Pediatric high-grade gliomas, H3-wildtype, IDH-wildtype: Refining diagnostic criteria and exploring clinical associations

Abstract

Background

Diffuse pediatric-type high-grade gliomas (pHGGs), H3-wildtype (H3 WT), and IDH-wildtype (IDH WT), represent a newly recognized, highly malignant brain tumor entity with unique molecular and epigenetic profiles. Despite their recent inclusion in the WHO 2021 classification, their histological and molecular diversity presents significant diagnostic and therapeutic challenges.

Methods

This study analyzed 12 cases of pHGGs, H3 WT and IDH WT, identified through whole-genome methylation profiling. Clinical, radiological, and histopathological evaluations were complemented by immunohistochemical profiling, employing an extended spectrum of antibodies, and molecular studies. Methylation profiling enabled precise tumor classification and correlation with known subtypes such as RTK-1, RTK-2, and MYCN.

Results

Among the 12 cases, 7 were classified under the RTK-1 subtype, 3 under RTK2 subtype, and a case each under MYCN, pHGGs, H3 WT, and IDH WT subtype B. Notably, 5/12 cases demonstrated a loss of H3K27me3 expression, contradicting the WHO 2021 recommendation for its retention as a diagnostic criterion. Furthermore, 6/7 RTK-1 subtype cases were linked to either mismatch-repair (MMR) deficiency or radiation-induced gliomas, highlighting an enrichment of these clinical scenarios within this subgroup. MGMT promoter methylation was observed in only 4/12 of these cases, consistent with its low prevalence in this tumor category.

Conclusions

This study provides a comprehensive characterization of pHGGs, H3 WT, IDH WT, and emphasizes the clinical and molecular complexity of this rare tumor entity. The findings challenge current WHO diagnostic criteria regarding H3K27me3 retention and demonstrate the critical role of molecular diagnostics, particularly methylation profiling, in refining classification and guiding clinical management. These results advocate for re-evaluation of existing diagnostic frameworks to better accommodate the observed variability and associations in this challenging tumor subtype.

Key Points.

• Discrepancy in current diagnostic criteria of pediatric-type high-grade gliomas H3-wildtype and IDH-wildtype.

• Molecular diagnostics are essential for refining tumor classification, highlighting the need to update frameworks to capture this rare tumor diversity.

Importance of the Study.

This study provides critical insights into the clinical and molecular complexity of diffuse pediatric-type high-grade gliomas, H3-wildtype, and IDH-wildtype, a newly recognized and highly aggressive brain tumor entity. Since its inclusion in the WHO CNS5 classification, limited studies have explored this tumor subtype, leaving significant gaps in understanding its diagnostic and therapeutic implications. By integrating whole-genome methylation profiling with detailed histopathological and immunohistochemical analyses, this study highlights a key discrepancy in current diagnostic criteria, particularly regarding H3K27me3 expression. The findings challenge the assumption that its retention is a defining feature, emphasizing the need for a more nuanced classification approach. Additionally, the study confirms the low prevalence of MGMT promoter methylation, reinforcing its limited role in therapeutic decision-making. The results highlight the essential role of molecular diagnostics in refining tumor classification and call for a re-evaluation of existing frameworks to better capture the biological diversity of this rare tumor subtype.

Diffuse pediatric-type high-grade glioma (pHGG), H3-wildtype (H3 WT) and IDH-wildtype (IDH WT), is a recently identified brain tumor entity defined by its unique DNA methylation profile. This tumor was introduced in the 2021 WHO classification of central nervous system (CNS) tumors as a highly malignant glioma (WHO grade 4) predominantly affecting children and adolescents.¹ Histologically, these tumors are characterized as diffusely infiltrating gliomas with high mitotic activity, lacking mutations in the IDH1, IDH2, and H3 genes. Retained H3K27me3 expression on immunohistochemistry (IHC) is one of the desirable diagnostic criteria as per WHO 2021.¹ Molecularly, it exhibits a methylation profile associated with subgroups such as pHGG RTK1, pHGG RTK2, or pHGG MYCN, and invariably shows alterations including PDGFRA amplification, EGFR alterations, or MYCN amplification.² Prognosis is generally poor, with a median overall survival (OS) of 17 months. Among the subtypes, pHGG RTK2 demonstrates the longest median survival of 44 months, followed by pHGG RTK1 at 21 months, and pHGG MYCN at 14 months.²

Radiation therapy (RT) remains a cornerstone of treatment for many cancers, significantly improving outcomes in pediatric malignancies, including CNS tumors and leukemia. However, radiation-induced malignancies can occur in a subset of long-term survivors, particularly pediatric patients, due to their extended follow-up periods.^3,4 Radiation-induced gliomas (RIGs) are aggressive secondary tumors that arise in a subset of patients receiving cranial RT for primary malignancies such as acute lymphoblastic leukemia (ALL) and medulloblastoma (MB). They typically manifest after a latency period ranging from 2.5 to 35 years post-irradiation.⁵ Recent studies have identified most RIGs as pHGG of the H3 WT IDH WT, RTK1 subtype, characterized by genetic alterations like chromosome 1q gain, CDKN2A/B loss, and PDGFRA amplification.^6,7 Unlike sporadic pHGG, RIGs rarely harbor mutations in histone H3 variants or IDH1/2.⁵

MMR-deficient pHGGs predominantly cluster within the methylation-defined pediatric RTK-1 subgroup, characterized by the absence of IDH1 and H3 mutations and large-scale copy number alterations.⁸ These tumors exhibit a unique CpG island demethylator phenotype (CIDP), in stark contrast to the hyper methylated phenotype seen in IDH-mutant gliomas, emphasizing their distinct epigenetic signature.⁸ Independent analyses confirmed their association with hypermutant profiles and highlighted their unique clustering pattern, aligning with the poorly understood “Wild type-C” group in other cohorts.⁸ These findings highlight the genomic and epigenetic specificity of MMR-deficient pHGGs, classifying them as H3 WT IDH WT gliomas within the RTK-1 subgroup.

Despite these advancements, significant challenges persist in the clinical translation of molecular insights, particularly in resource-limited settings. The variability in access to advanced diagnostic modalities necessitates the exploration of surrogate markers and streamlined workflows to bridge the gap in care. In this study, we evaluated the clinical, histopathological, immunohistochemical, MMR status, and radiological findings of 12 pHGG characterized as H3 WT and IDH WT on methylation profiling. Our study aims to provide a comprehensive characterization of these tumors by integrating multidisciplinary data and to contribute to the evolving framework of precision neuro-oncology, particularly for this challenging tumor subtype.

Materials and Methods

Histologically confirmed cases of diffuse high-grade gliomas (HGGs), grade 4, not otherwise specified (NOS), were retrieved from the archives of the Neuropathology Laboratory at AIIMS, New Delhi. The study was approved by the AIIMS Ethics Committee (IEC 200/04.03.2022). Informed written consent was obtained from all prospective patients at the time of surgery, while a waiver of consent was granted for retrospective cases involving deceased patients. Only cases with sufficient paraffin-embedded tissue and/or fresh tissue from consenting patients were included in the study. Histopathological features and immunohistochemical markers were reviewed by 2 independent pathologists (M.C.S. and V.S.), and a final diagnosis was made according to the 2021 WHO classification of CNS tumors.

Immunohistochemistry

IHC was performed on 5-μm-thick formalin-fixed, paraffin-embedded tumor sections using an automated immunostainer (Benchmark XT, Ventana Medical Systems) following standard protocols. Pretreatment included cell conditioning with Ventana’s Cell Conditioning 1 buffer for 52 min, along with standard Ventana signal amplification.

Commercially available primary antibodies were employed for IHC, including IDH1-R132H (clone H09, 1:100; Dianova), ATRX (1:400; Sigma), H3K27M (clone ABE419, 1:1000; Millipore), p53 (1:200; Santa Cruz Biotechnology), Olig2 (1:100, BioSB, RBT-OLIG2), H3G34R (1:200, Invitrogen, Anti-H3.3 G34W Recombinant Monoclonal [RM263]), H3G34V (1:500, Invitrogen Anti-H3.3 G34V onco-histone mutant Recombinant Monoclonal [RM307]), H3K27me3 (1:400, Cell Signaling Technology), MIB-1 (clone MIB-1/M7240, 1:200; Dako), BRAF V600E (RTU, Ventana, anti-BRAF V600E [VE1]), and EZHIP (Anti-CXorf67, Sigma Aldrich). For mismatch-repair protein analysis, MLH1, MSH2, MSH6, and PMS2 antibodies (Ventana, RTU) were used. Immune profiling was conducted using antibodies for PD-L1 (clone SP263, 1:100; Ventana) and CTLA-4 (1:75; Santa Cruz Biotechnology).

Interpretation of IDH1-R132H staining followed a previously established 3-tiered semi-quantitative system. ATRX expression was considered negative when there was complete nuclear absence of staining in tumor cells, with retained staining in endothelial cells serving as an internal control. Loss of ATRX expression was interpreted as indirect evidence of ATRX mutation. For p53, strong nuclear positivity in >50% of tumor cells was considered positive. BRAF V600E positivity was defined by strong, granular cytoplasmic staining, while EZHIP over-expression was identified by diffuse, intense nuclear immunopositivity. Deficiency in MMR proteins was defined as the complete absence of nuclear staining in tumor cells, either alone (Lynch-like) or in combination with endothelial cells (CMMRD-like), for any one or more of the MMR proteins (MLH1, MSH2, MSH6, and PMS2). PD-L1 positivity was defined as >1% of tumor cells exhibiting membrane staining, with or without cytoplasmic staining, of any intensity. PD-L1 expression was further categorized as weak (1%–5%), moderate (>5%–25%), or strong (>25%) based on the percentage of positive cells. Similarly, CTLA-4 expression in T cells was scored as negative (0%), weak (1%–5%), moderate (6%–25%), or strong (>25%) based on the percentage of lymphocytes showing positivity.

Sanger Sequencing

Sanger sequencing was performed for IDH1, and IDH2 genes as described in our previous studies.⁹ The IDH2 primer set included the following sequences: forward primer: 5’GCTGCAGTGGGACCACTATT3′ and reverse primer: 5′ TGTGGCCTTGTACTGCAGAG 3′.

Methylation Profiling

Whole-genome DNA methylation analysis was performed using the Infinium Methylation EPIC BeadChip V2 (935K) array, following the manufacturer’s recommended protocols. The resulting.idat files from the methylation array were processed for analysis. Tumor methylation classification was performed using the DKFZ (v12.8) and the Methylscape classifier, as described in our previous study.^10–12

Results

Case Cohort

Whole-genome methylation profiling was performed on 61 cases of pediatric and AYA diffuse HGGs, initially categorized as NOS based on routine histopathological and immunohistochemical evaluations. Of these, 12 cases (19.7%) were identified as diffuse pHGG, H3 WT, and IDH WT. The median age of patients was 11.5 (interquartile range [IQR]: 3.5 years, age range: 2–21 years), and a male-to-female ratio of 1:1. Postsurgical treatment strategies varied across cases and are not the primary focus of this study. Patient characteristics are summarized in Table 1.

Table 1.

Patient Characteristics of All 12 Cases of pHGG, H3 WT, IDH WT

Case ID	Age/sex	Location	Pleomorphism/endothelial proliferation/necrosis/ mitosis	MIB-LI (%)	Olig2	IDH1 R132H/ATRX/P53/H3K27M/H3G34R/H3G34V/EZHIP/BRAFV600E	H3K27 Me3	HPE report	Methylation profiling (CI score) [classified on DKFZ/NIH classifier]	MGMT promoter methylation	Overall survival (months)	Clinical correlation
AYA 1	12/M	Right frontal, temporal, and insula	Mild/+/−/−	20	+	−/ R/ −/ −/ −/ −/ −/ −	L	Diffuse HGG, NOS	Mc diffuse PHGG, RTK2 subtype, subclass A (novel) (0.99) [DKFZ]	UM	20.6, death	—
AYA 30	10/M	Left frontal	Marked/+/+/+	50	+	−/ R/ −/ −/ −/ −/ −/ −	R	Diffuse HGG, NOS	Mc Diffuse PHGG, RTK1 Subtype, Subclass A (novel) (0.99) [DKFZ]	UM	30.1, alive	MMR deficient, PMS2 loss
AYA 33	8/M	Left parieto-occipital	Marked/+/+/+	40	+	−/ L/ +/ −/ −/ −/ −/ −	R	Diffuse HGG, NOS	Mc Diffuse PHGG, RTK1 Subtype, Subclass A (novel) (0.33) [DKFZ]	UM	9.7, death	MMR deficient, PMS2 loss
AYA 43	11/M	Left lateral ventricle and thalamus	Moderate/−/−/+	25	+	−/ R/ +/ −/ −/ −/ −/ −	R	Diffuse HGG, NOS	Mc Diffuse PHGG, RTK2 Subtype, Subclass B (novel) (0.99) [DKFZ]	UM	13.1, death	—
AYA 55	12/F	Optic-chiasmatic	Marked/−/−/+	12	+	−/ R/ −/ −/ −/ −/ −/ −	R	Diffuse HGG, NOS	Mc diffuse PHGG, RTK1 subtype, subclass C (novel) (0.99) [DKFZ]	M	1.8, death	—
AYA 62	2/F	Left parietooccipital	Marked/+/+/+	40	+	−/ L/ +/ −/ −/ −/ −/ −	L	Diffuse HGG, NOS	pedHGG_RTK1A (0.80) [NIH]	M	Death on the day of surgery	MMR deficient, PMS2, MLH1 loss
AYA 71	11/M	Left middle frontal gyrus	Marked/+/−/+	40	+	−/ R/ −/−/ −/ −/ −/ −	R	Diffuse HGG, NOS	Mc Diffuse PHGG, RTK2 Subtype, Subclass A (novel) (0.99) [DKFZ]	UM	1.9, death	—
AYA 89	10/F	Left parietal	Moderate/−/+/+	70	+	−/ R/ +/−/ −/ −/ −/ −	R	Diffuse HGG, NOS	pedHGG_MYCN (0.99) [NIH]	UM	12.3, alive	—
AYA 92	16/F	Right frontal lobe and frontotemporal	Moderate/+/−/+	20	+	−/ R/ −/−/ −/ −/ −/ −	L	Diffuse HGG, NOS	Mc Diffuse PHGG, RTK1 Subtype, Subclass B (novel) (0.99) [DKFZ]	M	4, death	History of ALL
AYA 94	13/F	Right frontal	Marked/−/−/+	30	+	−/ R/−/ −/ −/ −/ −/ −	L	Diffuse HGG, NOS	Mc Diffuse PHGG, RTK1 Subtype, Subclass B (novel) (0.99) [DKFZ]	M	5, death	History of ALL
CAR 35	21/F	Right temporal and gangliothalamic	Mild/−/−/+	10	+	−/ R/−/−/ −/ −/ −/ −	R	Diffuse HGG, NOS	Mc Diffuse PHGG, H3 wildtype and IDH wild type, Subtype B (novel) (0.98) [DKFZ]	UM	8.06, alive	—
CAR 45	14/M	Left posterior frontal	Moderate/+/+/+	20	+	−/R/+/−/ −/ −/ −/ −	L	Diffuse HGG, NOS	MC Diffuse PHGG, RTK1 subtype, subclass B (novel) (0.49) [DKFZ]	UM	1.7, alive (lost to follow-up)	History of ALL

Abbreviations: ALL, acute lymphoblastic leukemia; F, female, HGG, high-grade glioma; HPE, histopathological examination; L, loss; M, male; MMR, mismatch repair; NOS, not otherwise specified; R, retained. +, Present; −, Negative.

Histology

Histopathological examination across all cases demonstrated high-grade morphology with hypercellularity, with moderate to marked nuclear pleomorphism, endothelial proliferation in (7/12) mitotic activity (12/12), and necrosis (5/12). The MIB-LI labeling index ranged from 10% to 70%.

Immunohistochemistry

Immunohistochemical analysis revealed strong nuclear positivity for Olig2 in all 12 cases (100%). IDH1 R132H, H3K27M, H3G34R, H3G34V, and BRAF V600E were negative in all 12 cases (100%). ATRX expression was retained in 10/12 (83.3%), while P53 was positive in 5/12 (41.7%) cases. Notably, H3K27Me3 exhibited a loss in 5/12 (41.6%) of the cases. None of the cases with H3K27Me3 loss showed overexpression of EZHIP (Supplementary Figure 1).

Three cases (25%) were deficient and showed loss of one or more MMR proteins. A single case showed loss of both PMS2 and MSH2, while 2 cases exhibited loss of PMS2. PD-L1 expression was seen in 4/12 (5%–25%), and CTLA-4 was positive in a single case.

Sanger Sequencing

None of the cases showed any mutation in the IDH1 and IDH2 genes as analyzed by Sanger sequencing.

Whole-Genome DNA Methylation Profiling

Twelve cases were classified as diffuse pHGG, H3 WT, and IDH WT, based on methylation profiling. Of these, 7 cases (58.3%) were identified as RTK1 subtype, including 3 cases each of subclass A and B, and a single case of subclass C. Three cases (25%) were classified under the RTK2 subtype, comprising 2 cases of subclass A and 1 of subclass B. One case (8.3%) was classified as the MYCN subtype, while another case (8.3%) was categorized as diffuse pHGG, H3 WT, and IDH WT, subtype B. The CI score of all the cases is mentioned in Table 1. CNV profiles are shown in Supplementary Figure 2.

Radiology

Out of the 12 cases, MRI scans are available for 8, while a CT scan is available for 1 patient. Lesions were primarily situated in the thalamus, temporal lobe, frontal lobe, and optic pathway areas. On T1-weighted imaging, signal characteristics varied from iso-hypointense to hypointense and iso-hyperintense. T2-weighted scans exhibited iso- to hyperintense signals, featuring patterns such as tiny intralesional cysts. Diffusion-weighted imaging demonstrated a range from no restriction to clear diffusion restriction, especially in solid enhancing components, signifying varying cellularity. Contrast enhancement patterns comprised no enhancement, peripheral enhancement surrounding necrosis, mild enhancement, and pronounced patchy uptake. Hemorrhage was absent in the majority of cases but observed in some instances, particularly within necrotic regions. Perilesional edema varied from absent to substantial, and lesion borders ranged from poorly defined to well defined.

Clinical Correlation

Out of the 3 cases classified as diffuse pHGG, H3 WT, and IDH WT, RTK1 subtype subclass A, all (100%) were found to be MMR deficient. One case demonstrated loss of both PMS2 and MSH2, while the remaining 2 showed isolated loss of PMS2. Notably, all 3 cases of the RTK1 subtype subclass B had a prior history of ALL and had been exposed to cranial irradiation.

Discussion

pHGGs are among the most aggressive and heterogeneous tumor groups in children, contributing significantly to cancer-related mortality.¹³ These tumors, distinct from their adult counterparts, have been reclassified based on molecular features in the WHO Classification of Tumors of the Central Nervous System, 2021.¹⁴ pHGGs now encompass several molecularly defined subgroups, including diffuse midline gliomas (DMGs; H3K27M-mutant), diffuse hemispheric gliomas (H3 G34R/V-mutant), and pHGGs with H3 WT/IDH WT.¹⁰

Pediatric-type HGGs with H3 WT/IDH WT form a diagnostically challenging subgroup characterized by diverse genetic and epigenetic profiles.¹⁵ These tumors lack hallmark mutations like H3K27M or IDH1/2 but often display aggressive clinical behavior.¹ Our study aimed to examine their histology, clinical features, and imaging to refine classification and biological understanding.

In 2015, Korshunov et al. classified 202 pediatric glioblastomas (pedGBMs) into 4 molecular subgroups via DNA methylation: H3.3 G34-mutant, H3.3/H3.1 K27-mutant, IDH1-mutant, and H3/IDH-wildtype, with the latter comprising 59 patients (median age: 12 years).² Further profiling in 2017 of 87 H3-/IDH-wt pedGBMs identified 3 subtypes—pedGBM_MYCN (MYCN amplification), pedGBM_RTK1 (PDGFRA amplification), and pedGBM_RTK2 (EGFR amplification), each with distinct molecular and prognostic features.¹⁵ These findings highlighted the heterogeneity of H3-/IDH-WT pedGBMs and the need for molecularly stratified therapies. Mackay et al.’s meta-analysis of over 1000 pHGGs and DIPGs further emphasized the molecular and clinical diversity of H3-/IDH1-wildtype tumors, reporting a 2-year survival of 23.5% and a median OS of 17.2 months.¹⁶ Methylation profiling identified distinct clusters, including PXA/LGG-like tumors (BRAF V600E mutations, CDKN2A/B deletions) with better outcomes and aggressive subgroups driven by EGFR, MYCN, CDK6, and PDGFRA/MET amplifications.¹⁶ Additionally, 71% of H3/IDH-wildtype tumors harbored amplified oncogenes, with only 19% showing MGMT promoter methylation, linking this subgroup to an intermediate prognosis.¹⁵ These studies highlight the need for refined molecular risk stratification to improve diagnosis and therapy in pedGBMs.

Bender et al. retrospectively analyzed 8 cases of diffuse pHGG (H3 WT/IDH WT) treated between 2015 and 2022, representing one of the first studies conducted after this entity was formally recognized in the CNS WHO 2021 classification.¹⁷ DNA methylation profiling proved essential for accurate diagnosis, revealing a broad age range (8–71 years), suggesting this tumor is not exclusive to pediatric patients. Imaging was nonspecific, with T2 hyperintensity, T1 hypo- to iso-intensity, and variable contrast enhancement. Clinical outcomes varied, with progression-free survival of 9–22 months and overall survival of 19–79 months; 3 patients died, while 4 remained alive at the last follow-up. The study reinforces the tumor’s molecular and clinical heterogeneity, the diagnostic value of DNA methylation profiling, and the potential involvement of MMR pathways, as 1 case showed PMS2 loss.¹⁷ A recent large cohort study of 207 H3 WT, IDH WT, HGGs in the teenage and young adult population showed that over half (86/158, 54.4%) of the methylation-classifiable cases were assigned to pediatric-type subgroups. These included pedHGG-RTK1 (51/158, 32.3%; RTK1A: 24/51, 47.1%; RTK1B: 16/51, 31.4%; RTK1C: 11/51, 21.6%), pedHGG-MYCN (13/158, 8.2%), pedHGG-RTK2A/B (12/158, 7.6%), and pedHGG-A/B (n = 10/158, 6.3%).¹⁸ While well established in other cohorts and supported by our findings, associations with cancer predisposition and RIGs were observed at lower frequencies in this study. Only 2/24 (8.3%) RTK1A cases were linked to tumor predisposition syndromes, and 2/16 (12.5%) RTK1B cases had a history of childhood cancer and cranial irradiation, suggestive of RIGs.¹⁸ Notably, both these comprehensive studies lack the assessment of H3K27Me3 status.^17,18

Recent studies have provided key insights into the molecular and clinical landscape of RIGs, particularly in pediatric cases. López et al. identified a distinct genomic profile with TP53 mutations, CDK4 amplifications, CDKN2A deletions, and receptor tyrosine kinase pathway alterations (PDGFRA, MET), along with highly aneuploid genomes.¹⁹ Deng et al. further refined this profile, clustering RIGs within the pedGBM_RTK1 subgroup due to frequent PDGFRA amplifications, CDKN2A/B deletions, and losses on chromosomes 1p, 13q, and 14q.⁵ DeSisto et al. identified distinct RIG subgroups, including stem-like and pro-inflammatory types, emphasizing the role of DNA methylation profiling and potential MAPK/ERK pathway targets.⁶ Grogan et al. reported rare cases of complete radiographic response to radiotherapy.⁷ Consistent with these findings, we report 3 cases of RIGs arising after prophylactic cranial radiotherapy for ALL, all classified as diffuse pHGGs, H3 WT, IDH WT, RTK-1 subtype, subclass B. Two of the patients (AYA 92 and AYA 94) succumbed to their illness within 4–5 months of diagnosis, while a patient (CAR 45) was lost to follow-up after 1.7 months of his surgery. These cases reinforce the critical role of methylation profiling in accurately classifying RIGs, distinguishing them from other gliomas, and providing insights into their distinct molecular and clinical features. Our findings highlight the need for vigilant long-term follow-up and early molecular diagnostics in ALL survivors receiving cranial radiotherapy.

Anirban Das et al. highlight the high prevalence of MMR deficiency in pHGGs in LMICs, accounting for 33%–50% of cases due to consanguinity.^20–22 These tumors have poor prognoses, with survival often under 3 months, and limited response to temozolomide.²³ However, their high tumor mutation burden and microsatellite instability make them responsive to immune checkpoint inhibitors, with 3-year survival rates of 39% in recurrent cases.²⁰ Dodgshun et al. studied 51 patients with replication repair-deficient (RRD) HGGs, identifying a unique CIDP. Most RRD HGGs lacking IDH1 R132 or H3F3A K27 mutations clustered within the Pediatric RTK1 and Wild Type-C methylation subgroups.^8,13,24 In our study also, the 3 MMR-deficient cases were classified as pHGGs, H3 WT, IDH WT, RTK-1 subtype, subclass A through methylation profiling. AYA 30 had no reported history of consanguinity; however, his elder brother was diagnosed with blood cancer and succumbed to the disease in early adolescence. AYA 33 had no family history of consanguinity or malignancy. In AYA 62, the parents were consanguineously married, although no other cancers were reported in the family.

Out of a cohort of 12 cases we analyzed, 7 were classified as RTK-1 subtype, of which 6 (85.7%) were associated with either MMR deficiency or RIGs. This indicates that the majority of RTK-1 subtype cases in our study were linked to these specific mechanisms.

Previous studies have demonstrated that MGMT promoter methylation is observed at low frequency in H3 WT, IDH WT pediatric gliomas. Korshunov et al. reported MGMT promoter methylation in only 19% of cases within the H3 WT subgroup. In a subsequent analysis in 2017, methylation was observed in 18% of RTK1 subtype cases, while it was absent in the RTK2 subtype.¹⁵ These findings highlight the general lack of MGMT promoter methylation in this category of tumors, which correlates with their poor response to temozolomide therapy. Consistent with these observations, our study also revealed that 8/12 cases (66.6%) exhibited an unmethylated MGMT promoter status, further emphasizing the aggressive biology and limited therapeutic responsiveness of these tumors.

H3K27M mutation and the associated loss of H3K27me3 are hallmark features of DMGs, defining their unique molecular and diagnostic identity. These alterations disrupt Polycomb Repressive Complex 2-mediated gene silencing, leading to widespread epigenetic dysregulation and aggressive clinical behavior, particularly in midline tumors such as those of the pons and thalamus.^2,16,25 These defining features underline the highly malignant nature of DMGs and their poor prognosis. In contrast, Tauziède-Espariat et al. analyzed pediatric supratentorial MYCN-amplified HGGs (HGG-MYCN) and found that all studied cases (n = 5) retained H3K27me3 expression.²⁵ This preservation differentiates them from H3K27M-mutant DMGs, which exhibit a loss of H3K27me3. The stability of H3K27me3 expression in HGG-MYCN serves as a critical diagnostic marker, emphasizing the importance of incorporating H3K27me3 analysis in the immunohistochemical diagnostic panel. This differentiation aids in accurate tumor classification and informs tailored therapeutic strategies for pediatric supratentorial HGGs.

However, in our study of 12 cases of pHGG, H3 WT, and IDH WT, 5 cases demonstrated a loss of H3K27me3 expression. According to the WHO classification, retention of H3K27me3 is a desirable criterion for H3 WT, IDH WT cases. Our findings, contrary to this guideline, suggest that the criterion of H3K27me3 retention may need re-evaluation in this context. These results highlight the importance of reconsidering the diagnostic criteria to better account for variability in H3K27me3 expression among pHGG subtypes.

Conclusion

Our study provides a detailed clinical, radiological, and immunohistochemical profile of pHGGs, H3 WT, IDH WT, offering valuable insights into this diagnostically complex subgroup. A noteworthy finding in our study is the observation that 5 out of 12 cases (41.6%) demonstrated a loss of H3K27me3 expression, contrary to the WHO classification, which considers retention of H3K27me3 in these tumors as a desirable criterion. This discrepancy highlights the potential need for revisiting this criterion to better reflect the variability observed in H3 WT, IDH WT gliomas, as it may indicate distinct biological mechanisms or subgroups within this classification. Additionally, our findings highlight an enrichment of the RTK-1 subtype with cases associated with either MMR deficiency or RIGs. This study contributes to the evolving landscape of pediatric glioma classification, providing evidence to refine existing diagnostic frameworks and offering deeper insights into the clinical and molecular complexity of H3 WT, IDH WT gliomas.

Supplementary material

Supplementary material is available online at Neuro-Oncology Practice (https://academic.oup.com/nop/).

Conflict of interest statement

The authors declare no competing interests.

Funding

The work was supported by a research grant from the Indian Council of Medical Research, India (F.No.: 5/13/1/2022/NCD-III and F.No. EM/Dev/CAR/01/0090/2024) to Dr. Vaishali Suri.

Author Contributions

Author contributions to the study and manuscript preparation include the following: Conception and design: S.B., J.S., S.S., C.S., and V.S. Sample collection: S.K. and A.S.

Acquisition of data: S.B., S.S., V.S., A.G., and M.C.S. Interpretation of data: S.B., J.S., S.S., S.V., C.R., H.J., A.G., M.C.S., and V.S. Drafting the article: S.B., J.S., and V.S. Critically revising the article: S.B., J.S., and V.S. Reviewed submitted version of manuscript: All authors approved the final version of the manuscript.

Ethics Statement

This study was approved by the AIIMS Ethics Committee (IEC 200/04.03.2022 and AIIMSA00916/05.03.2024).

Data Availability

The raw data supporting this study can be made available upon reasonable request by contacting the corresponding author.