iiinova — where your digital self keeps evolving

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Huge thanks to my co–first author Raphael Collot, our supervisors Anne Rios & Hendrik Stunnenberg, all co-authors, collaborators, funding agencies, patients & families who made this work possible. We’d love to hear your feedback and thoughts! 23/n

010

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

We are committed to making our data FAIR and openly accessible. Processed data are freely available on Zenodo! 🌐 Processed data (Zenodo): zenodo.org/deposit/1706... 💻 Code (coming soon): github.com/ccruizm/pHGG... 22/n

zenodo.org

Zenodo

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

We hope pHGGmap becomes a community resource for discovering additional vulnerabilities and uncovering new biology — similar in spirit to what GBmap has provided for adult GB😊 — especially in tumor-reactive states (PM/WR-like) and the RG-like complement axis driving invasion. 21/n

110

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Beyond what I could cover here, the study also contains interesting insights into epigenetic plasticity and state transitions. I encourage you to dive into the full manuscript for more details. 20/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Take-home: pHGG biology is organized into repeatable cancer–myeloid ecosystems, and successful therapy may require niche-aware immune modulation (lipid-laden TAMCs vs microglia vs complement-TAMCs) rather than “one-size-fits-all” immunotherapy. 19/n

110

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Inspired by evidence that TAMC programs can be reprogrammed, the patient received hydroxychloroquine as adjuvant therapy. This was associated with reduced peritoneal RG-like cells and increased cytotoxic CD8⁺ T cell recruitment/activation. 18/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

We validate the clinical relevance in a striking case of abdominal metastasis via VP shunt, where ascitic fluid was dominated by RG-like tumor cells and infiltrating myeloid cells that progress toward an immunosuppressive, complement-TAMC-like state. 17/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

This RG-like/complement-TAMCs ecosystem shares stress-, motility-, and immune-related gene signatures, suggesting a cooperative axis that supports invasion/adaptation. 16/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Importantly, we discovered a third niche, stable during disease progression, enriched with stem-like states dominated by RG-like cells and complement-high TAMCs, displaying the highest proliferation + invasiveness (primarily driven by RG-like cells! not OPC-like).🤯 15/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Importantly, despite changes in their proportions, both niches remain consistently present across patient samples, highlighting fundamental microenvironments that persist despite therapy-driven remodeling! 14/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Both single-cell and spatial data confirmed that tumor-reactive regions align with high hypoxia/stress, while astro-microglial niches sit in lower-stress regions. 13/n

110

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

An astro-microglial niche (abundant in treatment-naïve): AC-like tumor cells + homeostatic/pro-inflammatory TAMCs, which we recently identified in DMG, representing a vulnerable ecosystem to target. doi.org/10.1016/j.cc... 12/n

doi.org

Redirecting

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Having established the map of cellular states, we next asked how malignant and myeloid programs co-occur and organize together within tumors, moving from individual states to multicellular communities (“niches”), and whether these ecosystems persist or shift with therapy. 10/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

We characterize TAMCs' programs, defining both homeostatic and inflammatory microglia, as well as multiple programs linked to lipid metabolism, complement signaling, and immune suppression. 9/n

110

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

pHGGs are immunologically “cold” tumors, and the immune landscape is dominated by tumor-associated myeloid cells (TAMCs) rather than lymphocytes, making myeloid states central to tumor–immune interactions. 8/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Within tumor-reactive programs, we identify a wound response (WR) and a distinct primitive mesenchymal-like (PM-like) program that exhibits embryonic and primitive features with increased stemness. PM-like appears largely pediatric-specific, especially when compared to adult HGG. 7/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Beyond the classic OPC-/OC-/AC-like framework, our atlas resolves very early neurodevelopmental progenitors, including phenotypes previously assumed absent in DMG (e.g., nIPC-like), and delineates an RG-like malignant state that later emerges as a key TME-associated state. 👀 6/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

Using updated human developmental brain references, we refine malignant-state nomenclature and position tumor cells along an axis of (1) developmental mimicry ↔ (2) tumor-reactive, stress/TME-driven programs. 5/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

To tackle this, we built pHGGmap: a large, multimodal atlas integrating sc/snRNA-seq, snATAC, Multiome + spatial transcriptomics, across discovery + validation cohorts (over 800,000 analyzed cells from 136 patients). 4/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

We focused on two aggressive pediatric glioma subgroups: Diffuse Midline Glioma (DMG) and pediatric high-grade glioma H3-wild-type (pDG H3-WT), which arise in distinct brain locations and show unique growth patterns, making them particularly challenging to treat. 3/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

pHGGs are devastating and still understudied. A major bottleneck has been limited access to diverse samples at scale, which has restricted our ability to capture the full spectrum of cellular states across patients. 2/n

100

Cristian Ruiz-Moreno @ccruizm.bsky.social · 26/01/2026

I’m excited to finally share our research on pediatric high-grade gliomas (pHGG), the result of years of effort integrating single-cell and spatial datasets to offer a fresh perspective on their organization and potential treatment strategies doi.org/10.64898/202... 🧵👇🏼 1/n

131