3 . Epub 2023 Apr 3. PMID: 37012457. Zeng D, Li M, Zhou R, et al. Tumor Microenvironment Characterization in Gastric Cancer Identifies Prognostic and Immunotherapeutically Relevant Gene Signatures. Cancer Immunol Res. 2019;7(5):737–750. doi: 10.1158/2326-6066.CIR-18-0436 . Epub 2019 Mar 6. PMID: 30842092. Li T, Fu J, Zeng Z, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020;48(W1):W509-W514. doi: 10.1093/nar/gkaa407 . PMID: 32442275. Newman AM, Liu CL, Green MR, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12(5):453–7. doi: 10.1038/nmeth.3337 . Epub 2015 Mar 30. PMID: 25822800. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. doi: 10.1186/1471-2105-9-559 . PMID: 19114008. McDermaid A, Monier B, Zhao J, Liu B, Ma Q. Interpretation of differential gene expression results of RNA-seq data: review and integration. Brief Bioinform. 2019;20(6):2044–2054. doi: 10.1093/bib/bby067 . PMID: 30099484. Wilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics. 2010;26(12):1572–3. doi: 10.1093/bioinformatics/btq170 . Epub 2010 Apr 28. PMID: 20427518. Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7. doi: 10.1186/1471-2105-14-7 . PMID: 23323831. Zhou Y, Zhou B, Pache L, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523. doi: 10.1038/s41467-019-09234-6 . PMID: 30944313. Mayakonda A, Lin DC, Assenov Y, Plass C, Koeffler HP. Maftools: efficient and comprehensive analysis of somatic variants in cancer. Genome Res. 2018;28(11):1747–1756. doi: 10.1101/gr.239244.118 . Epub 2018 Oct 19. PMID: 30341162. Wang T, Dai L, Shen S, et al. Comprehensive Molecular Analyses of a Macrophage-Related Gene Signature With Regard to Prognosis, Immune Features, and Biomarkers for Immunotherapy in Hepatocellular Carcinoma Based on WGCNA and the LASSO Algorithm. Front Immunol. 2022;13:843408. doi: 10.3389/fimmu.2022.843408 . PMID: 35693827. Jiang H, Awuti G, Guo X. Construction of an Immunophenoscore-Related Signature for Evaluating Prognosis and Immunotherapy Sensitivity in Ovarian Cancer. ACS Omega. 2023;8(36):33017–33031. doi: 10.1021/acsomega.3c04856 . PMID: 37720747. Liu Z, Liu L, Weng S, et al. Machine learning-based integration develops an immune-derived lncRNA signature for improving outcomes in colorectal cancer. Nat Commun. 2022;13(1):816. doi: 10.1038/s41467-022-28421-6 . PMID: 35145098. Doncheva NT, Morris JH, Gorodkin J, Jensen LJ. Cytoscape StringApp: Network Analysis and Visualization of Proteomics Data. J Proteome Res. 2019;18(2):623–632. doi: 10.1021/acs.jproteome.8b00702 . Epub 2018 Dec 5. PMID: 30450911. Yan C, Niu Y, Li F, Zhao W, Ma L. System analysis based on the pyroptosis-related genes identifies GSDMC as a novel therapy target for pancreatic adenocarcinoma. J Transl Med. 2022;20(1):455. doi: 10.1186/s12967-022-03632-z . PMID: 36199146. Tan Z, Chen X, Zuo J, Fu S, Wang H, Wang J. Comprehensive analysis of scRNA-Seq and bulk RNA-Seq reveals dynamic changes in the tumor immune microenvironment of bladder cancer and establishes a prognostic model. J Transl Med. 2023;21(1):223. doi: 10.1186/s12967-023-04056-z . PMID: 36973787. Zhao S, Ye B, Chi H, Cheng C, Liu J. Identification of peripheral blood immune infiltration signatures and construction of monocyte-associated signatures in ovarian cancer and Alzheimer's disease using single-cell sequencing. Heliyon. 2023;9(7):e17454. doi: 10.1016/j.heliyon.2023.e17454 . PMID: 37449151. Addeo A, Friedlaender A, Banna GL, Weiss GJ. TMB or not TMB as a biomarker: That is the question. Crit Rev Oncol Hematol. 2021;163:103374. doi: 10.1016/j.critrevonc.2021.103374 . Epub 2021 Jun 2. PMID: 34087341. Shlien A, Malkin D. Copy number variations and cancer. Genome Med. 2009;1(6):62. doi: 10.1186/gm62 . PMID: 19566914. Johnson DE, Burtness B, Leemans CR, Lui VWY, Bauman JE, Grandis JR. Head and neck squamous cell carcinoma. Nat Rev Dis Primers. 2020;6(1):92. doi: 10.1038/s41572-020-00224-3 . Erratum in: Nat Rev Dis Primers. 2023;9(1):4. PMID: 33243986. Ghosh S, Shah PA, Johnson FM. Novel Systemic Treatment Modalities Including Immunotherapy and Molecular Targeted Therapy for Recurrent and Metastatic Head and Neck Squamous Cell Carcinoma. Int J Mol Sci. 2022;23(14):7889. doi: 10.3390/ijms23147889 . PMID: 35887235. Hsieh CY, Lin CC, Huang YW, et al. Macrophage secretory IL-1β promotes docetaxel resistance in head and neck squamous carcinoma via SOD2/CAT-ICAM1 signaling. JCI Insight. 2022;7(23):e157285. doi: 10.1172/jci.insight.157285 . PMID: 36264639. Wu J, Shen Y, Zeng G, Liang Y, Liao G. SPP1 + TAM subpopulations in tumor microenvironment promote intravasation and metastasis of head and neck squamous cell carcinoma. Cancer Gene Ther. 2023 Dec 5. doi: 10.1038/s41417-023-00704-0 . Epub ahead of print. PMID: 38052857. Wang Y, Jasinski-Bergner S, Wickenhauser C, Seliger B. Cancer Immunology: Immune Escape of Tumors-Expression and Regulation of HLA Class I Molecules and Its Role in Immunotherapies. Adv Anat Pathol. 2023;30(3):148–159. doi: 10.1097/PAP.0000000000000389. Epub 2022 Dec 15. PMID: 36517481. Zhang Y, Zheng J. Functions of Immune Checkpoint Molecules Beyond Immune Evasion. Adv Exp Med Biol. 2020;1248:201–226. doi: 10.1007/978-981-15-3266-5_9 . PMID: 32185712. Hao X, Zheng Z, Liu H, et al. Inhibition of APOC1 promotes the transformation of M2 into M1 macrophages via the ferroptosis pathway and enhances anti-PD1 immunotherapy in hepatocellular carcinoma based on single-cell RNA sequencing. Redox Biol. 2022;56:102463. doi: 10.1016/j.redox.2022.102463 . Epub 2022 Sep 2. PMID: 36108528. Zheng XJ, Chen WL, Yi J, et al. Apolipoprotein C1 promotes glioblastoma tumorigenesis by reducing KEAP1/NRF2 and CBS-regulated ferroptosis. Acta Pharmacol Sin. 2022;43(11):2977–2992. doi: 10.1038/s41401-022-00917-3 . Epub 2022 May 17. Erratum in: Acta Pharmacol Sin. 2022;: PMID: 35581292. Liang Z, Chen Y, Wang L, et al. CYP27A1 inhibits bladder cancer cells proliferation by regulating cholesterol homeostasis. Cell Cycle. 2019;18(1):34–45. doi: 10.1080/15384101.2018.1558868. Epub 2018 Dec 30. PMID: 30563407. Ke S, Guo J, Wang Q, et al. Netrin Family Genes as Prognostic Markers and Therapeutic Targets for Clear Cell Renal Cell Carcinoma: Netrin-4 Acts through the Wnt/β-Catenin Signaling Pathway. Cancers (Basel). 2023;15(10):2816. doi: 10.3390/cancers15102816 . PMID: 37345154. Tang X, Tang Q, Li S, Li M, Yang T. IGF2BP2 acts as a m6A modification regulator in laryngeal squamous cell carcinoma through facilitating CDK6 mRNA stabilization. Cell Death Discov. 2023;9(1):371. doi: 10.1038/s41420-023-01669-7 . PMID: 37816718. Yu GT, Bu LL, Zhao YY, et al. CTLA4 blockade reduces immature myeloid cells in head and neck squamous cell carcinoma. Oncoimmunology. 2016;5(6):e1151594. doi: 10.1080/2162402X.2016.1151594. PMID: 27471622. Kanai Y. Amino acid transporter LAT1 (SLC7A5) as a molecular target for cancer diagnosis and therapeutics. Pharmacol Ther. 2022;230:107964. doi: 10.1016/j.pharmthera.2021.107964 . Epub 2021 Aug 12. PMID: 34390745. Tao Z, Wang S, Wu C, et al. The repertoire of copy number alteration signatures in human cancer. Brief Bioinform. 2023;24(2):bbad053. doi: 10.1093/bib/bbad053 . PMID: 36806386. Jardim DL, Goodman A, de Melo Gagliato D, Kurzrock R. The Challenges of Tumor Mutational Burden as an Immunotherapy Biomarker. Cancer Cell. 2021;39(2):154–173. doi: 10.1016/j.ccell.2020.10.001 . Epub 2020 Oct 29. PMID: 33125859. Chan TA, Yarchoan M, Jaffee E, et al. Development of tumor mutation burden as an immunotherapy biomarker: utility for the oncology clinic. Ann Oncol. 2019;30(1):44–56. doi: 10.1093/annonc/mdy495 . PMID: 30395155. Wang Z, Strasser A, Kelly GL. Should mutant TP53 be targeted for cancer therapy? Cell Death Differ. 2022;29(5):911–920. doi: 10.1038/s41418-022-00962-9 . Epub 2022 Mar 24. PMID: 35332311. Donehower LA, Soussi T, Korkut A, et al. Integrated Analysis of TP53 Gene and Pathway Alterations in The Cancer Genome Atlas. Cell Rep. 2019;28(5):1370–1384.e5. doi: 10.1016/j.celrep.2019.07.001 . Erratum in: Cell Rep. 2019;28(11):3010. PMID: 31365877. Shi Y, Ren X, Cao S, et al.TP53 gain-of-function mutation modulates the immunosuppressive microenvironment in non-HPV-associated oral squamous cell carcinoma. J Immunother Cancer. 2023;11(8):e006666. doi: 10.1136/jitc-2023-006666 . PMID: 37604640. Liu Z, Zhao X, Wang R, et al. Heterogeneous pattern of gene expression driven by TTN mutation is involved in the construction of a prognosis model of lung squamous cell carcinoma. Front Oncol. 2023;13:916568. doi: 10.3389/fonc.2023.916568 . PMID: 37035196. Xie X, Tang Y, Sheng J, et al. Titin Mutation Is Associated With Tumor Mutation Burden and Promotes Antitumor Immunity in Lung Squamous Cell Carcinoma. Front Cell Dev Biol. 2021;9:761758. doi: 10.3389/fcell.2021.761758 . PMID: 34746153. Szeto GL, Finley SD. Integrative Approaches to Cancer Immunotherapy. Trends Cancer. 2019;5(7):400–410. doi: 10.1016/j.trecan.2019.05.010 . PMID: 31311655. Nussinov R, Tsai CJ, Jang H. Anticancer drug resistance: An update and perspective. Drug Resist Updat. 2021;59:100796. doi: 10.1016/j.drup.2021.100796 . Epub 2021 Dec 16. PMID: 34953682. Mehla K, Singh PK. Metabolic Regulation of Macrophage Polarization in Cancer. Trends Cancer. 2019;5(12):822–834. doi: 10.1016/j.trecan.2019.10.007 . Epub 2019 Nov 6. PMID: 31813459. Zhang J, Zhou X, Hao H. Macrophage phenotype-switching in cancer. Eur J Pharmacol. 2022;931:175229. doi: 10.1016/j.ejphar.2022.175229 . Epub 2022 Aug 21. PMID: 36002039. Additional Declarations No competing interests reported. Supplementary Files SUP1.tif SUP2.tif SUP3.tif Supplementaryfigurelegends.docx Cite Share Download PDF Status: Published Journal Publication published 22 Aug, 2024 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 13 Jun, 2024 Reviews received at journal 25 May, 2024 Reviewers agreed at journal 15 May, 2024 Reviews received at journal 24 Apr, 2024 Reviewers agreed at journal 17 Apr, 2024 Reviewers invited by journal 17 Apr, 2024 Editor assigned by journal 17 Apr, 2024 Editor invited by journal 14 Apr, 2024 Submission checks completed at journal 11 Apr, 2024 First submitted to journal 04 Apr, 2024 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-4219358","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":290962500,"identity":"d642d42f-6c72-4600-bb91-d6ae0700e0ba","order_by":0,"name":"Yao Wang","email":"","orcid":"","institution":"Yuhuangding Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yao","middleName":"","lastName":"Wang","suffix":""},{"id":290962501,"identity":"dc7d01d1-05b3-41d9-a740-87621298a1f8","order_by":1,"name":"Ya‐Kui Mou","email":"","orcid":"","institution":"Yuhuangding Hospital","correspondingAuthor":false,"prefix":"","firstName":"Ya‐Kui","middleName":"","lastName":"Mou","suffix":""},{"id":290962502,"identity":"5b477ca4-0d05-4480-abeb-123d8ee5fe6f","order_by":2,"name":"Wan-Chen Liu","email":"","orcid":"","institution":"Yuhuangding Hospital","correspondingAuthor":false,"prefix":"","firstName":"Wan-Chen","middleName":"","lastName":"Liu","suffix":""},{"id":290962503,"identity":"b309329d-94a6-4bc8-9168-d59ab6afe9ab","order_by":3,"name":"Han‐Rui Wang","email":"","orcid":"","institution":"Yuhuangding Hospital","correspondingAuthor":false,"prefix":"","firstName":"Han‐Rui","middleName":"","lastName":"Wang","suffix":""},{"id":290962504,"identity":"90c8b558-c17a-452c-992e-08728a041666","order_by":4,"name":"Xiao-Yu Song","email":"","orcid":"","institution":"Yuhuangding Hospital","correspondingAuthor":false,"prefix":"","firstName":"Xiao-Yu","middleName":"","lastName":"Song","suffix":""},{"id":290962505,"identity":"fb5705a8-5d4d-4eab-9b4c-df5515c572b8","order_by":5,"name":"Ting Yang","email":"","orcid":"","institution":"Yuhuangding Hospital","correspondingAuthor":false,"prefix":"","firstName":"Ting","middleName":"","lastName":"Yang","suffix":""},{"id":290962506,"identity":"bd4ff983-cdb2-42f0-a878-c10c9049faf0","order_by":6,"name":"Chao Ren","email":"","orcid":"","institution":"Yuhuangding Hospital","correspondingAuthor":false,"prefix":"","firstName":"Chao","middleName":"","lastName":"Ren","suffix":""},{"id":290962507,"identity":"7644dd67-9055-4a27-b9b4-4d2f0e8bec84","order_by":7,"name":"Xi-cheng Song","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAyElEQVRIiWNgGAWjYDCCAwwMzGAGe2Pjww+kaeE53GwsQZoWifQ2AR5idPDdPmP4ueCPTZ585MM2BgkGOzndBgJaJM+lJUvPbEsrNryd2PaggCHZ2OwAAS0GZ5iPMfM2HE7cODux3UCC4UDiNsJaGNuYef4Atcw82CbBQ5wWoC08bIcT50swEqlF8gxbsjRvW1riBp5EYCAbEOEXvjM8hp95/tgkzm8//vDhhwo7OYJaEC4EqzQgVjkIyDeQonoUjIJRMApGFAAAS25DJSseG0QAAAAASUVORK5CYII=","orcid":"","institution":"Yuhuangding Hospital","correspondingAuthor":true,"prefix":"","firstName":"Xi-cheng","middleName":"","lastName":"Song","suffix":""}],"badges":[],"createdAt":"2024-04-04 18:15:58","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4219358/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4219358/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-024-70430-6","type":"published","date":"2024-08-22T15:57:22+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":55005005,"identity":"ed58a2b3-756b-4ee1-b2aa-3fb549e41f06","added_by":"auto","created_at":"2024-04-19 18:47:54","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":501461,"visible":true,"origin":"","legend":"\u003cp\u003eThe flowchart of the design of this study.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/55a2bc60f470f5f3b7c68436.png"},{"id":55005092,"identity":"cfa3e910-8102-447c-94de-c6a984705cf3","added_by":"auto","created_at":"2024-04-19 18:47:57","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":866722,"visible":true,"origin":"","legend":"\u003cp\u003eConstruction of coexpression network and identification macrophage-related genes (MRGs) using weighted gene coexpression network analysis. \u003cstrong\u003e(A, B) \u003c/strong\u003eHeatmap demonstrating the correlation between module eigengenes and macrophages in TCGA-HNSCC and GSE146771 datasets. \u003cstrong\u003e(C)\u003c/strong\u003e The blue module had a significant correlation with macrophages in the TCGA-HNSCC dataset (Cor=0.79, \u003cem\u003ep\u003c/em\u003e\u0026lt;1e−200). \u003cstrong\u003e(D-F) \u003c/strong\u003eThe M0, M1, and M2 macrophages had the strongest correlation with the red, tan, and turquoise modules, respectively (M0: Cor=0.7, \u003cem\u003ep\u003c/em\u003e\u0026lt;1e−200, M1: Cor=0.79, \u003cem\u003ep\u003c/em\u003e=4.6e−49, M2: Cor=0.63, \u003cem\u003ep\u003c/em\u003e\u0026lt;1e−200).\u003cstrong\u003e (G)\u003c/strong\u003e The volcano plot showing the genes with significant differences in the top six positions in TCGA-HNSCC dataset.\u003cstrong\u003e (H) \u003c/strong\u003eVenn diagram displaying the macrophage-related selected intersection genes from different datasets. \u003cstrong\u003e(I-K)\u003c/strong\u003e Functional enrichment analysis on the 194 intersected MRGs using Kyoto Encyclopedia of Genes and Genomes (KEGG), Molecular Signatures Database (MisgDB), and Reactome databases.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/4c62d387eb14ee864d842246.png"},{"id":55005990,"identity":"a3610f72-243b-48ee-9073-f0b51909450d","added_by":"auto","created_at":"2024-04-19 18:55:54","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1423537,"visible":true,"origin":"","legend":"\u003cp\u003eCluster analysis of intersected MRGs in the TCGA cohort. \u003cstrong\u003e(A)\u003c/strong\u003eConsensus clustering identified two clusters of HNSCC with different macrophages infiltration characteristics. \u003cstrong\u003e(B) \u003c/strong\u003eConsensus clustering cumulative distribution function (CDF) for k = 2 to 6. \u003cstrong\u003e(C)\u003c/strong\u003e Relative change in the area under the CDF curve for k = 2 to 6. \u003cstrong\u003e(D) \u003c/strong\u003eThe 3D PCA plot demonstrated the two clusters could be easily identified based on the MRGs. \u003cstrong\u003e(E) \u003c/strong\u003eThe Kaplan–Meier curve survival analysis between different clusters.\u003cstrong\u003e (F) \u003c/strong\u003eHeatmap showing the distribution differences between clusters 1 and 2 in terms of clinicopathological features. \u003cstrong\u003e(G) \u003c/strong\u003eHeatmap displaying notable disparities between the two clusters in multiple biological processes via gene set variation analysis (GSVA). \u003cstrong\u003e(H) \u003c/strong\u003eThe bar plot of the KEGG pathways enriched on the differentially expressed genes (DEGs) between different clusters. \u003cstrong\u003e(I) \u003c/strong\u003eThe cluster plot of the Gene ontology (GO) pathways enriched on the DEGs between different clusters. *\u003cem\u003eP \u003c/em\u003e\u0026lt; 0.05; **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01; ***\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/00ce3ae69964cd6e8128a631.png"},{"id":55005065,"identity":"e3a387c3-59ff-4a58-9eb8-d03d42a9a552","added_by":"auto","created_at":"2024-04-19 18:47:55","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":490699,"visible":true,"origin":"","legend":"\u003cp\u003eImmune infiltration and tumor mutation analysis between different clusters. \u003cstrong\u003e(A)\u003c/strong\u003e The comparisons of stromal score, immune score, and estimated score between various clusters.\u003cstrong\u003e (B) \u003c/strong\u003eThe comparisons of tumor purity between distinct clusters. \u003cstrong\u003e(C) \u003c/strong\u003eThe box plot demonstrating the difference in the immune cell functions involved between the two clusters. \u003cstrong\u003e(D)\u003c/strong\u003e The box plot showing the difference in HLA expression between distinct clusters.\u003cstrong\u003e (E) \u003c/strong\u003eThe box plot displaying the difference in immune checkpoint genes between the two clusters. \u003cstrong\u003e(F-I)\u003c/strong\u003e Immunophenoscore (IPS) analysis of CTLA4\u003csup\u003e−\u003c/sup\u003e_PD1\u003csup\u003e−\u003c/sup\u003e, CTLA4\u003csup\u003e−\u003c/sup\u003e_PD1\u003csup\u003e+\u003c/sup\u003e, CTLA4\u003csup\u003e+\u003c/sup\u003e_PD1\u003csup\u003e+ \u003c/sup\u003eand CTLA4\u003csup\u003e+\u003c/sup\u003e_PD1\u003csup\u003e− \u003c/sup\u003egroups between various clusters. \u003cstrong\u003e(J) \u003c/strong\u003eThe macrophage-related DEGs in HNSCC, together with their mutation rates, were displayed in a waterfall plot. \u003cstrong\u003e(K)\u003c/strong\u003e The comparison of tumor mutation burden (TMB) between different clusters.\u003cstrong\u003e (L)\u003c/strong\u003e The Kaplan–Meier curve showed the survival analysis between high- and low TMB groups.\u003cstrong\u003e (M)\u003c/strong\u003e The Kaplan–Meier curve showed the survival analysis combining the cluster with the TMB risk group. *\u003cem\u003eP \u003c/em\u003e\u0026lt; 0.05; **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01; ***\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/6c21e3a7258883eeb826d4d2.png"},{"id":55005126,"identity":"1f275fd2-efef-4559-b0f5-76cd2e1b2685","added_by":"auto","created_at":"2024-04-19 18:47:59","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":959628,"visible":true,"origin":"","legend":"\u003cp\u003eUsing macrophage-related clusters as a basis, the macrophage-related signature (MRS) was developed and validated. \u003cstrong\u003e(A)\u003c/strong\u003e A total of 101 combinations of machine learning algorithms for the MRS via a 10-fold cross-validation framework. The Kaplan–Meier curve showed the survival analysis of HNSCC patients in TCGA training\u003cstrong\u003e (B)\u003c/strong\u003e, TCGA testing\u003cstrong\u003e (C)\u003c/strong\u003e, which was divided based on the genes in MRS, and GSE65858 cohort \u003cstrong\u003e(D)\u003c/strong\u003e. \u003cstrong\u003e(E-G)\u003c/strong\u003e The distribution of risk scores and survival statuses for patients with HNSCC in the two risk groups as determined by the TCGA training, TCGA test, and GSE65858 cohort. \u003cstrong\u003e(H-J) \u003c/strong\u003eThe ROC analysis showed the AUCs for 1-, 3-, and 5-year OS of patients with HNSCC in the TCGA training, TCGA testing, and GSE65858 cohort. *\u003cem\u003eP\u003c/em\u003e\u0026lt; 0.05; **\u003cem\u003eP \u003c/em\u003e\u0026lt; 0.01; ***\u003cem\u003eP\u003c/em\u003e\u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/660e01ea9d3ce38dfba20467.png"},{"id":55005064,"identity":"31e80c5f-4171-45a8-9214-ad15c2997040","added_by":"auto","created_at":"2024-04-19 18:47:55","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":617111,"visible":true,"origin":"","legend":"\u003cp\u003eConstruction and assessment of the survival prediction nomogram. \u003cstrong\u003e(A, B)\u003c/strong\u003e In the TCGA-training set, univariate and multivariate Cox regression analyses revealed that the risk score derived from clusters associated with macrophages is an independent prognostic factor that impacts the prognosis of patients with HNSCC. \u003cstrong\u003e(C, D)\u003c/strong\u003e In the TCGA testing set, univariate and multivariate Cox regression analyses revealed that the risk score derived from clusters associated with macrophages is an independent prognostic factor that impacts the prognosis of patients with HNSCC. \u003cstrong\u003e(E, F)\u003c/strong\u003e In the GSE65858 dataset, univariate and multivariate Cox regression analyses revealed that the risk score derived from clusters associated with macrophages is an independent prognostic factor that impacts the prognosis of patients with HNSCC. \u003cstrong\u003e(G)\u003c/strong\u003e The nomogram for the prediction of 1-, 3-, and 5-year OS of patients with HNSCC based on the risk score combined with other clinicopathological characteristics. \u003cstrong\u003e(H) \u003c/strong\u003eThe decision curve analysis was conducted to assess the net benefit of nomogram and other clinicopathological features for predicting patient OS over the range of clinical threshold.\u003cstrong\u003e (I) \u003c/strong\u003eThe calibration plot of nomogram exhibited strong consistence of patient OS between predicted and observed probabilities. \u003cstrong\u003e(J-R)\u003c/strong\u003e The boxplots showed expression differences of nine genes in MRS between tumor and adjacent normal tissues of patients with HNSCC in TCGA. *\u003cem\u003eP\u003c/em\u003e\u0026lt; 0.05; **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01; ***\u003cem\u003eP\u003c/em\u003e\u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/178b8ab428316c0b999edc59.png"},{"id":55005131,"identity":"29c7ca1a-0aab-4365-9305-9d217bc0028f","added_by":"auto","created_at":"2024-04-19 18:48:00","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":535206,"visible":true,"origin":"","legend":"\u003cp\u003eFunctional enrichment analysis on the basis of MRS model. \u003cstrong\u003e(A) \u003c/strong\u003eThe bar plot showed the functional analysis based on DEGs between the high- and low-risk groups using the Metascape database. \u003cstrong\u003e(B) \u003c/strong\u003eThe protein-protein interaction (PPI) network constructed using the Metascape database demonstrated a correlation between functional pathways at the protein level. \u003cstrong\u003e(C-F) \u003c/strong\u003eEnrichment plots were generated to analyze gene set enrichment in both high- and low-risk groups, depending on the risk score derived from macrophage-related clusters.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/2de7c7c876941a550deb82fb.png"},{"id":55005093,"identity":"9545d251-14a7-4147-9612-7dcd572b4be8","added_by":"auto","created_at":"2024-04-19 18:47:57","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":716421,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of hub genes and genetic variation characteristics in high- and low-risk groups. \u003cstrong\u003e(A) \u003c/strong\u003eThe PPI network for the DEGs between high- and low risk groups constructed based on the “cytoscape” software. \u003cstrong\u003e(B-D)\u003c/strong\u003e The top ten hub genes identified employing Degree, MCC and MNC algorithms, respectively. \u003cstrong\u003e(E) \u003c/strong\u003eThe Venn diagram illustrated the hub genes that were generated through the intersection of hub genes derived from the three algorithms mentioned earlier. \u003cstrong\u003e(F-I)\u003c/strong\u003eThe box plot showed how the expression of four hub genes changed in MOC22 tumor mouse models after anti-PD1 treatment was given.\u003cstrong\u003e (J)\u003c/strong\u003e The box plot showed frequencies of gain and loss of the four hub genes between high- and low risk groups. \u003cstrong\u003e(K) \u003c/strong\u003eCircus plot exhibit the distribution on chromosomes of the four hub genes between high- and low risk groups.\u003cstrong\u003e (L)\u003c/strong\u003e Comparison of TMB differences between high- and low risk groups. \u003cstrong\u003e(M) \u003c/strong\u003eThe correlation analysis revealed the relation among the expression levels of four hub genes, risk scores, and TMB. \u003cstrong\u003e(N) \u003c/strong\u003eThe Kaplan-Meier curve displayed the survival analysis of patients with HNSCC, categorised based on both TMB groups and risk score. \u003cstrong\u003e(O, P) \u003c/strong\u003eWaterfall plot displaying gene mutations in the high- and low-risk groups.\u003c/p\u003e","description":"","filename":"Figure8.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/83b74a6a3be24c5901c532de.png"},{"id":55005124,"identity":"d5293b47-34c4-4e74-81bd-a07cd948d058","added_by":"auto","created_at":"2024-04-19 18:47:58","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":842686,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of immune subtypes and response to immunotherapy between different risk groups. \u003cstrong\u003e(A) \u003c/strong\u003eThe Sankey diagram unveiled the potential correlation among macrophage-related cluster, risk score, and survival status. \u003cstrong\u003e(B) \u003c/strong\u003eExamining the variations in immune subtype between various risk categories. \u003cstrong\u003e(C)\u003c/strong\u003e The expression disparities of immune checkpoint genes between different risk groups. \u003cstrong\u003e(D)\u003c/strong\u003e The expression disparities of HLA genes between different risk groups. \u003cstrong\u003e(E) \u003c/strong\u003eComparison of TIDE score in the low- and high-risk groups. \u003cstrong\u003e(F)\u003c/strong\u003e Comparison of positive response rates to immunotherapy between high- and low risk groups. \u003cstrong\u003e(G)\u003c/strong\u003e Comparison of risk scores between different immune response groups.\u003cstrong\u003e (H, I) \u003c/strong\u003eThe composition differences of the proportion of inflammatory immune subtypes or tumor-infiltrating immune cells expressing PD-L1 between the low- and high-risk groups divided based on MRS signature of patients with metastatic urothelial carcinoma in the IMvigor210 cohort. \u003cstrong\u003e(J)\u003c/strong\u003e Comparison of risk scores between the CR/PR group and the SD/PD group in IMvigor210 dataset. \u003cstrong\u003e(H)\u003c/strong\u003eThe Kaplan-Meier curve displayed the survival analysis of patients in IMvigor210 dataset, categorised based on risk score.\u003cstrong\u003e (I) \u003c/strong\u003eThe ROC analysis showed the AUCs for 1-, 3-, and 5-year OS of patients in the IMvigor210 dataset. The specimens were categorised as immunohistochemistry IC0, IC1 or IC2+ based on the percentage of PD-L1 positive cells: less than 1%, 1% to less than 5%, or 5% or more, respectively. *\u003cem\u003eP\u003c/em\u003e\u0026lt; 0.05; **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01; ***\u003cem\u003eP\u003c/em\u003e\u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure9.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/6fc74e6865b60c1459a19a71.png"},{"id":55005006,"identity":"100500e3-e658-460c-a156-1708aeb24809","added_by":"auto","created_at":"2024-04-19 18:47:54","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":373872,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e(A-L)\u003c/strong\u003e Chemotherapeutic drugs sensitivity analysis between high- and low-risk groups.\u003c/p\u003e","description":"","filename":"Figure10.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/07a50f08b2eca0c15e08d74e.png"},{"id":55005067,"identity":"e82627a2-be10-4406-ba29-f872ab9dd7bb","added_by":"auto","created_at":"2024-04-19 18:47:55","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":569616,"visible":true,"origin":"","legend":"\u003cp\u003ePatterns of gene expression in MRS model at the single cell level. \u003cstrong\u003e(A) \u003c/strong\u003eThe uniform manifold approximation and projection (UMAP) plot showed all the cells in the GSE1822271 dataset can be classified into 30 clusters. \u003cstrong\u003e(B)\u003c/strong\u003eThe UMAP plot exhibit aforementioned 30 cell clusters can be annotated as 7 major cell lineages. \u003cstrong\u003e(C) \u003c/strong\u003eHeatmap displayed the differentially expressed top five marker genes in each cell type. \u003cstrong\u003e(D)\u003c/strong\u003e The UMAP plots visualized the distribution of nine genes in MRS across different cell types. \u003cstrong\u003e(E) \u003c/strong\u003eViolin plots compared the expression level of nine genes in MRS. \u003cstrong\u003e(F, G) \u003c/strong\u003eThe bar plot and UMAP plot showed the proportion of cells in distinct cell cycles. \u003cstrong\u003e(H, I) \u003c/strong\u003eCellular functional pathways enriched in different cell types and their inhibition or activation status. \u003cstrong\u003e(J) \u003c/strong\u003eBar plots showing the proportion of cell types in each sample. \u003cstrong\u003e(K)\u003c/strong\u003e Bar plots showing the proportion of cell types in patients with different HPV infection status. \u003cstrong\u003e(L, M) \u003c/strong\u003eThe UMAP plot and bar plot displayed the proportion of cell types in high- and low-expression group divided based on the expression levels of nine genes in MRS. *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05; **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01; ***\u003cem\u003eP\u003c/em\u003e\u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure11.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/ec00d23ed722511c4025294c.png"},{"id":55005091,"identity":"21a0b8ab-b766-4e26-b1a0-1c137cdcd8b4","added_by":"auto","created_at":"2024-04-19 18:47:56","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":1635230,"visible":true,"origin":"","legend":"\u003cp\u003eExperimental verification of expression levels of nine genes in MRS. (A) The box plots compared the levels of mRNA expression of genes in MRS between cancer tissue and its neighbouring tissue using RT-qPCR. (B) The immunohistochemistry pictures obtained from the HPA database visualized the protein expression levels of genes in MRS between cancer and normal tissue. *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05; **\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01; ***\u003cem\u003eP\u003c/em\u003e\u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure12.png","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/d3307853980abac397aa2cea.png"},{"id":63300187,"identity":"8aa0a9d1-5d0b-4f2b-b1b6-0aefb27437c9","added_by":"auto","created_at":"2024-08-26 16:12:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9336254,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/e1aee59b-fcc3-4183-a686-4776495114b8.pdf"},{"id":55005068,"identity":"f1bb7dbc-e85e-4417-95cb-5409ae252e85","added_by":"auto","created_at":"2024-04-19 18:47:55","extension":"tif","order_by":14,"title":"","display":"","copyAsset":false,"role":"supplement","size":15440868,"visible":true,"origin":"","legend":"","description":"","filename":"SUP1.tif","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/56ecd88b24edaab35f142b0a.tif"},{"id":55005090,"identity":"b60280d6-e5c7-4676-8248-8175499d17e1","added_by":"auto","created_at":"2024-04-19 18:47:56","extension":"tif","order_by":15,"title":"","display":"","copyAsset":false,"role":"supplement","size":20209884,"visible":true,"origin":"","legend":"","description":"","filename":"SUP2.tif","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/08aee02465879b75220af1fd.tif"},{"id":55005128,"identity":"296f8589-c6a2-428f-ad13-9838d35b5213","added_by":"auto","created_at":"2024-04-19 18:47:59","extension":"tif","order_by":16,"title":"","display":"","copyAsset":false,"role":"supplement","size":14663420,"visible":true,"origin":"","legend":"","description":"","filename":"SUP3.tif","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/97ad4ac977980f15530a2af2.tif"},{"id":55005991,"identity":"4f9d47cb-ecb3-485c-9168-a6c723c778c0","added_by":"auto","created_at":"2024-04-19 18:55:55","extension":"docx","order_by":17,"title":"","display":"","copyAsset":false,"role":"supplement","size":2535742,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryfigurelegends.docx","url":"https://assets-eu.researchsquare.com/files/rs-4219358/v1/68191d3ce19275e0d764fbcd.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Machine learning developed a macrophage signature for predicting prognosis, immune infiltration and immunotherapy features in head and neck squamous cell carcinoma","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eHead and neck squamous cell carcinoma (HNSCC) is the sixth most prevalent cancer worldwide, accounting for over 95% of all head and neck tumors [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. While the TNM classification system is the standard for prognosis and treatment determination, patient outcomes vary significantly even within the same TNM stages [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. This variability underscores the need for new biomarkers for an accurate survival prognosis and novel treatment targets to improve patient outcomes.\u003c/p\u003e \u003cp\u003eThe tumor microenvironment (TME) plays a crucial role in HNSCC progression by influencing recurrence, metastasis, and drug resistance [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Tumor-associated macrophages (TAMs), predominantly in the TME, either promote or hinder tumor growth depending on their M1 or M2 polarization [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. The dynamic balance between these macrophage types is critical for cancer progression, with a higher M2/M1 ratio often indicating advanced tumor stages [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIn this study, we explored macrophage-related genes (MRGs) by clustering HNSCC subtypes based on macrophage infiltration. We developed a macrophage-related risk-signature (MRS) model to predict patient outcomes and make diagnostic and therapeutic decisions. This study extended the examination of gene mutations, immune infiltration, and drug sensitivity, offering insights into the role of macrophages in HNSCC at the single-cell level. This research aimed to enhance targeted diagnosis and treatment strategies, fostering personalized patient care for HNSCC.\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Data collection and collation\u003c/h2\u003e \u003cp\u003eWe accessed The Cancer Genome Atlas (TCGA) and GSE65858 datasets in the Gene Expression Omnibus (GEO) database to gather clinicopathological information, gene expression, and genomic mutation data of HNSCC patients [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Single-cell RNA sequencing (scRNA-seq) profiles from nine HNSCC samples, comprising of 1,4087 cells, were retrieved from the GSE182227 database [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. The IMvigor210 dataset helped assess immunotherapy efficacy predictions by our risk model [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. A total of 484 and 270 patients with HNSCC were collected from TCGA and GEO databases, respectively. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows a flowchart of the design of this study.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Macrophage co-expression network construction\u003c/h2\u003e \u003cp\u003eTo estimate immune cell infiltration, HNSCC samples from the TCGA database were analyzed using the TIMER algorithm, while the CIBERSORT method was applied to patients in the GSE65858 dataset [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. We employed the weighted gene co-expression network analysis (WGCNA) algorithm to create a scale-free network by grouping highly correlated genes into modules [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Pearson\u0026rsquo;s test was used to ascertain gene significance (GS) and module membership (MM), linking module eigengenes (MEs) with macrophages.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Acquisition of DEGs related to macrophage\u003c/h2\u003e \u003cp\u003eWe employed the \u0026ldquo;DESeq2\u0026rdquo; R package to detect differentially expressed genes (DEGs) by contrasting HNSCC samples with paired healthy samples in the TCGA dataset [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. DEGs were selected based on the criteria |log2FC| \u0026gt; 1 and FDRs\u0026thinsp;\u0026lt;\u0026thinsp;0.05. We then cross-referenced these DEGs with genes related to macrophage modules, identifying 194 MRGs. Subsequent analyses focused on exploring the enriched cellular pathways of these MRGs, utilizing resources such as the Kyoto Encyclopedia of Genes and Genomes (KEGG), Molecular Signatures Database (MSigDB), and Reactome database.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Consensus Clustering Analysis\u003c/h2\u003e \u003cp\u003eUsing the \u0026ldquo;ConsensusClusterPlus\u0026rdquo; R package, we clustered HNSCC subtypes based on intersected MRGs [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Cluster numbers were determined by analyzing the consensus cumulative distribution function (CDF), with a focus on significant delta area reductions. The subtypes were visualized using 3D principal component analysis (PCA).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Functional Enrichment Analysis\u003c/h2\u003e \u003cp\u003eWe assessed the biological functions of prognostic genes using the \u0026ldquo;clusterProfiler\u0026rdquo; R package, analyzing Gene Ontology (GO) and KEGG pathways. Variations in gene enrichment were examined via gene set variation analysis (GSVA) using \u0026ldquo;GSVA\u0026rdquo; packages [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. The Metascape database provided additional insights into GO/KEGG enrichment [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Immune Landscape Construction\u003c/h2\u003e \u003cp\u003eThe \u0026ldquo;estimate\u0026rdquo; R package calculated stromal and immune scores, estimation scores, and tumor purity. Immune pathway activities were analyzed, and human leukocyte antigen (HLA) and immune checkpoint gene expression were compared between the groups.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7 Tumor Mutation Burden (TMB) Calculation\u003c/h2\u003e \u003cp\u003eTMB was calculated from somatic mutation data on the GDC website (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://portal.gdc.cancer.gov\u003c/span\u003e\u003cspan address=\"https://portal.gdc.cancer.gov\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), identifying genes with high mutation frequencies and visualizing them using \u0026ldquo;maftools\u0026rdquo; package [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. TMB levels in MRG-related subgroups and risk groups were compared using the Wilcoxon test.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.8 Immunotherapy Efficacy Prediction\u003c/h2\u003e \u003cp\u003eImmunophenoscore (IPS) and Tumor Immune Dysfunction and Exclusion (TIDE) scores have been used to predict the efficacy, which were obtained from TCIA (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://tcia.at/home\u003c/span\u003e\u003cspan address=\"https://tcia.at/home\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and TIDE (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://tide.dfci.harvard.edu\u003c/span\u003e\u003cspan address=\"http://tide.dfci.harvard.edu\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), respectively [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. The TISMO platform (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://tismo.cistrome.org\u003c/span\u003e\u003cspan address=\"http://tismo.cistrome.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was used to evaluate the predictive capacity of hub genes for immunotherapy responses in syngeneic mouse immunotherapy cohorts and tumor cell lines.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e2.9 MRS Model Construction\u003c/h2\u003e \u003cp\u003eA consensus MRS model was developed using ten machine learning techniques, integrating 101 algorithm permutations for optimal accuracy and stability [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. The TCGA and GEO patients were divided into training and testing cohorts, and Harrell\u0026rsquo;s concordance index (C-index) was used to evaluate the accuracy of the model.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e2.10 MRS Performance Assessment\u003c/h2\u003e \u003cp\u003eThe \u0026ldquo;survminer\u0026rdquo; and \u0026ldquo;timeROC\u0026rdquo; R packages compared survival outcomes and assessed the MRS\u0026rsquo;s predictive ability in HNSCC, determining the optimal \u0026ldquo;cut-off\u0026rdquo; value for risk groups.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e2.11 Clinical Prognostic Model Construction and Validation\u003c/h2\u003e \u003cp\u003eCox regression analyses evaluated the risk score as an independent factor in the TCGA and GEO cohorts. The \u0026ldquo;regplot\u0026rdquo; package developed a nomogram incorporating patient characteristics for survival assessment.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e2.12 PPI Network and Hub Gene Identification\u003c/h2\u003e \u003cp\u003eThe \u0026ldquo;Cytoscape\u0026rdquo; software, aided by the \u0026ldquo;CytoHubba\u0026rdquo; plugin, constructed a PPI network for DEGs and identified hub genes [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e2.13 Chemotherapeutic Drug Sensitivity Assessment\u003c/h2\u003e \u003cp\u003eThe \u0026ldquo;pRRophetic\u0026rdquo; and \u0026ldquo;ggplot2\u0026rdquo; packages analyzed chemotherapeutic drug sensitivity in HNSCC, comparing IC50 values between risk groups [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e2.14 scRNA-seq Analysis of HNSCC\u003c/h2\u003e \u003cp\u003eThe \u0026ldquo;Seruat\u0026rdquo; package processed scRNA-seq data, identifying distinct cell clusters visualized using UMAP. The \u0026ldquo;TriCycle\u0026rdquo; algorithm analyzed cell cycle correlations, and the \u0026ldquo;irGSEA\u0026rdquo; method explored functional pathways [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e2.15 RT-qPCR and Immunohistochemistry Staining analysis for MRS-related genes\u003c/h2\u003e \u003cp\u003eWe used VeZol Reagent (Vazyme, China) to process the collected 14 pairs of HNSCC tumor tissues and their adjacent normal tissues to extract total RNA, which was then reverse-transcribed into complementary DNA using HiScript QRT SuperMix for qPCR (Vazyme, China). RT-qPCR was performed on an FTC-3000P system (Funglyn Biotech, Canada) using qPCR SYBR Green Master Mix (Vazyme, China) as a fluorescent dye. GAPDH was used as an internal reference in the current research. The experiments were conducted separately and in duplicates. Immunohistochemistry (IHC) data were retrieved from the Human Protein Atlas (HPA, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.proteinatlas.org\u003c/span\u003e\u003cspan address=\"http://www.proteinatlas.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) to evaluate the protein expression levels of MRS-related genes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e2.16 Statistical Analysis\u003c/h2\u003e \u003cp\u003eStatistical analyses were performed using R software (version 4.1.0, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.r-project.org/\u003c/span\u003e\u003cspan address=\"https://www.r-project.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Continuous variables were compared using the Student\u0026rsquo;s t-test and Wilcoxon rank-sum test. Survival curves were constructed using the Kaplan-Meier algorithm, with the Pearson and Spearman methods for correlation analysis. Statistical significance was set at a P-value\u0026thinsp;=\u0026thinsp;0.05.\u003c/p\u003e \u003c/div\u003e"},{"header":"3 Results","content":"\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Macrophage infiltration into HNSCC: hub module identification and enrichment analysis\u003c/h2\u003e \u003cp\u003eIn the TCGA-HNSCC cohort, the blue module was significantly correlated with macrophages (\u003cem\u003eR\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.59, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5e-57) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). In the GSE146771 cohort, the M0, M1, and M2 macrophages had the strongest correlation with the red, tan, and turquoise modules, respectively (M0: \u003cem\u003eR\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.4, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;8e-12, M1: \u003cem\u003eR\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.59, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;4e-26, M2: \u003cem\u003eR\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.35, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6e-09) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). And the correlation values between macrophage module membership and gene significance were shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC-F. Subsequently, we conducted a paired difference analysis of cancerous and adjacent tissues from patients with HNSCC in the TCGA database (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eG). A total of 194 MRGs were identified by intersecting the hub modules and DEGs in the TCGA dataset \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eH\u003cb\u003e)\u003c/b\u003e. A functional enrichment analysis of the 194 MRGs revealed that their biological activities predominantly involved cellular immunity. Different databases emphasized different aspects of the immune system, such as \u0026ldquo;cytokine-cytokine receptor interaction\u0026rdquo; at the center of the KEGG database, while \u0026ldquo;inflammatory response\u0026rdquo; and \u0026ldquo;allograft rejection\u0026rdquo; were the focus of the MisgDB database, and maintenance of immune function and cytokine signaling transmission were the primary concerns of the Reactome database (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eI-K). Meaning that the function of MRGs is mainly related to the immune inflammatory response of cells.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Construction and prognostic analysis of molecular subtypes of MRGs in HNSCC patients\u003c/h2\u003e \u003cp\u003eTo delve deeper into the heterogeneity of macrophages among HNSCC tumors, unsupervised consensus analysis was used to classify the patients in TCGA-HNSCC into two distinct subtypes according to the 194 MRGs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA-D). Patients in cluster 2 had considerably superior overall survival (OS) than that in cluster 1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE). We presented distribution differences between clusters 1 and 2, in terms of clinicopathological features, and simultaneously visualized the top ten genes with upregulated and downregulated levels between the two clusters (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eGSVA revealed notable disparities between the two clusters in multiple functional pathways, and almost all immune and inflammatory response pathways were enriched in cluster 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eG). Finally, we performed KEGG and GO enrichment analyses of the DEGs between clusters 1 and 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eH-I). KEGG analysis suggested that the main pathways were involved in viral protein interactions with cytokine and cytokine receptors. GO analysis revealed that the DEGs were mostly enriched in the BP functional set, namely, leukocyte cell-cell adhesion and regulation of leukocyte proliferation, and were associated with the MF functional set. These findings indicated that MRGs may influence the interaction between cytokines and tumor cells.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003e\u003cb\u003e3.3 Evaluation of TME and biological characteristics of each macrophage-related cluster\u003c/b\u003e\u003c/h2\u003e \u003cp\u003eUsing the ESTIMATE algorithm, we computed the immune score, estimated score, stromal score, and tumor purity to compare the TME and activities of immune-related pathways between the two clusters. Cluster 1 exhibited a significantly lower immune score, stromal score, and estimated score than cluster 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA); conversely, cluster 1 demonstrated a substantially higher tumor purity than cluster 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). Based on these findings, it appeared the two clusters had totally distinct TME infiltration patterns. There were significant differences in the immune cell functions involved between the two clusters, with almost all immune pathways enriched in cluster 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC). Cluster 2 showed higher expression of HLA-related genes and immune checkpoint genes (\u003cem\u003eNRP1, IDO1, LGALS9, CD40, TNFRSF14, CD274\u003c/em\u003e, etc.) and lower expression of \u003cem\u003eCD44, TNFRSF18, TNFSF9\u003c/em\u003e, and \u003cem\u003eTNFSF18\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD-E). To assess the predictive power of the clusters associated with macrophages for ICI response, IPS analysis was performed on patients with HNSCC to ascertain their immunotherapeutic sensitivity. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eF-I, cluster 2 showed a higher IPS score in CTLA4\u003csup\u003e\u0026minus;\u003c/sup\u003e_PD1\u003csup\u003e\u0026minus;\u003c/sup\u003e, CTLA4\u003csup\u003e\u0026minus;\u003c/sup\u003e_PD1\u003csup\u003e+\u003c/sup\u003e, and CTLA4\u003csup\u003e+\u003c/sup\u003e_PD1\u003csup\u003e+\u003c/sup\u003e, which suggested that individuals in cluster 2 may potentially get more advantages from immunotherapy.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTMB, or non-synonymous variation, is strongly correlated with the infiltration of immune cells and activation of immunological responses [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. The somatic cell mutation frequency in patients is 94.18%, with the highest frequencies observed for \u003cem\u003eTP53, TTN, CDKN2A\u003c/em\u003e, and \u003cem\u003eFAT1\u003c/em\u003e mutations (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eJ). Comparison of the TMB scores between clusters 1 and 2 revealed that cluster 1 was significantly higher than cluster 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eK). Furthermore, according to the TMB score, patients were divided into high and low TMB groups; patients with high TMB had a poorer prognosis than those with low TMB (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eL). Additionally, we conducted a stratification study and found that combining the cluster with the TMB risk group could more accurately predict the prognosis of patients with HNSCC (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eM).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Integrative machine learning algorithms constructed an optimal prognostic MRS\u003c/h2\u003e \u003cp\u003eWe evaluated the MRGs and created a prognostic MRS, both accurately and stably, using an integrative approach combined with 10 machine-learning-based algorithms. Consequently, a total of 101 distinct prediction models were obtained, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA. The StepCox[both]\u0026thinsp;+\u0026thinsp;CoxBoost method yielded the optimum model, which consisted of \u003cem\u003eAPOC1, CTLA4, IGF2BP2, CYP27A1, NTN4, SLC7A5, PPP1R14C, KRT9\u003c/em\u003e, and \u003cem\u003eRAC2\u003c/em\u003e, as evidenced by an average C-index of 0.6469 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). We subsequently classified HNSCC cases into high- and low-risk categories based on their risk scores. Differences in age, sex, pathological grade, and clinical stage distribution among patients in the various risk groups are shown in \u003cb\u003eSupplementary Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eA-B\u003c/b\u003e. As anticipated, HNSCC patients with high-risk scores had a significantly poor OS rate in TCGA training (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001, Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB), TCGA testing (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.002, Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC), and GSE65858 (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.037, Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD) cohorts. Analysis of the risk curve revealed that patients in the high-risk group had a greater likelihood of death and a shorter duration of survival. The population of patients in the high-risk group expanded and mortality rates increased with increasing risk scores (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eE-G). Additionally, the AUCs for 1-, 3-, and 5-year OS were shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eH-J.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs shown in \u003cb\u003esupplementary Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e\u003c/b\u003e, age, pathological grade, T status, and N status were significantly associated with risk score in the TCGA training cohort. \u003cb\u003eSupplementary Figure \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e\u003c/b\u003e shows the ROC curves for forecasting patient survival probability. Subsequently, COX analysis suggested that the risk score was an independent prognostic factor for predicting survival and nomogram was constructed to predict the survival duration of patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA-I). Furthermore, we conducted a paired difference analysis to confirm the expression of nine genes comprising MRS, utilizing data from patients with HNSCC in TCGA (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eJ-R). Additionally, we performed functional enrichment analysis and gene set variation analysis (GSVA) on the basis of MRS model, which was shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003e3.5 Identification of hub genes and variation characteristics in high and low risk groups\u003c/h2\u003e \u003cp\u003eA PPI network for the DEGs between high and low risk groups was constructed based on the \u0026ldquo;cytoscape\u0026rdquo; software (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eA). We applied three different algorithms to identify the top 10 hub genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB-D). Further intersections were performed, resulting in the identification of four hub genes: \u003cem\u003eCD80, CTLA4, CCL2\u003c/em\u003e, and \u003cem\u003eIFNG\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eE). We employed TISMO to conduct a comparative analysis of the alterations in the expression of four hub genes in MOC22 tumor mouse models following the administration of anti-PD1 therapy. The findings indicated that the levels of CCL2 and IFNG were significantly increased in the group that responded positively to anti-PD1 therapy compared to the group that did not respond (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eF-I), which suggest that CCL2 and IFNG can be used as biomarkers to predict the responsiveness of HNSCC patients to anti-PD1 treatment.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eA copy number variant (CNV) is a DNA segment and is associated with the heterogeneity of the genome and molecular phenotype, which contributes to the development of tumors [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. We conducted a mutation frequency analysis and the results showed that IFNG had a greater occurrence of gain-of-function mutations in HNSCC, whereas \u003cem\u003eCTLA4\u003c/em\u003e displayed a higher frequency of loss-of-function mutations (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eJ-K). The TMB scores of the high- and low-risk groups were calculated and compared, and no significant differences were observed (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eL). Subsequently, we conducted a correlation analysis, which revealed that the expression levels of \u003cem\u003eCD80, CTLA4\u003c/em\u003e, and \u003cem\u003eCCL2\u003c/em\u003e were negatively correlated with the TMB score, whereas the expression levels of \u003cem\u003eIFNG\u003c/em\u003e and \u003cem\u003eCD80\u003c/em\u003e were positively correlated with the risk score. According to prior research, there is no notable link between risk and TMB scores. There was a possibility of a \u0026ldquo;chain effect\u0026rdquo; in the regulation of gene expression among the four hub genes, where their expression levels are positively associated with one another (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eM).\u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eN, the higher the TMB and risk score, the lower the probability of patient survival. In accordance with the TMB score, individuals classified as high-risk had a greater prevalence of mutations. \u003cem\u003eTP53\u003c/em\u003e had the highest mutation rate in the high-risk group, with a mutation frequency of 80%, whereas the mutation frequency was 60% in the low-risk group. Notably, \u003cem\u003eTTN\u003c/em\u003e had a higher mutation rate in the low-risk group (40%) than in the high-risk group (36%). The main types of \u003cem\u003eTP53\u003c/em\u003e and \u003cem\u003eTTN\u003c/em\u003e mutations were missense mutations and multiple hits (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eO-P).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003e3.6 The estimate of immunotherapy for MRS model\u003c/h2\u003e \u003cp\u003eWe conducted a comparative analysis of the proportions of the immunological subtypes of HNSCC across various risk groups. While the link between them may not be statistically significant, it was evident that clusters 1 and 2 exhibited a substantial association in terms of survival outcomes. Specifically, patients in cluster 1 predominantly exhibited survival results, whereas those in cluster 2 mainly died (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eA-B). We observed substantial disparities in the expression of immune checkpoint genes between high- and low-risk groups. \u003cem\u003eCD44, CD276, NRP1, CD274, LAIR1, HAVCR2, CD86, PDCD1LG2\u003c/em\u003e, and \u003cem\u003eCD80\u003c/em\u003e were predominantly overexpressed in the high-risk group. Conversely, \u003cem\u003eTNFRSF18, TNFRSF25\u003c/em\u003e, and \u003cem\u003eCD200\u003c/em\u003e were highly expressed in the low-risk group (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eC). In terms of HLA gene expression, except for \u003cem\u003eHLA-DOB\u003c/em\u003e, which was significantly overexpressed in the low-risk group, all other differentially expressed genes were enriched in the high-risk group (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eD).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eUsing the TIDE methodology, we found that the risk score was positively correlated with the TIDE score (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eE). The high-risk group exhibited a considerably lower response rate than the low-risk group. Conversely, the risk score of the immune response group was much lower than that of the non-response group (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eF-G). Patients with metastatic urothelial carcinoma in the IMvigor210 cohort were divided into high- and low-risk groups. In both groups, the proportions of inflammatory immune subtypes and tumor-infiltrating immune cells expressing PD-L1 differed significantly (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eH-I). Nevertheless, the treatment outcomes following immunotherapy did not differ considerably between two groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eJ, \u003cb\u003eSupplementary Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eC-D\u003c/b\u003e). It is worth mentioning that the survival probability of patients in the high-risk group was significantly lower than that in the low-risk group, which is consistent with the conclusion obtained from the HNSCC dataset (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eK-L). Additionally, the chemotherapy drugs sensitivity analysis was showed in Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section2\"\u003e \u003ch2\u003e3.7 MRS gene expression patterns at single cell resolution.\u003c/h2\u003e \u003cp\u003eTo further investigate the patterns of gene expression in the MRS model at the single-cell level, we analyzed the single-cell dataset GSE1822271. After quality control and dimensional reduction, a grand total of 30 cell clusters were obtained (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003eA-B). The top five differentially expressed marker genes in each cell type are shown as a heatmap (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003eC). We further visualized the distribution of nine genes in MRS across different cell types, among which \u003cem\u003eSLC7A5, RAC2\u003c/em\u003e, and \u003cem\u003eAPOC1\u003c/em\u003e were enriched in macrophages, with \u003cem\u003eAPOC1\u003c/em\u003e being the most highly expressed (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003eD-E).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eEvery cell type plays a role in the onset and progression of the disease, and follows a distinct cell cycle. Data analysis revealed that macrophages were mostly in the transitional phase from G2 to M, with the lowest cell count observed in the S phase, suggesting that macrophage division and proliferation were more active. Simultaneously, fibroblasts mostly resided in the G2 and S phases and were primarily tasked with duplicating genetic material and preparing for cellular division (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003eF-G). Functional enrichment analysis showed that epithelial mesenchymal transition was significantly upregulated in fibroblasts, but downregulated in macrophages, mast cells, plasma cells, and T cells. We also found that epithelial cells, endothelial cells, fibroblasts, and macrophages were involved in the activation of tumor-related pathways, whereas mast cells, plasma cells, and T cells were mainly involved in the inhibition of related pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003eH-I). It can be seen that macrophages, epithelial cells, and fibroblasts were more abundant in the HPV\u003csup\u003e\u0026minus;\u003c/sup\u003e group, while endothelial cells and T cells were more abundant in the HPV\u003csup\u003e+\u003c/sup\u003e group (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003eK). In addition, the nine genes in MRS were mainly enriched in macrophages, T cells, epithelial cells, and mast cells, with the highest proportion of high expression in macrophages and T cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003eL-M). To some extent, this confirms the effectiveness of our MRS prognostic model.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec27\" class=\"Section2\"\u003e \u003ch2\u003e3.8 Authentication of MRS-related genes via qRT-PCR and HPA platform\u003c/h2\u003e \u003cp\u003eWe conducted RT-qPCR to quantify mRNA expression levels of the nine genes in the MRS model to confirm our bioinformatic findings. Figure\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003eA showing our experimental findings, consistent with earlier findings, all genes except \u003cem\u003eCYP27A1\u003c/em\u003e and \u003cem\u003eNTN4\u003c/em\u003e showed elevated expression levels in cancer. Additionally, we used the HPA database to obtain immunohistochemical images to evaluate the protein expression of MRS-related genes. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003eB, IGF2BP2, PPP1R14C, KRT9, RAC2, SLC7A5, and APOC1 protein expression was substantially higher in tumor tissues than in normal tissues, whereas CYP27A1 and NTN4 showed lower expression levels. Thus, the experimental outcomes of our study effectively confirmed the bioinformatics-based conclusions, thereby strengthening the importance of our work. Validating our bioinformatics predictions is crucial, as it enhances the dependability and trustworthiness of our results. The congruity between experimental and bioinformatics data offers compelling evidence to substantiate MRS-related genes and their regulatory networks.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eHNSCC, the most prevalent malignant tumor of the head and neck, has increased in incidence owing to risk factors such as smoking, alcohol consumption, environmental pollutants, and pathogenic viruses such as HPV and EBV [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Immunotherapy has emerged as a promising treatment for HNSCC as it reactivates the ability of the immune system to recognize cancer cells [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. TAMs, as a crucial role in immune infiltration in HNSCC can activate pathways that increase tumor stemness and chemotherapy resistance, reduce chemotherapy sensitivity, and elevate the risk of recurrence in some HNSCC patients [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Single-cell studies have also revealed that TAMs promote angiogenesis, lymph node metastasis, and tumor invasion by secreting CCL8 and CXCL18, thus promoting HNSCC progression [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eWe quantitatively analyzed the infiltration abundance of macrophages in the two cohorts, identified key modules related to MRGs in HNSCC, ultimately obtaining 194 key MRGs. We classified HNSCC into two molecular subtypes and found that the OS of patients in cluster 1 was lower than that in cluster 2, which may be linked to the differences in enriched functional pathways: cluster 1 was enriched in metabolic pathways, while cluster 2 was mainly related to inflammatory response, which may be due to the strong \u0026ldquo;immune clearance effect\u0026rdquo; of the body\u0026rsquo;s immune system on tumor cells. In addition, cluster 2 had higher matrix, immune, and infiltration scores than cluster 1, and the functional pathways of immune cells and the expression of HLA molecules were enriched in cluster 2. Owing to immune editing, the absence or downregulation of HLA molecules is a common early event in cancer occurrence, which can lead to impaired recognition of tumor cells by cytotoxic T cells, resulting in immune escape and is closely associated with tumor recurrence and metastasis [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. The ICI gene was predominantly expressed in cluster 2, which was closely linked to malignant characteristics such as epithelial-mesenchymal transition, angiogenesis, and the dissemination and invasion of malignancies. We hypothesized that highly expressed ICI genes may not play a prominent role in cluster 2 [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. The tumor purity of cluster 2 was significantly lower than that of cluster1, and the treatment sensitivity against CTLA4 and PD1 was higher, suggesting that attention should be paid to the role of immunotherapy in the treatment of HNSCC patients in cluster 2, which can be combined with conventional methods to further improve patient survival.\u003c/p\u003e \u003cp\u003eTo better apply macrophage-related typing to the diagnosis and treatment of HNSCC, we constructed an MRS using a combination of 101 machine learning algorithms to further identify molecular targets related to the prognosis of HNSCC. A total of nine genes were found to comprise the MRS, which was utilized to stratify patients with HNSCC into high- and low-risk groups. The high-risk group exhibited a considerably poorer survival rate, and the risk score may serve as an independent risk factor for predicting patient survival outcomes.\u003c/p\u003e \u003cp\u003eAmong the nine genes in MRS, APOC1 has been shown that inhibition of its expression can induce ferroptosis to reverse M2 macrophages to M1 macrophages, leading to immune activation and increased sensitivity to PD1 therapy [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. More detailed studies have indicated that APOC1 inhibits KEAP1, promotes NRF2 nuclear translocation, and increases the expression of cystathionine-beta-synthase (CBS) to inhibit ferroptosis, thereby promoting tumor proliferation [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. CYP27A1 mainly regulates cholesterol homeostasis using the CYP27A1/27-hydroxycholesterol (27-HC) axis as a mediator to regulate cell apoptosis and the cell cycle and is associated with good clinicopathological characteristics and prognosis [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Similarly, NTN4 is also associated with better survival rates facilitated by Wnt/β- Catenin signal transduction, and is positively correlated with the degree of infiltration of immune cells, including macrophages and neutrophils, as well as the immunosuppressive condition of tumors [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. Previous studies have demonstrated that IGF2BP2 promotes the clearance, synthesis, metabolism, and growth of HNSCC cells and that its overexpression is a risk factor for poor prognosis in patients. Furthermore, by increasing the stability of CDK6, IGF2BP2 can upregulate its expression, which in turn promotes malignant characteristics [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. The CTLA4 molecule, which is mainly located on the surface of T cells, can effectively bind to the B7 protein to induce T cell dysfunction and immune suppression [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. SLC7A5 maintains the levels of essential amino acids in cancer cells through transcription and metabolic recording, which are crucial for the proliferation of cancer cells. Specific knockout of SLC7A5 significantly downregulates the mTORC1 signaling pathway in cancer cells, mobilizing the general amino acid control (GAAC) pathway to inhibit the synthesis of specific proteins, thereby hindering tumor progression [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. Other genes in the MRS, including \u003cem\u003eKRT9, PPP1R14C\u003c/em\u003e, and \u003cem\u003eRAC2\u003c/em\u003e, can influence to varying degrees the proliferation, invasion, metastasis, and angiogenesis of tumor cells, as well as the functions of immune cells in the TME. The specific mechanisms by which these genes regulate HNSCC warrant further investigation.\u003c/p\u003e \u003cp\u003eCNV is a driving force in cancer development and plays a critical role in oncogene activation and tumor suppressor gene inactivation [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. Data analysis revealed that CTLA4 is prone to copy number loss, whereas IFNG is prone to copy number amplification. TMB can reflect the number of gene mutations in tumor cells and predict the sensitivity and resistance of patients to ICI treatment; the more mutations, the more new antigens, and the higher the probability of triggering a T cell immune response [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. Our analysis revealed a significant correlation between the expression levels of hub genes \u003cem\u003eCD80\u003c/em\u003e, \u003cem\u003eCTLA4\u003c/em\u003e, and \u003cem\u003eCCL2\u003c/em\u003e and TMB. Previous studies have demonstrated that T cells in the TME, activated in a CTLA4 dependent manner, exhibit strong immunogenicity in high-TMB tumor cells, leading to a series of immune responses [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. In addition, \u003cem\u003eTP53\u003c/em\u003e showed a higher mutation frequency in the high-risk group, whereas \u003cem\u003eTTN\u003c/em\u003e showed a higher mutation frequency in the low-risk group. \u003cem\u003eTP53\u003c/em\u003e, as the gene with the highest mutation frequency in human cancer, plays an important role in inducing cell apoptosis, aging, cell cycle arrest, and DNA damage repair, and is associated with increased chromosomal instability [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. Previous studies have shown that \u003cem\u003eTP53\u003c/em\u003e mutations can increase the expression of relevant immune checkpoint molecules, and inhibit the infiltration of cytotoxic CD8\u003csup\u003e+\u003c/sup\u003eT cells, leading to immunosuppressive states that induce macrophage polarization towards the M2 phenotype of TAMs, resulting in sustained tumor progression and a higher mutation frequency in high-risk groups [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. TTN mutations can cause varied gene expression in cancers, boost tumor immune response, and increase susceptibility to immunotherapy [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. Mutations in \u003cem\u003eTTN\u003c/em\u003e are strongly associated with the conversion of macrophages into M1 type TAM, which could explain the increased occurrence of \u003cem\u003eTTN\u003c/em\u003e mutations in the low-risk group [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eCancer immunotherapy is an innovative strategy that seeks to augment the capacity of the immune system to combat cancer by performing radical cancer treatment and preventing recurrence [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. We observed that the vast majority of immune checkpoint genes and HLA genes were significantly upregulated in the high-risk group, which could potentially affect the level of immune infiltration and exacerbate tumor growth by recruiting various immune cells or cytokines. The TIDE score of tumors was positively correlated with the risk score of the MRS model, indicating that the tumor cells in the high-risk group have a stronger immune escape ability, whereas patients in the low-risk group may be more sensitive to ICI therapy and have a higher positive immune response rate. To confirm the reliability of the MRS model in accurately forecasting the effectiveness of immunotherapy and patient survival, we used clinical data from individuals with urothelial carcinoma in the IMvigor210 cohort to assess the predictive capacity of immunotherapy. We found the high-risk group had lower survival rates and the predictive power of the risk score for patient survival was higher. It is worth noting that the sensitivity of patients with HNSCC to chemotherapy drugs varied among different risk groups, which was related to tumor heterogeneity caused by mutations in drug resistance-related genes and crosstalk between different immune cells [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. Hence, the MRS model score can be utilized to determine specific chemotherapeutic medications for patients with HNSCC, thereby attaining precision in medicine.\u003c/p\u003e \u003cp\u003eFinally, our single-cell analysis indicated that SLC7A5, RAC2, and APOC1 were highly expressed in the macrophage subpopulation in the MRS. Macrophages in the TME mainly stagnated in G2.M phase, which is related to the downregulation of multiple functional pathways, with the inhibition of epithelial-mesenchymal transition being the most prominent and the highest abundance in the HPV\u003csup\u003e\u0026minus;\u003c/sup\u003e patient population, all of which confirm that macrophages themselves may play a favorable role in the progression of HNSCC. This phenomenon could be attributed to the scavenging functions of macrophages, which also regulate the immune response to tumor cells and preserve tissue homeostasis. By regulating the differentiation of TAM into M2 macrophages with cytokines or metabolites, the TME may induce anti-tumor immunity [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThis study aimed to categorize HNSCC patients into subtypes based on MRGs, identify DEGs among these clusters, and develop an MRS model for patient prognosis assessment to aid immunotherapy selection. Through comprehensive validation across various perspectives and databases, we developed an MRS model that effectively predicts patient survival and provides information on treatment strategies. However, this study has some limitations. HNSCC patient data, sourced solely from public databases, often lack critical details, such as HPV infection status and microvascular infiltration. However, the effects of these factors remain unclear. To authenticate and assess the diagnostic and therapeutic efficacy of the MRS model, future studies should include broader multicenter clinical cohorts. Additionally, our validation of MRS-related gene expression was restricted to the RNA and protein levels. The specific molecular mechanisms and pathways in HNSCC have been inferred from the literature and warrant further exploration through in vivo and in vitro experiments.\u003c/p\u003e \u003cp\u003eIn summary, this study identified characteristic genes associated with macrophages in HNSCC and established an MRS model that fully validated its diagnostic efficacy in predicting patient survival prognosis, clarified its potential relationship with tumor cell genome mutations, and provided a theoretical basis for immunotherapy and chemotherapy drug selection. These results will help us further explore the characteristics and related molecular mechanisms of macrophage immune infiltration in HNSCC to identify new molecular targets for personalized diagnosis and treatment of HNSCC patients in the future.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e: Not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e: The studies involving human participants were reviewed and approved by Ethics Committee of Yantai Yuhuangding Hospital. The patients/participants provided their written informed consent to participate in this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e: Not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e: The data underlying this article are available in the Gene Expression Omnibus (GEO) database at https://www.ncbi.nlm.nih.gov/geo/ and The Cancer Genome Atlas (TCGA) at https://portal.gdc.cancer.gov/, and can be accessed with GSE65858 and GSE182227.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e: The authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e This work was supported by Taishan Scholar Project (No.ts20190991), the Key R\u0026amp;D Project of Shandong Province (2022CXPT023) and the Natural Science Foundation of Shandong Province (No. ZR2019PH113).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e: X-CS and CR: conception, design and administrative support. YW, Y-KM and W-CL: study conception, data analysis and visualization. YW, Y-KM and H-RW, X-YS and TY : data collection and manuscript revision. YW, Y-KM and W-CL, CR and X-CS: data analysis and interpretation. All authors: manuscript writing and final approval of manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eGlobal Burden of Disease Cancer Collaboration; Fitzmaurice C, Allen C, et al. Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With Disability, and Disability-Adjusted Life-years for 32 Cancer Groups, 1990 to 2015: A Systematic Analysis for the Global Burden of Disease Study. JAMA Oncol. 2017;3(4):524\u0026ndash;548. doi: 10.1001/jamaoncol.2016.5688. Erratum in: JAMA Oncol. 2017;3(3):418. PMID: 27918777.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTakes RP, Rinaldo A, Silver CE, et al. Future of the TNM classification and staging system in head and neck cancer. Head Neck. 2010;32(12):1693\u0026thinsp;\u0026ndash;\u0026thinsp;711. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hed.21361\u003c/span\u003e\u003cspan address=\"10.1002/hed.21361\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 20191627.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin Y, Zheng X, Gao W, Wang B, Wu Y. Tumor microenvironment and immune-related therapies of head and neck squamous cell carcinoma. Mol Ther Oncolytics. 2021;20:342\u0026ndash;351. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.omto.2021.01.011\u003c/span\u003e\u003cspan address=\"10.1016/j.omto.2021.01.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 33614915.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMantovani A, Marchesi F, Malesci A, Laghi L, Allavena P. Tumour-associated macrophages as treatment targets in oncology. Nat Rev Clin Oncol. 2017;14(7):399\u0026ndash;416. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/nrclinonc.2016.217\u003c/span\u003e\u003cspan address=\"10.1038/nrclinonc.2016.217\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2017 Jan 24. PMID: 28117416.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYunna C, Mengru H, Lei W, Weidong C. Macrophage M1/M2 polarization. Eur J Pharmacol. 2020;877:173090. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ejphar.2020.173090\u003c/span\u003e\u003cspan address=\"10.1016/j.ejphar.2020.173090\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2020 Mar 29. PMID: 32234529.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWichmann G, Rosolowski M, Krohn K, et al. The role of HPV RNA transcription, immune response-related gene expression and disruptive TP53 mutations in diagnostic and prognostic profiling of head and neck cancer. Int J Cancer. 2015;137(12):2846\u0026ndash;57. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ijc.29649\u003c/span\u003e\u003cspan address=\"10.1002/ijc.29649\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2015 Jul 6. PMID: 26095926.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePuram SV, Mints M, Pal A, et al. Cellular states are coupled to genomic and viral heterogeneity in HPV-related oropharyngeal carcinoma. Nat Genet. 2023;55(4):640\u0026ndash;650. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41588-023-01357-3\u003c/span\u003e\u003cspan address=\"10.1038/s41588-023-01357-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2023 Apr 3. PMID: 37012457.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeng D, Li M, Zhou R, et al. Tumor Microenvironment Characterization in Gastric Cancer Identifies Prognostic and Immunotherapeutically Relevant Gene Signatures. Cancer Immunol Res. 2019;7(5):737\u0026ndash;750. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1158/2326-6066.CIR-18-0436\u003c/span\u003e\u003cspan address=\"10.1158/2326-6066.CIR-18-0436\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2019 Mar 6. PMID: 30842092.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi T, Fu J, Zeng Z, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020;48(W1):W509-W514. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/nar/gkaa407\u003c/span\u003e\u003cspan address=\"10.1093/nar/gkaa407\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 32442275.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewman AM, Liu CL, Green MR, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12(5):453\u0026ndash;7. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/nmeth.3337\u003c/span\u003e\u003cspan address=\"10.1038/nmeth.3337\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2015 Mar 30. PMID: 25822800.\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. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/1471-2105-9-559\u003c/span\u003e\u003cspan address=\"10.1186/1471-2105-9-559\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 19114008.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcDermaid A, Monier B, Zhao J, Liu B, Ma Q. Interpretation of differential gene expression results of RNA-seq data: review and integration. Brief Bioinform. 2019;20(6):2044\u0026ndash;2054. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bib/bby067\u003c/span\u003e\u003cspan address=\"10.1093/bib/bby067\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 30099484.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics. 2010;26(12):1572\u0026ndash;3. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bioinformatics/btq170\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btq170\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2010 Apr 28. PMID: 20427518.\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. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/1471-2105-14-7\u003c/span\u003e\u003cspan address=\"10.1186/1471-2105-14-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 23323831.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou Y, Zhou B, Pache L, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-019-09234-6\u003c/span\u003e\u003cspan address=\"10.1038/s41467-019-09234-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 30944313.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMayakonda A, Lin DC, Assenov Y, Plass C, Koeffler HP. Maftools: efficient and comprehensive analysis of somatic variants in cancer. Genome Res. 2018;28(11):1747\u0026ndash;1756. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/gr.239244.118\u003c/span\u003e\u003cspan address=\"10.1101/gr.239244.118\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2018 Oct 19. PMID: 30341162.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang T, Dai L, Shen S, et al. Comprehensive Molecular Analyses of a Macrophage-Related Gene Signature With Regard to Prognosis, Immune Features, and Biomarkers for Immunotherapy in Hepatocellular Carcinoma Based on WGCNA and the LASSO Algorithm. Front Immunol. 2022;13:843408. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fimmu.2022.843408\u003c/span\u003e\u003cspan address=\"10.3389/fimmu.2022.843408\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 35693827.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang H, Awuti G, Guo X. Construction of an Immunophenoscore-Related Signature for Evaluating Prognosis and Immunotherapy Sensitivity in Ovarian Cancer. ACS Omega. 2023;8(36):33017\u0026ndash;33031. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acsomega.3c04856\u003c/span\u003e\u003cspan address=\"10.1021/acsomega.3c04856\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 37720747.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Z, Liu L, Weng S, et al. Machine learning-based integration develops an immune-derived lncRNA signature for improving outcomes in colorectal cancer. Nat Commun. 2022;13(1):816. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-022-28421-6\u003c/span\u003e\u003cspan address=\"10.1038/s41467-022-28421-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 35145098.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDoncheva NT, Morris JH, Gorodkin J, Jensen LJ. Cytoscape StringApp: Network Analysis and Visualization of Proteomics Data. J Proteome Res. 2019;18(2):623\u0026ndash;632. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acs.jproteome.8b00702\u003c/span\u003e\u003cspan address=\"10.1021/acs.jproteome.8b00702\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2018 Dec 5. PMID: 30450911.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYan C, Niu Y, Li F, Zhao W, Ma L. System analysis based on the pyroptosis-related genes identifies GSDMC as a novel therapy target for pancreatic adenocarcinoma. J Transl Med. 2022;20(1):455. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12967-022-03632-z\u003c/span\u003e\u003cspan address=\"10.1186/s12967-022-03632-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 36199146.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTan Z, Chen X, Zuo J, Fu S, Wang H, Wang J. Comprehensive analysis of scRNA-Seq and bulk RNA-Seq reveals dynamic changes in the tumor immune microenvironment of bladder cancer and establishes a prognostic model. J Transl Med. 2023;21(1):223. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12967-023-04056-z\u003c/span\u003e\u003cspan address=\"10.1186/s12967-023-04056-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 36973787.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao S, Ye B, Chi H, Cheng C, Liu J. Identification of peripheral blood immune infiltration signatures and construction of monocyte-associated signatures in ovarian cancer and Alzheimer's disease using single-cell sequencing. Heliyon. 2023;9(7):e17454. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.heliyon.2023.e17454\u003c/span\u003e\u003cspan address=\"10.1016/j.heliyon.2023.e17454\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 37449151.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAddeo A, Friedlaender A, Banna GL, Weiss GJ. TMB or not TMB as a biomarker: That is the question. Crit Rev Oncol Hematol. 2021;163:103374. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.critrevonc.2021.103374\u003c/span\u003e\u003cspan address=\"10.1016/j.critrevonc.2021.103374\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2021 Jun 2. PMID: 34087341.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShlien A, Malkin D. Copy number variations and cancer. Genome Med. 2009;1(6):62. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/gm62\u003c/span\u003e\u003cspan address=\"10.1186/gm62\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 19566914.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJohnson DE, Burtness B, Leemans CR, Lui VWY, Bauman JE, Grandis JR. Head and neck squamous cell carcinoma. Nat Rev Dis Primers. 2020;6(1):92. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41572-020-00224-3\u003c/span\u003e\u003cspan address=\"10.1038/s41572-020-00224-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Erratum in: Nat Rev Dis Primers. 2023;9(1):4. PMID: 33243986.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhosh S, Shah PA, Johnson FM. Novel Systemic Treatment Modalities Including Immunotherapy and Molecular Targeted Therapy for Recurrent and Metastatic Head and Neck Squamous Cell Carcinoma. Int J Mol Sci. 2022;23(14):7889. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ijms23147889\u003c/span\u003e\u003cspan address=\"10.3390/ijms23147889\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 35887235.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHsieh CY, Lin CC, Huang YW, et al. Macrophage secretory IL-1β promotes docetaxel resistance in head and neck squamous carcinoma via SOD2/CAT-ICAM1 signaling. JCI Insight. 2022;7(23):e157285. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1172/jci.insight.157285\u003c/span\u003e\u003cspan address=\"10.1172/jci.insight.157285\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 36264639.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu J, Shen Y, Zeng G, Liang Y, Liao G. SPP1\u0026thinsp;+\u0026thinsp;TAM subpopulations in tumor microenvironment promote intravasation and metastasis of head and neck squamous cell carcinoma. Cancer Gene Ther. 2023 Dec 5. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41417-023-00704-0\u003c/span\u003e\u003cspan address=\"10.1038/s41417-023-00704-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub ahead of print. PMID: 38052857.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Y, Jasinski-Bergner S, Wickenhauser C, Seliger B. Cancer Immunology: Immune Escape of Tumors-Expression and Regulation of HLA Class I Molecules and Its Role in Immunotherapies. Adv Anat Pathol. 2023;30(3):148\u0026ndash;159. doi: 10.1097/PAP.0000000000000389. Epub 2022 Dec 15. PMID: 36517481.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Y, Zheng J. Functions of Immune Checkpoint Molecules Beyond Immune Evasion. Adv Exp Med Biol. 2020;1248:201\u0026ndash;226. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-981-15-3266-5_9\u003c/span\u003e\u003cspan address=\"10.1007/978-981-15-3266-5_9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 32185712.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHao X, Zheng Z, Liu H, et al. Inhibition of APOC1 promotes the transformation of M2 into M1 macrophages via the ferroptosis pathway and enhances anti-PD1 immunotherapy in hepatocellular carcinoma based on single-cell RNA sequencing. Redox Biol. 2022;56:102463. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.redox.2022.102463\u003c/span\u003e\u003cspan address=\"10.1016/j.redox.2022.102463\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2022 Sep 2. PMID: 36108528.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng XJ, Chen WL, Yi J, et al. Apolipoprotein C1 promotes glioblastoma tumorigenesis by reducing KEAP1/NRF2 and CBS-regulated ferroptosis. Acta Pharmacol Sin. 2022;43(11):2977\u0026ndash;2992. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41401-022-00917-3\u003c/span\u003e\u003cspan address=\"10.1038/s41401-022-00917-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2022 May 17. Erratum in: Acta Pharmacol Sin. 2022;: PMID: 35581292.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiang Z, Chen Y, Wang L, et al. CYP27A1 inhibits bladder cancer cells proliferation by regulating cholesterol homeostasis. Cell Cycle. 2019;18(1):34\u0026ndash;45. doi: 10.1080/15384101.2018.1558868. Epub 2018 Dec 30. PMID: 30563407.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKe S, Guo J, Wang Q, et al. Netrin Family Genes as Prognostic Markers and Therapeutic Targets for Clear Cell Renal Cell Carcinoma: Netrin-4 Acts through the Wnt/β-Catenin Signaling Pathway. Cancers (Basel). 2023;15(10):2816. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/cancers15102816\u003c/span\u003e\u003cspan address=\"10.3390/cancers15102816\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 37345154.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang X, Tang Q, Li S, Li M, Yang T. IGF2BP2 acts as a m6A modification regulator in laryngeal squamous cell carcinoma through facilitating CDK6 mRNA stabilization. Cell Death Discov. 2023;9(1):371. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41420-023-01669-7\u003c/span\u003e\u003cspan address=\"10.1038/s41420-023-01669-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 37816718.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu GT, Bu LL, Zhao YY, et al. CTLA4 blockade reduces immature myeloid cells in head and neck squamous cell carcinoma. Oncoimmunology. 2016;5(6):e1151594. doi: 10.1080/2162402X.2016.1151594. PMID: 27471622.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKanai Y. Amino acid transporter LAT1 (SLC7A5) as a molecular target for cancer diagnosis and therapeutics. Pharmacol Ther. 2022;230:107964. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.pharmthera.2021.107964\u003c/span\u003e\u003cspan address=\"10.1016/j.pharmthera.2021.107964\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2021 Aug 12. PMID: 34390745.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTao Z, Wang S, Wu C, et al. The repertoire of copy number alteration signatures in human cancer. Brief Bioinform. 2023;24(2):bbad053. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bib/bbad053\u003c/span\u003e\u003cspan address=\"10.1093/bib/bbad053\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 36806386.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJardim DL, Goodman A, de Melo Gagliato D, Kurzrock R. The Challenges of Tumor Mutational Burden as an Immunotherapy Biomarker. Cancer Cell. 2021;39(2):154\u0026ndash;173. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ccell.2020.10.001\u003c/span\u003e\u003cspan address=\"10.1016/j.ccell.2020.10.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2020 Oct 29. PMID: 33125859.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChan TA, Yarchoan M, Jaffee E, et al. Development of tumor mutation burden as an immunotherapy biomarker: utility for the oncology clinic. Ann Oncol. 2019;30(1):44\u0026ndash;56. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/annonc/mdy495\u003c/span\u003e\u003cspan address=\"10.1093/annonc/mdy495\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 30395155.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Z, Strasser A, Kelly GL. Should mutant TP53 be targeted for cancer therapy? Cell Death Differ. 2022;29(5):911\u0026ndash;920. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41418-022-00962-9\u003c/span\u003e\u003cspan address=\"10.1038/s41418-022-00962-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2022 Mar 24. PMID: 35332311.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDonehower LA, Soussi T, Korkut A, et al. Integrated Analysis of TP53 Gene and Pathway Alterations in The Cancer Genome Atlas. Cell Rep. 2019;28(5):1370\u0026ndash;1384.e5. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.celrep.2019.07.001\u003c/span\u003e\u003cspan address=\"10.1016/j.celrep.2019.07.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Erratum in: Cell Rep. 2019;28(11):3010. PMID: 31365877.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi Y, Ren X, Cao S, et al.TP53 gain-of-function mutation modulates the immunosuppressive microenvironment in non-HPV-associated oral squamous cell carcinoma. J Immunother Cancer. 2023;11(8):e006666. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/jitc-2023-006666\u003c/span\u003e\u003cspan address=\"10.1136/jitc-2023-006666\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 37604640.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Z, Zhao X, Wang R, et al. Heterogeneous pattern of gene expression driven by TTN mutation is involved in the construction of a prognosis model of lung squamous cell carcinoma. Front Oncol. 2023;13:916568. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fonc.2023.916568\u003c/span\u003e\u003cspan address=\"10.3389/fonc.2023.916568\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 37035196.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXie X, Tang Y, Sheng J, et al. Titin Mutation Is Associated With Tumor Mutation Burden and Promotes Antitumor Immunity in Lung Squamous Cell Carcinoma. Front Cell Dev Biol. 2021;9:761758. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fcell.2021.761758\u003c/span\u003e\u003cspan address=\"10.3389/fcell.2021.761758\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 34746153.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSzeto GL, Finley SD. Integrative Approaches to Cancer Immunotherapy. Trends Cancer. 2019;5(7):400\u0026ndash;410. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.trecan.2019.05.010\u003c/span\u003e\u003cspan address=\"10.1016/j.trecan.2019.05.010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 31311655.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNussinov R, Tsai CJ, Jang H. Anticancer drug resistance: An update and perspective. Drug Resist Updat. 2021;59:100796. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.drup.2021.100796\u003c/span\u003e\u003cspan address=\"10.1016/j.drup.2021.100796\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2021 Dec 16. PMID: 34953682.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMehla K, Singh PK. Metabolic Regulation of Macrophage Polarization in Cancer. Trends Cancer. 2019;5(12):822\u0026ndash;834. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.trecan.2019.10.007\u003c/span\u003e\u003cspan address=\"10.1016/j.trecan.2019.10.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2019 Nov 6. PMID: 31813459.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang J, Zhou X, Hao H. Macrophage phenotype-switching in cancer. Eur J Pharmacol. 2022;931:175229. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ejphar.2022.175229\u003c/span\u003e\u003cspan address=\"10.1016/j.ejphar.2022.175229\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2022 Aug 21. PMID: 36002039.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"HNSCC, Macrophage, Prognostic model, Risk score, Immunotherapy","lastPublishedDoi":"10.21203/rs.3.rs-4219358/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4219358/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eMacrophages played an important role in the progression and treatment of cancer. Nevertheless, there is a limited amount of research that has comprehensively elucidated the characteristics of macrophages associated genes in head and neck squamous cell carcinoma (HNSCC). We employed weighted gene co-expression network analysis (WGCNA) to identify macrophage-related genes (MRGs) and classify patients with HNSCC into two distinct subtypes. A macrophage-related risk signature (MRS) model, comprising nine genes: \u003cem\u003eIGF2BP2, PPP1R14C, SLC7A5, KRT9, RAC2, NTN4, CTLA4, APOC1\u003c/em\u003e, and \u003cem\u003eCYP27A1\u003c/em\u003e, was formulated by integrating 101 machine learning algorithm combinations. We observed lower overall survival (OS) in the high-risk group and the high-risk group showed elevated expression levels in most of the differentially expressed immune checkpoint and human leukocyte antigen (HLA) genes, suggesting a strong immune evasion capacity in these tumors. Correspondingly, TIDE score positively correlated with risk score, implying that high-risk tumors may resist immunotherapy more effectively. At the single-cell level, we noted macrophages in the TME predominantly stalled in the G2/M phase, potentially hindering epithelial-mesenchymal transition and playing a crucial role in the inhibition of tumor progression. Additionally, we validated MRS gene expression levels using RT-qPCR and immunohistochemistry (IHC). The current study constructed a novel MRS for HNSCC, which could serve as an indicator for predicting the prognosis, immune infiltration and immunotherapy benefits for HNSCC patients.\u003c/p\u003e","manuscriptTitle":"Machine learning developed a macrophage signature for predicting prognosis, immune infiltration and immunotherapy features in head and neck squamous cell carcinoma","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-04-19 18:47:47","doi":"10.21203/rs.3.rs-4219358/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-06-13T10:04:50+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-26T03:30:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"160453535162381951760846990140964843730","date":"2024-05-15T21:03:26+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-04-25T02:49:21+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"48c57424-9419-4014-a03a-8a03ca50c9ba","date":"2024-04-18T00:50:15+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-04-17T19:55:10+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-04-17T19:53:16+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2024-04-14T09:02:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-04-11T12:39:48+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2024-04-04T18:14:42+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d15b2d85-d457-4b80-89bc-2f578117974a","owner":[],"postedDate":"April 19th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":30648255,"name":"Biological sciences/Computational biology and bioinformatics/Machine learning"},{"id":30648256,"name":"Biological sciences/Biochemistry"},{"id":30648257,"name":"Biological sciences/Biological techniques"},{"id":30648258,"name":"Biological sciences/Cancer"},{"id":30648259,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":30648260,"name":"Health sciences/Biomarkers"}],"tags":[],"updatedAt":"2024-08-26T16:01:39+00:00","versionOfRecord":{"articleIdentity":"rs-4219358","link":"https://doi.org/10.1038/s41598-024-70430-6","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2024-08-22 15:57:22","publishedOnDateReadable":"August 22nd, 2024"},"versionCreatedAt":"2024-04-19 18:47:47","video":"","vorDoi":"10.1038/s41598-024-70430-6","vorDoiUrl":"https://doi.org/10.1038/s41598-024-70430-6","workflowStages":[]},"version":"v1","identity":"rs-4219358","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4219358","identity":"rs-4219358","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}