The link between MGMT expression and outcomes in H3 G34-mutant diffuse hemispheric glioma persists after accounting for pre-analytic and technical factors
-
By
-
Qunyu Zhang
-
Dana Tlais
-
Tong Lin
-
Xiaoyu Li
-
Arzu Onar-Thomas
-
Jason Chiang
-
October 14, 2026
Clinical Report: MGMT–outcome association in H3 G34-mutant glioma
Overview
In 110 centrally sequenced H3 G34-mutant diffuse hemispheric gliomas, normalization and batch correction removed release-year clustering and correlations attributable to sequencing depth. In the 25-patient outcome cohort, adjustment for four pre-analytic and technical factors did not eliminate the association between higher MGMT expression and shorter progression-free and overall survival. MGMT expression also remained associated with methylation at three gene-body/intronic CpG sites after adjustment.
Background
H3 G34-mutant diffuse hemispheric glioma is a rare pediatric-type high-grade glioma, and retrospective multi-institutional tumor cohorts may combine different collection periods, preservation methods, and technical workflows. These differences can introduce batch effects into RNA-sequencing analyses. The source article addresses whether such factors account for reported associations between MGMT expression, methylation, and clinical outcomes. It also describes steps used to assess and mitigate technical effects in this cohort.
Data Highlights
| Analysis | Result |
|---|---|
| RNA-seq cohort | 110 H3 G34-mutant tumors; centrally sequenced at St. Jude on NovaSeq 6000 |
| Release-year clustering before and after correction | Before correction: Kruskal–Wallis p < 0.001 for PC1 and PC2. After correction: p = 0.51 and 0.81. |
| Raw-count correlations of MGMT with CD34 and ACTB | r = 0.29 and 0.42 before correction; r = −0.04 and 0.01 after correction |
| ACTB counts by preservation method | Median raw counts were 4.6-fold higher in fresh/frozen than FFPE samples (p = 0.014). |
| MGMT expression and outcomes in 25 patients | Per doubling of expression: adjusted PFS HR 1.64 (95% CI 1.20–2.24; p = 0.002); adjusted OS HR 3.00 (95% CI 1.73–5.21; p < 0.001). |
| MGMT expression and methylation in 34 tumors | For cg15441483, cg18453665, and cg15782766, Spearman ρ values were 0.74, 0.75, and 0.71; after adjustment for four variables, ρ values were 0.58, 0.78, and 0.75 (all p < 0.001). |
| MGMT-STP27 promoter CpGs | Expression correlations were negative at cg12434587 (ρ = −0.48) and cg12981137 (ρ = −0.36). |
Key Findings
- The analysis included all 110 RNA-sequenced tumors from the initial report; four older-platform specimens identified as extreme by a reanalysis were excluded from both analyses.
- RNA-seq counts were TMM-normalized with edgeR, transformed to log2 counts per million, and adjusted for processing batch with limma removeBatchEffect.
- Release year distinguished samples before correction but not afterward. Four low-complexity FFPE samples remained outliers after correction; their exclusion did not change subsequent results.
- Raw-count correlations between MGMT and CD34 or ACTB were weaker in this cohort than those reported in a 20-tumor reanalysis, and correction eliminated the correlations in this cohort.
- Adjustment for collection year, preservation method, data release year, and tumor content did not eliminate the association between continuous MGMT expression and PFS or OS in the 25-patient outcome cohort; none of the four variables was significantly related to MGMT expression, PFS, or OS in the reported analyses.
- MGMT expression remained associated with methylation at three gene-body/intronic CpGs after adjustment; the authors note that methylation was measured using DNA on an independent array platform. Promoter CpGs showed negative correlations with expression.
Clinical Implications
The authors propose documenting pre-analytic variables, normalizing samples and correcting batch effects before comparison, excluding low-complexity libraries that cannot be rescued by correction, and testing whether adjustment changes marker–outcome associations. They also propose confirming key results with an independent assay and caution that correlation alone cannot demonstrate regulation of MGMT by gene-body methylation.
Conclusion
In this cohort, the reported association between MGMT expression and outcome persisted after accounting for four pre-analytic and technical factors. The authors report that normalization and batch correction addressed identified technical effects while low-complexity libraries remained outliers.
Related Resources & Content
- Source article, title and citation details not provided -- The link between MGMT expression and outcomes in H3 G34-mutant diffuse hemispheric glioma persists after accounting for pre-analytic and technical factors
- Related article, authors and year not provided -- Pediatric H3 G34-mutant diffuse hemispheric glioma: clinical, imaging and molecular prognostic factors, MGMT expression, and temozolomide response
- ASCO Publications — Analysis of MGMT promoter methylation status in high grade glioma patients with long term and conventional survival times: A retrospective study
- ASCO Publications — Analysis of MGMT promoter methylation status in high grade glioma patients with long term and conventional survival times: A retrospective study
- ASCO Publications — Predictive value of MGMT in glioblastoma: A multicenter study
- Pediatric H3 G34-mutant diffuse hemispheric glioma: clinical, imaging and molecular prognostic factors, MGMT expression, and temozolomide response
- Pediatric Central Nervous System Cancers, Version 2.2025, NCCN Clinical Practice Guidelines In Oncology
- Table 1. Molecular and immunohistochemical alterations listed as essential or desirable diagnostic criteria for gliomas, glioneuronal and neuronal tumours in the 2021 World Health Organization Classification of Tumours of the Central Nervous System.⁴#
- Pediatric Central Nervous System Cancers, Version 2.2025, NCCN Clinical Practice Guidelines In Oncology - PubMed
- Clinical Outcomes and Prognostic Features of Diffuse Hemispheric Glioma, H3 G34-mutant: An International Multi-Institutional Study | Clinical Cancer Research | American Association for Cancer Research
Based on findings from:
The association between MGMT expression and outcome in H3 G34-mutant diffuse hemispheric glioma is robust to adjustment for pre-analytic and technical variables
Qunyu Zhang, Dana Tlais, Tong Lin, Xiaoyu Li, Arzu Onar-Thomas, Jason Chiang. Acta Neuropathologica, 2026.
https://link.springer.com/article/10.1007/s00401-026-03105-3
This content is an AI-generated, fully rewritten summary based on a published scholarly article. It does not reproduce the original text and is not a substitute for the original publication. Readers are encouraged to consult the source for full context, data, and methodology.