A Comprehensive Treatment-Induced Resistance Atlas of Glioblastoma Reveals a Fibrotic Niche Shielding the Tumor from Immunotherapy
Wang, F.; Huang, R.; Ling, H.; Bai, Y.; Yang, C.; Zhao, G.; Wu, X.; Cao, W.; Lu, Y.; Zhang, Y.; Lu, D.; Qiu, Y.; Zhang, J.; Gao, B.; Ma, C.; Dang, H.; Zhang, Y.; Yang, Y.; Sun, T.; Chen, Z.; Wang, Z.
Show abstract
2Therapeutic resistance in IDH-wildtype glioblastoma (GBM) is driven by profound cellular plasticity and a structured immunosuppressive tumor microenvironment (TME). Here, we present the Glioblastoma Resistance Insights from Treatment Atlas (GRIT-Atlas), the most comprehensive single-cell resource to date, encompassing nearly one million cells from 296 samples across primary and recurrent cohorts, including those treated with immune checkpoint blockade (ICB) and anti-angiogenic combination therapy. We identify a convergent evolutionary trajectory where therapeutic pressure selects for a specific malignant state, cNMF7 (MES-like), characterized by a synergy of hypoxia, stemness, and inflammatory signaling. Integrating spatial transcriptomics across 48 patient sections, we define a "Spatial Resistance Triad"--a core functional unit composed of cNMF7 cells, differentiation-arrested E-MDSCs, and Type VI Collagen-secreting myCAFs. This triad specifically colonizes the hypoxic microvascular proliferation (MVP) and pseudopalisading necrosis (PAN) niches. Mechanistically, we show that myCAFs act as stromal architects, constructing a fibrotic scaffold through a Collagen/Fibronectin-CD44 signaling axis. This spatial infrastructure not only physically excludes cytotoxic T cells but also provides essential cues to sustain malignant plasticity and myeloid-mediated immunosuppression. Our findings across seven independent cohorts and pan-cancer validation underscore the clinical significance of this axis in driving immunotherapy failure. Collectively, the GRIT-Atlas provides a blueprint for dismantling the "immunosuppressive sanctuaries" of GBM to overcome therapeutic resistance.
Matching journals
The top 8 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Drug and single-cell gene expression integration identifies sensitive and resistant glioblastoma cell populations 97%
- Cellular senescence in malignant cells promotes tumor progression in mouse and patient Glioblastoma 97%
- Time-, tissue- and treatment-associated heterogeneity in tumour-residing migratory DCs 96%
Similar papers in this journal
- Single-nucleus and spatial landscape of the sub-ventricular zone in human glioblastoma 98%
- NRCAM variant defined by microexon skipping is a targetable cell surface proteoform in high-grade gliomas 96%
- The expression profile and tumorigenic mechanisms of CD97 (ADGRE5) in glioblastoma render it a targetable vulnerability 96%
Similar papers in this journal
- Single cell view of tumor microenvironment gradients in pleural mesothelioma 96%
- Multimodal Spatial Profiling Reveals Immune Suppression and Microenvironment Remodeling in Fallopian Tube Precursors to High-Grade Serous Ovarian Carcinoma 96%
- Precision combination therapies based on recurrent oncogenic co-alterations 95%
Similar papers in this journal
- A single-cell atlas characterizes dysregulation of the bone marrow immune microenvironment associated with outcomes in multiple myeloma 96%
- Cancer-associated fibroblast compositions change with breast cancer progression linking S100A4 and PDPN ratios with clinical outcome 95%
- Cell lineage as a predictor of immune response in neuroblastoma 95%
Similar papers in this journal
- Multi-omic landscape of human gliomas from diagnosis to treatment and recurrence 98%
- Single-cell integration and multi-modal profiling reveals phenotypes and spatial organization of neutrophils in colorectal cancer 96%
- MHC class II-restricted antigen presentation is required to prevent dysfunction of cytotoxic T cells by blood-borne myeloids in brain tumors 95%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.