Introduction:IgA nephropathy (IgAN) is the most common primary glomerulonephritis, and emerging evidence implicates the gut microbiome in its pathogenesis. Additional studies focusing on the molecular mechanisms linking gut microbial signals to intraglomerular changes are warranted. Methods:We performed 16S rRNA-based microbial profiling of fecal samples of 172 IgAN patients, 51 healthy controls, and other glomerular disease controls including 15 diabetic nephropathy, 35 minimal change disease, and 63 membranous nephropathy cases. Serum and fecal acetate levels were measured by liquid chromatography-mass spectrometry. Glomerular spatial transcriptomic profiling was performed with the GeoMx Digital Spatial Profiler. DESeq2 analysis was performed to identify differentially expressed genes, followed by gene ontology annotations. Results:Beta diversity differed significantly between IgAN and healthy controls (p = 0.001). While no single taxon showed consistent differences in abundance, the methanogenesis from acetate pathway was significantly enriched in IgAN, accompanied by an increased proportion of major acetate-producing gut microbial genera. Serum acetate levels were elevated in IgAN (p = 0.03), while fecal acetate levels were comparable to those in healthy controls. In glomerular transcriptomes, functional annotations of 1,227 upregulated and 1,078 downregulated genes in IgAN indicated decreased activities of G protein-coupled receptors, short-chain fatty acid transporters, and beta-1,3-galactosyltransferases. Discussion:IgAN is characterized by gut microbial enrichment in acetate metabolism and increased systemic acetate levels, along with altered intraglomerular expression of metabolic and signaling genes. These findings suggest a gut microbiome-glomerular signaling axis contributing to disease pathogenesis.

Fibroblasts are dynamic structural cells that direct both beneficial tissue repair and pathological organ fibrosis through interactions with tissue-resident type 2 lymphocytes (T2Ls) and type 3/17 lymphocytes (T3Ls). The cytokines interleukin-13 (IL-13) and IL-17A, produced by T2Ls and T3Ls, respectively, are linked to both tissue inflammation and fibrosis, but how their spatial positioning influences beneficial or pathological organ remodeling remains unclear. Using mouse models of liver injury and fibrosis, three-dimensional microscopy, and spatial transcriptomics, we found an accumulation of periportal and fibrotic tract T2Ls, predominantly group 2 innate lymphoid cells (ILC2s), positioned near T3Ls and niche adventitial fibroblasts and adjacent to discrete profibrotic myofibroblasts. Unexpectedly, T2L ablation worsened both carbon tetrachloride- and bile duct ligation-induced liver fibrosis, accompanied by increased IL-17A+ T3Ls, predominantly γδ T cells. In contrast, concurrent T2L and T3L ablation reduced liver fibrosis. Our work suggests a spatially associated cross-talk between liver lymphocytes and fibroblast niches that tunes liver repair but can go awry in pathological liver fibrosis.

Cell type annotation is essential for gaining biological insight from single-cell RNA sequencing data, yet manual labeling remains time-consuming and difficult to reproduce. Various computational approaches have been developed to automate this process, and recent studies suggest that large language models can infer cell types with promising accuracy in single-cell analysis. However, most workflows still rely on cluster-specific markers derived from gene expression alone or manual curation. As a result, marker selection can be sensitive to statistical criteria and dataset-dependent bias, which may lead to the selection of less informative genes or missing important markers, while providing limited biological context. To address this limitation, we introduce CELLIA, an LLM-based workflow for automated and robust cell type annotation. CELLIA employs an integrative evidence-knowledge marker selection strategy that combines statistical differential expression criteria with curated tissue-specific marker resources to identify informative marker genes. In benchmarking analyses of 102 cell types, this approach improved agreement with manual annotations. In addition, CELLIA achieved higher agreement in subtype-level analyses of closely related immune populations and was further evaluated in a non-immune stromal subtype setting, covering 25 cell types in total. By integrating evidence-knowledge from gene expression with curated biological prior knowledge, CELLIA provides a more stable marker selection and improves the reliability of LLM-cell type annotation.

Abstract Introduction Mycosis fungoides (MF), the prevalent subtype of cutaneous T-cell lymphoma (CTCL), primarily affects the skin and progresses systemically over time. Early-stage MF diagnosis relies on immunohistochemistry, T cell receptor (TCR) gene rearrangement testing, and clinical evaluation. However, in the clinic settings, we observe eczematous MF (eMF) patients who meet early-stage MF diagnostic criteria but exhibit phenotypes resembling atopic dermatitis (AD), and refractory to dupilumab. This study investigates the molecular characteristics of eMF in comparison to AD and classic MF, providing insights into its distinct pathophysiology and potential therapeutic implications. Methods Skin biopsies were collected from seven AD patients and eleven eMF patients. Paired single-cell RNA sequencing (scRNA-seq) and single-cell TCR sequencing (scTCR-seq) were conducted to evaluate transcriptional profiles and T-cell clonality. Publicly available MF and AD scRNA-seq data were used for comparative analysis. In addition, spatial transcriptomic profiling was performed to validate. Results Comparative transcriptomic analysis showed that the molecular profiles of T cells in eMF resembled those in AD rather than in conventional MF. Furthermore, in eMF, we observed oligoclonal T-cell expansions, which were not found in conventional MF both in scRNA-seq and spatial transcriptomic data. These expanded cells were predominantly Th22 cells, identified as the primary producers of IL-13, a key cytokine in AD pathophysiology. Conclusion eMF diagnostic phenotype resembles conventional MF, and AD primarily due to the hyperproliferation of Th22 cells and their expression of IL-13. In this context, this mechanism underscores the superior efficacy of Upadacitinib, which could target Th22 differentiation, over Dupilumab. As a result, our findings indicate that eMF is not cancer, but a Th22-mediated AD endotype. Funding Source n/a Topic Categories Immune Mechanisms of Human Disease (HUM)

Bronchiectasis describes chronic airway inflammation involving various immune cells; however, little information is available regarding cell-type-specific pathogenic changes that influence disease development of bronchiectasis. We aimed to investigate immune dysregulation in bronchiectasis through single-cell RNA sequencing (scRNA-seq) of peripheral blood mononuclear cells (PBMCs). PBMCs from eight bronchiectasis patients and eight healthy controls were isolated and subjected to scRNA-seq using the 10X Genomics platform. Frequencies of immune cell subsets were compared between groups, and functional implications were inferred based on transcriptional signatures. The overall innate immune cell composition was similar between bronchiectasis patients and healthy controls, but significant subset-level alterations were observed. Bronchiectasis patients exhibited increased CD4 + and CD8 + effector memory T cells, suggesting chronic inflammatory activated status of T cells. Notably, FCER1G + NK cells were significantly reduced in bronchiectasis patients, accompanied by decreased expression of chemokines such as CCL3, CCL4, XCL1, and XCL2. In bronchiectasis patients, pro-inflammatory CD14 + monocytes tended to be decreased, showing reduced CXCR4 expression. Our findings reveal distinct immune alterations in bronchiectasis, especially involving NK cells and monocytes. The depletion of FCER1G + NK cells and downregulation of CXCR4 in monocytes suggest a disrupted innate immune cascade that may contribute to disease progression in bronchiectasis.

Abstract Introduction Inflammatory skin diseases are a heterogeneous group of chronic disorders characterized by persistent or recurrent inflammation with complex and multifactorial immunopathomechanisms. These diseases share overlapping clinical features and some overlapping molecular mechanisms, despite heterogeneity and complex pathophysiology, although their pathophysiology remains only partially understood. To guide treatment strategies and predict therapeutic response, characterization of shared or disease-specific pathomechanism is needed. In addition, how the clinical and immunological phenotypes of these diseases differ by ancestry remains underexplored. Methods Here, we newly generated a single-cell atlas of inflammatory skin diseases, atopic dermatitis (AD), generalized pustular psoriasis (GPP), plaque psoriasis (PSO), hidradenitis suppurativa (HS) and vitiligo (VIT). Our dataset includes single-cell RNA sequencing of 77 tissue biopsies and 85 peripheral blood mononuclear cell (PBMC) samples from Korean patients, and single-cell resolution spatial transcriptomics on 35 tissue biopsies, with a custom add-on panel to refine immune signatures. Results The tissue dataset enabled cross-disease stratification at the sample level, where heterogeneity in keratinocytes shaped by disease-specific cytokine milieus is a major determining factor in disease stratification. In addition, we identified some disease-specific features, especially in immune cell subsets including resident memory T cells and macrophage subsets, which we show can inform potential druggable targets. In parallel, we have profiled PBMCs using mass cytometry for protein level validation of immune phenotypes. Conclusion These integrated approaches will enable discovery of unique molecular signatures of inflammatory skin diseases, offering insights for the development of treatment strategies and predictive biomarkers for inflammatory skin diseases with overlapping phenotypes. Funding Source Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number : RS-2024-00403375) Topic Categories Computational and Systems Immunology (COMP)

[This corrects the article DOI: 10.3389/fimmu.2025.1680437.].

BACKGROUND:Early-stage mycosis fungoides (MF) often presents diagnostic challenges because of its clinical overlap with atopic dermatitis (AD). In clinical practice, we encountered a subset of patients with severe AD who fulfilled the MF diagnostic criteria yet remained clinically indistinguishable from AD and presented refractoriness to advanced therapies. We termed this ambiguous entity "mycosis fungoides-like AD" (mfAD) and sought to determine whether it represents malignant transformation or a distinct inflammatory endotype of AD. METHODS:Skin biopsies were obtained from 7 patients with AD and 11 patients with mfAD. We performed paired single-cell RNA sequencing and single-cell T-cell receptor sequencing analyses. Publicly available MF and AD datasets were integrated for comparative analysis. Spatial transcriptomic profiling was used to contextualize single-cell findings within the tissue architecture. RESULTS:Comparative transcriptomic analysis revealed that T cells in mfAD were aligned with those in AD and lacked genomic instability. High-resolution profiling showed that mfAD was characterized by oligoclonal Th22 expansion rather than a single dominant malignant clone. Notably, all patients with mfAD achieved rapid clinical remission with selective JAK1 inhibition, indicating the therapeutic response characteristics of inflammatory dermatoses. CONCLUSION:Our findings demonstrate that mfAD is not a true malignancy, but rather a Th22-driven inflammatory endotype of AD. These results redefine mfAD as an inflammatory subtype within the AD spectrum, providing a mechanistic explanation for both the "pseudo-monoclonality" that leads to MF misdiagnosis and the failure of dupilumab. This study establishes a rationale for the use of JAK inhibitors in precision medicine for this patient population.

Key Points Plasma proteome profiling identified distinct signatures across biopsy-proven primary glomerular disease subtypes. An elastic net model using 93 proteins classified primary glomerular disease subtypes and controls, with external validation. Integrating proteomics with machine learning yields biologically interpretable insights in primary glomerular diseases. Background Primary GN is a heterogeneous group of kidney disorders where understanding of their pathophysiology remains incomplete. Despite the diagnostic potential of high-throughput proteomics, constrained proteomic depth and a reliance on binary comparisons have left the feasibility of using systemic signatures to differentiate multiple GN subtypes largely unexplored. Methods To identify protein signatures that noninvasively differentiate major primary glomerular disease subtypes and provide mechanistic insights, we performed large-scale systemic proteome profiling of 5416 plasma proteins via Olink Explore HT in a discovery cohort ( n =147) and an external validation cohort ( n =85) of Korean participants (mean age, 41±13 years; 46% female). The study population included patients with four GN subtypes—focal segmental glomerulosclerosis, IgA nephropathy, minimal change disease, and membranous nephropathy—alongside healthy controls. We developed a machine learning (ML) model using logistic regression with elastic net regularization to classify disease groups based on proteomic profiles and evaluated its performance in the independent validation cohort. Results Plasma proteome profiles were distinct among disease subtypes, emerging as a significant source of data variation independent of conventional markers such as eGFR or proteinuria levels. The ML model performed robustly in both the discovery and validation cohorts, achieving an area under the receiver operating characteristic curve >0.8 for differentiating minimal change disease, membranous nephropathy, and IgA nephropathy. The model, even without clinical information, correctly identified 93% of minimal change disease cases (14 of 15) and 63% of IgA nephropathy cases (20 of 32), but its performance was limited for focal segmental glomerulosclerosis, with only 21% of cases (three of 14) correctly classified. Functional analysis of key proteins highlighted distinct biologic pathways, such as hemostasis in minimal change disease. Conclusions We identified distinct systemic proteome signatures for primary glomerular diseases, where disease subtype served as a major determinant of proteomic variance alongside conventional clinical markers. ML models demonstrated robust discriminatory performance for minimal change disease, membranous nephropathy, and IgA nephropathy, underscoring the potential for proteome-based classification.

Abstract Elucidating the roles of intra-tumoral macrophage subtypes that promote a high-TIL microenvironment in triple-negative breast cancer. High levels of tumor-infiltrating lymphocytes (TILs) within the stromal compartment are associated with favorable prognosis and improved response to immune checkpoint blockade in triple-negative breast cancer (TNBC). To elucidate the mechanisms underlying the divergence between TIL-high and TIL-low tumor microenvironments (TMEs), we analyzed treatment-naïve, early-stage TNBC samples. We profiled the TME of 21 patients using single-cell-resolution spatial transcriptomics on FFPE tissue microarray blocks. Patients were classified into high TIL group with over 60% ratio of TIL in stroma. sTILs (0-100%) were evaluated on H&E-stained slides, and tumors with sTIL ≥ 60% and ≤ 10% were classified as high and low TIL, respectively. These thresholds are stricter than the commonly used sTIL ≥ 50% and < 30% cutoffs associated with distant RFS. Using this classification, 11 patients were assigned to the TIL-high group and 10 to the TIL-low group. We clustered 40 cell types for TNBC samples using Xenium 5K platform and identified 12 spatial domains by uncovering recurrent cellular neighborhoods (RCNs) that share similar spatial patterns using SCIMAP(v2.2.11). Among those 12 spatial domains, domain 9 and domain 0 represented tumor center and tumor edge, respectively. We identified 10 myeloid cell subtypes and discovered macrophage subtype MC2 expressing LDHA and PLAUR are enriched specifically in tumor domains consisting of domain 9 and 0. Interestingly, the MC2(LDHA_PLAUR) subtype present low PLAUR expression in TIL-high group compared to TIL-low group in tumor domains. Although the proportion of infiltrating MC2 (LDHA_PLAUR) macrophages within tumor domains was higher in the TIL-high group compared with the TIL-low group, the expression level of PLAUR within MC2 cells was paradoxically lower in the TIL-high group. Additionally, the MC2(LDHA_PLAUR) expressed more IFN-stimulated genes such as CXCL9, GBP5, and IRF1 in TIL-high group compared to TIL-low group. PLAUR has been known for inducing HIF1A highly in tumor microenvironment. As HIF1A hinders IFNG-signaling important to establish TIL-high environments, high expression of PLAUR promotes inhibition of IFNG-signaling in macrophages. So, low expression of PLAUR for MC2 in TIL-high group helps normal IFNG-signaling pathways in macrophages. Taken together, these findings indicate MC2 macrophages located in both the cancer center and edge regions showed reduced PLAUR expression, rendering them less susceptible to HIF1A-driven suppression and allowing sustained IFNG-signaling in high-TIL tumors. This suggests that these MC2 cells help establish an intra-tumoral environment conducive to TIL infiltration by continuously supplying CXCL9, CXCL10, and CXCL11. Citation Format: Sowon Choi, Hee Jin Lee, Byung-Kwan Jeong, Hyun Je Kim. Identification of tumor microenvironment for high infiltration of lymphocytes in triple negative breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 4986.

Abstract Introduction Endometriosis affects approximately 10% of women of reproductive age and remains a major clinical challenge due to its association with infertility and reduced quality of life. The disease is difficult both to diagnose and to treat, as current approaches–including laparoscopic removal and hormonal therapy–are time-consuming and often fail to achieve a complete cure. Previous studies have demonstrated impaired NK cell function in patients with endometriosis, while murine experiments have suggested a potential role for NK cells in disease regulation; however, the mechanisms underlying this dysfunction remain unclear. This study aims to investigate whether and how aberrant NK cell migration and differentiation contribute to immune dysregulation in endometriosis. Methods A murine endometriosis model induced by intraperitoneal injection of dissociated uterine horn tissue and spatial transcriptomic profiling of patient-derived lesions were utilized to characterize NK cell subsets and their transcriptional signatures within endometriotic tissues in mice and humans. Results Antibody-mediated NK cell depletion using an NK1.1 monoclonal antibody resulted in significantly larger lesions, indicating that NK cells contribute to lesion control. Moreover, lesion weight was negatively correlated with the abundance of tissue-resident NK cells. Using a congenic mouse model, we confirmed that CD49a+ tissue-resident NK cells were not derived from donor uteri, but originated from the host, suggesting that NK cell migration and differentiation are essential for their function in regulating lesion growth. Conclusion Ongoing studies are expanding on these preliminary findings to elucidate how NK cell migration and differentiation shape the immune landscape of endometriosis. A deeper understanding of these mechanisms may provide a foundation for developing novel diagnostic and therapeutic strategies targeting NK cell—mediated immune regulation. Funding Source This research was supported by a grant of the Boston-Korea Innovative Research Project through the Korea Health Industry Development Institute(KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant no. : RS-2024-00403047) Topic Categories Immune Mechanisms of Human Disease (HUM)

Abstract Background Engineered synthetic RNAs enable cellular control by sensing and responding to intracellular biomolecules. Recently developed sense-edit-switch RNAs (sesRNAs) based on Adenosine Deaminase Acting on RNA (ADAR)—which edits a stop codon to switch on custom payload translation in the presence of a target RNA—are consistently functional across different species. The ability of sesRNAs to couple bespoke payload translation to the presence of cell type-specific transcripts will usher in an era of precise cell-targeted biotechnological interventions. Results To expedite the generation of sesRNAs, we develop ADAR-Sense—a universal web tool for automated sensor design based on user-defined sensor length, sensor-target RNA mismatch number, mismatch proximity to the ADAR-editable stop codon, and targeted custom element inclusion for improved ADAR recruitment and subsequent payload induction. Conclusions Compared to current tools, the simplicity and flexibility of ADAR-Sense will streamline the design and screening of sesRNAs in new cell types and conditions, supporting the swift adoption of this sensing platform in both basic and translational research.

6043 Background: Nivolumab has demonstrated meaningful survival benefit in patients with refractory R/M HNSCC. Nevertheless, reliable predictive biomarkers remain scarce—particularly those capable of identifying long-term survivors—underscoring the need for translational studies to uncover immune correlates of durable response. In this prospective phase II study (NCT04603248), we sought to define dynamic circulating immune cell–based biomarkers predicting durable clinical outcomes with nivolumab in patients with R/M HNSCC. Methods: Patients with R/M HNSCC who had prior failure of or intolerance to platinum-based chemotherapy were treated with nivolumab (3 mg/kg) intravenously every 2 weeks until disease progression or unacceptable toxicity occurred. Clinical outcomes were correlated with single-cell transcriptomic profiles of circulating immune cells at baseline (cycle 1 day 1) and on-treatment (cycle 2 day 1). To validate the transcriptomic findings at the protein level, mass cytometry by time-of-flight (CyTOF) analysis was additionally performed. Results: A total of 48 patients were enrolled. The objective response rate was 22.9%, and the disease control rate was 62.5%. The median progression-free survival (PFS) and overall survival (OS) were 4.4 and 13.3 months, respectively. Single-cell transcriptomic analysis revealed a significant expansion of circulating NKG7⁺ cytotoxic CD4⁺ T cells in long-term responders (PFS > 48 months; n=6) compared with early progressors (PFS < 2 months; n=6) at cycle 2 day 1. T cell receptor analysis further demonstrated that nivolumab induced marked clonal expansion of these NKG7⁺ cytotoxic CD4⁺ T cells, particularly in long-term responders. Their sustained presence was confirmed in blood samples collected one year after treatment initiation in long-term responders. CyTOF analysis (n=37) revealed that expansion of NKG7⁺ cytotoxic CD4⁺ T cells at cycle 2 day 1 was significantly associated with both PFS and OS, supporting their potential role as predictive biomarkers of response to PD-1 blockade. Conclusions: Expansion and clonal amplification of circulating NKG7⁺ cytotoxic CD4⁺ T cells represent a key immune correlate of favorable outcomes with nivolumab in refractory R/M HNSCC. These findings highlight their potential as predictive biomarkers of durable response to PD-1 blockade in R/M HNSCC and implicate this immune subset as a promising target for future immunotherapeutic strategies. Clinical trial information: NCT04603248 .

Abstract Introduction CD4+ regulatory T cells (Tregs) are essential for maintaining immune homeostasis and preventing autoimmunity. Tregs primarily develop in the thymus, but can also arise from naïve CD4+ T cells in the periphery (pTregs) or be generated in vitro (iTregs). However, a major limitation of Treg-based therapies is the instability and plasticity of pTregs and iTregs. In contrast, tTregs have been shown to exhibit stable suppressive capacity, largely due to thymic-derived signals that epigenetically reinforce the Treg program. Elucidating the mechanisms governing Treg differentiation, stability, and function is therefore critical for improving Treg-based therapies. Methods We generated single-cell multiome data from human fetal and pediatric thymuses. We developed various analytical frameworks for unravelling gene regulatory networks (GRNs) involved in Treg lineage commitment. We identified candidate transcription factors (TFs) involved in thymic Treg differentiation and validated these TFs using a CRISPR-Cas9 KO system in primary human thymocytes. Results GRN analysis revealed key driver TFs, such as FOXP3, REL and IKZF2 within CD4+ Tregs. Comparison of GRNs between mature CD4+ Tregs and conventional CD4+ T cells further reveals TFs related to TCR signaling and other novel TFs. Finally, candidate TFs including IRF4, REL, FOXO1, BATF and others were validated utilizing a CRISPR-Cas9 KO system. Conclusion We generated a single cell multiome atlas of fetal and pediatric thymuses and unravel GRNs involved in Treg lineage-specific differentiation. We develop novel analytical frameworks to identify lineage-specific driver TFs in Tregs and validated these by KO of primary human thymocytes. This framework provides an important mapping of GRNs involved in thymic T cell differentiation, particularly focusing on Tregs and will serve as an important basis for understanding Treg biology and improving Treg-based therapies. Funding Source Creative-Pioneering Researchers Program (800-20230490) Seoul National University Topic Categories Hematopoiesis and Immune System Development (HEM)

BackgroundNon-small cell lung cancer (NSCLC) patients with oncogenic driver mutations such as EGFR, ALK or ROS1 (mutant-type [MT]) exhibit poor responses to PD-1/PD-L1 immune checkpoint inhibitors (ICIs) compared to wild-type (WT) patients. The mechanisms underlying this limited response to ICIs in MT patients remain unclear. This study aimed to identify key immune biomarkers and elucidate immune cell dynamics in peripheral blood contributing to ICI resistance in MT-NSCLC.MethodsA total of 262 NSCLC patients who received PD-1/PD-L1 inhibitor monotherapy between February 2018 and July 2024 were included. Of these, 43 patients were assigned to the discovery cohort, where immune profiling was performed on peripheral blood mononuclear cells using Cytometry by Time-of-Flight (CyTOF) at baseline and cycle 2, day 1 (C2D1). The findings were validated in an independent cohort (n=57) using flow cytometry.ResultsBaseline CXCR3+ CD127+ effector CD8+ T cells were significantly elevated in WT compared to MT patients (P = 0.0120) and were a robust predictor of favorable response (AUC = 0.745). Patients with CXCR3+ CD127+ CD8+ T cell frequencies above 3.54% exhibited superior progression-free survival (PFS, P = 0.0005) and overall survival (OS, P = 0.0012). In contrast, MT patients demonstrated a distinct reduction in CD27+ PD-1⁻ effector memory CD4+ T cells during treatment, correlating with poor outcomes (AUC = 0.716). This reduction was associated with diminished conventional dendritic cell (cDC) abundance, suggesting impaired T-cell differentiation and function in MT patients. These findings were consistently validated using flow cytometry in an independent cohort.ConclusionDistinct immune cell profiles highlight elevated baseline CXCR3+ CD127+ effector CD8+ T cells predicting favorable outcomes and impaired cDC-CD4+ T cell dynamics as critical contributors to anti-PD-1/PD-L1 resistance in MT-NSCLC.

Vessels encapsulating tumor clusters (VETC) is a distinct angiogenic pattern in hepatocellular carcinoma (HCC). While VETC-positive HCC is associated with poor prognosis, it demonstrates improved responses to vascular-targeted therapies. However, the molecular signatures that define VETC-positive HCC remain elusive. This study employed spatial transcriptomics and single-cell RNA sequencing to characterize the molecular landscape of VETC-positive HCC and identify candidate biomarkers. Using GeoMx Digital Spatial Profiling, we analyzed tumor and adjacent liver tissues from 24 HCC patients, focusing on VETC-positive, VETC-negative, or non-neoplastic liver regions. Differential gene expression and pathway enrichment analyses were performed. Candidate genes were further evaluated using immunohistochemistry, The Cancer Genome Atlas (TCGA) dataset, and single-cell RNA sequencing. Spatial transcriptomic analysis identified 39 genes upregulated in VETC-positive regions, with enrichment of angiogenesis, WNT/β-catenin, Hedgehog, and p53 signaling, epithelial–mesenchymal transition, and fatty acid metabolism pathways, and suppression of immune-related pathways. Among these, LAMTOR2, DPP4, ZNHIT1, and MLXIPL were consistently upregulated in VETC-positive regions compared to both VETC-negative and normal liver regions. LAMTOR2 demonstrated tumor-specific localization in TCGA dataset. Immunohistochemistry confirmed that peripheral accentuation pattern of LAMTOR2 expression was restricted to tumor cells and strongly correlated with RNA expression and VETC positivity. Single-cell RNA sequencing further revealed LAMTOR2 upregulation in malignant hepatocytes, particularly within VETC-high subpopulations enriched for lipid degradation, fatty acid oxidation, and detoxification pathways. VETC-positive HCC exhibits a distinct transcriptomic and metabolic profile. LAMTOR2, which is consistently upregulated in VETC-positive HCCs, may contribute to angiogenic and metabolic reprogramming, offering potential implications for therapeutic stratification in HCC.