Meningeal immune convergence: IFN-γ–driven DC–NK axis suppresses leptomeningeal tumor

Recent studies have increasingly focused on modulating immune responses in cancer therapy, particularly regarding the complex immune circuits involved. For instance, in a recently published article in Nature, Remsik et al. highlighted the pivotal role of interferon-gamma (IFN-γ) in orchestrating a distinct meningeal anti-tumor immune circuit, presenting a promising immunotherapeutic strategy for this lethal disease.¹ This study revealed that IFN-γ modulates meningeal immune responses through the dendritic cell-natural killer (DC-NK) cell axis, providing a new theoretical basis for treating leptomeningeal metastasis (LM). While mouse models effectively demonstrate the role of IFN-γ, they cannot fully replicate the immune microenvironment of human LM, which limits their clinical translation. Moreover, systemic administration of IFN-γ may cause central nervous system toxicity, and the complexity of the immune microenvironment presents challenges in formulating treatment strategies. Nevertheless, future research should explore the safety and efficacy of local IFN-γ delivery through large-scale prospective clinical trials and optimize immunotherapy strategies to bridge the gap between basic research and clinical translation. These studies are expected to provide new directions for clinical treatment of LM.

1. Immune response and the role of IFN-γ in LM

LM is a fatal complication of solid tumors, characterized by the dissemination of cancer cells to the meninges bathed in cerebrospinal fluid (CSF).² Clinically, LM often mimics infectious meningitis, marked by extensive immune cell infiltration in the CSF—a sharp contrast to the immune-excluded microenvironment of parenchymal brain metastases, where immune cells are largely kept out. Yet in LM, despite abundant immune cells and a robust inflammatory response, tumor cells persist and proliferate. This paradox—where active immune responses coexist with persistent tumor burden—has remained a central enigma in the field.³ Previous research has suggested that inflammation may, counterintuitively, support tumor survival by triggering adaptive responses. Inflammatory cytokines, for instance, can induce the expression of iron-binding proteins and promote blood–CSF barrier disruption through complement activation, thereby creating a microenvironment favorable to cancer cell persistence.^4,5 In this study, Remsik et al. identified markedly elevated levels of IFN-γ in the CSF of LM patients, implicating this cytokine as a central regulator of the meningeal immune response. Through a combination of clinical sample analysis and innovative murine models, they further revealed that T cell-derived IFN-γ activates a distinct DC–NK cell axis, which in turn orchestrates a protective antitumor immune response.

2. The remodeling role of IFN-γ in the immune microenvironment of LM

To provide an unbiased evaluation of the leptomeningeal anti-tumor response, Remsik et al. collected CSF from breast and lung cancer patients both with and without LM, then performed single-cell transcriptomic profiling together with targeted proteomic analysis. A marked remodeling of the CSF immune landscape was observed in LM patients, characterized by the presence of diverse myeloid and lymphoid populations in place of the predominantly CD4⁺ T-cell milieu seen in non-metastatic samples, and by a broad upregulation of inflammatory ligands, most notably IFN-γ, which was elevated across tumor types. Interestingly, analysis of clinical cohort samples revealed significantly elevated IFN-γ levels in the cerebrospinal fluid of LM patients. Notably, those with higher CSF IFN-γ at diagnosis exhibited longer overall survival and increased numbers of lymphoid and myeloid cells. These findings suggest that IFN-γ plays a crucial role in the anti-tumor response within the leptomeninges.

3. IFN-γ regulation of the innate immune response in LM

Systematic investigation of immune–tumor cell interactions in LM has been hindered by the lack of suitable immunocompetent animal models. To overcome this and to elucidate downstream IFN-γ effectors and their roles within the leptomeninges, the research team developed six immunocompetent mouse models of LM across two genetic backgrounds. These models, derived from lung cancer, breast cancer, and melanoma through iterative in vivo selection, faithfully reproduce key features of human LM. They found that mice lacking IFN-γ or its receptor Ifngr1 could not control tumor growth in the meninges, leading to rapid disease progression. By contrast, meningeal-specific overexpression of Ifng via adeno-associated virus (AAV) markedly restricted tumor spread, even in immunodeficient mice lacking adaptive immunity. These findings demonstrate that IFN-γ acts locally in the meninges through innate immune pathways. Following mechanistic studies showed that meningeal T cells are the primary source of IFN-γ, with NK cells contributing to a lesser extent. Conditional knockout of IFN-γ in T cells prevented recruitment and activation of peripheral myeloid cells−particularly conventional dendritic cells−in the meninges. Finally, IFN-γ was essential for driving conventional dendritic cell 2 (cDC2) maturation into migratory CCR7⁺ dendritic cells, a critical step for mounting an effective immune response.

4. The potential and clinical application prospects of IFN-γ in LM

This commentary summarizes the research of Remsik et al. on the role of IFN-γ in LM, focusing on its profound impact on meningeal immune responses. Remsik et al. were the first to map the IFN-γ-dependent immunoregulatory network, revealing the central role of IFN-γ in coordinating meningeal antitumor immunity and identifying it as a promising therapeutic target. However, translating these findings into clinical application still faces numerous challenges, particularly in terms of mechanistic understanding and clinical trials. Although the study provides new insights into the role of IFN-γ in meningeal metastasis immune responses, the conclusions are mainly based on murine models, which may not fully replicate the complex pathological process of human LM. Therefore, how to translate these findings into effective clinical treatment strategies remains a significant challenge.

Although the research of Remsik et al. has provided new insights into the immune responses in LM, their conclusions are primarily based on murine models. While these models are crucial for uncovering relevant pathways, they may not fully replicate the pathological features of human LM, and the differences between murine models and human pathology complicate the clinical translation of these findings. In particular, systemic IFN-γ administration may cause severe side effects, especially central nervous system toxicity. Therefore, developing local targeted delivery platforms for brain membrane metastasis, to reduce systemic side effects and improve therapeutic efficacy, is a key issue that needs to be addressed. Localized delivery can precisely direct IFN-γ to the leptomeningeal microenvironment, avoid systemic side effects, and enhance the treatment’s specificity and safety. AAV vectors and other targeted delivery systems offer potential solutions for this.

Furthermore, the research of Remsik et al. primarily focuses on the role of T cells and natural killer (NK) cells in the IFN-γ-mediated immune response, but the role of other immune cells, particularly macrophages, in LM has not been fully explored. Future research should further investigate the specific functions of these immune cells in LM to provide a more comprehensive theoretical foundation for the development of immune therapies. Specifically, the mechanism by which CCR7⁺ dendritic cells activate NK cells under IFN-γ influence remains unclear, and this critical process requires further research. Additionally, understanding immune cell interactions, immune tolerance, and their support of anti-tumor immunity within the leptomeningeal microenvironment should also be important directions for future studies.

Regarding human data, although Remsik et al.’s retrospective cohort study showed encouraging preliminary results, these data have not yet been validated by large-scale prospective clinical trials. Therefore, conducting prospective clinical trials to validate the efficacy and safety of IFN-γ in clinical treatments will be a key focus for future research.

Looking forward, combining IFN-γ modulation with other immunotherapies, such as NK cell activators or immune checkpoint inhibitors, could provide new opportunities to enhance anti-tumor efficacy. For example, using NK cell activators to boost DC function, or combining IFN-γ with immune checkpoint inhibitors, could generate synergistic effects, enhancing both the intensity and duration of the immune response. Meanwhile, developing robust and neuron-safe IFN-γ delivery platforms will be key to bridging the gap between basic research and clinical application. Innovative delivery technologies may maximize the therapeutic potential of IFN-γ while minimizing side effects, thus advancing clinical treatment.

In conclusion, Remsik et al. highlighted a unique IFN-γ-mediated mechanism that suppresses tumor growth within the leptomeningeal microenvironment and emphasized the critical role of the DC-NK cell axis. They found that IFN-γ derived from T cells in the meninges recruited myeloid cells to act on cDCs, thereby promoting the generation of CCR7⁺ DCs and supporting NK cell function, ultimately leading to the efficient clearance of intrameningeal tumor cells. This work reveals a distinctive, CNS-specific immune surveillance pathway operating in the leptomeningeal space−one that differs from classical adaptive immunity−and offers new insight into the specialized immune response in LM. Undoubtedly, despite the numerous challenges, IFN-γ remains a promising therapeutic target for LM. By decoding the unique immunological features of the leptomeningeal microenvironment, this study not only deepens our understanding of lethal LM but also lays a solid theoretical foundation for developing novel immunotherapeutic strategies, offering new treatment options for LM patients. Fig. 1.

Fig. 1.

Interferon-γ drives a new mechanism of leptomeningeal anti-tumor immunity. LM are associated with increased lymphoid and myeloid infiltration and elevated IFN-γ, which promotes DC2 maturation and NK cell–mediated tumor control via IL-12 and IL-15. Loss of IFN-γ signaling impairs NK cell maintenance and facilitates metastatic outgrowth.

CRediT authorship contribution statement

Yicheng Zhi: Visualization and Writing – original draft. Chenyang Ye, Yi Sun: Supervision and Writing – review & editing. Ji Wang: Visualization, Supervision Writing – original draft and Writing – review & editing.