Bone-Eating Cells Sabotage the Brain’s Drainage System to Help Tumors Invade the Dura

When cancer spreads to the dura mater, the tough outermost membrane that envelops the brain and spinal cord, the prognosis is grim and the biology has remained stubbornly opaque. Dural metastasis, a devastating complication of advanced breast and lung cancers, has long been treated clinically as a cousin of brain metastasis, yet it behaves in ways that standard models of intracerebral spread fail to explain. A new study published in Nature now reveals a startling mechanism: tumor cells do not invade the dura alone. They recruit the skeleton’s own demolition crew, the osteoclasts, and those bone-resorbing cells physically dismantle the meningeal lymphatic vessels that the brain relies on for immune surveillance and fluid drainage. The finding reframes dural metastasis as a disease of corrupted lymphatic plumbing, and it points to a therapeutic strategy that already has approved drugs in the clinic.

The research team, led by Rui Zeng, Jiaxu Zhao, Haibao Peng and senior author Yudan Chi at Fudan University in Shanghai, began with an observation from the clinic. Reviewing magnetic resonance imaging and computed tomography scans of patients with breast and lung cancer, they noticed that dural metastases produce a distinctive pattern of cranial erosion, with thinning of the skull bone directly above the tumor lesions. This signature was absent in patients with brain metastases, leptomeningeal metastases or meningiomas, suggesting that something unique was happening at the interface between the skull and the dura when tumor cells settled there. The erosion was not incidental damage; it was a clue that bone-resorbing cells were active participants in the disease process.

To interrogate that clue, the researchers built a mouse model of dural metastasis in which fluorescently labeled tumor cells home specifically to the dura mater. Using bioluminescence imaging and histological analysis, they confirmed that the model faithfully recapitulates the human pattern of dural colonization and skull erosion. They then turned their attention to the meningeal lymphatic vessels, the network of thin-walled channels discovered to line the dura and drain fluid and immune cells from the central nervous system to the deep cervical lymph nodes. Since their characterization in 2015, these vessels have been recognized as a critical highway for anti-tumor immunity, allowing T cells primed in the meninges to mount responses against brain tumors. In the dural metastasis model, that highway was in ruins.

Through intravital two-photon microscopy, which allows researchers to watch living tissue in real time through a surgically prepared window, the team documented dramatic impairments in lymphatic drainage as tumors progressed. The vessels adjacent to tumor lesions became dilated, developed blebs, and suffered local structural injuries. Functional tracer studies, in which fluorescent dextran and protein tracers were injected into the cerebrospinal fluid and tracked to the deep cervical lymph nodes, confirmed that drainage capacity collapsed in tumor-bearing animals. Notably, the researchers found that the tumors did not simply sprout new lymphatic vessels or prune existing ones wholesale; instead, the architecture of the vessels was disrupted in a way that suggested mechanical or biochemical sabotage rather than simple lymphangiogenesis.

The identity of the saboteurs emerged from single-cell RNA sequencing of the dural tumor microenvironment. The analysis revealed a strongly immunosuppressive landscape, with T cells showing clear molecular signatures of exhaustion, including elevated expression of inhibitory receptors such as PD1 and TIM3 and diminished cytotoxic function. Among the immune and stromal populations, cancer-associated osteoclasts stood out. Using a genetic reporter system that labels cathepsin K-expressing cells, the team traced two distinct osteoclast populations: large osteoclasts that localized specifically at the skull-tumor interface, where they carved resorption cavities into the bone, and smaller osteoclasts that permeated the intratumoral regions. Dynamic imaging showed these two subsets were interrelated, with the interface population appearing to seed the intratumoral one.

The causal link between osteoclasts and lymphatic destruction was tested pharmacologically. When the researchers depleted osteoclasts with zoledronic acid, a bisphosphonate drug already approved for the treatment of bone metastases and osteoporosis, the meningeal lymphatic vessels recovered their structure and their drainage function. Tracer flow to the deep cervical lymph nodes was restored, vessel dilation and injury diminished, and, critically, the exhausted T cell population in the dura rebounded. Tumor growth slowed, and survival improved in the treated animals. The effect depended on intact lymphatic drainage: when the researchers surgically ligated the drainage pathways in zoledronic acid-treated mice, the anti-tumor benefit was largely abolished, demonstrating that the drug was working not by directly killing tumor cells but by reopening the immune corridor that the tumors had shut down.

The mechanistic culprit turned out to be angiopoietin-2, a signaling molecule secreted by the osteoclasts themselves. Angiopoietin-2 is known to destabilize blood vessels, but its role in lymphatic biology is context-dependent, determined in part by the endothelial phosphatase VEPTP. In the dural metastasis model, osteoclast-derived angiopoietin-2 was sufficient to disrupt lymphatic vessel integrity even in the absence of tumors: administering the protein to healthy mice or to zoledronic acid-treated tumor-bearing mice reproduced the drainage defects. Conversely, knocking down Angpt2 expression specifically in cathepsin K-expressing cells using an adeno-associated viral vector restored lymphatic structure and drainage, reduced T cell exhaustion, slowed tumor growth and extended survival. The team also ruled out a parallel candidate, PDGF-BB, which showed no effect on lymphatic drainage or tumor progression, sharpening the case for angiopoietin-2 as the key mediator.

The implications extend beyond dural metastasis. Meningeal lymphatic vessels have been implicated in glioblastoma immunity, in the efficacy of radiotherapy, and in the response to checkpoint blockade immunotherapy, and recent work has shown that osteoclasts can undermine immunotherapy in bone metastases through other secreted factors. The new study connects these threads, positioning the skull-dura interface as a battleground where bone-resorbing cells, acting under tumor influence, disable the very drainage routes that anti-tumor T cells need. It also fits within a broader rethinking of skull bone marrow as an active immunological compartment, connected to the brain surface through specialized channels and responsive to pathological signals from the central nervous system and its coverings.

Translational prospects are tantalizing but cautious. Zoledronic acid is already in widespread clinical use, and patients with bone metastases routinely receive it, yet dural metastasis remains a clinical challenge, suggesting that timing, dosing and delivery will matter enormously. The selective angiopoietin-2 knockdown strategy used in mice is far from bedside readiness, though anti-angiopoietin-2 agents have been explored in oncology trials for other purposes. What the study provides, above all, is a conceptual framework: dural metastasis is not merely tumor growth in an inconvenient location, but a coordinated hijacking of the meningeal lymphatic system by tumor-corrupted osteoclasts, a process that can in principle be reversed. If that reversal can be achieved safely in patients, the brain’s own drainage highways might once again carry the immune cells needed to fight back.

Subject of Research: The role of osteoclast-derived angiopoietin-2 in disrupting meningeal lymphatic drainage during dural metastasis

Article Title: Osteoclasts intercept meningeal lymphatic paths for dural metastasis

Article References: Zeng, R., Zhao, J., Peng, H., Ma, J., Hua, S., Zeng, Y., Li, X., Huang, Y., Li, C., Yang, J., Chen, H., Feng, X., Zhang, Q., Chen, L., & Chi, Y. (2026). Osteoclasts intercept meningeal lymphatic paths for dural metastasis. Nature. https://doi.org/10.1038/s41586-026-11082-6

Image Credits: AI Generated

DOI: 10.1038/s41586-026-11082-6

Keywords: dural metastasis, osteoclasts, meningeal lymphatic vessels, angiopoietin-2, zoledronic acid, T cell exhaustion, tumor microenvironment, single-cell RNA sequencing, breast cancer, lung cancer, skull erosion, cancer immunotherapy