Glioma Constructs Retain Patient Features, but Cell Expansion Shifts Drug Response
Hyaluronan-based patient-derived tumor constructs preserved patient-associated expression patterns better than two-dimensional cultures. Rebuilding constructs after cell expansion could either increase or reduce temozolomide sensitivity.
An Engineered Glioma Model with a Passage Comparison
Researchers at The Ohio State University and Wake Forest used hyaluronan-based hydrogels to study patient-derived glioma cells. Their Micromachines paper asked whether the resulting tumor constructs preserved features of the original tumor and whether prior cell expansion changed drug response. The model consisted of tumor cells embedded in crosslinked hyaluronan and gelatin.
The team procured 29 biospecimens and fabricated constructs from 16. Nine glioma cases formed the analyzed collection: five IDH-wild-type grade 4 glioblastomas, three IDH-mutant astrocytomas spanning grades 2–3 and one oligodendroglioma. Insufficient cell yield, banking and other selection steps limited how many specimens reached the reported experiments.
Fresh and Expanded Cells Could Give Different Answers
The investigators compared constructs made directly from freshly dissociated tumor cells, designated P0, with constructs rebuilt from cells that had first undergone one or two passages in two-dimensional culture. For P1 and P2, the passages occurred in two dimensions before new three-dimensional constructs were fabricated.
After seven days without drug, constructs received temozolomide at 100 μM, 500 μM or 1 mM for 72 hours. Separate experiments used 10 or 100 μM NSC59984, a compound studied as a p53-pathway activator. ATP-based viability and live/dead imaging assessed the treatment response.
Two-dimensional expansion changed temozolomide sensitivity in both directions. In two cases, fresh-cell constructs did not respond, whereas constructs remade from later-passage cells became sensitive to 1 mM temozolomide. Another case showed the opposite pattern: fresh-cell constructs responded at all tested doses, P1 constructs responded only at 1 mM and P2 constructs showed no response. Returning cells to a three-dimensional matrix did not restore a uniform original response.
Responses Varied Across Glioma Cases
Among the five glioblastoma cases, two were sensitive to temozolomide and three were not under the tested conditions. All three astrocytoma cases responded at 1 mM, but only one responded at lower concentrations. The oligodendroglioma construct showed no temozolomide response. These small groups describe heterogeneity within the collection rather than comparative efficacy estimates for the diagnoses.
NSC59984 also produced variable effects across patient-derived models. In an accompanying A172 cell-line experiment, increased p21 supported engagement of a p53-associated response. That finding provides mechanistic context, while leaving the compound’s complete contribution to cell death unresolved in each patient construct.
The millimolar temozolomide condition is especially relevant to interpretation. The culture system provides a direct drug exposure and contains no blood–brain barrier. Its response at a high concentration cannot establish that the same exposure or effect is achievable in a patient’s brain tumor.
Three-Dimensional Expression Patterns Retained Patient Identity
RNA sequencing compared parent tumors, hydrogel constructs and two-dimensional cultures using the 1,000 most variable genes. Tumors and three-dimensional constructs tended to cluster by patient. The two-dimensional cultures instead resembled one another more strongly and were all assigned to a mesenchymal category using the study’s expression-classification scheme.
This result supports the defined matrix as a means of retaining patient-associated expression patterns, but pharmacological stability requires its own check. Drug-response changes after expansion show why transcriptional resemblance alone cannot certify that a model preserves the original sensitivity.
The collection was small and depended on resected material with enough viable cells for culture. The defined hydrogel also contains fewer growth factors and cellular components than a tumor environment. No clinical response comparison was performed. For drug research, the practical finding is that cell history belongs alongside matrix composition and tumor identity when results from patient-derived glioma constructs are compared.