Journal for Immunotherapy of Cancer最新文献

Background: While adipose tissue constitutes a substantial proportion of breast composition, the functional characteristics and pathological relevance of the adipocyte microenvironment in breast carcinogenesis remain undercharacterized. This study employs single-nucleus RNA sequencing (snRNA-seq) to establish a comprehensive cellular atlas of adipocyte heterogeneity across molecular subtypes of breast cancer, aiming to elucidate subtype-specific adipocyte contributions to tumor microenvironment modulation.

Methods: snRNA-seq was performed on breast adipose tissues isolated from individuals without cancer and treatment-naïve breast cancer. Various adipocyte and pre-adipocyte subclusters were identified. Comparative analyses of cellular distribution and transcriptional profiles were performed across disease states and molecular subtypes. Pseudotime, cell communication, and immunofluorescence analyses were further implemented to investigate cellular dynamics and microenvironment interactions.

Results: snRNA-seq data of 86,529 nuclei were obtained. Three adipocyte and seven pre-adipocyte subclusters were identified, of which Adi_LDLR, Pre_Adi_LDLR, and Pre_Adi_LGR4_TGFBR1 were restricted to cancer-associated adipose (CAAs). Adi_LDLR and Pre_Adi_LDLR were enriched in estrogen receptor (ER)-positive CAAs and related to cell senescence and immunosuppression. Pre_Adi_LGR4_TGFBR1 was predominantly present in triple-negative breast cancer, functionally pro-proliferative, immunosuppressive, and lacked normal adipose function. The immunofluorescence intensity of LDLR (p=0.031) and TGFBR1 (p=0.038) was positively associated with disease recurrence, suggesting the formation of immunosuppressive niches by these cancer-specific adipose subsets in both subtypes. Cell communication analyses revealed a specific (pre-) adipocyte-macrophage interaction via ligand-receptor pairs involved in stromal remodeling and tumor migration for ER-positive tumors, whereas tumor proliferation and metastasis for triple-negative ones likely contribute to tumor progression.

Conclusions: This study delineated a distinct adipocyte landscape in breast cancer and subtype-specific immunosuppressive niches fostered by CAAs and (pre-) adipocyte-macrophage interactions. These findings provide novel therapeutic targets for microenvironment-directed interventions in breast oncology.

Sarcomas are rare malignancies of mesenchymal origin, characterized by significant biological and clinical heterogeneity. Many subtypes demonstrate limited sensitivity to standard systemic treatments, including immune checkpoint inhibitors. Cell therapy has emerged as a promising strategy, with the potential of durable clinical responses seen with genetically-engineered T-cell receptor T-cell therapies (TCR-T) such as those targeting the cancer-testis antigen MAGE-A4 in synovial sarcoma, leading to the US Food and Drug Administration approval of afamitresgene autoleucel in 2024. This constituted only the second approval of a cell therapy in a solid tumor following lifileucel in melanoma and demonstrated the potential of cell therapies in sarcomas. This review provides the current landscape and growing potential of cell therapies in sarcomas, including TCR-T, chimeric antigen receptor-T cells, tumor-infiltrating lymphocytes, natural killer (NK) cells, and mesenchymal stromal cells. However, the broader application of these therapies is hindered by the lack of targetable sarcoma-restricted immunogenic epitopes, spatiotemporal intratumoral heterogeneity, and a profoundly immunosuppressive tumor microenvironment that impedes effector-cell trafficking, expansion and persistence. While cell therapies hold promise for integration into precision medicine approaches for sarcomas, their successful implementation will require careful evaluation of clinical feasibility, logistical considerations and cost-effectiveness to optimize patient outcomes.

Purpose: Chemoradiotherapy (CRT) combined with anti-PD-1 for locally advanced esophageal squamous cell carcinoma (ESCC) has shown promising efficacy but lack the predictive biomarkers to identify patients who could benefit from this therapy. The predictive value of serum cytokines in ESCC patients remains unclear. We aimed to identify cytokine-based biomarkers for treatment response and survival in this setting.

Experimental design: Exploratory analyses were conducted on 81 ESCC patients from two phase II trials treated with CRT plus toripalimab, with validation in an independent prospective cohort (n=61). Nineteen serum cytokines were assessed at baseline, during, and post-CRT plus anti-PD-1 antibody. A cytokine-based risk score model (CYTOscore) was constructed. Multi-omics profiling including RNA-seq, WES, and spatial transcriptomics were performed to explore potential differences in tumor microenvironments.

Results: Cox analyses identified Interleukin-8 (IL-8), C-C motif chemokine ligand 3 (CCL3), and C-C motif chemokine ligand 4 (CCL4) as potential biomarkers and were used to constructed the CYTOscore. Patients stratified by baseline CYTOscore showed significantly longer OS (HR, 0.31; 95%CI, 0.16-0.62; p= 0.00045) and PFS (HR, 0.33; 95%CI, 0.17-0.62; p= 0.00036) in the low-risk group, which also had higher complete response (CR) rates (66% vs 35%, p=0.014). These finding were next validated in the external cohort, with the low-risk group demonstrating higher CR rates (66% vs 27%, p=0.039) and longer OS (HR 0.30, 95% CI 0.09-0.99, p=0.045). A nomogram incorporating baseline CYTOscore and clinical characteristics showed promising predictive accuracy in 1-, 2-, and 3-year OS (AUC=0.77, 0.78, and 0.76). Multi-omics analysis revealed enriched interferon-γ/α signaling in B cells within low-risk patients.

Conclusions: The CYTOscore based on IL-8, CCL3, and CCL4 effectively predicts treatment response and survival in ESCC patients receiving CRT plus anti-PD-1 antibody.

Background: Although immune checkpoint inhibitors (ICIs) have significantly improved outcomes for patients with certain cancers, their efficacy is largely confined to "hot" tumors characterized by robust infiltration of tumor-specific CD8⁺ T cells. Conversely, tumors expressing B7-H4 often exhibit an immunologically "cold" tumor microenvironment with poor T cell infiltration, contributing to primary resistance to programmed cell death protein 1 (PD-1) blockade.

Methods: We evaluated the association between B7-H4 expression and clinical outcomes in ICI-treated patients using public immunotherapy datasets. The role of B7-H4 in mediating resistance to PD-1 therapy was examined in mouse tumor models. A fully human anti-B7-H4 monoclonal antibody (clone A8) was generated via phage display screening from a non-immunized human single-chain variable fragment library. In vitro assays assessed antibody-induced tumor cell death and immune activation, while in vivo efficacy was tested in MC38-mH4 and SKOV3-hH4 tumor models, as well as human colorectal cancer organoids. Statistical analyses included Student's t-test, one-way analysis of variance, and Kaplan-Meier survival analysis, with p<0.05 considered significant.

Results: High B7-H4 expression was associated with inferior prognosis in patients receiving ICI therapy. In MC38-mH4 tumors, B7-H4 expression conferred resistance to anti-PD-1 treatment. We identified A8, a novel antibody targeting the IgV-like domain of B7-H4, with cross-reactivity to both human and mouse B7-H4. A8-hIgG1 and its Fab fragment induced dynamin-dependent endocytosis of B7-H4, resulting in lysosomal accumulation, altered lysosomal membrane permeabilization and intracellular acidification, ultimately triggering ferroptosis, a form of immunogenic cell death. A8 binding was enhanced under acidic conditions (pH 5.5), promoting lysosome-dependent degradation of B7-H4. A8-induced ferroptosis enhanced dendritic cell maturation, macrophage phagocytosis, and T cell activation. In vivo, A8 promoted CD8⁺ T cell and HER2 chimeric antigen receptor-T cell infiltration, inhibited tumor growth, and synergized with PD-1 blockade to overcome primary resistance in multiple preclinical models. This immunogenic and lysosome-dependent cell death mechanism was unique to A8 among the anti-B7-H4 antibodies tested.

Conclusions: Our study identifies a novel mechanism by which a fully human anti-B7-H4 antibody induces lysosome-dependent immunogenic tumor cell death. These findings support the therapeutic potential of A8 as a single agent or in combination with PD-1 blockade to overcome immune resistance in B7-H4-expressing "cold" tumors.

Background: Granzyme K (GZMK) is a serine protease known for its perforin-dependent cytotoxicity. However, the non-cytotoxic role of GZMK in lung adenocarcinoma (LUAD) remains largely elusive.

Methods: Multiomics datasets were integrated to investigate the clinical relevance of GZMK and its association with programmed death-ligand 1 (PD-L1) in LUAD. Recombinant human GZMK (rhGzmK) was applied in tumor-CD8⁺ T cell co-culture systems, with its effects on PD-L1 expression and CD8⁺ T-cell function evaluated via flow cytometry. Key signaling proteins were analyzed by Western blotting. To evaluate the therapeutic potential of GZMK inhibition, a selective GZMK inhibitor was combined with anti-programmed cell death protein 1 (anti-PD-1) therapy in both C57BL/6 and human peripheral blood mononuclear cells (huPBMC)-reconstituted NVSG humanized mouse models. Finally, multiplex immunofluorescence analysis was conducted on paired pretreatment and post-treatment specimens from a clinical cohort of patients with LUAD receiving immunotherapy to assess the spatial dynamics of GZMK expression in response to treatment.

Results: GZMK upregulated PD-L1 expression on tumor cells and enhanced PD-L1/PD-1 binding. Furthermore, GZMK promoted CD8⁺ T-cell dysfunction through the induction of apoptosis, the promotion of CD8⁺ T-cell exhaustion and suppression of proliferation. Mechanistically, cleavage of F2R-like trypsin receptor 1 (F2RL1) by GZMK activated the AKT Serine/Threonine Kinase (AKT) /glycogen synthase kinase-3β/β-catenin and Janus kinase 2/signal transducer and activator of transcription 1 (JAK2/STAT1) pathways, triggering nuclear accumulation of β-catenin and phosphorylated STAT1, which ultimately drove PD-L1 transcription. Additionally, F2RL1 signaling upregulated COPS8, stabilizing PD-L1 through inhibition of its ubiquitin-mediated degradation. In vivo, pharmacological inhibition of GZMK synergized with anti-PD-1 therapy to suppress tumor growth and enhance CD8⁺ T-cell infiltration and function. Clinically, high baseline GZMK expression correlated with an improved response to immunotherapy, and anti-PD-1 treatment modulated the spatial distribution of GZMK within the tumor microenvironment.

Conclusion: In the absence of perforin, GZMK acquires an immunosuppressive function through F2RL1 activation on tumor cells, which in turn promotes the formation of an immune-suppressive niche. Accordingly, combined targeting of the GZMK/F2RL1 axis and the PD-1/PD-L1 pathway represents a promising synergistic strategy to overcome immune evasion in LUAD.

Background: Engineering chimeric antigen receptor (CAR) T cells with logic-gated synthetic Notch (synNotch) receptor circuits can enhance specificity and mitigate on-target/off-tumor toxicity. However, the conventional synNotch system uses two lentiviral vectors encoding the synNotch receptor and inducible CAR, requiring dual transduction and cell sorting, which limits clinical translation. Integrating the synNotch-CAR circuit into a single lentiviral vector could overcome this limitation, yet manufacturing CAR T cells with large transgenes remains challenging, as increasing transgene size drastically reduces lentiviral titers and T cell transduction efficiency. Current production workflows compensate for low transduction efficiency by sorting transduced cells, further impeding clinical translation. Consequently, these constraints have limited the broader development of synNotch-CAR T cell therapies.

Methods: We engineered a single-vector synNotch (svsNotch) system that integrates all components of the conventional dual-vector circuit into one lentiviral vector to facilitate clinical translation. To overcome the low lentiviral titers and T cell transduction efficiency caused by the large svsNotch transgene, we established an optimized CAR T cell production workflow for effector T cells with large lentiviral transgenes.

Results: Our optimized workflow increased T cell transduction rates by up to 14.8-fold and enabled the production of effector T cells with lentiviral transgenes exceeding the effective packaging capacity limit of 9.2 kb. As a proof of concept, we engineered human epidermal growth factor receptor 2 (HER2)-mesothelin (MSLN) svsNotch (9.2 kb), in which a synNotch receptor targeting HER2 regulates the expression of a second-generation 4-1BBζ CAR against MSLN to enable selective targeting of double-positive HER2⁺MSLN⁺ ovarian tumors. In vitro, HER2-MSLN svsNotch T cells demonstrated superior specificity to conventional dual-vector synNotch-CAR T cells, with selective cytotoxicity against HER2⁺MSLN⁺ but not HER2^koMSLN⁺ tumor cells. To enable in vivo monitoring, we engineered HER2-MSLN-click beetle green (CBG) svsNotch (10.1 kb) incorporating CBG luciferase. In mouse models using constitutive CAR T cells as controls, HER2-MSLN-CBG svsNotch T cells exhibited minimal cytotoxicity in the absence of HER2 and superior efficacy against HER2^lowMSLN^high and HER2^highMSLN^high tumors.

Conclusion: These data establish a framework for engineering logic-gated single-vector immunotherapies and provide an optimized workflow for generating CAR T cells with transgenes that exceed current size limitations.

Adoptive cell therapy (ACT) has demonstrated curative potential in select cancers, but its translation to solid tumors such as ovarian cancer (OC) has been hindered by multiple factors, including tumor heterogeneity, immune exclusion, and a profoundly immunosuppressive tumor microenvironment. This review provides a comprehensive analysis of current ACT modalities, including tumor-infiltrating lymphocytes, T cell receptor-engineered, and chimeric antigen receptor-T cell therapies, as well as emerging approaches such as bispecific T cell engager (BiTE)-secreting T cells, dual-targeting platforms, and synthetic antigen receptors. We examine their application in OC and contextualize relevant findings using insights from other solid tumors. Key barriers, including limited T cell persistence, antigen escape, and T cell exhaustion, are explored alongside strategies to enhance efficacy through cytokine armoring, checkpoint modulation, metabolic reprogramming, and gene editing. We further highlight innovations in safety engineering, including logic-gated and self-regulating synthetic circuits, to mitigate toxicity and improve precision. Additional attention is given to the evolving role of allogeneic products and in vivo engineering as scalable solutions. Finally, we emphasize the critical value of integrating high-dimensional tools such as spatial transcriptomics, single-cell profiling, and machine learning to refine ACT design, identify biomarkers of response, and support patient selection and stratification. Collectively, these advances offer a roadmap for overcoming the unique immunologic barriers to ACT in OC and accelerating the development of more potent, durable, and personalized T cell-based strategies.

Peptide presentation on human leukocyte antigens (HLAs) is essential for initiating T-cell responses and all consequences of this presentation including anticancer immunity or immune escape. Many studies have relied on in silico prediction tools rather than biological measurement of HLA presentation to study these effects. To better assess the frequency and consequences of neoantigen presentation, we overexpressed 125 combinations of full-length neoantigens and one HLA class I allele to experimentally validate presentation of mutated and non-mutated HLA ligands through HLA ligand isolation followed by tandem mass spectrometry. A successful presentation was observed only in 22% of predicted cases with strong implications on previously described downstream effects. For example, the association of HLA loss of heterozygosity with predicted neoepitopes was challenged for 58% (73/125) of combinations. Furthermore, when testing 51 sequences used for personalized messenger RNA neoepitope vaccines, we observed that clinical responses were independent of the presentation status of the neoepitopes. Even a presumably neoepitope-specific and strongly expanded T cell receptor clone from a neoantigen vaccination study could not be linked to a successfully presented neoepitope. Overall, these data highlight the importance of validating the presentation of neoepitopes to fully understand our interpretation of clinical mutation-specific responses and their related effects, including immune evasion.

Background: Oncolytic herpes simplex virus (oHSV) therapy is a live virus-based immunotherapy that lyses tumor cells which release antigens and activate antitumor immunity. oHSV therapy has been shown to increase ATP production and release of extracellular ATP (eATP). In the extracellular tumor microenvironment, eATP functions as an immune-activating damage-associated molecular pattern but is hydrolyzed to extracellular adenosine (eADO), which can be immune-suppressive. eADO is generated by the sequential action of ectoenzymes CD39 and CD73 (NT5E). Here, we examined the role of immunosuppressive eADO signaling in regulating antitumor immune efficacy of oHSV.

Methods: We evaluated changes in eADO signaling in vitro and in patient specimens after virotherapy. A genetic CD73 knock-out mouse model and blocking antibodies were used to assess the impact of CD73 on virotherapy in two different solid tumor models. Single-cell RNA sequencing was employed to assess changes in immune cell infiltration and communication. Flow cytometric immunophenotyping and immunofluorescent imaging were utilized to confirm single-cell sequencing predicted changes in tumor microenvironment.

Results: Transcriptomic analysis of patient tumors pre-virotherapy and post-virotherapy with CAN-3110 revealed increased expression of the adenosine receptor gene ADORA2B after treatment. High NT5E gene expression, as well as gene signatures suggestive of adenosine signaling, correlated with a significantly worse prognosis for patients with solid tumors. Single-cell sequencing of immune cells recruited to tumor-bearing brain hemispheres in CD73 knockout mice revealed an increase in macrophage-mediated antigen presentation and CD4⁺ T cell cross-communication. Intracranial tumor-bearing CD73 knock-out mice treated with oHSV showed significant therapeutic improvement as the result of oHSV compared with wild-type mice. Combination of virotherapy with CD73 antibody blockade also resulted in enhanced antitumor efficacy.

Conclusions: Here, we identify that immunosuppressive eADO signaling in the TME is a major barrier to oHSV therapy and CD73 blockade prevents tumor immune escape. The combination of oHSV with CD73 blockade supports the development of an antitumor immune memory response in solid tumors. This study supports clinical development of this combination strategy.