Association of Adverse Transcriptomic Subtypes of Non-WNT/Non-SHH Medulloblastoma with the Dominance of Proliferative and Progenitor-like Subpopulations

Research Background

Medulloblastoma (MB) is one of the most common malignant tumors of the central nervous system in children. Based on molecular characteristics, the medical community typically classifies MB into four major subtypes: WNT, SHH, Group 3 (GRP3), and Group 4 (GRP4). While the molecular mechanisms of the WNT and SHH subtypes are relatively well understood, the molecular characteristics and clinical relevance of the non-WNT/non-SHH (GRP3/GRP4) subtypes are not fully identified. In recent years, researchers have found that GRP3/GRP4 MB can be further subdivided into eight second-generation subgroups (SGS). These secondary divisions aid in more precise risk stratification, but their internal cellular composition and relation to prognosis are not yet fully elucidated.

Paper Information

This paper, co-authored by Konstantin Okonechnikov, Daniel Schrimpf, Jan Koster, and others, is affiliated with institutions including the German Cancer Research Center (DKFZ), Heidelberg University Hospital, and others. The paper was published in the 2024 issue of the journal Acta Neuropathologica.

Research Methods

The study explores the internal and external cellular components of non-WNT/non-SHH medulloblastoma (MB) and their relationship with clinical prognosis through deconvolution analysis. The primary research methods are as follows:

• Dataset Selection and Processing: The study utilizes RNA transcriptome data of 435 previously generated non-WNT/non-SHH MB samples and single-cell RNA sequencing (scRNA-seq) reference datasets. Additionally, 168 samples from the ICGC cohort were used for comparison and validation.

• Deconvolution Analysis: Using the BayesPrism tool, combined with bulk RNA and single-cell RNA data, deconvolution analysis was conducted to infer the proportion of different cellular populations within tumor samples. UMAP visualization was used to display heterogeneity in cellular composition.

• Statistical Analysis: T-tests and Benjamini-Hochberg correction were used to measure differences in cellular type proportions across different subgroups and prognosis-associated transcriptomic subtypes. Kaplan-Meier methods and Cox regression models were used for multivariable analysis and survival correlation analysis.

Research Results

Cellular Composition within Tumor Subgroups

The study found that different subgroups within GRP3/GRP4 MB consist of various tumor cell subpopulations. These subpopulations include the neuron-associated axonal dendritic GP3-C1 and glutamatergic GP4-C1 subgroups, the progenitor GP3-B2 subgroup prominently featured in aggressive SGS II MB, and the photoreceptor/visual perception GP3/4-C2 cells typically present in SGS III/IV MB.

Moreover, the results indicated significant differences in the proportion of cellular subpopulations across clinically relevant transcriptomic subtypes. Adverse subtypes were often rich in cell cycle and progenitor-like cellular subpopulations, whereas favorable subtypes were primarily composed of neuronally differentiated cellular components.

Prognosis-related Cellular Types and Survival Analysis

The study found that proportions of proliferative and progenitor subpopulations higher than the median were associated with the shortest survival times in GRP3 and GRP4 MB. Most second-generation subgroups, with the exception of SGS IV MB, exhibited similar survival associations across all clinical subtypes. Notably, the adverse transcriptomic subtype of SGS VIII MB also mixed GRP3-related neuron GP3-C1 cell subpopulations, which were associated with extremely poor survival.

Validation Results

Independent validation using GSVA (Gene Set Variation Analysis) confirmed the deconvolution analysis results. The results demonstrated consistent gene expression patterns in tumor cell subpopulations across different SGS MB subtypes, further validating the accuracy of the deconvolution results.

Conclusion

Through deconvolution analysis, this study revealed the cellular composition and clinical relevance of different second-generation subgroups within non-WNT/non-SHH MB. Major findings include:

1. Cellular Heterogeneity of Tumor Subgroups: Significant differences in cellular composition across different subgroups, particularly based on specific distribution and proportion differences of cell types.

2. Identification of Prognosis-related Cellular Types: High proportions of proliferative and progenitor cell subpopulations were associated with adverse clinical prognosis, whereas neuronally differentiated cells showed the opposite relationship. This finding aids in further understanding the impact of intratumoral heterogeneity on clinical outcomes.

3. Validation and Reliability: The reliability of the study’s results was confirmed through multiple computational methods and datasets.

This cell composition-based research approach provides a new perspective for future targeted therapy studies. Future research should continue to verify the heterogeneity of GRP3/GRP4 MB and explore its clinical significance in conjunction with single-cell techniques.

Research Significance and Highlights

The highlight of this paper lies in its detailed analysis of the cellular composition and clinical relevance of non-WNT/non-SHH MB through the integration of single-cell RNA sequencing data. This approach not only reveals the impact of tumor heterogeneity on prognosis but also provides new research directions for molecular classification and precision medicine. By deeply understanding the cellular composition of tumors, more effective and targeted treatment strategies can be developed for different MB subtypes, ultimately improving treatment outcomes and the quality of life for affected children.