Multi-omics analysis reveals the role of tumor-infiltrating CD4+CCR7+ T cells in EGFR antibody resistance and prognosis of hepatocellular carcinoma

Multi-omics analysis reveals the role of tumor-infiltrating CD4+CCR7+ T cells in EGFR antibody resistance and prognosis of hepatocellular 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 Research Article Multi-omics analysis reveals the role of tumor-infiltrating CD4+CCR7+ T cells in EGFR antibody resistance and prognosis of hepatocellular carcinoma Zizhong Yang, Lupeng Qiu, Guhe Jia, Zhuoya Sun, Yixin Gong, Yin Chen, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7189781/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 29 Nov, 2025 Read the published version in BMC Cancer → Version 1 posted 14 You are reading this latest preprint version Abstract Background Despite the crucial involvement of the EGFR pathway in hepatocellular carcinoma (HCC), the clinical efficacy of EGFR antibodies in HCC remains uncertain. While existing evidence suggests that immune dysfunction and tumor microenvironment alterations may contribute to treatment resistance, the precise mechanisms underlying this phenomenon in HCC warrant further investigation. Methods In this study, we employed patient-derived xenograft (PDX) models generated from 14 HCC patients enrolled in the REHOPE301 cohort to evaluate the sensitivity to nimotuzumab, a humanized anti-EGFR monoclonal antibody. Whole-exome sequencing (WES) and single-cell RNA sequencing were performed on tumor tissues and tumor-infiltrating lymphocytes (TILs) to elucidate the association between TIL characteristics and EGFR antibody response. A predictive risk score and nomogram were subsequently developed using LASSO regression analysis. The prognostic performance of this model was evaluated using 2 external datasets (ICGC-JP and GSE141202) through receiver operator characteristic (ROC) curves and calibration curves analyses. Results Nimotuzumab demonstrated a 50% response rate (7/14) in PDX models. Immune profiling revealed distinct TIL patterns between responders and non-responders. Notably, CD4 + CCR7 + T cells were significantly enriched in resistant tumors (p < 0.001) and negatively correlated with the nimotuzumab response (r = -0.767 p = 0.02). In non-responsive tumors, CD4 + CCR7 + T cells exhibited interactions with macrophages and CD8 + PDCD1 + T cells. A reduced infiltration of CD4 + CCR7 + T cells was associated with improved prognosis and enhanced EGFR antibody efficacy across multiple cancer types. Furthermore, a nine-gene signature related to CD4 + CCR7 + T cells was identified as a strong prognostic factor in HCC (HR = 5.19, 95% CI: 3.18–8.46, P < 0.001), and was used to construct a nomogram. WES confirmed prognostic gene mutations (VCAN, CAMK4, and CD226) potentially influencing nimotuzumab response. Conclusions Our findings indicate that increased infiltration of central memory CD4 + CCR7 + T cells in HCC may reflect an immunosuppressive tumor microenvironment, thereby impairing EGFR antibody efficacy and worsening patient prognosis. Hepatocellular carcinoma Nimotuzumab CC-chemokine receptor 7 EGFR-antibody PDX-model Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Background Primary hepatocellular carcinoma (HCC), ranks as the third leading cause of cancer-related mortality worldwide [ 1 ], with approximately 529,000 new cases and 483,800 deaths reported annually [ 2 ]. HCC patients exhibit limited responsiveness to conventional chemotherapy, resulting in few effective systemic treatment options for patients with advanced-stage disease. Tyrosine kinase inhibitors (TKIs), such as sorafenib and lenvatinib [ 3 ], have demonstrated notable efficacy in HCC management [ 4 ]. Among the targets of TKI, the epidermal growth factor receptor (EGFR) is overexpressed in 68–96% of HCC cases [ 5 ]. Although aberrant EGFR signaling is implicated in tumor initiation and progression, clinical trials evaluating various anti-EGFR agents have yielded inconsistent outcomes [ 6 ]. Beyond EGFR expression and mutation status [ 7 ], multiple additional factors may influence the efficacy of EGFR-targeted therapies and remain to be fully elucidated. In addition to small-molecule EGFR TKIs, EGFR monoclonal antibodies [ 8 , 9 ] are commonly used to treat solid tumors with EGFR-overexpression. However, their efficacy in HCC also remains controversial [ 10 ]. Emerging evidence indicates that activated T cells [ 11 ], natural killer (NK) cells [ 12 ], and macrophages [ 13 ] may enhance the antitumor effect of EGFR antibodies through mechanisms such as antibody-dependent cellular cytotoxicity (ADCC). Our previous studies on the EGFR antibody responses across HCC cell lines demonstrated that therapeutic efficacy was more closely associated with extracellular matrix characteristics and Hoshida subclass than EGFR pathway gene expression, suggesting a critical role of tumor microenvironment in modulating the treatment response [ 14 ]. Notably, in tumors such as head and neck squamous cell carcinoma (HNSCC) and colorectal cancer (CRC), EGFR antibodies have also been shown to modulate tumor-infiltrating lymphocyte (TIL) infiltration [ 15 ] and alter intratumoral T cell clonality [ 16 ], further highlighting the importance of immune components in shaping treatment efficacy. Patient-derived xenograft (PDX) models, which preserve the cellular heterogeneity of the original tumors, offer a valuable platform for predicting drug responses [ 17 ]. When combined with single-cell sequencing, these models enable detailed analysis of micro-environmental and molecular features. Prior studies have shown that inhibiting the differentiation of naïve CD4 + T cells into regulatory T cells (Tregs) can enhance the efficacy of EGFR antibodies in PDX models [ 18 , 19 ]. However, such investigations remain limited in HCC. Nimotuzumab, a humanized monoclonal antibody targeting EGFR, has been approved for the treatment of HNSCC and pancreatic cancer [ 20 ] in several countries [ 21 ]. Previous studies have also suggested potential therapeutic benefits of nimotuzumab in patients with HCC [ 14 , 22 ]. Unlike other EGFR inhibitors, nimotuzumab had an intermediate binding affinity for EGFR, enabling preferential accumulation in EGFR-overexpressing tissues [ 23 , 24 ], thereby supporting its favorable safety profile and enabling us to investigate the mechanisms of EGFR antibody action in the context of HCC. In this study, we sought to explore the responsiveness to anti-EGFR monoclonal antibodies in HCC using PDX models. By integrating single-cell sequencing, we profiled the TIL composition and its association with nimotuzumab sensitivity [ 25 ]. Additionally, our findings were validated using external datasets involving other anti-EGFR antibodies. Bioinformatics analysis revealed a negative association between CD4⁺CCR7⁺ T cell infiltration and response to EGFR-targeted therapy. Based on genes associated with these cells, we developed a clinical predictive model, which was further supported by whole-exome sequencing (WES). Overall, this study provides new insights into the immune-mediated mechanisms underlying anti-EGFR resistance in HCC and offers potential avenues to optimize therapeutic strategies. Methods Data and sample collection This study utilized patient specimens from the REHOPE301 cohort [ 26 , 27 ], a prospective investigation designed to evaluate TIL characteristics in HCC via multi-omics sequencing and assess their prognostic relevance. HCC diagnoses were pathological confirmed and no enrolled patients received chemotherapy or radiotherapy prior to surgery. Details clinical characteristics and study design are illustrated in Fig. S1 A. Following hepatectomy, a portion of tumor tissue was processed for pathological analysis, while the remaining samples were dissociated for downstream PDX model establishment and nimotuzumab sensitivity testing. Tissue dissociation and PDX drugsensitivity assay To increase the success rate and efficiency of model establishment and drugsensitivity testing, we adapted the PDX protocol reported by Wen et al. by implanting patient tumor tissue into hollowfiber capsules [ 28 ]. Fresh tumor samples were rinsed with Hank’s balanced salt solution to remove necrotic and nonsolid material. Dissociation was performed using the gentleMACS Human Tumor Dissociation Kit (Miltenyi Biotec, Germany), followed by centrifugation and filtration to eliminate blood and debris. Cellmax hollowfiber cartridges (Spectrum, Netherlands) were sterilized after soaking in absolute ethanol for 30 minutes and rinsing with ultrapure water [ 29 ]. A tumorcell suspension (2 × 10 5 cells/mL) was loaded into 10mm fiber segments, which were heatsealed at both ends. Three segments were implanted subcutaneously into each 6weekold female NPG mouse (Charles River, China) to establish HCC PDX models. Prior to implantation, animals were anesthetized via intraperitoneal injection of 1% sodium pentobarbital (50 mg/kg), and satisfactory anesthesia was achieved. After 24 hours, successfully engrafted mice were randomly assigned to the treatment (nimotuzumab, 2.5 mg/mL; 20 mg/kg, i.v., Q4D; Biotech Pharma, China), or control (saline) group. On day 7, the fibers were removed from mice, the encapsulated cells were flushed out with saline, enriched using the Human Tumor Cell Isolation Kit (Miltenyi Biotech, Germany), and subjected to the CellTiterGlo Luminescent Cell Viability Assay (Promega, USA). Tumor growth inhibition (TGI) was calculated as \(\:\text{T}\text{G}\text{I}\:\left(\text{%}\right)\:=\:[1-\:(\text{T}\text{r}\text{e}\text{a}\text{t}\text{m}\text{e}\text{n}\text{t}\:\text{v}\text{i}\text{a}\text{b}\text{i}\text{l}\text{i}\text{t}\text{y}\:/\:\text{C}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}\:\text{v}\text{i}\text{a}\text{b}\text{i}\text{l}\text{i}\text{t}\text{y}\left)\right]\:\times\:\:100\) . CD45 + TILs sorting and single-cell sequencing Remaining tumor not used in PDX experiments were processed for CD45 + TIL isolation and single-cell RNA sequencing as previously described [ 26 ]. Briefly, cells were magnetically sorted using anti-human CD45 microbeads (130-118-780, Miltenyi Biotec, Germany) and stained with anti-CD45 antibodies (304008, BioLegend, USA) to evaluate sorting efficiency by flow cytometry. The TILs were subsequently processed using the 10x Chromium System (10x Genomics, USA). Single-cell data processing The Raw fastq files were processed using Cell Ranger (version 7.1.0). Subsequent analyses were conducted with the “Seurat” R package (version 4.2.3). During quality control, genes expressed in > 3 cells and cells expressing > 200 genes were retained. Cells with > 20% mitochondrial or > 0.1% hemoglobin gene expression were excluded. Potential doublets were removed using “DoubletFinder” (version 2.0.4) [ 30 ]. The remaining cells underwent normalization, scaling, variable gene selection, principal component analysis (PCA), batch effect correction using “Harmony” (version 1.2.0), and clustering (dims = 30, resolution = 0.5). Cell clusters were annotated using canonical markers and top-ranked differentially expressed genes (DEGs) (adjusted p-values 1). Cell cycle states were inferred using “CellCycleScoring”. T and NK cell subsets were further characterized through subclustering following the aforementioned pipeline. To assess their tissue distribution preferences between nimotuzumab-responsive (R) and non-responsive (NR) groups, odds ratios (ORs) were calculated as previously described by Zhang et al. [ 31 ]. DEGs within each subcluster between R and NR patients were identified using the “FindMarkers” and functional enrichment analyses of DEGs were performed with “clusterProfiler” (version 3.21). Additionally, to delineate the functional states of immune cells, we applied the “AddModuleScore” function to quantify 19 immune-related gene signatures defined by Chu et.al. [ 32 ] and compared scores between R and NR groups. Analysis of cell-cell communication Intercellular communication among TIL subclusters was inferred using the “CellChat” R package (version 1.5.0). Non-immune clusters (e.g., “Hepatocyte/Cancer Cell”, “Proliferating Cell”, “Endothelial Cell”, “Fibroblast”) were excluded to avoid bias. CellChat objects were grouped by treatment response, and the “rankNet” function was used to compare ligand–receptor interactions and signaling pathways between groups [ 33 ]. Immune Infiltration and Treatment Correlation Analyses We used the single-sample Gene Set Enrichment Analysis (ssGSEA) to quantify the enriching levels and relative abundances of immunocytes in samples from different HCC datasets using R package “GSVA” (version 1.48.3). The samples were stratified into high- and low- infiltration groups based on the median infiltration score. In addition, tumor microenvironment characteristics were assessed using ESTIMATE scores calculated via the R package “estimate” (version 1.0.13). Prognostic gene identification and nomogram construction Using TCGA-LIHC as the training dataset, we developed a prognostic risk score model and validated it in two independent HCC cohorts (ICGC-JP and GSE141202) [ 34 ]. DEGs were first identified between tumor and normal tissues within the TCGA-LIHC cohort. DEGs between tumor and normal tissues (|log₂FC| >1, FDR < 0.05) were overlapped with CD4 + CCR7 + T cell-related genes (adjusted p 0.4). Prognostic genes were identified via univariate Cox regression and refined using least absolute shrinkage and selection operator (LASSO) regression method [ 35 ]. The risk scores were calculated via the following formula: \(\:\:\text{S}\text{c}\text{o}\text{r}\text{e}\:=\:{\Sigma\:}{\beta\:}\text{ᵢ}\:\times\:\:\text{E}\text{x}\text{p}\text{ᵢ}\) (β i : regression coefficient; Exp i : gene expression level). The predictive performance of the risk score was evaluated through Kaplan–Meier and receiver operating characteristic (ROC) curve analyses across both the training and validation cohorts. Univariate and multivariate Cox regression analyses were conducted incorporating the risk score and clinical parameters. Variables with p < 0.05 in multivariate analysis were used to construct a nomogram using the “rms” R package (version 6.4.0). Predictive accuracy was assessed using ROC curves and calibration plots in both the training and validation sets. WES analysis To confirm the mutations status of prognosis-related genes. DNA was extracted from formalin-fixed, paraffin-embedded (FFPE) HCC tissues using the Maxwell 16 FFPE Plus LEV DNA Purification Kit (Promega, USA) and sequenced on the NovaSeq 6000 platform (Illumina, USA). After sequencing, quality control of the raw FASTQ files was preformed using “FastQC” (version 0.12.1). The sequence data were processed according to “GATK” (version 3.5) best practices. Somatic single-nucleotide variants (SNVs) and indels were detected using MuTect2, filtered via “VCFtools” (version 0.1.16) and annotated using “SnpEff” (version 5.1f). Mutation profiles were visualized using the R package “maftools” (version 2.16.0) [ 33 ]. Statistical analysis Statistical comparisons were conducted using Student’s t-test or the Wilcoxon rank-sum test, depending on data distribution. One-way ANOVA was used for multiple group comparisons. Pearson and Spearman correlations were applied for continuous and ordinal variables, respectively. Survival analysis was performed using the log-rank test and visualized with Kaplan–Meier plots. All analyses were conducted in R software, with p < 0.05 considered statistically significant. Results TIL landscape in PDX models stratified by nimotuzumab response We successfully established PDX models from 14 patients in the REHOPE cohort, with TIL single-cell RNA sequencing conducted on 10 of TILs these patients (Fig. 1 A, S1A). In the PDX drug sensitivity assay, 7 samples exhibited reduced tumor cell viability following nimotuzumab treatment and were classified as responders, with maximum inhibition rate of 71.46%. The remaining 7 samples showed no significant response and were classified as non-responders, resulting in an overall nimotuzumab response rate of 50% (Fig. 1 B, S1B). Baseline characteristics, including age and BMI, were comparable between responders and non-responders (Table 1 ). Notably, patients 001, 017, and 026 demonstrated over a two-fold increase in tumor cell viability post-treatment, indicating strong resistance to nimotuzumab. Table 1 Brief clinical characteristics of HCC patients included for the establishment of PDX models. Characteristic Level Responder(n = 7) Non-responder(n = 7) P.value Sex (%) Male 7 (100.0) 7 (100.0) Age of diagnose (y/o, mean (SD)) 64.29 (11.86) 58.71 (11.80) 0.396 Weight (kg, mean (SD)) 73.83 (13.11) 73.29 (12.72) 0.939 Height (cm, mean (SD)) 168.71 (4.89) 172.57 (6.48) 0.232 BMI (mean (SD)) 25.97 (4.69) 24.47 (2.77) 0.480 Tumor diameter (cm, mean (SD)) 6.25 (1.94) 5.66 (3.16) 0.698 Stage (%) T2N0M0 2 (28.6) 4 (57.1) 0.589 T3N0M0 5 (71.4) 3 (42.9) HBV (%) - 2 (28.6) 0 ( 0.0) 0.445 + 5 (71.4) 7 (100.0) HCV (%) - 6 (85.7) 7 (100.0) 0.999 + 1 (14.3) 0 ( 0.0) Cirrhosis (%) - 5 (71.4) 5 (71.4) 0.999 + 2 (28.6) 2 (28.6) MVI (%) 0 6 (85.7) 5 (71.4) 0.999 1 1 (14.3) 2 (28.6) Ki-67 (median [IQR]) 0.10 [0.08, 0.25] 0.20 [0.10, 0.25] 0.870 CA199 (U/mL, median [IQR]) 16.82 [13.47, 18.22] 10.23 [7.14, 11.54] 0.142 AFP (ng/mL, median [IQR]) 18.66 [3.63, 50.28] 8.02 [3.42, 1062.52] 0.949 PVKII (ng/mL, median [IQR]) 235.00 [81.00, 2224.50] 849.00 [117.75, 2390.25] 0.668 CEA (ng/mL, median [IQR]) 3.16 [1.98, 3.30] 2.17 [1.68, 2.89] 0.482 NLR (mean (SD)) 2.00 (1.28) 2.12 (1.17) 0.856 ALB (g/L, mean (SD)) 40.57 (1.76) 40.96 (4.44) 0.834 GLB (g/L, mean (SD)) 28.37 (5.27) 26.96 (7.66) 0.694 TBIL (µmol/L, mean (SD)) 13.47 (3.28) 11.49 (4.14) 0.339 The blood test was performed before surgery. BMI: body-mass index, HBV: hepatitis B virus, HCV: hepatitis C virus, MVI: micro-vascular invasion, AFP: alpha fetoprotein, PVKII: protein induced by vitamin K absence or antagonist II, CEA: carcinoembryonic antigen, NLR: neutrophil-to-lymphocyte ratio, ALB: albumin, GLB: globulin, TBIL: total bilirubin. Based on PDX response, scRNA-seq samples were stratified into two groups. After quality control, 86,828 single cells were retained. UMAP-based clustering identified major immune subsets (Fig. 1 C, F) including lymphoid-derived: T cells, NK/NKT cells, dendritic cells (DCs), B cells, and plasma cells, as well as macrophages, myeloid-derived suppressor cells (MDSCs), and neutrophils. The macrophages were further subdivided into three subsets: monocyte-derived macrophages (Mφ-Monocytes), liver-resident Kupffer cells (Mφ-Kupffer), and tumor-associated macrophages (TAMs). A proliferating cell cluster characterized by high expression of cell cycle–related genes were identified (Fig. 1 D). Despite CD45 + enrichment, non-immune cells—such as hepatocytes/HCC tumor cells, fibroblasts, and endothelial cells—were detected, primarily from non-responders (Fig. 1 E), suggesting a potentially higher proportion of non-immune components in these tumors. Although the relative abundance of TIL subsets varied across samples, T cells, B cells, NK cells, and neutrophils were generally enriched in responder samples, indicating a greater immune infiltration (Fig. 1 G). PCA-based unsupervised clustering stratified the samples into three distinct groups (Fig. 1 H), suggesting that single-cell profiles can differentiate between response phenotypes. Of note, patients with mild resistance (011, 015, 016) and those with marked resistance (017, 026) were located distantly from each other in PCA space. Distinct T and NK cell features between different response groups Given the critical role of T and NK (TNK) cells in antitumor immunity, we performed subclustering of TNK populations. This analysis identified CD4⁺, CD8⁺, mucosal-associated invariant T (MAIT), and three NK subsets, along with smaller populations of proliferating T cells, innate lymphoid cells (ILCs), and mast-like cells (Fig. 2 A, E). Unlike the TIL composition, TNK subsets displayed more consistent distribution across samples but showed marked differences between response groups (Fig. 2 B-C). Tissue preference analysis [ 31 ] revealed that activated T and NK subsets were predominantly enriched in responder samples (Fig. 2 D), indicative of enhanced immune activation. Conversely, non-responders showed enrichment of the ILCs (in minimal numbers), CD4-FOXP3 (Tregs), and CD4-CCR7 (Tcm) subsets. Among these, only the CD4-CCR7 proportion was significantly negatively correlated with nimotuzumab response (Fig. 2 F), while CD4-FOXP3 cells were not (Fig. 2 G). Differential gene expression analysis of TNK subsets (Fig. 2 H) revealed three notable features: 1) cells from the responses exhibited high levels of activation markers (CD69, FOS, TOX) and effector genes (GZMK, CCL4, NKG7); 2) cells from non-response exhibited upregulation of stress-related proteins (HSP90AB1, DNAJB1, BAG3), suggestive of a stress/damage phenotype (Fig. 2 I); and 3) certain T cell subsets (e.g., CD4-CCL5, CD8-PDCD1) expressed clonally enriched TCR VDJ genes (e.g., TRAV13-1, TRBV5-6), suggesting clonal expansion of specific T cell populations. Further analysis of the CD4-CCR7 subset showed higher enrichment in cytokine production, metabolic activity, and TCR signaling scores among responders (Fig. 2 J–K), suggesting increased functional activation in this population. Enhanced intercellular communication of CCR7 + CD4 + T cells in non-responders To investigate cell–cell communication among TILs, we applied CellChat. While the overall number of interactions was similar between groups, interaction strength was markedly higher in non-responders, especially among the CD4-CCR7, Mφ-Monocyte, and CD8-PDCD1 subsets (Fig. 3 A, 3 C). Mφ-Monocytes consistently showed the strongest interactions across groups, whereas CD4-CCR7 T cells exhibited strong interactions specifically in non-responders (Fig. 3 B). Input-output (IO) analysis confirmed elevated signaling activity in CD4-CCR7 and CD8-PDCD1 subsets in the non-responsive group that surpassed the average levels observed in the other clusters (Fig. 3 D). Comparison of ligand–receptor interactions (Fig. 3 E–F) revealed enrichment of inflammatory (TNF, Chemerin [ 36 ]) and immunosuppressive pathways (CD30, IL10 [ 37 ]) in non-responders (Fig. 3 G), whereas responder samples were enriched in cytokine- and immune activation–related pathways, including CCL, CSF, and CLEC [ 38 ] signaling (Fig. 3 H). Notably, CD4-CCR7 subset was a key contributor to the signaling networks of non-responders (Fig. 3 G, I). Associations between CD4 + CCR7 + T cells and EGFR antibodies response To investigate the association between CD4 + CCR7 + T cells and treatment response across different anti-EGFR regimens and tumor types, we performed ssGSEA using CD4 + CCR7 + T-cell marker genes across the TCGA and GEO datasets. In the TCGA-LIHC cohort, tumor tissues exhibited significantly lower CD4 + CCR7 + T-cell scores compared to adjacent normal tissues (Fig. 4 A), and higher infiltration of these cells was negatively correlated with patient survival (Fig. 4 B). We next analyzed GSE102995 [ 39 ] dataset, which includes clinical data from panitumumab-treated patients. CD4 + CCR7 + T cell scores were significantly negatively associated with progression-free survival (PFS) (Fig. 4 C). To further assess whether these cells fluctuate during treatment, we examined GSE180480, which contains course gene expression data from patient-derived organoids. Compared to the baseline, CD4 + CCR7 + T-cell scores significantly decreased at the first emergence of drug resistance, suggesting reduced infiltration during treatment response (Fig. 4 D). A similar result was found in the GSE108277 [ 40 ] dataset, which includes a PDX model treated with cetuximab, samples exhibiting partial responses showed significantly lower CD4 + CCR7 + T cell scores than non-responders did (Fig. 4 E). We further investigated whether these associations extended to other TKI. In the GSE109211 dataset [ 41 ], which studied sorafenib treatment in HCC, responders had significantly lower CD4⁺CCR7⁺ T-cell scores than non-responders (Fig. 4 F), with the difference more pronounced than that seen in EGFR antibody datasets. Construction of a CD4 + CCR7 + T cell-associated risk signature To evaluate the prognostic value of CD4⁺CCR7⁺ T-cell infiltration in HCC, we developed a risk score based on gene expression associated with these cells. Differential expression analysis between 371 tumor and 51 normal samples in the TCGA-LIHC cohort identified 13,581 DEGs (Fig. 5 A), comprising 1,354 upregulated and 11,227 downregulated genes. Among these, 410 were significant associations with the CD4 + CCR7 + T cell cluster (Fig. 5 B). Univariate Cox regression identified 43 prognostically relevant genes among them, with 37 associated with increased risk and 6 with protective effects. Functional enrichment analysis of these 410 genes revealed significant involvement of GO terms including cell motility, angiogenesis, membrane structures and antigen binding (Fig. 5 C). KEGG pathway analysis revealed associations with cytokine signaling, TNF signaling, the PI3K-Akt pathway, cytoskeletal regulation, and specific infection-related pathways, consistent with the KEGG pathways identified from single-cell CD4 + CCR7 + T cell cluster DEGs (Fig. 2 K). We used LASSO regression to refine the model, selecting 9 key genes at log(λ) = − 0.31 (Fig. 5 D-E), including four protective genes (N4BP3, UAP1L1, MTHFD1L, MSC) and five risk-associated genes (LTK, PGF, CAMK4, SPRED3, TNFRSF18). The final multivariate model was constructed using stepwise selection, resulting in the following risk score formula (Fig. 5 F): \(\:\text{R}\text{i}\text{s}\text{k}\:\text{S}\text{c}\text{o}\text{r}\text{e}=1.4\times\:\text{N}4\text{B}\text{P}3+0.096\times\:\text{U}\text{A}\text{P}1\text{L}1+0.056\times\:\text{M}\text{T}\text{H}\text{F}\text{D}1\text{L}+0.032\times\:\text{M}\text{S}\text{C}-0.85\times\:\text{S}\text{P}\text{R}\text{E}\text{D}3-0.22\times\:\text{C}\text{A}\text{M}\text{K}4-0.12\times\:\text{L}\text{T}\text{K}-0.045\times\:\text{P}\text{G}\text{F}-0.005\times\:\text{T}\text{N}\text{F}\text{R}\text{S}\text{F}18\) . Risk scores were calculated for each TCGA sample, and patients were stratified into high- and low-risk groups. Survival analysis revealed significantly poorer survival outcomes in the high-risk group (Fig. 5 G, 5 I), with area under the curve (AUC) values of 0.705, 0.700, and 0.730 for 1-, 3-, and 5-year survival, respectively (Fig. 5 H). External validation using the ICGC-JP (Fig. 5 J) and GSE141202 (Fig. 5 K) cohorts confirmed the model's predictive power, as high-risk patients consistently had poorer survival outcomes. Construction and validation of a prognostic nomogram To determine whether the CD4 + CCR7 + T cell-associated risk score serves as an independent prognostic factor, we conducted both univariate (Fig. 6 A) and multivariate (Fig. 6 B) Cox regression analyses incorporating baseline characteristics sex, age, clinical stage, hepatitis B (HBV) infection, alcohol consumption, and tumor grade. HBV status, clinical stage, and the risk score were identified as independent predictors of overall HCC survival. A prognostic nomogram integrating these variables was developed to predict overall survival (Fig. 6 C). Kaplan–Meier analysis demonstrated significantly worse outcomes in the nomogram high-risk group (Fig. 6 D). Calibration curves and ROC analysis indicated strong predictive performance, with AUCs ranging from 0.669 to 0.793 (Fig. 6 E-F). The nomogram was further validated in the GSE141202 dataset, showing comparable AUCs (0.685–0.824) for 1- to 3-year survival predictions (Fig. 6 G–H), supporting its robustness. Association between the CD4 + CCR7 + T cell-associated risk score and immune infiltration To investigate the immunological context of the risk score, we examined its correlation with immune infiltration in the GSE196576 dataset [ 42 ]. The CD4 + CCR7 + T level were positively correlated with the expression of the 9-gene risk signature (Fig. 7 A). Using the MCPcounter algorithm, we confirmed that higher risk scores were associated with increased CD4 + CCR7 + T-cell infiltration in both the TCGA (Fig. 7 B) and GSE196576 (Fig. 7 C) cohorts. We next applied the ESTIMATE algorithm to assess stromal and immune infiltration. In GSE196576, the low-risk group had significantly higher ESTIMATE scores (p = 0.001; Fig. 7 D), stromal scores (p = 0.003; Fig. 7 E), and immune scores (p = 0.003; Fig. 7 F), indicating a greater non-tumor and immune component in the tumor microenvironment. These findings suggest that patients with lower risk scores have higher immune cell infiltration. WES analyses and validation of the CD4 + CCR7 + T cell-associated prognostic genes To validate the presence of genetic alterations in CD4 + CCR7 + T cell-associated risk genes, we performed WES on 13 tumor samples corresponding to the PDX models. While no significant differences in variant types or SNV classes were observed between responders and non-responders, the non-responders exhibited a higher median mutation burden (Fig. 8 A-B). We subsequently examined genetic mutations within the 43 risk signature genes. Among these genes, three (VCAN, CD226, and CAMK4) were mutated exclusively in non-responders (Fig. 8 C), while seven genes (STRA6, NTM, GLP2R, ADAM12, CTHRC1, HTRA3, and SCG2) were mutated only in responders (Fig. 8 D), suggesting distinct mutational profiles between the two groups. Notably, VCAN mutations were present in 60% (3/5) of non-responders (Fig. 8 E). Survival analysis in the TCGA-LIHC cohort revealed that patients with VCAN mutations had significantly poorer survival (p = 0.015; Fig. 8 F). All three VCAN-mutant patients harbored SNVs at distinct sites (Fig. 8 G). In summary, our WES result confirmed the presence of mutations in CD4 + CCR7 + T cell-related risk genes, with distinct mutational patterns between responders and non-responders to nimotuzumab. Discussion As a pivotal receptor orchestrating various proliferative and survival pathways, EGFR is associated with aggressive tumor behavior, metastasis, and poor patient survival. However, clinical trials evaluating erlotinib or cetuximab monotherapy in advanced HCC patients have not demonstrated clinical benefits. Nevertheless, preclinical studies and case reports suggest that EGFR antibodies may yield favorable outcomes in certain HCC patients [ 22 ]. Recent trials focusing on combining EGFR antibodies with chemotherapy [ 43 ], ICIs [ 11 , 44 ] or mTOR inhibitors [ 45 ], had shown promising results. Unlike small-molecule TKIs, EGFR antibodies also exert antitumor effects through ADCC and immune modulation [ 46 ]. Therefore, their efficacy is influenced by the infiltration and functional states of immune cells in the HCC microenvironment. In this study, we employed PDX models and single-cell RNA sequencing to analyze TILs from HCC samples treated with nimotuzumab, a humanized anti-EGFR monoclonal antibody. Our funding revealed a relationship between TIL composition and the therapeutic efficacy of EGFR antibody treatment. The PDX model demonstrated that nimotuzumab achieved a 50% response rate in HCC samples, outerforming our previous in-vitro studies using HCC cell lines and clinical trials involving cetuximab. Notably, the therapeutic response varied considerably, ranging from 70% tumor inhibition to more than 10-fold tumor growth. Single-cell sequencing revealed that nimotuzumab-resistant samples harbored a higher proportion of tumor cells and tumor-associated stromal components, consistent with previous studies suggesting that increased tumor burden and heterogeneity negatively correlate with EGFR antibody efficacy. Moreover, the responsive samples showed elevated neutrophil and B cell infiltration, whereas T cell, NK cell, and macrophage levels remained relatively unchanged. These findings suggest that general variations in these immune subsets may not directly determine nimotuzumab efficacy. The increased neutrophil and B cell content in responders may reflect enhanced humoral immunity and non-specific inflammation, consistent with prior reports [ 47 , 48 ]. PCA further revealed distinct transcriptional patterns in highly resistant samples, indicating the presence of multiple resistance mechanisms. Previous studies have demonstrated that EGFR antibodies exert antitumor effects through T and NK cell-mediated mechanisms, particularly via ADCC [ 12 , 49 ]. Compared with the overall TIL distribution, T and NK cells show relatively uniform abundance across samples, allowing us to explore tissue-specific functional subsets. Our analysis revealed enrichment of effector clusters such as NK-CCL3, CD4-GZMB, CD4-CCL5, and MAIT-GZMK in responders, suggesting that immune activation contributes to the efficacy of EGFR antibodies. Notably, there were no significant differences in CD8 + T cell populations between groups. Previous studies have shown that EGFR signaling promotes CD4 + T cell activation and apoptosis via Glut1-mediated pathways [ 50 ]. Activated CD4 + T cells, in turn, support NK cell survival and enhance ADCC. In contrast, non-responders were enriched in CD4 + CCR7 + T cells and Tregs. While Tregs have been reported to impair EGFR antibody efficacy [ 51 ], the role of ILCs and CD4 + CCR7 + T cells remains poorly defined. Our study identified CD4 + CCR7 + T cells as markers of EGFR antibody resistance for the first time. CCR7 is a hallmark of central memory T cells. The role of CD4 + memory T cells in tumors is remains controversial. Prior studies suggest that pembrolizumab can inhibit the differentiation of T cells into CCR7 + phenotypes and enhance immune responses. In our study, non-responders exhibited not only higher quantities of CD4-CCR7 subsets but also different functional characteristics. CD4-CCR7 cells from responders expressed higher levels of effector and activation markers such as CCL4, NKG7, and IL32. In contrast, T cells from non-responders expressed higher levels of naïve and anti-apoptotic markers, indicating a dysfunctional or exhausted state that may impair ADCC [ 52 ]. Additionally, other immune cell subsets in non-responders also showed increased expression of stress response genes like HSP90, indicating a stressed immune microenvironment. Chu et al. previously showed that HSP gene overexpression can suppress specific TIL activation [ 32 ]. Overall, non-responders exhibited reduced effector TILs, increased CD4⁺ memory T cells, and signs of impaired immune function. To further explore the interactions between CD4-CCR7 cells and other immune cells, we performed CellChat analysis. Monocyte-derived macrophages showed the strongest overall intercellular communication; however, they did not significantly differ between groups. Interestingly, CD4-CCR7 cells exhibited enhanced intercellular communication specifically in non-responders, particularly with CD8-PDCD1 cells and macrophages. Previous studies suggest a negative correlation between CD4 + CCR7 + central memory T cells and CD8 + PDCD1 + effector memory re-expressing CD45RA (TEMRA) cells [ 53 ], potentially linked to protection against immune-related adverse events. However, the interaction between CD4⁺CCR7⁺ T cells and CD8⁺ T during treatment resistance has not been previously reported. Subsequent pathway-specific communication analysis revealed distinct signaling patterns between groups. Responders were enriched in CCL4 and CLEC family pathways, involving activated MAIT, naïve CD4⁺ T, and NK cells, consistent with prior studies showing CLEC4 downregulation during tumor progression and positive correlations with CD4⁺ T and B cell infiltration. In contrast, non-responders showed enrichment of immunosuppressive (IL-10) and pro-inflammatory (CHEMERIN, TNF) pathways. Notably, macrophages in non-responsive samples exhibited high IL-10 expression, a marker of TAM polarization [ 54 ]. IL-10 has been shown to limit the proliferation of effector CD8 T cells following anti-PD-1 therapy, while reducing immune-related adverse events [ 55 ]. These results suggest that non-responders display increased immunosuppression and non-specific inflammation, which may contribute to poor prognosis. Based on the ssGSEA results, we infer the CD4 + CCR7 + T cell infiltration levels in multiple anti-EGFR bulk RNA-seq datasets to further validate our findings. Lower CD4 + CCR7 + T levels were consistently correlated with better outcomes in treatments involving cetuximab, panitumumab, and even sorafenib. Additionally, reduced abundance of CD4 + CCR7 + T was observed in tumor compared to normal tissues, in line with prior studies [ 56 ]. In the GSE180480 dataset, CD4 + CCR7 + T scores decreased at effective treatment timepoints compared to baseline. As a memory phenotype, a reduction in CD4 + CCR7 + T may indicate increased CD4 + T cell activation. Supporting this, studies have reported elevated CD4 + CCR7 + T cells levels in both relapsed breast cancer [ 57 ], and positive lymph nodes [ 58 ]. Thus, while CD4⁺CCR7⁺ T cell abundance negatively correlates with EGFR therapy response, they may not directly promote tumor progression. Instead, their accumulation may indicate insufficient immune activation raised by EGFR antibodies. Interestingly, CD4 + CCR7 + T cells have also been associated with better ICI outcomes [ 59 ], suggesting that combining ICIs with EGFR antibodies may improve therapeutic efficacy in HCC. Given the limited feasibility of routine scRNA-seq or immune profiling in clinical practice, we identified CD4 + CCR7 + T cell-associated genes in tumor tissue to construct a prognostic model. Therefore, we aimed to identify CD4 + CCR7 + T cell-associated genes within tumor tissues to construct a prognostic model. The risk model comprises nine genes: N4BP3, UAP1L1, MTHFD1L, MSC, LTK, PGF, CAMK4, SPRED3, and TNFRSF18. MTHFD1L and UAP1L1 are involved in cellular metabolism; CAMK4, LTK, TNFRSF18, SPRED3, and MSC play roles in immune signaling and regulation. Notably, SPRED3 and LTK may suppress CD4 + T cell responses via the MAPK and JAK-STAT pathways, and MSC and TNFRSF18 are implicated in the activation of Tregs and T cell exhaustion. PGF may effect the microenvironment remodeling and angiogenesis [ 60 ]. Together, these genes reflect immune suppression and tumor adaptation. This model is the first HCC prognostic framework based on CD4 + CCR7 + T cells. Compared with previous models focusing on exhausted CD4 + T cells [ 61 ], it demonstrates superior prognostic performance. The risk score also negatively correlated with immune infiltration. Patients with lower risk scores exhibit higher non-tumor components and immune cell infiltration, markers of favorable prognosis [ 62 ]. These results are consistent with our single-cell findings and highlight the model’s ability to predict immune status as well as the clinical outcomes. WES analysis showed no EGFR mutations in either group, suggesting resistance was not driven by EGFR alterations [ 7 ]. Higher tumor mutation burden (TMB) was demonstrated to be favorable factor for cetuximab therapy in CRC patients. However, in this study, higher median mutations were found in non-responders[ 63 ]. We hypothesized that higher TMB may indicate increased malignancy or EGFR-independent signaling activation. Among the nine model genes, only CAMK4 was mutated. While CAMK4 inhibition suppresses tumor growth, it also regulates Th17 cell development [ 64 ]. Notably, VCAN mutations were detected in 60% of resistant samples but absent in responders. VCAN encodes a major extracellular matrix component and has been implicated in HCC progression and poor prognosis [ 65 ]. It could influence EGFR signaling via the ADAMTS1–VCAN–EGFR axis in renal cancer [ 66 ], and this study is the first to suggest its role in anti-EGFR resistance in HCC. This study has several limitations. First, scRNA-seq was performed only on pre-treatment samples, limiting our ability to assess dynamic changes in CD4⁺CCR7⁺ T cells during therapy. Second, due to a lack of fresh tissue, we could only evaluate gene mutations by WES, without RNA or protein-level validation. Third, although our findings were supported by external datasets involving different EGFR-targeted therapies, further studies are needed to determine whether other EGFR antibodies or TKIs have distinct immune biomarkers or mechanisms. Conclusion Elevated proportions of CD4 + CCR7 + T cells in HCC TILs are associated with resistance to nimotuzumab treatment, characterized by a reduction in effector immune cells and attenuated immune activation. A prognostic model based on CD4 + CCR7 + T cell-associated genes effectively predicts survival risk in HCC patients and inversely correlates with the extent of immune cell infiltration within the tumor microenvironment. Abbreviations EGFR Epidermal Growth Factor Receptor HCC Hepatocellular Carcinoma PDX Patient-Derived Xenograft WES Whole-Exome Sequencing LASSO Least Absolute Shrinkage and Selection Operator ADCC Antibody-Dependent Cellular Cytotoxicity HNSCC Head and Neck Squamous Cell Carcinoma CRC Colorectal Cancer TILs Tumor-Infiltrating Lymphocytes PCA Principal Component Analysis UMAP Uniform Manifold Approximation and Projection NR Non-Responder DEG Differentially Expressed Gene ssGSEA Single Sample Gene Set Enrichment Analysis GSVA Gene Set Variation Analysis TCGA The Cancer Genome Atlas ICGC International Cancer Genome Consortium FDR False Discovery Rate ROC Receiver Operating Characteristic (curve) GATK Genome Analysis Toolkit MVI Microvascular Invasion MAIT Mucosal-Associated Invariant T (cell) KEGG Kyoto Encyclopedia of Genes and Genomes PFS Progression-Free Survival PR Partial Response GO Gene Ontology AUC Area Under the Curve HR Hazard Ratio SNV Single Nucleotide Variant TEMRA Terminally Differentiated Effector Memory CD8⁺ T Cell Re-expressing CD45RA TAM Tumor-Associated Macrophage ICI Immune Checkpoint Inhibitor TMB Tumor Mutational Burden Declarations Ethical approval and consent to participate This study complies with the Helsinki Declaration. All participants provided written informed consent prior to enrollment. All animal experiments comply with the ARRIVE guidelines. Animals were purchased from Charles River Laboratories with informed consent and maintained under standard laboratory conditions. Mice were euthanized by cervical dislocation while under deep anesthesia after experiment. Both the clinical study and animal experiments protocol received approval from the Ethics Committee of PLA General Hospital (approval number S2022-313-02, dated July 27, 2022.). Consent for publication Not applicable. Funding Not applicable. Author Contribution ZY, LQ and GJ contributed equally to this research and share the first authorship.ZY and LQ wrote the main manuscript text and figures. GJ, ZY, YG, ZS and YW performed the laboratory work and assisted with the manuscript. YC, ZS and LS performed single cell RNA and whole-exome sequencing. ZY and LS performed the statistical and bioinformatics analyses. ZS and XZ collected surgical samples. XZ and SJ conceptualized the project and edited the manuscript.All the authors have read and approved the contents of the manuscript prior to submission. Acknowledgement We thank Beijing DCTY Biotech Co., Ltd. for support of this research. Data Availability The sequencing data used in this study have been deposited in the China National Center for Bioinformation (https://ngdc.cncb.ac.cn/) GSA human database with the primary accession code HRA011337 (single cell sequence) and HRA012698 (whole-exome sequence). The public sequencing data used to support this study were obtained from public databases TCGA (https://portal.gdc.cancer.gov), ICGC (https://dcc.icgc.org) and GEO (GSE108277, GSE102995, GSE180480, GSE109211, GSE141202 and GSE196576, https://www.ncbi.nlm.nih.gov/gds). Other datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. References Hwang SY, Danpanichkul P, Agopian V, Mehta N, Parikh ND, Abou-Alfa GK, Singal AG, Yang JD. Hepatocellular carcinoma: updates on epidemiology, surveillance, diagnosis and treatment. Clin Mol Hepatol. 2025;31(Suppl):S228–54. Danpanichkul P, Suparan K, Tothanarungroj P, Dejvajara D, Rakwong K, Pang Y, Barba R, Thongpiya J, Fallon MB, Harnois D, et al. Epidemiology of gastrointestinal cancers: a systematic analysis from the Global Burden of Disease Study 2021. Gut. 2024;74(1):26–34. Kudo M, Finn RS, Qin S, Han KH, Ikeda K, Piscaglia F, Baron A, Park JW, Han G, Jassem J, et al. Lenvatinib versus sorafenib in first-line treatment of patients with unresectable hepatocellular carcinoma: a randomised phase 3 non-inferiority trial. Lancet (London England). 2018;391(10126):1163–73. Kim HD, Jung S, Lim HY, Ryoo BY, Ryu MH, Chuah S, Chon HJ, Kang B, Hong JY, Lee HC, et al. Regorafenib plus nivolumab in unresectable hepatocellular carcinoma: the phase 2 RENOBATE trial. Nat Med. 2024;30(3):699–707. Qiang Z, Wan J, Chen X, Wang H. Mechanisms and therapeutic targets of ErbB family receptors in hepatocellular carcinoma: a narrative review. Translational cancer Res. 2024;13(6):3156–78. Haber PK, Puigvehí M, Castet F, Lourdusamy V, Montal R, Tabrizian P, Buckstein M, Kim E, Villanueva A, Schwartz M, et al. Evidence-Based Management of Hepatocellular Carcinoma: Systematic Review and Meta-analysis of Randomized Controlled Trials (2002–2020). Gastroenterology. 2021;161(3):879–98. Sueangoen N, Tantiwetrueangdet A, Panvichian R. HCC-derived EGFR mutants are functioning, EGF-dependent, and erlotinib-resistant. Cell bioscience. 2020;10:41. Van Cutsem E, Köhne CH, Hitre E, Zaluski J, Chang Chien CR, Makhson A, D'Haens G, Pintér T, Lim R, Bodoky G, et al. Cetuximab and chemotherapy as initial treatment for metastatic colorectal cancer. N Engl J Med. 2009;360(14):1408–17. Patil VM, Noronha V, Joshi A, Agarwal J, Ghosh-Laskar S, Budrukkar A, Murthy V, Gupta T, Mahimkar M, Juvekar S, et al. A randomized phase 3 trial comparing nimotuzumab plus cisplatin chemoradiotherapy versus cisplatin chemoradiotherapy alone in locally advanced head and neck cancer. Cancer. 2019;125(18):3184–97. Zhu AX, Stuart K, Blaszkowsky LS, Muzikansky A, Reitberg DP, Clark JW, Enzinger PC, Bhargava P, Meyerhardt JA, Horgan K, et al. Phase 2 study of cetuximab in patients with advanced hepatocellular carcinoma. Cancer. 2007;110(3):581–9. Wang X, Li T, Slebos RJC, Chaudhary R, Guevara-Patino JA, Bonomi M, Saba NF, Chung CH. Clinical significance of peripheral T-cell repertoire in head and neck squamous cell carcinoma treated with cetuximab and nivolumab. Cancer Immunol immunotherapy: CII. 2025;74(4):142. Mahgoub S, Abosalem H, Emara M, Kotb N, Maged A, Soror S. Restoring NK cells functionality via cytokine activation enhances cetuximab-mediated NK-cell ADCC: A promising therapeutic tool for HCC patients. Mol Immunol. 2021;137:221–7. Zaryouh H, De Pauw I, Baysal H, Melis J, Van den Bossche V, Hermans C, Lau HW, Lambrechts H, Merlin C, Corbet C, et al. Establishment of head and neck squamous cell carcinoma mouse models for cetuximab resistance and sensitivity. Cancer drug Resist (Alhambra Calif). 2023;6(4):709–28. Wang Y, Zhang M, Gong Y, Wu Q, Zhang L, Jiao S. Bioinformatic Analysis of Hepatocellular Carcinoma Cell Lines to the Efficacy of Nimotuzumab. Int J Gen Med. 2021;14:2611–21. Wu X, Yan H, Qiu M, Qu X, Wang J, Xu S, Zheng Y, Ge M, Yan L, Liang L. Comprehensive characterization of tumor microenvironment in colorectal cancer via molecular analysis. eLife 2023, 12. Ge H, Ferris RL, Wang JH. Cetuximab Responses in Patients with HNSCC Correlate to Clonal Expansion Feature of Peripheral and Tumor-Infiltrating T Cells with Top T-Cell Receptor Clonotypes. Clin cancer research: official J Am Association Cancer Res. 2023;29(3):647–58. Su T, Huang M, Liao J, Lin S, Yu P, Yang J, Cai Y, Zhu S, Xu L, Peng Z, et al. Insufficient Radiofrequency Ablation Promotes Hepatocellular Carcinoma Metastasis Through N6-Methyladenosine mRNA Methylation-Dependent Mechanism. Hepatology (Baltimore MD). 2021;74(3):1339–56. Boreddy SR, Nair R, Pandey PK, Kuriakose A, Marigowda SB, Dey C, Banerjee A, Kulkarni H, Sagar M, Krishn SR, et al. BCA101 Is a Tumor-Targeted Bifunctional Fusion Antibody That Simultaneously Inhibits EGFR and TGFβ Signaling to Durably Suppress Tumor Growth. Cancer Res. 2023;83(11):1883–904. Pircher A, Gamerith G, Amann A, Reinold S, Popper H, Gächter A, Pall G, Wöll E, Jamnig H, Gastl G, et al. Neoadjuvant chemo-immunotherapy modifies CD4(+)CD25(+) regulatory T cells (Treg) in non-small cell lung cancer (NSCLC) patients. Lung cancer (Amsterdam Netherlands). 2014;85(1):81–7. Qin S, Li J, Bai Y, Wang Z, Chen Z, Xu R, Xu J, Zhang H, Chen J, Yuan Y, et al. Nimotuzumab Plus Gemcitabine for K-Ras Wild-Type Locally Advanced or Metastatic Pancreatic Cancer. J Clin oncology: official J Am Soc Clin Oncol. 2023;41(33):5163–73. Ramakrishnan MS, Eswaraiah A, Crombet T, Piedra P, Saurez G, Iyer H, Arvind AS. Nimotuzumab, a promising therapeutic monoclonal for treatment of tumors of epithelial origin. mAbs. 2009;1(1):41–8. Song P, Yang J, Li X, Huang H, Guo X, Zhou G, Xu X, Cai Y, Zhu M, Wang P, et al. Hepatocellular carcinoma treated with anti-epidermal growth factor receptor antibody nimotuzumab: A case report. Medicine. 2017;96(39):e8122. Garrido G, Tikhomirov IA, Rabasa A, Yang E, Gracia E, Iznaga N, Fernández LE, Crombet T, Kerbel RS, Pérez R. Bivalent binding by intermediate affinity of nimotuzumab: a contribution to explain antibody clinical profile. Cancer Biol Ther. 2011;11(4):373–82. Crombet T, Osorio M, Cruz T, Roca C, del Castillo R, Mon R, Iznaga-Escobar N, Figueredo R, Koropatnick J, Renginfo E, et al. Use of the humanized anti-epidermal growth factor receptor monoclonal antibody h-R3 in combination with radiotherapy in the treatment of locally advanced head and neck cancer patients. J Clin oncology: official J Am Soc Clin Oncol. 2004;22(9):1646–54. Liu WN, Harden SL, Tan SLW, Tan RJR, Fong SY, Tan SY, Liu M, Karnik I, Shuen TWH, Toh HC, et al. Single-cell RNA sequencing reveals anti-tumor potency of CD56(+) NK cells and CD8(+) T cells in humanized mice via PD-1 and TIGIT co-targeting. Mol therapy: J Am Soc Gene Therapy. 2024;32(11):3895–914. Sun Z, Gao B, Song L, Wang B, Li J, Jiang H, Li X, Yu Y, Zhou Z, Yang Z, et al. Single-cell RNA sequencing reveals intratumoral heterogeneity and multicellular community in primary hepatocellular carcinoma underlying microvascular invasion. Heliyon. 2024;10(18):e37233. Gao B, Wang Y, Lu S. Cellular senescence affects energy metabolism, immune infiltration and immunotherapeutic response in hepatocellular carcinoma. Sci Rep. 2023;13(1):1137. Zhang F, Wang W, Long Y, Liu H, Cheng J, Guo L, Li R, Meng C, Yu S, Zhao Q, et al. Characterization of drug responses of mini patient-derived xenografts in mice for predicting cancer patient clinical therapeutic response. Cancer Commun (London England). 2018;38(1):60. Neuhaus W, Lauer R, Oelzant S, Fringeli UP, Ecker GF, Noe CR. A novel flow based hollow-fiber blood-brain barrier in vitro model with immortalised cell line PBMEC/C1-2. J Biotechnol. 2006;125(1):127–41. McGinnis CS, Murrow LM, Gartner ZJ. DoubletFinder: Doublet Detection in Single-Cell RNA Sequencing Data Using Artificial Nearest Neighbors. Cell Syst. 2019;8(4):329–e337324. Zheng L, Qin S, Si W, Wang A, Xing B, Gao R, Ren X, Wang L, Wu X, Zhang J, et al. Pan-cancer single-cell landscape of tumor-infiltrating T cells. Sci (New York NY). 2021;374(6574):abe6474. Chu Y, Dai E, Li Y, Han G, Pei G, Ingram DR, Thakkar K, Qin JJ, Dang M, Le X, et al. Pan-cancer T cell atlas links a cellular stress response state to immunotherapy resistance. Nat Med. 2023;29(6):1550–62. Yang C, Cheng X, Gao S, Pan Q. Integrating bulk and single-cell data to predict the prognosis and identify the immune landscape in HNSCC. J Cell Mol Med. 2024;28(1):e18009. Hsu CL, Ou DL, Bai LY, Chen CW, Lin L, Huang SF, Cheng AL, Jeng YM, Hsu C. Exploring Markers of Exhausted CD8 T Cells to Predict Response to Immune Checkpoint Inhibitor Therapy for Hepatocellular Carcinoma. Liver cancer. 2021;10(4):346–59. Ma J, Chen Z, Hou L. Revealing a cancer-associated fibroblast-based risk signature for pancreatic adenocarcinoma through single-cell and bulk RNA-seq analysis. Aging. 2024;16(18):12525–42. Zhao L, Yang W, Yang X, Lin Y, Lv J, Dou X, Luo Q, Dong J, Chen Z, Chu Y, et al. Chemerin suppresses murine allergic asthma by inhibiting CCL2 production and subsequent airway recruitment of inflammatory dendritic cells. Allergy. 2014;69(6):763–74. Yang W, Liu H, Xu L, Yu T, Zhao X, Yao S, Zhao Q, Barnes S, Cohn SM, Dann SM, et al. GPR120 Inhibits Colitis Through Regulation of CD4(+) T Cell Interleukin 10 Production. Gastroenterology. 2022;162(1):150–65. Lopez Robles MD, Pallier A, Huchet V, Le Texier L, Remy S, Braudeau C, Delbos L, Moreau A, Louvet C, Brosseau C, et al. Cell-surface C-type lectin-like receptor CLEC-1 dampens dendritic cell activation and downstream Th17 responses. Blood Adv. 2017;1(9):557–68. Siano M, Espeli V, Mach N, Bossi P, Licitra L, Ghielmini M, Frattini M, Canevari S, De Cecco L. Gene signatures and expression of miRNAs associated with efficacy of panitumumab in a head and neck cancer phase II trial. Oral Oncol. 2018;82:144–51. Lindner AU, Carberry S, Monsefi N, Barat A, Salvucci M, O'Byrne R, Zanella ER, Cremona M, Hennessy BT, Bertotti A, et al. Systems analysis of protein signatures predicting cetuximab responses in KRAS, NRAS, BRAF and PIK3CA wild-type patient-derived xenograft models of metastatic colorectal cancer. Int J Cancer. 2020;147(10):2891–901. Pinyol R, Montal R, Bassaganyas L, Sia D, Takayama T, Chau GY, Mazzaferro V, Roayaie S, Lee HC, Kokudo N, et al. Molecular predictors of prevention of recurrence in HCC with sorafenib as adjuvant treatment and prognostic factors in the phase 3 STORM trial. Gut. 2019;68(6):1065–75. Yazdani A, Lenz HJ, Pillonetto G, Mendez-Giraldez R, Yazdani A, Sanoff H, Hadi R, Samiei E, Venook AP, Ratain MJ, et al. Gene signatures derived from transcriptomic-causal networks stratify colorectal cancer patients for effective targeted therapy. Commun Med. 2025;5(1):9. Asnacios A, Fartoux L, Romano O, Tesmoingt C, Louafi SS, Mansoubakht T, Artru P, Poynard T, Rosmorduc O, Hebbar M, et al. Gemcitabine plus oxaliplatin (GEMOX) combined with cetuximab in patients with progressive advanced stage hepatocellular carcinoma: results of a multicenter phase 2 study. Cancer. 2008;112(12):2733–9. Reusch U, Sundaram M, Davol PA, Olson SD, Davis JB, Demel K, Nissim J, Rathore R, Liu PY, Lum LG. Anti-CD3 x anti-epidermal growth factor receptor (EGFR) bispecific antibody redirects T-cell cytolytic activity to EGFR-positive cancers in vitro and in an animal model. Clin cancer research: official J Am Association Cancer Res. 2006;12(1):183–90. Geng J, Li X, Lang X, Qiao C, Hu M, Yang J, Feng J, Lv M. Combination of cetuximab and rapamycin enhances the therapeutic efficacy in hepatocellular carcinoma. Technol Cancer Res Treat. 2014;13(4):377–85. Huether A, Höpfner M, Baradari V, Schuppan D, Scherübl H. EGFR blockade by cetuximab alone or as combination therapy for growth control of hepatocellular cancer. Biochem Pharmacol. 2005;70(11):1568–78. Vera-Yunca D, Parra-Guillen ZP, Girard P, Trocóniz IF, Terranova N. Relevance of primary lesion location, tumour heterogeneity and genetic mutation demonstrated through tumour growth inhibition and overall survival modelling in metastatic colorectal cancer. Br J Clin Pharmacol. 2022;88(1):166–77. Turner RJ, Guy TV, Geraghty NJ, Splitt A, Watson D, Brungs D, Carolan MG, Miller AA, de Leon JF, Aghmesheh M et al. Low Pretreatment CD4(+):CD8(+) T Cell Ratios and CD39(+)CD73(+)CD19(+) B Cell Proportions Are Associated with Improved Relapse-Free Survival in Head and Neck Squamous Cell Carcinoma. Int J Mol Sci 2023, 24(16). Cheng Y, Borcherding N, Ogunsakin A, Lemke-Miltner CD, Gibson-Corley KN, Rajan A, Choi AB, Wongpattaraworakul W, Chan CHF, Salem AK, et al. The anti-tumor effects of cetuximab in combination with VTX-2337 are T cell dependent. Sci Rep. 2021;11(1):1535. Huang L, Zhang X, Fan J, Liu X, Luo S, Cao D, Liu Y, Xia Z, Zhong H, Chen C, et al. EGFR promotes the apoptosis of CD4(+) T lymphocytes through TBK1/Glut1 induced Warburg effect in sepsis. J Adv Res. 2023;44:39–51. Okuyama K, Naruse T, Yanamoto S. Tumor microenvironmental modification by the current target therapy for head and neck squamous cell carcinoma. J experimental Clin cancer research: CR. 2023;42(1):114. von Witzleben A, Fehn A, Grages A, Ezić J, Jeske SS, Puntigam LK, Brunner C, Kraus JM, Kestler HA, Doescher J, et al. Prospective longitudinal study of immune checkpoint molecule (ICM) expression in immune cell subsets during curative conventional therapy of head and neck squamous cell carcinoma (HNSCC). Int J Cancer. 2021;148(8):2023–35. Yu J, Long B, Li Z, Tian X, Li D, Long J, Wang Y, Chen Y, Zhang F, Liu H, et al. Central memory CD4 + T cells play a protective role against immune checkpoint inhibitor-associated myocarditis. Cardiovascular Res. 2024;120(12):1442–55. Zhang LN, Chen JY, Liu YX, Zhang Y, Hong LL, Li XX, Liu SH, Chen SQ, Peng L, Huang YT. Identification of lncRNA dual targeting PD-L1 and PD-L2 as a novel prognostic predictor for gastric cancer. Front Oncol. 2024;14:1341056. Rosnev S, Sterner B, Schiele P, Kolling S, Martin M, Flörcken A, Erber B, Wittenbecher F, Kofla G, Kurreck A et al. Reduced monocytic IL10 expression in PD1 inhibitor-treated patients is a harbinger of severe immune-related adverse events. European journal of cancer (Oxford, England : 1990) 2025, 217:115252. Chen D, Li X, Li H, Wang K, Tian X. Identification of Immune-Related Prognostic mRNA and lncRNA in Patients with Hepatocellular Carcinoma. J Oncol. 2022;2022:5313149. Nardone V, Botta C, Caraglia M, Martino EC, Ambrosio MR, Carfagno T, Tini P, Semeraro L, Misso G, Grimaldi A, et al. Tumor infiltrating T lymphocytes expressing FoxP3, CCR7 or PD-1 predict the outcome of prostate cancer patients subjected to salvage radiotherapy after biochemical relapse. Cancer Biol Ther. 2016;17(11):1213–20. Vahidi Y, Faghih Z, Talei AR, Doroudchi M, Ghaderi A. Memory CD4(+) T cell subsets in tumor draining lymph nodes of breast cancer patients: A focus on T stem cell memory cells. Cell Oncol (Dordrecht Netherlands). 2018;41(1):1–11. Inomata M, Matsumoto M, Takata N, Hayashi K, Seto Z, Hirai T, Tokui K, Taka C, Okazawa S, Kambara K, et al. Peripheral CD4 memory T cells predict the efficacy of immune checkpoint inhibitor therapy in patients with non-small cell lung cancer. Sci Rep. 2023;13(1):10807. Miao YD, Tang XL, Wang JT, Mi DH. Prognostic role of expression of angiogenesis markers in hepatocellular carcinoma: A bioinformatics analysis. World J Gastroenterol. 2022;28(30):4221–6. Zuo A, Lv J, Jia W, Ba Y, Liu S, Zhang Y, Weng S, Xu H, Liu L, Wang L, et al. High ratio of resident to exhausted CD4 + T cells predicts favorable prognosis and potentially better immunotherapeutic efficacy in hepatocellular carcinoma. BMC Cancer. 2024;24(1):1152. Yu L, Shen N, Shi Y, Shi X, Fu X, Li S, Zhu B, Yu W, Zhang Y. Characterization of cancer-related fibroblasts (CAF) in hepatocellular carcinoma and construction of CAF-based risk signature based on single-cell RNA-seq and bulk RNA-seq data. Front Immunol. 2022;13:1009789. Innocenti F, Ou FS, Qu X, Zemla TJ, Niedzwiecki D, Tam R, Mahajan S, Goldberg RM, Bertagnolli MM, Blanke CD, et al. Mutational Analysis of Patients With Colorectal Cancer in CALGB/SWOG 80405 Identifies New Roles of Microsatellite Instability and Tumor Mutational Burden for Patient Outcome. J Clin Oncol. 2019;37(14):1217–27. Zhan CS, Chen J, Chen J, Zhang LG, Liu Y, Du HX, Wang H, Zheng MJ, Yu ZQ, Chen XG, et al. CaMK4-dependent phosphorylation of Akt/mTOR underlies Th17 excessive activation in experimental autoimmune prostatitis. FASEB journal: official publication Federation Am Soc Experimental Biology. 2020;34(10):14006–23. Sun G, Zheng W, Tan P, Zhou J, Tang W, Cao H, Liu L, Shi X, Li Z, Zhang W. Comprehensive Analysis of VCAN Expression Profiles and Prognostic Values in HCC. Front Genet. 2022;13:900306. Wen YC, Lin YW, Ho KH, Yang YC, Lai FR, Chu CY, Chen JQ, Lee WJ, Chien MH. The oncogenic ADAMTS1-VCAN-EGFR cyclic axis drives anoikis resistance and invasion in renal cell carcinoma. Cell Mol Biol Lett. 2024;29(1):126. Additional Declarations No competing interests reported. Supplementary Files FigureS1.pdf Cite Share Download PDF Status: Published Journal Publication published 29 Nov, 2025 Read the published version in BMC Cancer → Version 1 posted Editorial decision: Revision requested 22 Aug, 2025 Reviews received at journal 22 Aug, 2025 Reviews received at journal 20 Aug, 2025 Reviews received at journal 18 Aug, 2025 Reviews received at journal 14 Aug, 2025 Reviewers agreed at journal 13 Aug, 2025 Reviewers agreed at journal 13 Aug, 2025 Reviewers agreed at journal 13 Aug, 2025 Reviewers agreed at journal 13 Aug, 2025 Reviewers invited by journal 13 Aug, 2025 Editor assigned by journal 13 Aug, 2025 Editor invited by journal 11 Aug, 2025 Submission checks completed at journal 10 Aug, 2025 First submitted to journal 10 Aug, 2025 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-7189781","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":503279333,"identity":"1abf9c04-27eb-43b0-8c06-db8202a2c299","order_by":0,"name":"Zizhong Yang","email":"","orcid":"","institution":"Nankai University","correspondingAuthor":false,"prefix":"","firstName":"Zizhong","middleName":"","lastName":"Yang","suffix":""},{"id":503279334,"identity":"e19f5df8-ee0b-4c94-9f3a-10bc536c6bb8","order_by":1,"name":"Lupeng Qiu","email":"","orcid":"","institution":"The First Medical Center of Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Lupeng","middleName":"","lastName":"Qiu","suffix":""},{"id":503279335,"identity":"364398b1-34a0-49da-a430-42b48fe4cc3d","order_by":2,"name":"Guhe Jia","email":"","orcid":"","institution":"Nankai University","correspondingAuthor":false,"prefix":"","firstName":"Guhe","middleName":"","lastName":"Jia","suffix":""},{"id":503279336,"identity":"c42ff0ed-e93b-4a08-b259-222895bd7838","order_by":3,"name":"Zhuoya Sun","email":"","orcid":"","institution":"The First Medical Center of Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Zhuoya","middleName":"","lastName":"Sun","suffix":""},{"id":503279337,"identity":"9f9fb103-9003-406c-bbf8-505cac24f0a6","order_by":4,"name":"Yixin Gong","email":"","orcid":"","institution":"Beijing DCTY Biotech Co., Ltd.","correspondingAuthor":false,"prefix":"","firstName":"Yixin","middleName":"","lastName":"Gong","suffix":""},{"id":503279338,"identity":"7fe77efc-950c-462a-810c-5de7a710ca0e","order_by":5,"name":"Yin Chen","email":"","orcid":"","institution":"Beijing DCTY Biotech Co., Ltd.","correspondingAuthor":false,"prefix":"","firstName":"Yin","middleName":"","lastName":"Chen","suffix":""},{"id":503279340,"identity":"0f18037a-f03c-4a89-95dd-c2289180c177","order_by":6,"name":"Yu Wang","email":"","orcid":"","institution":"The First Medical Center of Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yu","middleName":"","lastName":"Wang","suffix":""},{"id":503279342,"identity":"7ba7faaf-87a6-4577-8d7e-89cc86c66b75","order_by":7,"name":"Lai Song","email":"","orcid":"","institution":"Beijing DCTY Biotech Co., Ltd.","correspondingAuthor":false,"prefix":"","firstName":"Lai","middleName":"","lastName":"Song","suffix":""},{"id":503279344,"identity":"64cf7587-29ae-4689-b2d6-6eb35bc461e9","order_by":8,"name":"Xiao Zhao","email":"","orcid":"","institution":"The Fifth Medical Center of Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Xiao","middleName":"","lastName":"Zhao","suffix":""},{"id":503279346,"identity":"fe70fc19-7cfc-4838-afb3-0e992546575f","order_by":9,"name":"Shunchang Jiao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzUlEQVRIiWNgGAWjYHCChAMMDBJy8uyNjQ8/kKDFwtiw53CzsQQJNlUkMtxIbxPgIUatfPuBhwd+VEgkMM582MYgwWAnp9tAQIvBmYSEgz1nJPLYpRPbHhQwJBubHSCkRQLoF942iWLG2YntQM6BxG2EtMjPYEg4+LdNIrHh5sE2CR5itDDcYEg4zAvScoORSC0gvxyWOSMBDOREYCAbEOEX+fYzyR/fVNQBo/L4w4cfKuzkCGphYOBJQLaUoHIQYCds6igYBaNgFIxwAAA7GkXZ/JYNzgAAAABJRU5ErkJggg==","orcid":"","institution":"Nankai University","correspondingAuthor":true,"prefix":"","firstName":"Shunchang","middleName":"","lastName":"Jiao","suffix":""}],"badges":[],"createdAt":"2025-07-22 18:23:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7189781/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7189781/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12885-025-15276-5","type":"published","date":"2025-11-29T15:58:10+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":89566304,"identity":"d79e8de4-d235-496c-b244-942d23d71be4","added_by":"auto","created_at":"2025-08-21 10:56:01","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":504680,"visible":true,"origin":"","legend":"\u003cp\u003eStudy design, PDX results and identification of cell clusters via single-cell analysis. (\u003cstrong\u003eA\u003c/strong\u003e) Workflow showing the collection and processing of HCC tissue for PDX, scRNA-seq and WES. (\u003cstrong\u003eB\u003c/strong\u003e) Waterfall plot demonstrating the tumor response to nimotuzumab treatment across PDX models, ordered by changes in cell viability (\u003cstrong\u003eC\u003c/strong\u003e) Uniform Manifold Approximation and Projection (UMAP) visualization of single-cell transcriptomes from 10 HCC patients, colored by annotated cell types. (\u003cstrong\u003eD-E\u003c/strong\u003e) UMAP of cells colored by G2/M phase proliferation score (\u003cstrong\u003eD\u003c/strong\u003e) and nimotuzumab response groups (\u003cstrong\u003eE\u003c/strong\u003e). (\u003cstrong\u003eF\u003c/strong\u003e) The average expression of marker genes in different cell types. (\u003cstrong\u003eG\u003c/strong\u003e) Stacked bar plot quantifying the proportional distribution of cell types within each sample. (\u003cstrong\u003eH\u003c/strong\u003e) Principal component analysis (PCA) of scRNA-seq samples showing separation along the first two principal components, with samples colored by nimotuzumab response status.\u003c/p\u003e","description":"","filename":"Fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7189781/v1/eb95a99f8f74ec2498a203dd.jpg"},{"id":89565250,"identity":"6f0b3835-43f5-495b-ac31-0eaf516836e9","added_by":"auto","created_at":"2025-08-21 10:40:01","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":806228,"visible":true,"origin":"","legend":"\u003cp\u003eSingle-cell characterization of T-NK cell subpopulations associated with different nimotuzumab responses. (\u003cstrong\u003eA\u003c/strong\u003e) UMAP showing a projection of cell types from cells in the T-NK cluster. (\u003cstrong\u003eB-C\u003c/strong\u003e) UMAP of the T-NK cells colored according to the nimotuzumab response groups (\u003cstrong\u003eB\u003c/strong\u003e) and donors (\u003cstrong\u003eC\u003c/strong\u003e). (\u003cstrong\u003eD\u003c/strong\u003e) Tissue distribution preference analysis heatmap, with color intensity representing the tissue distribution odds ratio for different subsets (** P\u0026lt;0.01, *** P\u0026lt;0.001, **** P\u0026lt;0.0001). (\u003cstrong\u003eE\u003c/strong\u003e) The average expression of cell type marker genes in different T-NK subclusters. (\u003cstrong\u003eF-G\u003c/strong\u003e) Spearman correlation of the relative tumor survival rate with (\u003cstrong\u003eF\u003c/strong\u003e) CD4-CCR7 and (\u003cstrong\u003eG\u003c/strong\u003e) CD4-FOXP3 subpopulation frequencies. (\u003cstrong\u003eH\u003c/strong\u003e) Multi-cluster volcano plot highlighting differentially expressed genes (DEGs) between response groups (red: response-upregulated; blue: non-response-upregulated) (\u003cstrong\u003eI\u003c/strong\u003e) Violin plots showing expression distributions of four key DEGs in the CD4-CCR7 subsets. (\u003cstrong\u003eJ\u003c/strong\u003e) Functional signature heatmap of the CD4-CCR7 clusters, scaled by module score significance. (\u003cstrong\u003eK\u003c/strong\u003e) KEGG pathway enrichment bubble plots comparing CD4-CCR7 upregulated genes between response groups.\u003c/p\u003e","description":"","filename":"Fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7189781/v1/c7fedb8d38c2f83e9ddd0710.jpg"},{"id":89563826,"identity":"a2a5c80e-249b-4c1b-aea0-0b7ae164ce70","added_by":"auto","created_at":"2025-08-21 10:32:01","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1108831,"visible":true,"origin":"","legend":"\u003cp\u003eDifferential cell–cell communication networks in the nimotuzumab Response (R) and Non-response (NR) groups. (\u003cstrong\u003eA\u003c/strong\u003e) Circos plots comparing interaction quantity (left) and strength (right) between responders (R, n=5) and non-responders (NR, n=5). (\u003cstrong\u003eB\u003c/strong\u003e) Group-specific interaction Circos plots showing differential communication patterns. (\u003cstrong\u003eC\u003c/strong\u003e) Columns showing the interactions number (left) and interaction strength (right) of the cell–cell communication. (\u003cstrong\u003eD\u003c/strong\u003e) IO plots revealing the communication hierarchy across clusters. (\u003cstrong\u003eE\u003c/strong\u003e) Information flow plots showing the differences in ligand–receptor pairs between the two groups. (\u003cstrong\u003eF\u003c/strong\u003e) Pathway-specific crosstalk heatmap between CD4-CCR7 and other cell types. (\u003cstrong\u003eG\u003c/strong\u003e) Circos plots of NR-enriched pathways: CD30 (left) and TNF (right) interactions. (\u003cstrong\u003eH\u003c/strong\u003e) Circos plots of R-enriched pathways: CLEC (left) and CCL (right) interactions. (\u003cstrong\u003eI\u003c/strong\u003e) Differential CD45 pathway engagement between groups.\u003c/p\u003e","description":"","filename":"Fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7189781/v1/7c9b480e5f388535e97ed4ac.jpg"},{"id":89565611,"identity":"e7fbe50e-1678-4e77-8b9b-2fdcca282701","added_by":"auto","created_at":"2025-08-21 10:48:01","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":251871,"visible":true,"origin":"","legend":"\u003cp\u003eRelationships between CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell infiltration and anti-EGFR treatment response. (A) CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell infiltration levels in tumor vs. normal tissues (TCGA-LIHC). (B) K-M curve comparing overall survival (OS) between the high- and low- infiltration groups. (C) Correlation between CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell infiltration and panitumumab PFS (GSE102995). (D) Boxplot comparing CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell infiltration at baseline and resistant-onset timepoints (GSE180480). (E) Violin plot visualizing CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell infiltration between cetuximab Partial Response (PR) and NPR groups (GSE108277). (F) Violin plot visualizing the CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell infiltration between sorafenib responsive and non-responsive groups (GSE109211).\u003c/p\u003e","description":"","filename":"Fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7189781/v1/acf7712bfe1b81e05b2d9018.jpg"},{"id":89563828,"identity":"47a3b44e-3a28-4df0-b5d7-d051b64f34bd","added_by":"auto","created_at":"2025-08-21 10:32:01","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":515582,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-associated genes and establishment of a risk model. (A) Volcano plot illustrating DEGs between tumor and normal tissues in the TCGA-LIHC cohort. (B) Volcano plot showing the CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-associated DEGs identified through Spearman regression. (C) Functional enrichment analysis of CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-related DEGs, including Gene Ontology (BP, CC, and MF) and KEGG pathway analyses. (D) Trajectory plot of independent variables with varying λ values in the LASSO model for TCGA-LIHC. (E) LASSO coefficient profiles plotted against the log(λ) sequence. (F) Cox coefficients of the 9 genes selected by the LASSO model. (G) Dot plots representing patient risk stratification based on risk scores in the TCGA-LIHC cohort. (H) ROC curves demonstrating the 1-, 3-, and 5-year AUC values of the risk score. (I-K) Kaplan‒Meier plots of the high- and low-risk groups in the TCGA cohort (I) and validation cohorts ICGC-JP (J) and GSE141202 (K).\u003c/p\u003e","description":"","filename":"Fig5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7189781/v1/b959ebf7eb321fba1c91c6ab.jpg"},{"id":89565253,"identity":"8706edf1-13c4-4201-af25-fd484b85aaa7","added_by":"auto","created_at":"2025-08-21 10:40:02","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":409765,"visible":true,"origin":"","legend":"\u003cp\u003eEstablishment and validation of a nomogram model. (\u003cstrong\u003eA-B\u003c/strong\u003e) Forest plots presenting hazard ratios (HRs) for clinical characteristics and CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+ \u003c/sup\u003eT cell-associated risk scores derived from univariate (\u003cstrong\u003eA\u003c/strong\u003e) and multivariate (\u003cstrong\u003eB\u003c/strong\u003e) Cox analyses. (\u003cstrong\u003eC\u003c/strong\u003e) Nomogram incorporating significant prognostic factors and risk score. (\u003cstrong\u003eD\u003c/strong\u003e) K‒M survival curves comparing the high- and low-risk groups stratified by the nomogram score. (\u003cstrong\u003eE-F\u003c/strong\u003e) ROC curve (\u003cstrong\u003eE\u003c/strong\u003e) and calibration plot (\u003cstrong\u003eF\u003c/strong\u003e) demonstrating the nomogram's predictive accuracy for survival in the TCGA cohort. (\u003cstrong\u003eG-H\u003c/strong\u003e) ROC curve (\u003cstrong\u003eG\u003c/strong\u003e) and calibration plot (\u003cstrong\u003eH\u003c/strong\u003e) validating the nomogram's predictive performance in the GSE141202 cohort.\u003c/p\u003e","description":"","filename":"Fig6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7189781/v1/a98124f5f6a8b2468d2c550b.jpg"},{"id":89563834,"identity":"b3764322-706c-4581-8799-d3df148847b0","added_by":"auto","created_at":"2025-08-21 10:32:02","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":278063,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation analysis between the risk score and immune infiltration characteristics. (\u003cstrong\u003eA\u003c/strong\u003e) Spearman correlation plot depicting the association of CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+ \u003c/sup\u003eT cell signature with the risk score in the GSE196576 cohort. (\u003cstrong\u003eB-C\u003c/strong\u003e) MCPcounter analysis quantifying differences in CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+ \u003c/sup\u003eT infiltration between the high- and low-risk groups in TCGA (\u003cstrong\u003eB\u003c/strong\u003e) and GSE196576 (\u003cstrong\u003eC\u003c/strong\u003e) cohorts. (\u003cstrong\u003eD-F\u003c/strong\u003e) Violin plots comparing tumor microenvironment scores derived from the ESTIMATE algorithm: microenvironment score (\u003cstrong\u003eD\u003c/strong\u003e), stromal score (\u003cstrong\u003eE\u003c/strong\u003e) and immune score (\u003cstrong\u003eF\u003c/strong\u003e) in the high- versus low-risk groups of GSE196576.\u003c/p\u003e","description":"","filename":"Fig7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7189781/v1/ae7ff744aa01899f28ce395c.jpg"},{"id":89563841,"identity":"e3c07cbe-4473-4306-90d0-f029aa3307e5","added_by":"auto","created_at":"2025-08-21 10:32:02","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":485493,"visible":true,"origin":"","legend":"\u003cp\u003eGene mutation landscape in nimotuzumab responsive versus non-responsive samples. (\u003cstrong\u003eA-B\u003c/strong\u003e) Distribution of variant classifications, variant types, single-nucleotide variant (SNV) categories, and mutation loads per sample in responsive (\u003cstrong\u003eA\u003c/strong\u003e) and non-responsive (\u003cstrong\u003eB\u003c/strong\u003e) cohorts. (\u003cstrong\u003eC-D\u003c/strong\u003e) Oncoprint visualization of mutations in 43 risk genes for the non-response (\u003cstrong\u003eC\u003c/strong\u003e) and responsive (\u003cstrong\u003eD\u003c/strong\u003e) groups, with sidebar plots indicating mutation type frequencies. (\u003cstrong\u003eE\u003c/strong\u003e) Bar plot comparing the top frequently mutated risk genes between the response and non-response groups. (\u003cstrong\u003eF\u003c/strong\u003e) Kaplan‒Meier survival analysis for the VCAN gene-unaltered group versus VCAN altered groups in the TCGA cohort. (\u003cstrong\u003eG\u003c/strong\u003e) Lollipop plot mapping the VCAN protein domains and mutation sites in altered samples.\u003c/p\u003e","description":"","filename":"Fig8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7189781/v1/3117ac7d9777d52a26ce64a1.jpg"},{"id":97179876,"identity":"043b182b-b812-4a86-b93a-bd2423478cb8","added_by":"auto","created_at":"2025-12-01 16:17:07","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5239602,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7189781/v1/91d54598-dfae-4643-8b28-55b139a9dbc3.pdf"},{"id":89563831,"identity":"cd9d8ea9-f949-4db6-8fd5-a0a0a6731cde","added_by":"auto","created_at":"2025-08-21 10:32:02","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3075672,"visible":true,"origin":"","legend":"","description":"","filename":"FigureS1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7189781/v1/a8a564d0a2cc4c367b634d02.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Multi-omics analysis reveals the role of tumor-infiltrating CD4+CCR7+ T cells in EGFR antibody resistance and prognosis of hepatocellular carcinoma","fulltext":[{"header":"Background","content":"\u003cp\u003ePrimary hepatocellular carcinoma (HCC), ranks as the third leading cause of cancer-related mortality worldwide [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e], with approximately 529,000 new cases and 483,800 deaths reported annually [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. HCC patients exhibit limited responsiveness to conventional chemotherapy, resulting in few effective systemic treatment options for patients with advanced-stage disease. Tyrosine kinase inhibitors (TKIs), such as sorafenib and lenvatinib [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], have demonstrated notable efficacy in HCC management [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Among the targets of TKI, the epidermal growth factor receptor (EGFR) is overexpressed in 68–96% of HCC cases [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Although aberrant EGFR signaling is implicated in tumor initiation and progression, clinical trials evaluating various anti-EGFR agents have yielded inconsistent outcomes [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Beyond EGFR expression and mutation status [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], multiple additional factors may influence the efficacy of EGFR-targeted therapies and remain to be fully elucidated.\u003c/p\u003e\u003cp\u003eIn addition to small-molecule EGFR TKIs, EGFR monoclonal antibodies [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e] are commonly used to treat solid tumors with EGFR-overexpression. However, their efficacy in HCC also remains controversial [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Emerging evidence indicates that activated T cells [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], natural killer (NK) cells [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], and macrophages [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] may enhance the antitumor effect of EGFR antibodies through mechanisms such as antibody-dependent cellular cytotoxicity (ADCC). Our previous studies on the EGFR antibody responses across HCC cell lines demonstrated that therapeutic efficacy was more closely associated with extracellular matrix characteristics and Hoshida subclass than EGFR pathway gene expression, suggesting a critical role of tumor microenvironment in modulating the treatment response [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Notably, in tumors such as head and neck squamous cell carcinoma (HNSCC) and colorectal cancer (CRC), EGFR antibodies have also been shown to modulate tumor-infiltrating lymphocyte (TIL) infiltration [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e] and alter intratumoral T cell clonality [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e], further highlighting the importance of immune components in shaping treatment efficacy.\u003c/p\u003e\u003cp\u003ePatient-derived xenograft (PDX) models, which preserve the cellular heterogeneity of the original tumors, offer a valuable platform for predicting drug responses [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. When combined with single-cell sequencing, these models enable detailed analysis of micro-environmental and molecular features. Prior studies have shown that inhibiting the differentiation of naïve CD4 + T cells into regulatory T cells (Tregs) can enhance the efficacy of EGFR antibodies in PDX models [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. However, such investigations remain limited in HCC.\u003c/p\u003e\u003cp\u003eNimotuzumab, a humanized monoclonal antibody targeting EGFR, has been approved for the treatment of HNSCC and pancreatic cancer [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] in several countries [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Previous studies have also suggested potential therapeutic benefits of nimotuzumab in patients with HCC [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Unlike other EGFR inhibitors, nimotuzumab had an intermediate binding affinity for EGFR, enabling preferential accumulation in EGFR-overexpressing tissues [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e], thereby supporting its favorable safety profile and enabling us to investigate the mechanisms of EGFR antibody action in the context of HCC.\u003c/p\u003e\u003cp\u003eIn this study, we sought to explore the responsiveness to anti-EGFR monoclonal antibodies in HCC using PDX models. By integrating single-cell sequencing, we profiled the TIL composition and its association with nimotuzumab sensitivity [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Additionally, our findings were validated using external datasets involving other anti-EGFR antibodies. Bioinformatics analysis revealed a negative association between CD4⁺CCR7⁺ T cell infiltration and response to EGFR-targeted therapy. Based on genes associated with these cells, we developed a clinical predictive model, which was further supported by whole-exome sequencing (WES). Overall, this study provides new insights into the immune-mediated mechanisms underlying anti-EGFR resistance in HCC and offers potential avenues to optimize therapeutic strategies.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eData and sample collection\u003c/p\u003e\u003cp\u003eThis study utilized patient specimens from the REHOPE301 cohort [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e], a prospective investigation designed to evaluate TIL characteristics in HCC via multi-omics sequencing and assess their prognostic relevance. HCC diagnoses were pathological confirmed and no enrolled patients received chemotherapy or radiotherapy prior to surgery. Details clinical characteristics and study design are illustrated in Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eA. Following hepatectomy, a portion of tumor tissue was processed for pathological analysis, while the remaining samples were dissociated for downstream PDX model establishment and nimotuzumab sensitivity testing.\u003c/p\u003e\u003cp\u003eTissue dissociation and PDX drugsensitivity assay\u003c/p\u003e\u003cp\u003eTo increase the success rate and efficiency of model establishment and drugsensitivity testing, we adapted the PDX protocol reported by Wen et al. by implanting patient tumor tissue into hollowfiber capsules [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Fresh tumor samples were rinsed with Hank’s balanced salt solution to remove necrotic and nonsolid material. Dissociation was performed using the gentleMACS Human Tumor Dissociation Kit (Miltenyi Biotec, Germany), followed by centrifugation and filtration to eliminate blood and debris. Cellmax hollowfiber cartridges (Spectrum, Netherlands) were sterilized after soaking in absolute ethanol for 30 minutes and rinsing with ultrapure water [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. A tumorcell suspension (2 × 10\u003csup\u003e5\u003c/sup\u003e cells/mL) was loaded into 10mm fiber segments, which were heatsealed at both ends. Three segments were implanted subcutaneously into each 6weekold female NPG mouse (Charles River, China) to establish HCC PDX models. Prior to implantation, animals were anesthetized via intraperitoneal injection of 1% sodium pentobarbital (50 mg/kg), and satisfactory anesthesia was achieved. After 24 hours, successfully engrafted mice were randomly assigned to the treatment (nimotuzumab, 2.5 mg/mL; 20 mg/kg, i.v., Q4D; Biotech Pharma, China), or control (saline) group. On day 7, the fibers were removed from mice, the encapsulated cells were flushed out with saline, enriched using the Human Tumor Cell Isolation Kit (Miltenyi Biotech, Germany), and subjected to the CellTiterGlo Luminescent Cell Viability Assay (Promega, USA). Tumor growth inhibition (TGI) was calculated as \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\text{T}\\text{G}\\text{I}\\:\\left(\\text{%}\\right)\\:=\\:[1-\\:(\\text{T}\\text{r}\\text{e}\\text{a}\\text{t}\\text{m}\\text{e}\\text{n}\\text{t}\\:\\text{v}\\text{i}\\text{a}\\text{b}\\text{i}\\text{l}\\text{i}\\text{t}\\text{y}\\:/\\:\\text{C}\\text{o}\\text{n}\\text{t}\\text{r}\\text{o}\\text{l}\\:\\text{v}\\text{i}\\text{a}\\text{b}\\text{i}\\text{l}\\text{i}\\text{t}\\text{y}\\left)\\right]\\:\\times\\:\\:100\\)\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e\u003cp\u003eCD45\u003csup\u003e+\u003c/sup\u003eTILs sorting and single-cell sequencing\u003c/p\u003e\u003cp\u003eRemaining tumor not used in PDX experiments were processed for CD45 + TIL isolation and single-cell RNA sequencing as previously described [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Briefly, cells were magnetically sorted using anti-human CD45 microbeads (130-118-780, Miltenyi Biotec, Germany) and stained with anti-CD45 antibodies (304008, BioLegend, USA) to evaluate sorting efficiency by flow cytometry. The TILs were subsequently processed using the 10x Chromium System (10x Genomics, USA).\u003c/p\u003e\u003cp\u003eSingle-cell data processing\u003c/p\u003e\u003cp\u003eThe Raw fastq files were processed using Cell Ranger (version 7.1.0). Subsequent analyses were conducted with the “Seurat” R package (version 4.2.3). During quality control, genes expressed in \u0026gt; 3 cells and cells expressing \u0026gt; 200 genes were retained. Cells with \u0026gt; 20% mitochondrial or \u0026gt; 0.1% hemoglobin gene expression were excluded. Potential doublets were removed using “DoubletFinder” (version 2.0.4) [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. The remaining cells underwent normalization, scaling, variable gene selection, principal component analysis (PCA), batch effect correction using “Harmony” (version 1.2.0), and clustering (dims = 30, resolution = 0.5). Cell clusters were annotated using canonical markers and top-ranked differentially expressed genes (DEGs) (adjusted p-values \u0026lt; 0.05 and log2FC \u0026gt; 1). Cell cycle states were inferred using “CellCycleScoring”.\u003c/p\u003e\u003cp\u003eT and NK cell subsets were further characterized through subclustering following the aforementioned pipeline. To assess their tissue distribution preferences between nimotuzumab-responsive (R) and non-responsive (NR) groups, odds ratios (ORs) were calculated as previously described by Zhang et al. [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. DEGs within each subcluster between R and NR patients were identified using the “FindMarkers” and functional enrichment analyses of DEGs were performed with “clusterProfiler” (version 3.21). Additionally, to delineate the functional states of immune cells, we applied the “AddModuleScore” function to quantify 19 immune-related gene signatures defined by Chu et.al. [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e] and compared scores between R and NR groups.\u003c/p\u003e\u003cp\u003eAnalysis of cell-cell communication\u003c/p\u003e\u003cp\u003eIntercellular communication among TIL subclusters was inferred using the “CellChat” R package (version 1.5.0). Non-immune clusters (e.g., “Hepatocyte/Cancer Cell”, “Proliferating Cell”, “Endothelial Cell”, “Fibroblast”) were excluded to avoid bias. CellChat objects were grouped by treatment response, and the “rankNet” function was used to compare ligand–receptor interactions and signaling pathways between groups [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eImmune Infiltration and Treatment Correlation Analyses\u003c/p\u003e\u003cp\u003e We used the single-sample Gene Set Enrichment Analysis (ssGSEA) to quantify the enriching levels and relative abundances of immunocytes in samples from different HCC datasets using R package “GSVA” (version 1.48.3). The samples were stratified into high- and low- infiltration groups based on the median infiltration score. In addition, tumor microenvironment characteristics were assessed using ESTIMATE scores calculated via the R package “estimate” (version 1.0.13).\u003c/p\u003e\u003cp\u003ePrognostic gene identification and nomogram construction\u003c/p\u003e\u003cp\u003eUsing TCGA-LIHC as the training dataset, we developed a prognostic risk score model and validated it in two independent HCC cohorts (ICGC-JP and GSE141202) [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. DEGs were first identified between tumor and normal tissues within the TCGA-LIHC cohort. DEGs between tumor and normal tissues (|log₂FC| \u0026gt;1, FDR \u0026lt; 0.05) were overlapped with CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-related genes (adjusted p \u0026lt; 0.05, |correlation coefficient| \u0026gt;0.4). Prognostic genes were identified via univariate Cox regression and refined using least absolute shrinkage and selection operator (LASSO) regression method [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. The risk scores were calculated via the following formula:\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\:\\text{S}\\text{c}\\text{o}\\text{r}\\text{e}\\:=\\:{\\Sigma\\:}{\\beta\\:}\\text{ᵢ}\\:\\times\\:\\:\\text{E}\\text{x}\\text{p}\\text{ᵢ}\\)\u003c/span\u003e\u003c/span\u003e (β\u003csub\u003ei\u003c/sub\u003e: regression coefficient; Exp\u003csub\u003ei\u003c/sub\u003e: gene expression level). The predictive performance of the risk score was evaluated through Kaplan–Meier and receiver operating characteristic (ROC) curve analyses across both the training and validation cohorts.\u003c/p\u003e\u003cp\u003eUnivariate and multivariate Cox regression analyses were conducted incorporating the risk score and clinical parameters. Variables with p \u0026lt; 0.05 in multivariate analysis were used to construct a nomogram using the “rms” R package (version 6.4.0). Predictive accuracy was assessed using ROC curves and calibration plots in both the training and validation sets.\u003c/p\u003e\u003cp\u003eWES analysis\u003c/p\u003e\u003cp\u003eTo confirm the mutations status of prognosis-related genes. DNA was extracted from formalin-fixed, paraffin-embedded (FFPE) HCC tissues using the Maxwell 16 FFPE Plus LEV DNA Purification Kit (Promega, USA) and sequenced on the NovaSeq 6000 platform (Illumina, USA).\u003c/p\u003e\u003cp\u003eAfter sequencing, quality control of the raw FASTQ files was preformed using “FastQC” (version 0.12.1). The sequence data were processed according to “GATK” (version 3.5) best practices. Somatic single-nucleotide variants (SNVs) and indels were detected using MuTect2, filtered via “VCFtools” (version 0.1.16) and annotated using “SnpEff” (version 5.1f). Mutation profiles were visualized using the R package “maftools” (version 2.16.0) [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e].\u003c/p\u003e\u003ch2\u003eStatistical analysis\u003c/h2\u003e\u003cp\u003eStatistical comparisons were conducted using Student’s t-test or the Wilcoxon rank-sum test, depending on data distribution. One-way ANOVA was used for multiple group comparisons. Pearson and Spearman correlations were applied for continuous and ordinal variables, respectively. Survival analysis was performed using the log-rank test and visualized with Kaplan–Meier plots. All analyses were conducted in R software, with p \u0026lt; 0.05 considered statistically significant.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eTIL landscape in PDX models stratified by nimotuzumab response\u003c/p\u003e\u003cp\u003eWe successfully established PDX models from 14 patients in the REHOPE cohort, with TIL single-cell RNA sequencing conducted on 10 of TILs these patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA, S1A). In the PDX drug sensitivity assay, 7 samples exhibited reduced tumor cell viability following nimotuzumab treatment and were classified as responders, with maximum inhibition rate of 71.46%. The remaining 7 samples showed no significant response and were classified as non-responders, resulting in an overall nimotuzumab response rate of 50% (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB, S1B). Baseline characteristics, including age and BMI, were comparable between responders and non-responders (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Notably, patients 001, 017, and 026 demonstrated over a two-fold increase in tumor cell viability post-treatment, indicating strong resistance to nimotuzumab.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eBrief clinical characteristics of HCC patients included for the establishment of PDX models.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCharacteristic\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLevel\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eResponder(n\u0026thinsp;=\u0026thinsp;7)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-responder(n\u0026thinsp;=\u0026thinsp;7)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eP.value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSex (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e7 (100.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e7 (100.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge of diagnose (y/o, mean (SD))\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e64.29 (11.86)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e58.71 (11.80)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.396\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWeight (kg, mean (SD))\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e73.83 (13.11)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e73.29 (12.72)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.939\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHeight (cm, mean (SD))\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e168.71 (4.89)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e172.57 (6.48)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.232\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBMI (mean (SD))\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e25.97 (4.69)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e24.47 (2.77)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.480\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTumor diameter (cm, mean (SD))\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e6.25 (1.94)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.66 (3.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.698\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStage (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eT2N0M0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2 (28.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4 (57.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.589\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eT3N0M0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e5 (71.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e3 (42.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHBV (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2 (28.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0 ( 0.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.445\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e+\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e5 (71.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e7 (100.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHCV (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e6 (85.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e7 (100.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.999\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e+\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1 (14.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0 ( 0.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCirrhosis (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e5 (71.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5 (71.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.999\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e+\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2 (28.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e2 (28.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMVI (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e6 (85.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5 (71.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.999\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1 (14.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e2 (28.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKi-67 (median [IQR])\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.10 [0.08, 0.25]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.20 [0.10, 0.25]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.870\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCA199 (U/mL, median [IQR])\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e16.82 [13.47, 18.22]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e10.23 [7.14, 11.54]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.142\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAFP (ng/mL, median [IQR])\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e18.66 [3.63, 50.28]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e8.02 [3.42, 1062.52]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.949\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePVKII (ng/mL, median [IQR])\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e235.00 [81.00, 2224.50]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e849.00 [117.75, 2390.25]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.668\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCEA (ng/mL, median [IQR])\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e3.16 [1.98, 3.30]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e2.17 [1.68, 2.89]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.482\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNLR (mean (SD))\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2.00 (1.28)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e2.12 (1.17)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.856\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eALB (g/L, mean (SD))\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e40.57 (1.76)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e40.96 (4.44)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.834\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGLB (g/L, mean (SD))\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e28.37 (5.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e26.96 (7.66)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.694\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTBIL (\u0026micro;mol/L, mean (SD))\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e13.47 (3.28)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e11.49 (4.14)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.339\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe blood test was performed before surgery. BMI: body-mass index, HBV: hepatitis B virus, HCV: hepatitis C virus, MVI: micro-vascular invasion, AFP: alpha fetoprotein, PVKII: protein induced by vitamin K absence or antagonist II, CEA: carcinoembryonic antigen, NLR: neutrophil-to-lymphocyte ratio, ALB: albumin, GLB: globulin, TBIL: total bilirubin.\u003c/p\u003e\u003cp\u003eBased on PDX response, scRNA-seq samples were stratified into two groups. After quality control, 86,828 single cells were retained. UMAP-based clustering identified major immune subsets (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC, F) including lymphoid-derived: T cells, NK/NKT cells, dendritic cells (DCs), B cells, and plasma cells, as well as macrophages, myeloid-derived suppressor cells (MDSCs), and neutrophils. The macrophages were further subdivided into three subsets: monocyte-derived macrophages (Mφ-Monocytes), liver-resident Kupffer cells (Mφ-Kupffer), and tumor-associated macrophages (TAMs). A proliferating cell cluster characterized by high expression of cell cycle\u0026ndash;related genes were identified (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD). Despite CD45\u003csup\u003e+\u003c/sup\u003e enrichment, non-immune cells\u0026mdash;such as hepatocytes/HCC tumor cells, fibroblasts, and endothelial cells\u0026mdash;were detected, primarily from non-responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE), suggesting a potentially higher proportion of non-immune components in these tumors. Although the relative abundance of TIL subsets varied across samples, T cells, B cells, NK cells, and neutrophils were generally enriched in responder samples, indicating a greater immune infiltration (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eG). PCA-based unsupervised clustering stratified the samples into three distinct groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eH), suggesting that single-cell profiles can differentiate between response phenotypes. Of note, patients with mild resistance (011, 015, 016) and those with marked resistance (017, 026) were located distantly from each other in PCA space.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eDistinct T and NK cell features between different response groups\u003c/p\u003e\u003cp\u003eGiven the critical role of T and NK (TNK) cells in antitumor immunity, we performed subclustering of TNK populations. This analysis identified CD4⁺, CD8⁺, mucosal-associated invariant T (MAIT), and three NK subsets, along with smaller populations of proliferating T cells, innate lymphoid cells (ILCs), and mast-like cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA, E). Unlike the TIL composition, TNK subsets displayed more consistent distribution across samples but showed marked differences between response groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB-C). Tissue preference analysis [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e] revealed that activated T and NK subsets were predominantly enriched in responder samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD), indicative of enhanced immune activation. Conversely, non-responders showed enrichment of the ILCs (in minimal numbers), CD4-FOXP3 (Tregs), and CD4-CCR7 (Tcm) subsets. Among these, only the CD4-CCR7 proportion was significantly negatively correlated with nimotuzumab response (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF), while CD4-FOXP3 cells were not (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eG).\u003c/p\u003e\u003cp\u003eDifferential gene expression analysis of TNK subsets (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eH) revealed three notable features: 1) cells from the responses exhibited high levels of activation markers (CD69, FOS, TOX) and effector genes (GZMK, CCL4, NKG7); 2) cells from non-response exhibited upregulation of stress-related proteins (HSP90AB1, DNAJB1, BAG3), suggestive of a stress/damage phenotype (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eI); and 3) certain T cell subsets (e.g., CD4-CCL5, CD8-PDCD1) expressed clonally enriched TCR VDJ genes (e.g., TRAV13-1, TRBV5-6), suggesting clonal expansion of specific T cell populations. Further analysis of the CD4-CCR7 subset showed higher enrichment in cytokine production, metabolic activity, and TCR signaling scores among responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eJ\u0026ndash;K), suggesting increased functional activation in this population.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eEnhanced intercellular communication of CCR7\u003csup\u003e+\u003c/sup\u003eCD4\u003csup\u003e+\u003c/sup\u003e T cells in non-responders\u003c/p\u003e\u003cp\u003eTo investigate cell\u0026ndash;cell communication among TILs, we applied CellChat. While the overall number of interactions was similar between groups, interaction strength was markedly higher in non-responders, especially among the CD4-CCR7, Mφ-Monocyte, and CD8-PDCD1 subsets (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA, \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). Mφ-Monocytes consistently showed the strongest interactions across groups, whereas CD4-CCR7 T cells exhibited strong interactions specifically in non-responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). Input-output (IO) analysis confirmed elevated signaling activity in CD4-CCR7 and CD8-PDCD1 subsets in the non-responsive group that surpassed the average levels observed in the other clusters (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD).\u003c/p\u003e\u003cp\u003eComparison of ligand\u0026ndash;receptor interactions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE\u0026ndash;F) revealed enrichment of inflammatory (TNF, Chemerin [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]) and immunosuppressive pathways (CD30, IL10 [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]) in non-responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eG), whereas responder samples were enriched in cytokine- and immune activation\u0026ndash;related pathways, including CCL, CSF, and CLEC [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e] signaling (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eH). Notably, CD4-CCR7 subset was a key contributor to the signaling networks of non-responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eG, I).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAssociations between CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cells and EGFR antibodies response\u003c/p\u003e\u003cp\u003eTo investigate the association between CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cells and treatment response across different anti-EGFR regimens and tumor types, we performed ssGSEA using CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T-cell marker genes across the TCGA and GEO datasets. In the TCGA-LIHC cohort, tumor tissues exhibited significantly lower CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T-cell scores compared to adjacent normal tissues (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA), and higher infiltration of these cells was negatively correlated with patient survival (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB).\u003c/p\u003e\u003cp\u003eWe next analyzed GSE102995 [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e] dataset, which includes clinical data from panitumumab-treated patients. CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell scores were significantly negatively associated with progression-free survival (PFS) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC). To further assess whether these cells fluctuate during treatment, we examined GSE180480, which contains course gene expression data from patient-derived organoids. Compared to the baseline, CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T-cell scores significantly decreased at the first emergence of drug resistance, suggesting reduced infiltration during treatment response (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD).\u003c/p\u003e\u003cp\u003eA similar result was found in the GSE108277 [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e] dataset, which includes a PDX model treated with cetuximab, samples exhibiting partial responses showed significantly lower CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell scores than non-responders did (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE). We further investigated whether these associations extended to other TKI. In the GSE109211 dataset [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e], which studied sorafenib treatment in HCC, responders had significantly lower CD4⁺CCR7⁺ T-cell scores than non-responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eF), with the difference more pronounced than that seen in EGFR antibody datasets.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eConstruction of a CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-associated risk signature\u003c/p\u003e\u003cp\u003eTo evaluate the prognostic value of CD4⁺CCR7⁺ T-cell infiltration in HCC, we developed a risk score based on gene expression associated with these cells. Differential expression analysis between 371 tumor and 51 normal samples in the TCGA-LIHC cohort identified 13,581 DEGs (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA), comprising 1,354 upregulated and 11,227 downregulated genes. Among these, 410 were significant associations with the CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell cluster (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). Univariate Cox regression identified 43 prognostically relevant genes among them, with 37 associated with increased risk and 6 with protective effects.\u003c/p\u003e\u003cp\u003eFunctional enrichment analysis of these 410 genes revealed significant involvement of GO terms including cell motility, angiogenesis, membrane structures and antigen binding (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC). KEGG pathway analysis revealed associations with cytokine signaling, TNF signaling, the PI3K-Akt pathway, cytoskeletal regulation, and specific infection-related pathways, consistent with the KEGG pathways identified from single-cell CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell cluster DEGs (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eK).\u003c/p\u003e\u003cp\u003eWe used LASSO regression to refine the model, selecting 9 key genes at log(λ) = \u0026minus;\u0026thinsp;0.31 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD-E), including four protective genes (N4BP3, UAP1L1, MTHFD1L, MSC) and five risk-associated genes (LTK, PGF, CAMK4, SPRED3, TNFRSF18). The final multivariate model was constructed using stepwise selection, resulting in the following risk score formula (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eF): \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\text{R}\\text{i}\\text{s}\\text{k}\\:\\text{S}\\text{c}\\text{o}\\text{r}\\text{e}=1.4\\times\\:\\text{N}4\\text{B}\\text{P}3+0.096\\times\\:\\text{U}\\text{A}\\text{P}1\\text{L}1+0.056\\times\\:\\text{M}\\text{T}\\text{H}\\text{F}\\text{D}1\\text{L}+0.032\\times\\:\\text{M}\\text{S}\\text{C}-0.85\\times\\:\\text{S}\\text{P}\\text{R}\\text{E}\\text{D}3-0.22\\times\\:\\text{C}\\text{A}\\text{M}\\text{K}4-0.12\\times\\:\\text{L}\\text{T}\\text{K}-0.045\\times\\:\\text{P}\\text{G}\\text{F}-0.005\\times\\:\\text{T}\\text{N}\\text{F}\\text{R}\\text{S}\\text{F}18\\)\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eRisk scores were calculated for each TCGA sample, and patients were stratified into high- and low-risk groups. Survival analysis revealed significantly poorer survival outcomes in the high-risk group (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eG, \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eI), with area under the curve (AUC) values of 0.705, 0.700, and 0.730 for 1-, 3-, and 5-year survival, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eH). External validation using the ICGC-JP (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eJ) and GSE141202 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eK) cohorts confirmed the model's predictive power, as high-risk patients consistently had poorer survival outcomes.\u003c/p\u003e\u003cp\u003eConstruction and validation of a prognostic nomogram\u003c/p\u003e\u003cp\u003eTo determine whether the CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-associated risk score serves as an independent prognostic factor, we conducted both univariate (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA) and multivariate (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB) Cox regression analyses incorporating baseline characteristics sex, age, clinical stage, hepatitis B (HBV) infection, alcohol consumption, and tumor grade. HBV status, clinical stage, and the risk score were identified as independent predictors of overall HCC survival.\u003c/p\u003e\u003cp\u003eA prognostic nomogram integrating these variables was developed to predict overall survival (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC). Kaplan\u0026ndash;Meier analysis demonstrated significantly worse outcomes in the nomogram high-risk group (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eD). Calibration curves and ROC analysis indicated strong predictive performance, with AUCs ranging from 0.669 to 0.793 (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eE-F). The nomogram was further validated in the GSE141202 dataset, showing comparable AUCs (0.685\u0026ndash;0.824) for 1- to 3-year survival predictions (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eG\u0026ndash;H), supporting its robustness.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAssociation between the CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-associated risk score and immune infiltration\u003c/p\u003e\u003cp\u003eTo investigate the immunological context of the risk score, we examined its correlation with immune infiltration in the GSE196576 dataset [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. The CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T level were positively correlated with the expression of the 9-gene risk signature (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eA). Using the MCPcounter algorithm, we confirmed that higher risk scores were associated with increased CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T-cell infiltration in both the TCGA (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB) and GSE196576 (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eC) cohorts. We next applied the ESTIMATE algorithm to assess stromal and immune infiltration. In GSE196576, the low-risk group had significantly higher ESTIMATE scores (p\u0026thinsp;=\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eD), stromal scores (p\u0026thinsp;=\u0026thinsp;0.003; Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eE), and immune scores (p\u0026thinsp;=\u0026thinsp;0.003; Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eF), indicating a greater non-tumor and immune component in the tumor microenvironment. These findings suggest that patients with lower risk scores have higher immune cell infiltration.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWES analyses and validation of the CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-associated prognostic genes\u003c/p\u003e\u003cp\u003eTo validate the presence of genetic alterations in CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-associated risk genes, we performed WES on 13 tumor samples corresponding to the PDX models. While no significant differences in variant types or SNV classes were observed between responders and non-responders, the non-responders exhibited a higher median mutation burden (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eA-B). We subsequently examined genetic mutations within the 43 risk signature genes. Among these genes, three (VCAN, CD226, and CAMK4) were mutated exclusively in non-responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eC), while seven genes (STRA6, NTM, GLP2R, ADAM12, CTHRC1, HTRA3, and SCG2) were mutated only in responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eD), suggesting distinct mutational profiles between the two groups. Notably, VCAN mutations were present in 60% (3/5) of non-responders (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eE). Survival analysis in the TCGA-LIHC cohort revealed that patients with VCAN mutations had significantly poorer survival (p\u0026thinsp;=\u0026thinsp;0.015; Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eF). All three VCAN-mutant patients harbored SNVs at distinct sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eG). In summary, our WES result confirmed the presence of mutations in CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-related risk genes, with distinct mutational patterns between responders and non-responders to nimotuzumab.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eAs a pivotal receptor orchestrating various proliferative and survival pathways, EGFR is associated with aggressive tumor behavior, metastasis, and poor patient survival. However, clinical trials evaluating erlotinib or cetuximab monotherapy in advanced HCC patients have not demonstrated clinical benefits. Nevertheless, preclinical studies and case reports suggest that EGFR antibodies may yield favorable outcomes in certain HCC patients [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Recent trials focusing on combining EGFR antibodies with chemotherapy [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e], ICIs [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e] or mTOR inhibitors [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e], had shown promising results. Unlike small-molecule TKIs, EGFR antibodies also exert antitumor effects through ADCC and immune modulation [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. Therefore, their efficacy is influenced by the infiltration and functional states of immune cells in the HCC microenvironment. In this study, we employed PDX models and single-cell RNA sequencing to analyze TILs from HCC samples treated with nimotuzumab, a humanized anti-EGFR monoclonal antibody. Our funding revealed a relationship between TIL composition and the therapeutic efficacy of EGFR antibody treatment.\u003c/p\u003e\u003cp\u003eThe PDX model demonstrated that nimotuzumab achieved a 50% response rate in HCC samples, outerforming our previous in-vitro studies using HCC cell lines and clinical trials involving cetuximab. Notably, the therapeutic response varied considerably, ranging from 70% tumor inhibition to more than 10-fold tumor growth. Single-cell sequencing revealed that nimotuzumab-resistant samples harbored a higher proportion of tumor cells and tumor-associated stromal components, consistent with previous studies suggesting that increased tumor burden and heterogeneity negatively correlate with EGFR antibody efficacy. Moreover, the responsive samples showed elevated neutrophil and B cell infiltration, whereas T cell, NK cell, and macrophage levels remained relatively unchanged. These findings suggest that general variations in these immune subsets may not directly determine nimotuzumab efficacy. The increased neutrophil and B cell content in responders may reflect enhanced humoral immunity and non-specific inflammation, consistent with prior reports [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. PCA further revealed distinct transcriptional patterns in highly resistant samples, indicating the presence of multiple resistance mechanisms.\u003c/p\u003e\u003cp\u003ePrevious studies have demonstrated that EGFR antibodies exert antitumor effects through T and NK cell-mediated mechanisms, particularly via ADCC [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. Compared with the overall TIL distribution, T and NK cells show relatively uniform abundance across samples, allowing us to explore tissue-specific functional subsets. Our analysis revealed enrichment of effector clusters such as NK-CCL3, CD4-GZMB, CD4-CCL5, and MAIT-GZMK in responders, suggesting that immune activation contributes to the efficacy of EGFR antibodies. Notably, there were no significant differences in CD8\u003csup\u003e+\u003c/sup\u003e T cell populations between groups. Previous studies have shown that EGFR signaling promotes CD4\u003csup\u003e+\u003c/sup\u003e T cell activation and apoptosis via Glut1-mediated pathways [\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. Activated CD4\u003csup\u003e+\u003c/sup\u003e T cells, in turn, support NK cell survival and enhance ADCC. In contrast, non-responders were enriched in CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cells and Tregs. While Tregs have been reported to impair EGFR antibody efficacy [\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e], the role of ILCs and CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cells remains poorly defined. Our study identified CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cells as markers of EGFR antibody resistance for the first time.\u003c/p\u003e\u003cp\u003eCCR7 is a hallmark of central memory T cells. The role of CD4\u003csup\u003e+\u003c/sup\u003e memory T cells in tumors is remains controversial. Prior studies suggest that pembrolizumab can inhibit the differentiation of T cells into CCR7\u003csup\u003e+\u003c/sup\u003e phenotypes and enhance immune responses. In our study, non-responders exhibited not only higher quantities of CD4-CCR7 subsets but also different functional characteristics. CD4-CCR7 cells from responders expressed higher levels of effector and activation markers such as CCL4, NKG7, and IL32. In contrast, T cells from non-responders expressed higher levels of na\u0026iuml;ve and anti-apoptotic markers, indicating a dysfunctional or exhausted state that may impair ADCC [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e]. Additionally, other immune cell subsets in non-responders also showed increased expression of stress response genes like HSP90, indicating a stressed immune microenvironment. Chu et al. previously showed that HSP gene overexpression can suppress specific TIL activation [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Overall, non-responders exhibited reduced effector TILs, increased CD4⁺ memory T cells, and signs of impaired immune function.\u003c/p\u003e\u003cp\u003eTo further explore the interactions between CD4-CCR7 cells and other immune cells, we performed CellChat analysis. Monocyte-derived macrophages showed the strongest overall intercellular communication; however, they did not significantly differ between groups. Interestingly, CD4-CCR7 cells exhibited enhanced intercellular communication specifically in non-responders, particularly with CD8-PDCD1 cells and macrophages. Previous studies suggest a negative correlation between CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e central memory T cells and CD8\u003csup\u003e+\u003c/sup\u003ePDCD1\u003csup\u003e+\u003c/sup\u003e effector memory re-expressing CD45RA (TEMRA) cells [\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e], potentially linked to protection against immune-related adverse events. However, the interaction between CD4⁺CCR7⁺ T cells and CD8⁺ T during treatment resistance has not been previously reported.\u003c/p\u003e\u003cp\u003eSubsequent pathway-specific communication analysis revealed distinct signaling patterns between groups. Responders were enriched in CCL4 and CLEC family pathways, involving activated MAIT, na\u0026iuml;ve CD4⁺ T, and NK cells, consistent with prior studies showing CLEC4 downregulation during tumor progression and positive correlations with CD4⁺ T and B cell infiltration. In contrast, non-responders showed enrichment of immunosuppressive (IL-10) and pro-inflammatory (CHEMERIN, TNF) pathways. Notably, macrophages in non-responsive samples exhibited high IL-10 expression, a marker of TAM polarization [\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e]. IL-10 has been shown to limit the proliferation of effector CD8 T cells following anti-PD-1 therapy, while reducing immune-related adverse events [\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e]. These results suggest that non-responders display increased immunosuppression and non-specific inflammation, which may contribute to poor prognosis.\u003c/p\u003e\u003cp\u003eBased on the ssGSEA results, we infer the CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell infiltration levels in multiple anti-EGFR bulk RNA-seq datasets to further validate our findings. Lower CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T levels were consistently correlated with better outcomes in treatments involving cetuximab, panitumumab, and even sorafenib. Additionally, reduced abundance of CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T was observed in tumor compared to normal tissues, in line with prior studies [\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e]. In the GSE180480 dataset, CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T scores decreased at effective treatment timepoints compared to baseline. As a memory phenotype, a reduction in CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T may indicate increased CD4\u003csup\u003e+\u003c/sup\u003e T cell activation. Supporting this, studies have reported elevated CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cells levels in both relapsed breast cancer [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e], and positive lymph nodes [\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e]. Thus, while CD4⁺CCR7⁺ T cell abundance negatively correlates with EGFR therapy response, they may not directly promote tumor progression. Instead, their accumulation may indicate insufficient immune activation raised by EGFR antibodies. Interestingly, CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cells have also been associated with better ICI outcomes [\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e], suggesting that combining ICIs with EGFR antibodies may improve therapeutic efficacy in HCC.\u003c/p\u003e\u003cp\u003eGiven the limited feasibility of routine scRNA-seq or immune profiling in clinical practice, we identified CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-associated genes in tumor tissue to construct a prognostic model. Therefore, we aimed to identify CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-associated genes within tumor tissues to construct a prognostic model. The risk model comprises nine genes: N4BP3, UAP1L1, MTHFD1L, MSC, LTK, PGF, CAMK4, SPRED3, and TNFRSF18. MTHFD1L and UAP1L1 are involved in cellular metabolism; CAMK4, LTK, TNFRSF18, SPRED3, and MSC play roles in immune signaling and regulation. Notably, SPRED3 and LTK may suppress CD4\u003csup\u003e+\u003c/sup\u003e T cell responses via the MAPK and JAK-STAT pathways, and MSC and TNFRSF18 are implicated in the activation of Tregs and T cell exhaustion. PGF may effect the microenvironment remodeling and angiogenesis [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]. Together, these genes reflect immune suppression and tumor adaptation.\u003c/p\u003e\u003cp\u003eThis model is the first HCC prognostic framework based on CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cells. Compared with previous models focusing on exhausted CD4\u003csup\u003e+\u003c/sup\u003e T cells [\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e], it demonstrates superior prognostic performance. The risk score also negatively correlated with immune infiltration. Patients with lower risk scores exhibit higher non-tumor components and immune cell infiltration, markers of favorable prognosis [\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e]. These results are consistent with our single-cell findings and highlight the model\u0026rsquo;s ability to predict immune status as well as the clinical outcomes.\u003c/p\u003e\u003cp\u003eWES analysis showed no EGFR mutations in either group, suggesting resistance was not driven by EGFR alterations [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Higher tumor mutation burden (TMB) was demonstrated to be favorable factor for cetuximab therapy in CRC patients. However, in this study, higher median mutations were found in non-responders[\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e]. We hypothesized that higher TMB may indicate increased malignancy or EGFR-independent signaling activation. Among the nine model genes, only CAMK4 was mutated. While CAMK4 inhibition suppresses tumor growth, it also regulates Th17 cell development [\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. Notably, VCAN mutations were detected in 60% of resistant samples but absent in responders. VCAN encodes a major extracellular matrix component and has been implicated in HCC progression and poor prognosis [\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e]. It could influence EGFR signaling via the ADAMTS1\u0026ndash;VCAN\u0026ndash;EGFR axis in renal cancer [\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e], and this study is the first to suggest its role in anti-EGFR resistance in HCC.\u003c/p\u003e\u003cp\u003eThis study has several limitations. First, scRNA-seq was performed only on pre-treatment samples, limiting our ability to assess dynamic changes in CD4⁺CCR7⁺ T cells during therapy. Second, due to a lack of fresh tissue, we could only evaluate gene mutations by WES, without RNA or protein-level validation. Third, although our findings were supported by external datasets involving different EGFR-targeted therapies, further studies are needed to determine whether other EGFR antibodies or TKIs have distinct immune biomarkers or mechanisms.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eElevated proportions of CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cells in HCC TILs are associated with resistance to nimotuzumab treatment, characterized by a reduction in effector immune cells and attenuated immune activation. A prognostic model based on CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+\u003c/sup\u003e T cell-associated genes effectively predicts survival risk in HCC patients and inversely correlates with the extent of immune cell infiltration within the tumor microenvironment.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eEGFR\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eEpidermal Growth Factor Receptor\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eHCC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eHepatocellular Carcinoma\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePDX\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003ePatient-Derived Xenograft\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eWES\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eWhole-Exome Sequencing\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eLASSO\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eLeast Absolute Shrinkage and Selection Operator\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eADCC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eAntibody-Dependent Cellular Cytotoxicity\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eHNSCC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eHead and Neck Squamous Cell Carcinoma\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eCRC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eColorectal Cancer\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eTILs\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eTumor-Infiltrating Lymphocytes\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePCA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003ePrincipal Component Analysis\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eUMAP\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eUniform Manifold Approximation and Projection\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eNR\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eNon-Responder\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eDEG\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eDifferentially Expressed Gene\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003essGSEA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eSingle Sample Gene Set Enrichment Analysis\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eGSVA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eGene Set Variation Analysis\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eTCGA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eThe Cancer Genome Atlas\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eICGC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eInternational Cancer Genome Consortium\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eFDR\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eFalse Discovery Rate\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eROC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eReceiver Operating Characteristic (curve)\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eGATK\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eGenome Analysis Toolkit\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eMVI\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eMicrovascular Invasion\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eMAIT\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eMucosal-Associated Invariant T (cell)\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eKEGG\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eKyoto Encyclopedia of Genes and Genomes\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePFS\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eProgression-Free Survival\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePR\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003ePartial Response\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eGO\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eGene Ontology\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eAUC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eArea Under the Curve\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eHR\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eHazard Ratio\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eSNV\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eSingle Nucleotide Variant\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eTEMRA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eTerminally Differentiated Effector Memory CD8⁺ T Cell Re-expressing CD45RA\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eTAM\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eTumor-Associated Macrophage\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eICI\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eImmune Checkpoint Inhibitor\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eTMB\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eTumor Mutational Burden\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthical approval and consent to participate\u003c/strong\u003e\u003cp\u003eThis study complies with the Helsinki Declaration. All participants provided written informed consent prior to enrollment. All animal experiments comply with the ARRIVE guidelines. Animals were purchased from Charles River Laboratories with informed consent and maintained under standard laboratory conditions. Mice were euthanized by cervical dislocation while under deep anesthesia after experiment. Both the clinical study and animal experiments protocol received approval from the Ethics Committee of PLA General Hospital (approval number S2022-313-02, dated July 27, 2022.).\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003cp\u003eNot applicable.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eNot applicable.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eZY, LQ and GJ contributed equally to this research and share the first authorship.ZY and LQ wrote the main manuscript text and figures. GJ, ZY, YG, ZS and YW performed the laboratory work and assisted with the manuscript. YC, ZS and LS performed single cell RNA and whole-exome sequencing. ZY and LS performed the statistical and bioinformatics analyses. ZS and XZ collected surgical samples. XZ and SJ conceptualized the project and edited the manuscript.All the authors have read and approved the contents of the manuscript prior to submission.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe thank Beijing DCTY Biotech Co., Ltd. for support of this research.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe sequencing data used in this study have been deposited in the China National Center for Bioinformation (https://ngdc.cncb.ac.cn/) GSA human database with the primary accession code HRA011337 (single cell sequence) and HRA012698 (whole-exome sequence). The public sequencing data used to support this study were obtained from public databases TCGA (https://portal.gdc.cancer.gov), ICGC (https://dcc.icgc.org) and GEO (GSE108277, GSE102995, GSE180480, GSE109211, GSE141202 and GSE196576, https://www.ncbi.nlm.nih.gov/gds). Other datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eHwang SY, Danpanichkul P, Agopian V, Mehta N, Parikh ND, Abou-Alfa GK, Singal AG, Yang JD. Hepatocellular carcinoma: updates on epidemiology, surveillance, diagnosis and treatment. Clin Mol Hepatol. 2025;31(Suppl):S228\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDanpanichkul P, Suparan K, Tothanarungroj P, Dejvajara D, Rakwong K, Pang Y, Barba R, Thongpiya J, Fallon MB, Harnois D, et al. Epidemiology of gastrointestinal cancers: a systematic analysis from the Global Burden of Disease Study 2021. Gut. 2024;74(1):26\u0026ndash;34.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKudo M, Finn RS, Qin S, Han KH, Ikeda K, Piscaglia F, Baron A, Park JW, Han G, Jassem J, et al. Lenvatinib versus sorafenib in first-line treatment of patients with unresectable hepatocellular carcinoma: a randomised phase 3 non-inferiority trial. Lancet (London England). 2018;391(10126):1163\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKim HD, Jung S, Lim HY, Ryoo BY, Ryu MH, Chuah S, Chon HJ, Kang B, Hong JY, Lee HC, et al. Regorafenib plus nivolumab in unresectable hepatocellular carcinoma: the phase 2 RENOBATE trial. Nat Med. 2024;30(3):699\u0026ndash;707.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQiang Z, Wan J, Chen X, Wang H. Mechanisms and therapeutic targets of ErbB family receptors in hepatocellular carcinoma: a narrative review. Translational cancer Res. 2024;13(6):3156\u0026ndash;78.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHaber PK, Puigveh\u0026iacute; M, Castet F, Lourdusamy V, Montal R, Tabrizian P, Buckstein M, Kim E, Villanueva A, Schwartz M, et al. Evidence-Based Management of Hepatocellular Carcinoma: Systematic Review and Meta-analysis of Randomized Controlled Trials (2002\u0026ndash;2020). Gastroenterology. 2021;161(3):879\u0026ndash;98.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSueangoen N, Tantiwetrueangdet A, Panvichian R. HCC-derived EGFR mutants are functioning, EGF-dependent, and erlotinib-resistant. Cell bioscience. 2020;10:41.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVan Cutsem E, K\u0026ouml;hne CH, Hitre E, Zaluski J, Chang Chien CR, Makhson A, D'Haens G, Pint\u0026eacute;r T, Lim R, Bodoky G, et al. Cetuximab and chemotherapy as initial treatment for metastatic colorectal cancer. N Engl J Med. 2009;360(14):1408\u0026ndash;17.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePatil VM, Noronha V, Joshi A, Agarwal J, Ghosh-Laskar S, Budrukkar A, Murthy V, Gupta T, Mahimkar M, Juvekar S, et al. A randomized phase 3 trial comparing nimotuzumab plus cisplatin chemoradiotherapy versus cisplatin chemoradiotherapy alone in locally advanced head and neck cancer. Cancer. 2019;125(18):3184\u0026ndash;97.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhu AX, Stuart K, Blaszkowsky LS, Muzikansky A, Reitberg DP, Clark JW, Enzinger PC, Bhargava P, Meyerhardt JA, Horgan K, et al. Phase 2 study of cetuximab in patients with advanced hepatocellular carcinoma. Cancer. 2007;110(3):581\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang X, Li T, Slebos RJC, Chaudhary R, Guevara-Patino JA, Bonomi M, Saba NF, Chung CH. Clinical significance of peripheral T-cell repertoire in head and neck squamous cell carcinoma treated with cetuximab and nivolumab. Cancer Immunol immunotherapy: CII. 2025;74(4):142.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMahgoub S, Abosalem H, Emara M, Kotb N, Maged A, Soror S. Restoring NK cells functionality via cytokine activation enhances cetuximab-mediated NK-cell ADCC: A promising therapeutic tool for HCC patients. Mol Immunol. 2021;137:221\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZaryouh H, De Pauw I, Baysal H, Melis J, Van den Bossche V, Hermans C, Lau HW, Lambrechts H, Merlin C, Corbet C, et al. Establishment of head and neck squamous cell carcinoma mouse models for cetuximab resistance and sensitivity. Cancer drug Resist (Alhambra Calif). 2023;6(4):709\u0026ndash;28.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang Y, Zhang M, Gong Y, Wu Q, Zhang L, Jiao S. Bioinformatic Analysis of Hepatocellular Carcinoma Cell Lines to the Efficacy of Nimotuzumab. Int J Gen Med. 2021;14:2611\u0026ndash;21.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWu X, Yan H, Qiu M, Qu X, Wang J, Xu S, Zheng Y, Ge M, Yan L, Liang L. Comprehensive characterization of tumor microenvironment in colorectal cancer via molecular analysis. eLife 2023, 12.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGe H, Ferris RL, Wang JH. Cetuximab Responses in Patients with HNSCC Correlate to Clonal Expansion Feature of Peripheral and Tumor-Infiltrating T Cells with Top T-Cell Receptor Clonotypes. Clin cancer research: official J Am Association Cancer Res. 2023;29(3):647\u0026ndash;58.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSu T, Huang M, Liao J, Lin S, Yu P, Yang J, Cai Y, Zhu S, Xu L, Peng Z, et al. Insufficient Radiofrequency Ablation Promotes Hepatocellular Carcinoma Metastasis Through N6-Methyladenosine mRNA Methylation-Dependent Mechanism. Hepatology (Baltimore MD). 2021;74(3):1339\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBoreddy SR, Nair R, Pandey PK, Kuriakose A, Marigowda SB, Dey C, Banerjee A, Kulkarni H, Sagar M, Krishn SR, et al. BCA101 Is a Tumor-Targeted Bifunctional Fusion Antibody That Simultaneously Inhibits EGFR and TGFβ Signaling to Durably Suppress Tumor Growth. Cancer Res. 2023;83(11):1883\u0026ndash;904.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePircher A, Gamerith G, Amann A, Reinold S, Popper H, G\u0026auml;chter A, Pall G, W\u0026ouml;ll E, Jamnig H, Gastl G, et al. Neoadjuvant chemo-immunotherapy modifies CD4(+)CD25(+) regulatory T cells (Treg) in non-small cell lung cancer (NSCLC) patients. Lung cancer (Amsterdam Netherlands). 2014;85(1):81\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQin S, Li J, Bai Y, Wang Z, Chen Z, Xu R, Xu J, Zhang H, Chen J, Yuan Y, et al. Nimotuzumab Plus Gemcitabine for K-Ras Wild-Type Locally Advanced or Metastatic Pancreatic Cancer. J Clin oncology: official J Am Soc Clin Oncol. 2023;41(33):5163\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRamakrishnan MS, Eswaraiah A, Crombet T, Piedra P, Saurez G, Iyer H, Arvind AS. Nimotuzumab, a promising therapeutic monoclonal for treatment of tumors of epithelial origin. mAbs. 2009;1(1):41\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSong P, Yang J, Li X, Huang H, Guo X, Zhou G, Xu X, Cai Y, Zhu M, Wang P, et al. Hepatocellular carcinoma treated with anti-epidermal growth factor receptor antibody nimotuzumab: A case report. Medicine. 2017;96(39):e8122.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGarrido G, Tikhomirov IA, Rabasa A, Yang E, Gracia E, Iznaga N, Fern\u0026aacute;ndez LE, Crombet T, Kerbel RS, P\u0026eacute;rez R. Bivalent binding by intermediate affinity of nimotuzumab: a contribution to explain antibody clinical profile. Cancer Biol Ther. 2011;11(4):373\u0026ndash;82.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCrombet T, Osorio M, Cruz T, Roca C, del Castillo R, Mon R, Iznaga-Escobar N, Figueredo R, Koropatnick J, Renginfo E, et al. Use of the humanized anti-epidermal growth factor receptor monoclonal antibody h-R3 in combination with radiotherapy in the treatment of locally advanced head and neck cancer patients. J Clin oncology: official J Am Soc Clin Oncol. 2004;22(9):1646\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu WN, Harden SL, Tan SLW, Tan RJR, Fong SY, Tan SY, Liu M, Karnik I, Shuen TWH, Toh HC, et al. Single-cell RNA sequencing reveals anti-tumor potency of CD56(+) NK cells and CD8(+) T cells in humanized mice via PD-1 and TIGIT co-targeting. Mol therapy: J Am Soc Gene Therapy. 2024;32(11):3895\u0026ndash;914.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSun Z, Gao B, Song L, Wang B, Li J, Jiang H, Li X, Yu Y, Zhou Z, Yang Z, et al. Single-cell RNA sequencing reveals intratumoral heterogeneity and multicellular community in primary hepatocellular carcinoma underlying microvascular invasion. Heliyon. 2024;10(18):e37233.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGao B, Wang Y, Lu S. Cellular senescence affects energy metabolism, immune infiltration and immunotherapeutic response in hepatocellular carcinoma. Sci Rep. 2023;13(1):1137.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang F, Wang W, Long Y, Liu H, Cheng J, Guo L, Li R, Meng C, Yu S, Zhao Q, et al. Characterization of drug responses of mini patient-derived xenografts in mice for predicting cancer patient clinical therapeutic response. Cancer Commun (London England). 2018;38(1):60.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNeuhaus W, Lauer R, Oelzant S, Fringeli UP, Ecker GF, Noe CR. A novel flow based hollow-fiber blood-brain barrier in vitro model with immortalised cell line PBMEC/C1-2. J Biotechnol. 2006;125(1):127\u0026ndash;41.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcGinnis CS, Murrow LM, Gartner ZJ. DoubletFinder: Doublet Detection in Single-Cell RNA Sequencing Data Using Artificial Nearest Neighbors. Cell Syst. 2019;8(4):329\u0026ndash;e337324.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZheng L, Qin S, Si W, Wang A, Xing B, Gao R, Ren X, Wang L, Wu X, Zhang J, et al. Pan-cancer single-cell landscape of tumor-infiltrating T cells. Sci (New York NY). 2021;374(6574):abe6474.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChu Y, Dai E, Li Y, Han G, Pei G, Ingram DR, Thakkar K, Qin JJ, Dang M, Le X, et al. Pan-cancer T cell atlas links a cellular stress response state to immunotherapy resistance. Nat Med. 2023;29(6):1550\u0026ndash;62.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang C, Cheng X, Gao S, Pan Q. Integrating bulk and single-cell data to predict the prognosis and identify the immune landscape in HNSCC. J Cell Mol Med. 2024;28(1):e18009.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHsu CL, Ou DL, Bai LY, Chen CW, Lin L, Huang SF, Cheng AL, Jeng YM, Hsu C. Exploring Markers of Exhausted CD8 T Cells to Predict Response to Immune Checkpoint Inhibitor Therapy for Hepatocellular Carcinoma. Liver cancer. 2021;10(4):346\u0026ndash;59.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMa J, Chen Z, Hou L. Revealing a cancer-associated fibroblast-based risk signature for pancreatic adenocarcinoma through single-cell and bulk RNA-seq analysis. Aging. 2024;16(18):12525\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhao L, Yang W, Yang X, Lin Y, Lv J, Dou X, Luo Q, Dong J, Chen Z, Chu Y, et al. Chemerin suppresses murine allergic asthma by inhibiting CCL2 production and subsequent airway recruitment of inflammatory dendritic cells. Allergy. 2014;69(6):763\u0026ndash;74.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang W, Liu H, Xu L, Yu T, Zhao X, Yao S, Zhao Q, Barnes S, Cohn SM, Dann SM, et al. GPR120 Inhibits Colitis Through Regulation of CD4(+) T Cell Interleukin 10 Production. Gastroenterology. 2022;162(1):150\u0026ndash;65.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLopez Robles MD, Pallier A, Huchet V, Le Texier L, Remy S, Braudeau C, Delbos L, Moreau A, Louvet C, Brosseau C, et al. Cell-surface C-type lectin-like receptor CLEC-1 dampens dendritic cell activation and downstream Th17 responses. Blood Adv. 2017;1(9):557\u0026ndash;68.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSiano M, Espeli V, Mach N, Bossi P, Licitra L, Ghielmini M, Frattini M, Canevari S, De Cecco L. Gene signatures and expression of miRNAs associated with efficacy of panitumumab in a head and neck cancer phase II trial. Oral Oncol. 2018;82:144\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLindner AU, Carberry S, Monsefi N, Barat A, Salvucci M, O'Byrne R, Zanella ER, Cremona M, Hennessy BT, Bertotti A, et al. Systems analysis of protein signatures predicting cetuximab responses in KRAS, NRAS, BRAF and PIK3CA wild-type patient-derived xenograft models of metastatic colorectal cancer. Int J Cancer. 2020;147(10):2891\u0026ndash;901.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePinyol R, Montal R, Bassaganyas L, Sia D, Takayama T, Chau GY, Mazzaferro V, Roayaie S, Lee HC, Kokudo N, et al. Molecular predictors of prevention of recurrence in HCC with sorafenib as adjuvant treatment and prognostic factors in the phase 3 STORM trial. Gut. 2019;68(6):1065\u0026ndash;75.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYazdani A, Lenz HJ, Pillonetto G, Mendez-Giraldez R, Yazdani A, Sanoff H, Hadi R, Samiei E, Venook AP, Ratain MJ, et al. Gene signatures derived from transcriptomic-causal networks stratify colorectal cancer patients for effective targeted therapy. Commun Med. 2025;5(1):9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAsnacios A, Fartoux L, Romano O, Tesmoingt C, Louafi SS, Mansoubakht T, Artru P, Poynard T, Rosmorduc O, Hebbar M, et al. Gemcitabine plus oxaliplatin (GEMOX) combined with cetuximab in patients with progressive advanced stage hepatocellular carcinoma: results of a multicenter phase 2 study. Cancer. 2008;112(12):2733\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eReusch U, Sundaram M, Davol PA, Olson SD, Davis JB, Demel K, Nissim J, Rathore R, Liu PY, Lum LG. Anti-CD3 x anti-epidermal growth factor receptor (EGFR) bispecific antibody redirects T-cell cytolytic activity to EGFR-positive cancers in vitro and in an animal model. Clin cancer research: official J Am Association Cancer Res. 2006;12(1):183\u0026ndash;90.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGeng J, Li X, Lang X, Qiao C, Hu M, Yang J, Feng J, Lv M. Combination of cetuximab and rapamycin enhances the therapeutic efficacy in hepatocellular carcinoma. Technol Cancer Res Treat. 2014;13(4):377\u0026ndash;85.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHuether A, H\u0026ouml;pfner M, Baradari V, Schuppan D, Scher\u0026uuml;bl H. EGFR blockade by cetuximab alone or as combination therapy for growth control of hepatocellular cancer. Biochem Pharmacol. 2005;70(11):1568\u0026ndash;78.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVera-Yunca D, Parra-Guillen ZP, Girard P, Troc\u0026oacute;niz IF, Terranova N. Relevance of primary lesion location, tumour heterogeneity and genetic mutation demonstrated through tumour growth inhibition and overall survival modelling in metastatic colorectal cancer. Br J Clin Pharmacol. 2022;88(1):166\u0026ndash;77.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTurner RJ, Guy TV, Geraghty NJ, Splitt A, Watson D, Brungs D, Carolan MG, Miller AA, de Leon JF, Aghmesheh M et al. Low Pretreatment CD4(+):CD8(+) T Cell Ratios and CD39(+)CD73(+)CD19(+) B Cell Proportions Are Associated with Improved Relapse-Free Survival in Head and Neck Squamous Cell Carcinoma. Int J Mol Sci 2023, 24(16).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCheng Y, Borcherding N, Ogunsakin A, Lemke-Miltner CD, Gibson-Corley KN, Rajan A, Choi AB, Wongpattaraworakul W, Chan CHF, Salem AK, et al. The anti-tumor effects of cetuximab in combination with VTX-2337 are T cell dependent. Sci Rep. 2021;11(1):1535.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHuang L, Zhang X, Fan J, Liu X, Luo S, Cao D, Liu Y, Xia Z, Zhong H, Chen C, et al. EGFR promotes the apoptosis of CD4(+) T lymphocytes through TBK1/Glut1 induced Warburg effect in sepsis. J Adv Res. 2023;44:39\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOkuyama K, Naruse T, Yanamoto S. Tumor microenvironmental modification by the current target therapy for head and neck squamous cell carcinoma. J experimental Clin cancer research: CR. 2023;42(1):114.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003evon Witzleben A, Fehn A, Grages A, Ezić J, Jeske SS, Puntigam LK, Brunner C, Kraus JM, Kestler HA, Doescher J, et al. Prospective longitudinal study of immune checkpoint molecule (ICM) expression in immune cell subsets during curative conventional therapy of head and neck squamous cell carcinoma (HNSCC). Int J Cancer. 2021;148(8):2023\u0026ndash;35.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYu J, Long B, Li Z, Tian X, Li D, Long J, Wang Y, Chen Y, Zhang F, Liu H, et al. Central memory CD4\u0026thinsp;+\u0026thinsp;T cells play a protective role against immune checkpoint inhibitor-associated myocarditis. Cardiovascular Res. 2024;120(12):1442\u0026ndash;55.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang LN, Chen JY, Liu YX, Zhang Y, Hong LL, Li XX, Liu SH, Chen SQ, Peng L, Huang YT. Identification of lncRNA dual targeting PD-L1 and PD-L2 as a novel prognostic predictor for gastric cancer. Front Oncol. 2024;14:1341056.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRosnev S, Sterner B, Schiele P, Kolling S, Martin M, Fl\u0026ouml;rcken A, Erber B, Wittenbecher F, Kofla G, Kurreck A et al. Reduced monocytic IL10 expression in PD1 inhibitor-treated patients is a harbinger of severe immune-related adverse events. \u003cem\u003eEuropean journal of cancer (Oxford, England\u003c/em\u003e: 1990) 2025, 217:115252.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen D, Li X, Li H, Wang K, Tian X. Identification of Immune-Related Prognostic mRNA and lncRNA in Patients with Hepatocellular Carcinoma. J Oncol. 2022;2022:5313149.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNardone V, Botta C, Caraglia M, Martino EC, Ambrosio MR, Carfagno T, Tini P, Semeraro L, Misso G, Grimaldi A, et al. Tumor infiltrating T lymphocytes expressing FoxP3, CCR7 or PD-1 predict the outcome of prostate cancer patients subjected to salvage radiotherapy after biochemical relapse. Cancer Biol Ther. 2016;17(11):1213\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVahidi Y, Faghih Z, Talei AR, Doroudchi M, Ghaderi A. Memory CD4(+) T cell subsets in tumor draining lymph nodes of breast cancer patients: A focus on T stem cell memory cells. Cell Oncol (Dordrecht Netherlands). 2018;41(1):1\u0026ndash;11.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eInomata M, Matsumoto M, Takata N, Hayashi K, Seto Z, Hirai T, Tokui K, Taka C, Okazawa S, Kambara K, et al. Peripheral CD4 memory T cells predict the efficacy of immune checkpoint inhibitor therapy in patients with non-small cell lung cancer. Sci Rep. 2023;13(1):10807.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMiao YD, Tang XL, Wang JT, Mi DH. Prognostic role of expression of angiogenesis markers in hepatocellular carcinoma: A bioinformatics analysis. World J Gastroenterol. 2022;28(30):4221\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZuo A, Lv J, Jia W, Ba Y, Liu S, Zhang Y, Weng S, Xu H, Liu L, Wang L, et al. High ratio of resident to exhausted CD4\u0026thinsp;+\u0026thinsp;T cells predicts favorable prognosis and potentially better immunotherapeutic efficacy in hepatocellular carcinoma. BMC Cancer. 2024;24(1):1152.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYu L, Shen N, Shi Y, Shi X, Fu X, Li S, Zhu B, Yu W, Zhang Y. Characterization of cancer-related fibroblasts (CAF) in hepatocellular carcinoma and construction of CAF-based risk signature based on single-cell RNA-seq and bulk RNA-seq data. Front Immunol. 2022;13:1009789.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eInnocenti F, Ou FS, Qu X, Zemla TJ, Niedzwiecki D, Tam R, Mahajan S, Goldberg RM, Bertagnolli MM, Blanke CD, et al. Mutational Analysis of Patients With Colorectal Cancer in CALGB/SWOG 80405 Identifies New Roles of Microsatellite Instability and Tumor Mutational Burden for Patient Outcome. J Clin Oncol. 2019;37(14):1217\u0026ndash;27.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhan CS, Chen J, Chen J, Zhang LG, Liu Y, Du HX, Wang H, Zheng MJ, Yu ZQ, Chen XG, et al. CaMK4-dependent phosphorylation of Akt/mTOR underlies Th17 excessive activation in experimental autoimmune prostatitis. FASEB journal: official publication Federation Am Soc Experimental Biology. 2020;34(10):14006\u0026ndash;23.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSun G, Zheng W, Tan P, Zhou J, Tang W, Cao H, Liu L, Shi X, Li Z, Zhang W. Comprehensive Analysis of VCAN Expression Profiles and Prognostic Values in HCC. Front Genet. 2022;13:900306.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWen YC, Lin YW, Ho KH, Yang YC, Lai FR, Chu CY, Chen JQ, Lee WJ, Chien MH. The oncogenic ADAMTS1-VCAN-EGFR cyclic axis drives anoikis resistance and invasion in renal cell carcinoma. Cell Mol Biol Lett. 2024;29(1):126.\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":"bmc-cancer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bcan","sideBox":"Learn more about [BMC Cancer](http://bmccancer.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bcan/default.aspx","title":"BMC Cancer","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Hepatocellular carcinoma, Nimotuzumab, CC-chemokine receptor 7, EGFR-antibody, PDX-model","lastPublishedDoi":"10.21203/rs.3.rs-7189781/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7189781/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e Despite the crucial involvement of the EGFR pathway in hepatocellular carcinoma (HCC), the clinical efficacy of EGFR antibodies in HCC remains uncertain. While existing evidence suggests that immune dysfunction and tumor microenvironment alterations may contribute to treatment resistance, the precise mechanisms underlying this phenomenon in HCC warrant further investigation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e In this study, we employed patient-derived xenograft (PDX) models generated from 14 HCC patients enrolled in the REHOPE301 cohort to evaluate the sensitivity to nimotuzumab, a humanized anti-EGFR monoclonal antibody. Whole-exome sequencing (WES) and single-cell RNA sequencing were performed on tumor tissues and tumor-infiltrating lymphocytes (TILs) to elucidate the association between TIL characteristics and EGFR antibody response. A predictive risk score and nomogram were subsequently developed using LASSO regression analysis. The prognostic performance of this model was evaluated using 2 external datasets (ICGC-JP and GSE141202) through receiver operator characteristic (ROC) curves and calibration curves analyses.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e Nimotuzumab demonstrated a 50% response rate (7/14) in PDX models. Immune profiling revealed distinct TIL patterns between responders and non-responders. Notably, CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+ \u003c/sup\u003eT cells were significantly enriched in resistant tumors (p \u0026lt; 0.001) and negatively correlated with the nimotuzumab response (r = -0.767 p = 0.02). In non-responsive tumors, CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+ \u003c/sup\u003eT cells exhibited interactions with macrophages and CD8\u003csup\u003e+\u003c/sup\u003ePDCD1\u003csup\u003e+ \u003c/sup\u003eT cells. A reduced infiltration of CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+ \u003c/sup\u003eT cells was associated with improved prognosis and enhanced EGFR antibody efficacy across multiple cancer types. Furthermore, a nine-gene signature related to CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+ \u003c/sup\u003eT cells was identified as a strong prognostic factor in HCC (HR = 5.19, 95% CI: 3.18–8.46, P \u0026lt; 0.001), and was used to construct a nomogram. WES confirmed prognostic gene mutations (VCAN, CAMK4, and CD226) potentially influencing nimotuzumab response.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions\u003c/strong\u003e Our findings indicate that increased infiltration of central memory CD4\u003csup\u003e+\u003c/sup\u003eCCR7\u003csup\u003e+ \u003c/sup\u003eT cells in HCC may reflect an immunosuppressive tumor microenvironment, thereby impairing EGFR antibody efficacy and worsening patient prognosis.\u003c/p\u003e","manuscriptTitle":"Multi-omics analysis reveals the role of tumor-infiltrating CD4+CCR7+ T cells in EGFR antibody resistance and prognosis of hepatocellular carcinoma","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-21 10:31:57","doi":"10.21203/rs.3.rs-7189781/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-08-22T18:57:56+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-22T06:28:27+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-20T18:25:24+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-18T22:54:20+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-14T05:12:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"237989824721658411296077568417822142955","date":"2025-08-14T02:35:04+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"114133081108660049407748332623025317248","date":"2025-08-14T02:08:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"219672848358389699923159404342076377162","date":"2025-08-13T20:35:04+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"324743151337516839465693765075661108342","date":"2025-08-13T17:25:53+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-13T16:03:28+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-13T15:54:40+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-08-11T05:40:13+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-10T07:59:30+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Cancer","date":"2025-08-10T07:55:32+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-cancer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bcan","sideBox":"Learn more about [BMC Cancer](http://bmccancer.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bcan/default.aspx","title":"BMC Cancer","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f07c5d93-1833-44b6-8ea9-000324857060","owner":[],"postedDate":"August 21st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-12-01T16:15:28+00:00","versionOfRecord":{"articleIdentity":"rs-7189781","link":"https://doi.org/10.1186/s12885-025-15276-5","journal":{"identity":"bmc-cancer","isVorOnly":false,"title":"BMC Cancer"},"publishedOn":"2025-11-29 15:58:10","publishedOnDateReadable":"November 29th, 2025"},"versionCreatedAt":"2025-08-21 10:31:57","video":"","vorDoi":"10.1186/s12885-025-15276-5","vorDoiUrl":"https://doi.org/10.1186/s12885-025-15276-5","workflowStages":[]},"version":"v1","identity":"rs-7189781","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7189781","identity":"rs-7189781","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}