Isocitrate dehydrogenase (IDH‑mutant) gliomas make up the majority of lower‑grade diffuse gliomas in young adults and typically behave indolently relative to IDH‑wildtype tumours, yet they undergo inevitable malignant transformation. The mechanisms that drive progression remain poorly defined. This study used longitudinal matched samples from two patients (WHO grades 2, 3 and 4 per patient) and applied spatial transcriptomics together with expression and enrichment analyses to identify transcriptional programs associated with progression.
The investigation used a unique longitudinal cohort comprising two patients with IDH‑mutant diffuse gliomas. For each patient, tumours that had been classified as WHO grade 2, grade 3 and grade 4 at different timepoints were profiled, enabling within‑patient comparison across malignant transformation.
Spatial gene expression was measured using two complementary platforms: NanoString GeoMx and 10X Visium. Analytical methods reported by the authors included differential expression analysis, identification of spatially variable genes, and gene ontology enrichment to interpret spatial transcriptional programs underlying progression.
Contrary to expectations, the authors observed that higher‑grade tumours showed reduced proliferation at the infiltrative margin when measured by Ki67 across both patients. This spatially resolved decrease in margin proliferation contrasts with the conventional view that higher grade necessarily equates to greater proliferative activity at invasive fronts, and emphasizes the value of spatially localized measurement.
Spatial transcriptomic analyses identified an enrichment of invasion‑associated genes with increasing tumour grade. The spatially derived transcriptional program emphasized pathways related to migration, cytoskeletal dynamics, and stress responses. These invasion‑related signatures were spatially localized and became more prominent as tumours progressed from lower to higher WHO grades within the longitudinal samples.
The authors evaluated the spatially discovered program in an independent bulk RNA‑seq cohort. In that bulk dataset, the expression of the spatially derived program did not outperform a model based on grade alone for prediction of progression‑free survival (PFS) or overall survival (OS). This result highlights a limitation of bulk expression profiling to capture spatially resolved evolutionary features that may be biologically meaningful but are diluted in bulk measurements.
The combined observations—decreased Ki67 at the infiltrative margin in higher‑grade samples and enrichment of invasion‑associated transcriptional programs with grade—support a model in which malignant progression in IDH‑mutant glioma involves spatially localized shifts toward invasion, cytoskeletal remodeling and stress pathways rather than uniform increases in proliferation across the tumour.
These findings underscore the need to develop new metrics that explicitly account for spatial evolution during progression, since such changes can escape detection in bulk RNA analyses. The authors propose prioritizing migration‑, cytoskeletal‑ and stress‑associated programs for further mechanistic and translational investigation as candidate drivers or biomarkers of malignant transformation.
The authors provide links to supporting data and code repositories. Accession identifiers and resources reported include a GEO accession, a BioProject entry, a cBioPortal study summary and a GitHub repository containing analysis code. Specific accession identifiers were listed in the source material.
The manuscript is a preprint and has not undergone peer review. The authors declared specific competing interests for several investigators (financial relationships and equity interests were listed in the source). Funding sources acknowledged by the authors include institutional and philanthropic support and grants as detailed in the original report.
This spatially resolved study of longitudinal, within‑patient samples of IDH‑mutant glioma demonstrates that malignant progression can be accompanied by increased expression of invasion‑related programs and reduced proliferative labeling at invasive margins. The work highlights the utility of spatial transcriptomics for revealing tumor‑microenvironment interactions that bulk profiling may miss and motivates development of spatially informed biomarkers and further study of migration and cytoskeletal pathways in progression.