The research, led by Rong Zhang, Feng Chen, and Wenbin Li of Beijing Tiantan Hospital, Capital Medical University, together with colleagues at Tsinghua University and the Beijing Institute of Neurosurgery, set out to answer three linked questions. Could fresh surgical specimens from adult ependymoma patients be reliably converted into three-dimensional organoids? Would those organoids preserve the histological and molecular identity of the tumors from which they came? And would they retain the malignant capacity to seed new tumors when transplanted into animals? The answers, drawn from a cohort of twenty-two patients, are cautiously encouraging and candidly nuanced in equal measure.

Organoids are self-organizing, three-dimensional cell cultures that recapitulate key architectural and molecular features of the tissue or tumor from which they are derived. Unlike traditional two-dimensional cell lines, which often shed the very characteristics that make a tumor clinically distinctive, organoids preserve much of the original cellular complexity. For brain tumors, this matters enormously: ependymomas are defined by perivascular pseudorosettes, ependymal rosettes, and the expression of markers such as glial fibrillary acidic protein and epithelial membrane antigen, features that are notoriously difficult to maintain in conventional culture.

The team processed fresh tumor tissue from twenty-two adult patients, most of whom harbored tumors in the posterior fossa, the compartment at the back of the skull that houses the cerebellum and brainstem. The success rate was striking: organoids were established from twenty of the twenty-two specimens, a 90.9 percent efficiency that compares favorably with organoid platforms developed for more common brain tumors. Under the microscope, the organoids broadly retained the morphology of their parental tumors, and immunofluorescence staining confirmed the continued expression of major ependymoma-associated markers as well as neural and progenitor cell markers. Importantly, selected morphologic and immunophenotypic features remained detectable even during prolonged culture, with six organoid models maintained to at least passage forty still showing GFAP and EMA expression comparable to early passages.

To probe fidelity at the molecular level, the researchers performed RNA sequencing on three matched tumor-organoid pairs. The global expression profiles were highly similar, with Pearson correlation coefficients ranging from 0.84 to 0.88, indicating that the organoids had not drifted wholesale away from their tumors of origin. Yet the analysis also revealed something the authors describe as substantial culture-associated remodeling, a transcriptomic adaptation to life in a dish that anyone hoping to use these models for drug discovery will need to take seriously.

The remodeling followed recognizable patterns. Endothelial signatures and stromal scores derived from the ESTIMATE algorithm were reduced in all three organoid pairs, an expected consequence of removing tumor cells from their vascular and stromal surroundings. Other immune and stromal signatures, however, changed in a patient-specific manner, suggesting that each tumor responds to culture conditions in its own idiosyncratic way. Paired differential-expression and gene-set enrichment analyses further identified a consistent enrichment of ribosome and translation programs and oxidative phosphorylation in the organoids, accompanied by a reduction in cilia- and axoneme-associated programs. That last finding is particularly intriguing for ependymoma specialists, because cilia, the hair-like projections that define ependymal cells, have been linked to tumor aggressiveness in previous work, with decreased FOXJ1 expression and ciliogenesis programs reported in aggressive ependymoma and choroid plexus tumors.

Perhaps the most consequential experiment came next. Selected organoid models were transplanted into mice, both under the skin and directly into the brain. The organoid-derived cells demonstrated tumorigenic capacity in both settings, generating xenografts that confirmed the cultures had not lost their malignant identity. This in vivo validation matters because the field has seen cautionary tales: a 2023 study reported loss of tumorigenicity in a patient-derived orthotopic xenograft model of ependymoma, underscoring that not every culture system preserves the biological essence of the original tumor. The Beijing team’s organoids passed that test, at least for the selected models evaluated.

The platform arrives at a moment when the molecular classification of ependymoma has matured considerably. The 2021 WHO Classification of Tumors of the Central Nervous System and subsequent large-cohort studies have established that ependymomas are best understood as a family of molecularly defined subgroups, including posterior fossa groups A and B, spinal ependymomas, and entities not elsewhere classifiable, each with distinct demographics, genetics, and prognosis. DNA methylation profiling has become central to diagnosis in adults, and single-cell RNA sequencing studies in pediatric ependymoma have revealed cellular hierarchies and developmental trajectories within these tumors. Yet therapeutic progress has been slow; clinical trials of targeted agents such as dose-dense temozolomide combined with lapatinib have yielded limited advances, and the EANO guidelines still anchor treatment in surgery and radiotherapy. A renewable, patient-specific model system is precisely the kind of tool that could accelerate preclinical drug screening for a disease where trial enrollment is inherently constrained by rarity.

The authors are careful about what their platform does and does not show. The transcriptomic remodeling they document is not a flaw to be hidden but a boundary condition to be respected: organoids faithfully carry selected tumor-associated features while undergoing measurable transcriptional and microenvironment-associated adaptation to ex vivo culture. For experiments probing cell-autonomous tumor biology, such as intrinsic drug sensitivity or proliferation signaling, the models appear well suited. For questions that hinge on endothelial interactions, immune infiltration, or stromal crosstalk, researchers may need to complement organoids with co-culture systems, organotypic slices, or in vivo xenografts, an approach echoed in the broader glioblastoma organoid literature, where three-dimensional cultures have been shown to recapitulate hypoxic gradients and cancer stem cell heterogeneity while still requiring careful interpretation of microenvironmental signals.

Practical accessibility may prove to be the platform’s most viral quality. The RNA sequencing dataset has been deposited in the Gene Expression Omnibus under accession number GSE347879, and the established organoid models may be obtained from the corresponding authors upon reasonable request, subject to institutional approval and a material transfer agreement. The work was approved by the Institutional Review Board of Beijing Tiantan Hospital and conducted with written informed consent from all participants, with animal procedures approved by the Beijing Neurosurgical Institute. As living biobanks of pediatric high-grade glioma and ependymoma have already demonstrated, shared model systems can transform a rare tumor field almost overnight, converting isolated surgical specimens into cumulative, comparable experimental resources. For adult ependymoma, a tumor that has watched other neuro-oncology subfields sprint ahead, twenty patient-derived organoids that keep their shape, their markers, and their malignant spark may be the starting line the field has been waiting for.

Subject of Research: Patient-derived organoid models of adult posterior fossa ependymoma

Article Title: A patient-derived organoid platform for predominantly posterior fossa adult ependymoma: histopathologic preservation, culture-associated transcriptomic remodeling, and in vivo tumorigenicity

Article References: Zhang, R., Li, S., Sun, Y., Ma, S., Wang, C., Kang, Z., Yang, X., Zhang, B., Xu, N., Guo, A., Wu, Z., Chen, F., & Li, W. (2026). A patient-derived organoid platform for predominantly posterior fossa adult ependymoma: histopathologic preservation, culture-associated transcriptomic remodeling, and in vivo tumorigenicity. Journal of Neuro-Oncology, 179(3), Article 105. https://doi.org/10.1007/s11060-026-05814-x

Image Credits: AI Generated

DOI: 10.1007/s11060-026-05814-x

Keywords: ependymoma, patient-derived organoids, brain tumor, posterior fossa, organoid culture, transcriptomics, xenograft, tumor modeling, neuro-oncology, RNA sequencing, tumor microenvironment, precision medicine