Distinct Clinicogenomic Contexts of BRAF‐Altered Gliomas and Glioneuronal Tumors: A Nationwide Real‐World Study Authors: Seiji Yamada, Tetsuya Takimoto, Sayaka Ueno, Sachiko Minamiguchi, Shigeo Ohba, Yuichi Hirose, Hideyuki Saya, Tamotsu Sudo Published: 2026 ABSTRACT ABSTRACT BRAF alterations occur across diverse gliomas and glioneuronal tumors, but their alteration‐specific clinicogenomic contexts in real‐world cancer genomic profiling cohorts remain incompletely characterized. We analyzed BRAF‐altered brain tumors registered in the Japanese Center for Cancer Genomics and Advanced Therapeutics database from June 2019 to March 2026. Operational diagnostic harmonization was performed using available clinicopathological and genomic information. Grade‐stratified analyses focused on BRAF p.V600E‐positive gliomas and glioneuronal tumors. Among 413 BRAF‐altered brain tumors, 344 were classified as gliomas and glioneuronal tumors after harmonization. Within this cohort, the major alteration contexts included BRAF p.V600E‐positive tumors (n = 166), KIAA1549::BRAF fusion‐positive tumors (n = 60), amplification‐only tumors (n = 15), and tumors with other BRAF alterations (n = 109), with overlap in a subset. KIAA1549::BRAF fusion‐positive tumors were mainly associated with low‐grade glioma contexts, particularly pilocytic astrocytoma. Amplification‐only tumors showed marked enrichment of MET alterations, all representing MET amplification, together with frequent EGFR and TP53 co‐alterations, supporting a receptor tyrosine kinase‐amplified, copy‐number‐rich context. Among 118 grade‐evaluable BRAF p.V600E‐positive gliomas and glioneuronal tumors, CDKN2A/B homozygous deletion and TP53/ATRX alterations were enriched in higher‐grade groups, and high‐grade tumors showed increased tumor mutational burden, although hypermutator phenotypes were rare. BRAF alteration classes define distinct clinicogenomic contexts in gliomas and glioneuronal tumors. These findings highlight both the opportunities and methodological challenges of nationwide real‐world cancer genomic profiling datasets in central nervous system tumor research.Keywords: BRAF, cancer genomic profiling, C‐CAT, glioma, MET amplification Abbreviations Center for Cancer Genomics and Advanced Therapeutics cancer genomic profiling copy‐number alteration central nervous system diffuse low‐grade glioma with MAPK pathway alterations dysembryoplastic neuroepithelial tumor European Association of Neuro‐Oncology epithelioid glioblastoma FoundationOne CDx FoundationOne Liquid CDx glioblastoma ganglioglioma high‐grade glioma interquartile range low‐grade glioma mitogen‐activated protein kinase OncoGuide NCC Oncopanel System not otherwise specified pilocytic astrocytoma pediatric‐type low‐grade glioma polymorphous low‐grade neuroepithelial tumor of the young pleomorphic xanthoastrocytoma receptor tyrosine kinase single‐nucleotide variant structural variant tumor mutational burden World Health Organization Introduction BRAF alterations are recurrent events across a broad spectrum of central nervous system (CNS) tumors [1, 2]. These alterations include KIAA1549::BRAF fusion‐positive low‐grade gliomas and BRAF p.V600E‐positive gliomas [1, 2]. Although these tumors share MAPK pathway activation as a common biological feature, accumulating evidence suggests that distinct BRAF alteration classes arise within substantially different clinicogenomic contexts. In particular, BRAF p.V600E‐positive gliomas and glioneuronal tumors represent a biologically heterogeneous group encompassing both circumscribed and diffuse tumor types with variable clinical behavior [2, 3]. These include pilocytic astrocytoma (PA), ganglioglioma (GG), pleomorphic xanthoastrocytoma (PXA), diffuse low‐grade glioma with MAPK pathway alterations (DLGG‐MAPK) [1, 4], polymorphous low‐grade neuroepithelial tumor of the young (PLNTY) [5], epithelioid glioblastoma (E‐GBM), and other diffuse high‐grade gliomas [2, 3]. Most previous studies of BRAF‐altered CNS tumors have primarily focused on canonical alteration classes, particularly KIAA1549::BRAF fusion‐positive low‐grade gliomas and BRAF p.V600E‐positive gliomas [6, 7]. In contrast, the broader clinicogenomic contexts of less common BRAF alteration classes, including BRAF amplification, non‐p.V600E mutations, and miscellaneous rearrangements, remain poorly characterized, particularly in large real‐world nationwide datasets. Previous studies have suggested that additional genomic events, including CDKN2A homozygous deletion, TERT promoter alteration, TP53/ATRX alterations, and broader copy‐number complexity, may contribute to malignant progression in BRAF p.V600E‐positive gliomas [8, 9]. However, the broader landscape of grade‐associated co‐occurring genomic alterations remains incompletely characterized, particularly in large real‐world clinical cohorts. In our previous institutional study using the PleSSision Rapid platform, we demonstrated that low‐grade BRAF p.V600E‐mutant gliomas and glioneuronal tumors generally exhibit relatively quiet genomic profiles, whereas higher‐grade tumors tend to acquire broader genomic alterations, particularly copy‐number changes [10]. However, the study was limited by its relatively small sample size and sensitivity to individual outlier tumors. Large‐scale nationwide cancer genomic profiling databases such as the Japanese Center for Cancer Genomics and Advanced Therapeutics (C‐CAT) now provide opportunities for investigating rare molecular tumor subsets across real‐world clinical settings [11, 12]. At the same time, real‐world nationwide cancer genomic profiling (CGP) datasets inevitably contain substantial diagnostic heterogeneity reflecting differences in institutional practice, evolving WHO criteria, historical terminology, and incomplete molecular information. The present study therefore sought not only to characterize grade‐associated genomic patterns in BRAF p.V600E‐positive gliomas and glioneuronal tumors, but also to investigate biologically distinct genomic contexts across BRAF alteration classes in a nationwide real‐world cohort. By integrating operational WHO‐based harmonization with large‐scale genomic analyses, we aimed to clarify both the biological heterogeneity of BRAF‐altered gliomas and glioneuronal tumors and the methodological challenges associated with real‐world nationwide molecular datasets. Materials and Methods Cohort Selection This study utilized brain tumor cases registered in the Japanese nationwide C‐CAT database between June 2019 and March 2026. A total of 3714 brain tumor cases were screened for BRAF alterations using clinical CGP data. CGP data were generated using FoundationOne CDx (F1CDx), GenMineTOP, OncoGuide NCC Oncopanel System (NOP), or FoundationOne Liquid CDx (F1LCDx). F1CDx interrogates 324 cancer‐related genes and detects short variants, copy‐number alterations (CNAs), and selected rearrangements. GenMineTOP integrates a 737‐gene DNA panel with an RNA panel capable of detecting fusions involving 455 genes. NOP is a tumor‐normal paired panel; depending on the version used during the study period, 114–124 genes were assessed for mutations and CNAs and 12–13 genes for rearrangements/fusions [13, 14, 15]. BRAF alterations were categorized into four operational groups: BRAF p.V600E‐positive tumors; BRAF p.V600E‐positive tumors; KIAA1549::BRAF fusion‐positive tumors; KIAA1549::BRAF fusion‐positive tumors; BRAF amplification‐positive tumors; BRAF amplification‐positive tumors; Tumors harboring other BRAF alterations, including non‐V600E mutations and other rearrangement/fusion events. Tumors harboring other BRAF alterations, including non‐V600E mutations and other rearrangement/fusion events. Several tumors harbored overlapping BRAF alteration classes. Subsequent grade‐stratified analyses focused on BRAF p.V600E‐positive gliomas and glioneuronal tumors. Operational Diagnostic Harmonization Because the C‐CAT database reflects real‐world clinical practice across multiple institutions and time periods, substantial diagnostic heterogeneity was anticipated. Therefore, operational WHO‐based harmonization was performed using available histological terminology, genomic alterations, age, and clinical context while intentionally avoiding unsupported reassignment of molecularly defined entities. Importantly, diagnoses requiring essential molecular criteria, such as oligodendroglioma or diffuse midline glioma, were not reassigned unless sufficient supporting molecular evidence was available within the CGP dataset. For grade‐stratified analyses of BRAF p.V600E‐positive gliomas and glioneuronal tumors, tumors were operationally categorized into Low‐grade, High‐grade, and Indeterminate groups based on integrated diagnostic, histopathological, and molecular features. The Low‐grade group consisted of tumors reasonably aligned with circumscribed low‐grade glioma/glioneuronal tumor contexts, whereas the High‐grade group included tumors compatible with diffuse high‐grade glioma phenotypes. The Indeterminate group comprised tumors with overlapping biological features or insufficient clinicopathological integration for definitive low−/high‐grade assignment, most of which represented PXA‐spectrum neoplasms. For transparency, the harmonized diagnostic categories used for operational grade stratification were predefined. Tumors assigned to the Low‐grade group included PA, GG, DNT, PLNTY, DLGG‐MAPK, pLGG, and LGG, NOS without high‐grade annotations. Tumors assigned to the High‐grade group included HGG, NOS and tumors otherwise compatible with diffuse high‐grade glioma phenotypes. PXA‐spectrum tumors and tumors with insufficient clinicopathological information for reliable low−/high‐grade assignment were assigned to the Indeterminate group. These categories were used solely for operational grade‐stratified analysis of real‐world CGP data and were not intended to replace formal CNS WHO grading. For an additional analysis of age distribution, cases with a harmonized diagnosis of PA were excluded from the KIAA1549::BRAF fusion‐positive group, and the age distribution of the remaining fusion‐positive tumors was compared with that of BRAF p.V600E‐positive tumors. Cases for which no usable diagnostic information was available in the C‐CAT dataset were designated as “diagnosis unavailable.” Cases for which diagnostic information was available but which could not be assigned to one of the predefined harmonized glioma or glioneuronal tumor categories were designated as “insufficient diagnostic information.” These cases, together with non‐glioma tumors, were excluded from the harmonized glioma/glioneuronal tumor cohort. Genomic Analysis Co‐occurring genomic alterations were analyzed using genes commonly represented across CGP platforms. Core alterations evaluated in the present study included CDKN2A/B, TERT promoter, TP53, ATRX, PTEN, EGFR, NF1, PDGFRA, and MET alterations. TERT promoter mutations were defined as canonical hotspot mutations (C228T [−124C>T] and C250T [−146C>T]). Alterations were treated as binary variables based on CGP annotations. Within the harmonized glioma and glioneuronal tumor cohort, BRAF amplification‐only tumors were defined as tumors harboring BRAF amplification without concurrent BRAF p.V600E mutation or KIAA1549::BRAF fusion. Tumors with concomitant other BRAF alterations, such as non‐p.V600E mutations or miscellaneous rearrangements, were retained in the amplification‐only group. To assess whether the accumulation of co‐occurring genomic alterations observed in high‐grade BRAF p.V600E‐positive tumors was specific to BRAF‐altered tumors, we performed an additional within‐tumor‐type analysis restricted to tumors registered as GBM, IDH‐wildtype. The frequencies and spectrum of core co‐occurring genomic alterations were compared between BRAF p.V600E‐positive and BRAF‐wild‐type GBMs. BRAF‐wild‐type tumors were defined as tumors without any reported BRAF alteration. For detailed characterization of the other BRAF alteration group, non‐p.V600E SNV/indels and non‐KIAA1549::BRAF rearrangements/structural variants (SVs) were reviewed at the case level. Fusion/rearrangement partners were recorded as reported in the CGP annotations, and duplicate or bidirectional annotations representing the same rearrangement within a tumor were consolidated as a single case‐level event. Because CDKN2A/B homozygous deletions are not routinely reported by GenMineTOP, a sensitivity analysis was performed after excluding tumors analyzed using this platform. In this analysis, only CDKN2A/B copy‐number loss events reported as CNAs were considered, excluding SNV/indel alterations. In the C‐CAT dataset, these events were annotated as copy‐number “loss”; all positive CDKN2A/B copy‐number loss calls in this sensitivity cohort were derived from F1CDx, for which such loss calls correspond to homozygous deletions rather than hemizygous loss. Tumor mutational burden (TMB) values were extracted from the C‐CAT dataset and analyzed across grade groups. Statistical Analysis Categorical variables were compared using Fisher's exact test or chi‐square analysis, as appropriate. Continuous variables, including TMB, were analyzed using the Mann–Whitney U test or Kruskal‐Wallis test. Statistical analyses were performed using Python. Two‐sided p values < 0.05 were considered statistically significant. Results Overall Landscape of BRAF Alterations Among 3714 brain tumor cases registered in the C‐CAT database, 413 harbored BRAF alterations (Figure 1). Among these 413 BRAF‐altered brain tumors, 200 harbored BRAF p.V600E alterations, 74 harbored KIAA1549::BRAF fusions, 26 harbored BRAF amplification, and 127 harbored other BRAF alterations, including non‐V600E mutations and other rearrangement/fusion events. Several tumors harbored overlapping BRAF alteration classes, including KIAA1549::BRAF fusion plus BRAF amplification (4 cases), KIAA1549::BRAF fusion plus other BRAF alterations (4 cases), and BRAF amplification plus other BRAF alterations (5 cases Table 1). CGP was performed using F1CDx in 280 cases (67.8%), GenMineTOP in 102 (24.7%), NOP in 30 (7.3%), and F1LCDx in one (0.2%). Within the harmonized cohort of 344 gliomas and glioneuronal tumors, the corresponding distributions were 220 (64.0%), 98 (28.5%), 25 (7.3%), and one (0.3%), respectively. Study flowchart and cohort composition of BRAF‐altered brain tumors in the nationwide C‐CAT dataset. Among 3714 brain tumor cases registered in the Japanese Center for Cancer Genomics and Advanced Therapeutics (C‐CAT) database, 413 tumors harbored BRAF alterations. BRAF alteration classes included BRAF p.V600E mutation, KIAA1549::BRAF fusion, BRAF amplification, and other BRAF alterations, including non‐p.V600E mutations and miscellaneous rearrangements/fusions. Several tumors demonstrated overlapping BRAF alteration classes. After operational harmonization and exclusion of non‐glioma or diagnostically insufficient cases, 344 tumors were categorized as gliomas and glioneuronal tumors. Subsequent grade‐stratified analyses focused on 118 BRAF p.V600E‐positive tumors categorized into Low‐grade, High‐grade, and Indeterminate groups. TABLE 1Clinicogenomic characteristics of BRAF alteration classes in the C‐CAT cohort.Group n Age, median (range)MaleFemaleTMB, median (IQR)LowIndeterminateHighGrade excluded CDKN2A loss CDKN2B loss TERTp hotspot TP53 ATRX PTEN EGFR NF1 PDGFRA MET All BRAF‐altered tumors41329 (0–86)1832301.26 (0.50–3.00)9314152154112/413 (27.1%)110/413 (26.6%)54/413 (13.1%)98/413 (23.7%)66/413 (16.0%)43/413 (10.4%)31/413 (7.5%)53/413 (12.8%)36/413 (8.7%)41/413 (9.9%) BRAF p.V600E20026.5 (0–80)791211.10 (0.50–2.10)3714678273/200 (36.5%)71/200 (35.5%)20/200 (10.0%)23/200 (11.5%)25/200 (12.5%)14/200 (7.0%)6/200 (3.0%)7/200 (3.5%)5/200 (2.5%)5/200 (2.5%) KIAA1549::BRAF fusion7410 (0–61)34400.00 (0.00–1.26)4406247/74 (9.5%)7/74 (9.5%)0/74 (0.0%)7/74 (9.5%)5/74 (6.8%)0/74 (0.0%)1/74 (1.4%)1/74 (1.4%)1/74 (1.4%)3/74 (4.1%) BRAF amplification2633.5 (2–80)12141.26 (0.00–4.00)4012106/26 (23.1%)6/26 (23.1%)5/26 (19.2%)13/26 (50.0%)5/26 (19.2%)4/26 (15.4%)6/26 (23.1%)5/26 (19.2%)4/26 (15.4%)15/26 (57.7%)Other BRAF alterations12746 (0–86)65623.78 (1.00–7.57)120704530/127 (23.6%)30/127 (23.6%)30/127 (23.6%)62/127 (48.8%)34/127 (26.8%)25/127 (19.7%)21/127 (16.5%)42/127 (33.1%)28/127 (22.0%)22/127 (17.3%) Clinicogenomic characteristics of BRAF alteration classes in the C‐CAT cohort. After operational harmonization and exclusion of non‐glioma or diagnostically insufficient cases, 344 tumors were classified as gliomas and glioneuronal tumors (Table 2). The overall genomic landscape demonstrated marked biological heterogeneity across BRAF alteration classes. Distinct age distributions were observed across major BRAF alteration classes within harmonized gliomas and glioneuronal tumors (Kruskal‐Wallis test, p = 1.10 × 10−23 Figure 2). KIAA1549::BRAF fusion‐positive tumors demonstrated marked pediatric predominance (median age, 10.5 years; interquartile range [IQR], 5–16.25 years). In contrast, the remaining alteration classes demonstrated broad and partially overlapping age distributions, including BRAF p.V600E‐positive tumors (median age, 25 years; IQR, 14–39.75 years), BRAF amplification‐positive tumors (median age, 51 years; IQR, 31.5–64.5 years), and tumors harboring other BRAF alterations (median age, 48 years; IQR, 38–64.5 years). To assess whether the younger age distribution of KIAA1549::BRAF fusion‐positive tumors was attributable solely to the predominance of PA, we performed an additional analysis excluding tumors with a harmonized diagnosis of PA from the fusion‐positive group. Even after exclusion of PA, KIAA1549::BRAF fusion‐positive tumors remained significantly younger than BRAF p.V600E‐positive tumors (median age, 14 vs. 25 years; Mann–Whitney U test, p = 0.036). TABLE 2Clinicogenomic characteristics of harmonized gliomas and glioneuronal tumors with BRAF alterations.Group n Age, median (range)MaleFemaleTMB, median (IQR)LowIndeterminateHighGrade excluded CDKN2A loss CDKN2B loss TERTp hotspot TP53 ATRX PTEN EGFR NF1 PDGFRA MET All glioma and glioneuronal tumors34430 (0–86)1541901.10 (0.50–3.00)93141528594/344 (27.3%)93/344 (27.0%)52/344 (15.1%)85/344 (24.7%)59/344 (17.2%)38/344 (11.0%)26/344 (7.6%)50/344 (14.5%)33/344 (9.6%)35/344 (10.2%) BRAF p.V600E16625 (0–80)641021.10 (0.50–2.00)3714674858/166 (34.9%)57/166 (34.3%)20/166 (12.0%)17/166 (10.2%)22/166 (13.3%)12/166 (7.2%)4/166 (2.4%)6/166 (3.6%)5/166 (3.0%)3/166 (1.8%) KIAA1549::BRAF fusion6010.5 (0–61)29310.00 (0.00–1.00)4406106/60 (10.0%)6/60 (10.0%)0/60 (0.0%)6/60 (10.0%)4/60 (6.7%)0/60 (0.0%)0/60 (0.0%)1/60 (1.7%)1/60 (1.7%)3/60 (5.0%) BRAF amplification only1941 (2–80)1091.13 (0.00–3.75)201165/19 (26.3%)5/19 (26.3%)4/19 (21.1%)11/19 (57.9%)5/19 (26.3%)4/19 (21.1%)5/19 (26.3%)5/19 (26.3%)3/19 (15.8%)13/19 (68.4%)Other BRAF alterations10948 (0–86)57524.00 (1.00–8.00)120702728/109 (25.7%)28/109 (25.7%)29/109 (26.6%)57/109 (52.3%)30/109 (27.5%)22/109 (20.2%)18/109 (16.5%)40/109 (36.7%)25/109 (22.9%)22/109 (20.2%) Clinicogenomic characteristics of harmonized gliomas and glioneuronal tumors with BRAF alterations. Age distributions across major BRAF alteration classes in harmonized gliomas and glioneuronal tumors. KIAA1549::BRAF fusion‐positive tumors demonstrated marked pediatric predominance, whereas BRAF amplification‐positive tumors and tumors harboring other BRAF alterations tended to occur in older patients. Horizontal bars indicate median age values. Distinct Genomic Contexts Across BRAF Alteration Classes The overall clinicogenomic landscape across major BRAF alteration classes in the cohort of 413 BRAF‐altered brain tumors is summarized in Figure 3. Among the harmonized cohort of 344 gliomas and glioneuronal tumors, 166 harbored BRAF p.V600E alterations. Of these, 118 tumors could be operationally categorized for detailed analyses, including 37 Low‐grade tumors, 67 High‐grade tumors, and 14 Indeterminate tumors (Table 3). Indeterminate tumors were predominantly composed of PXA‐spectrum neoplasms with overlapping low−/high‐grade biological features and insufficient integrated information for definitive operational categorization. Low‐grade tumors generally exhibited relatively quiet genomic profiles. In contrast, High‐grade tumors demonstrated accumulation of multiple co‐occurring genomic alterations, particularly involving CDKN2A/B, TERT promoter hotspot, TP53, and ATRX. CDKN2A alterations were identified in 8.1% of Low‐grade tumors, compared with 53.7% of High‐grade tumors and 71.4% of Indeterminate tumors (p = 1.96 × 10−6). Similar findings were observed for CDKN2B alterations (p = 4.33 × 10−6). Because GenMineTOP does not routinely report CDKN2A/B homozygous deletion, we performed a sensitivity analysis excluding tumors analyzed using this platform and restricting the analysis to copy‐number loss events. Among the remaining BRAF p.V600E‐positive tumors, CDKN2A copy‐number loss remained significantly more frequent in High‐grade than in Low‐grade tumors (35/49 [71.4%] vs. 3/19 [15.8%], Fisher's exact test, p = 7.03 × 10−5). Similarly, CDKN2B copy‐number loss was significantly enriched in High‐grade tumors (34/49 [69.4%] vs. 3/19 [15.8%], p = 9.20 × 10−5). TERT promoter hotspot alterations demonstrated a trend toward enrichment in High‐grade tumors (19.4%) compared with Low‐grade (5.4%) and Indeterminate tumors (14.3%), although the difference did not reach statistical significance (p = 0.150). TP53 and ATRX alterations were also significantly associated with higher‐grade groups. One tumor additionally harbored an IDH1 p.R132H mutation, further illustrating the molecular heterogeneity observed within BRAF‐altered gliomas and glioneuronal tumors. High‐grade tumors demonstrated significantly increased TMB compared with Low‐grade tumors (p = 0.00079). However, overall TMB values remained relatively modest across the cohort, and hypermutator phenotypes were rare. Only a single tumor demonstrated TMB ≥ 10 mutations/Mb (Figure 4; Table 3). Oncoplot of clinicogenomic alterations across major BRAF alteration classes in BRAF‐altered brain tumors. Comprehensive oncoplot summarizing major genomic alterations, harmonized diagnostic categories, and operational grade groups across 413 BRAF‐altered brain tumors in the nationwide C‐CAT cohort. Columns represent individual tumors and rows represent genomic alterations or clinicopathological annotations. BRAF alteration classes included BRAF p.V600E‐positive tumors, KIAA1549::BRAF fusion‐positive tumors, BRAF amplification‐positive tumors, and tumors harboring other BRAF alterations, with overlap in a subset of tumors. Genomic alteration types are distinguished by color as SNV/indel, amplification, homozygous deletion/loss, and rearrangement/fusion. Cells containing multiple alteration types within the same gene are subdivided accordingly. Gray cells indicate alterations not routinely reported by the assay used. Distinct distributions of harmonized diagnostic categories and major co‐alterations were observed across BRAF alteration classes. DLGG‐MAPK, diffuse low‐grade glioma with MAPK pathway alterations; DNT, dysembryoplastic neuroepithelial tumor; GG, ganglioglioma; Glioma, NOS, glioma, not otherwise specified; HGG, NOS, high‐grade glioma, not otherwise specified; indel, insertion/deletion; LGG, NOS, low‐grade glioma, not otherwise specified; N/A, not available; PA, pilocytic astrocytoma; pLGG, pediatric‐type low‐grade glioma; PLNTY, polymorphous low‐grade neuroepithelial tumor of the young; PMA, pilomyxoid astrocytoma; PXA, pleomorphic xanthoastrocytoma; SNV, single‐nucleotide variant. TABLE 3Clinicogenomic comparison across operational grade groups in BRAF p.V600E‐positive gliomas and glioneuronal tumors.FeatureLowIndeterminateHigh p Additional p n 371467TMB, median (IQR)0.65 (0.50–1.02)1.00 (1.00–1.10)1.26 (0.50–2.41)Kruskal‐Wallis: 0.002High vs. low mann–whitney: 7.92E‐04 CDKN2A loss3/37 (8.1%)10/14 (71.4%)36/67 (53.7%)1.96E‐06 CDKN2B loss3/37 (8.1%)10/14 (71.4%)34/67 (50.7%)4.33E‐06 TERTp hotspot2/37 (5.4%)2/14 (14.3%)13/67 (19.4%)0.15 TP53 0/37 (0.0%)2/14 (14.3%)12/67 (17.9%)0.025 ATRX 0/37 (0.0%)2/14 (14.3%)13/67 (19.4%)0.017 PTEN 1/37 (2.7%)1/14 (7.1%)7/67 (10.4%)0.362 EGFR 0/37 (0.0%)1/14 (7.1%)2/67 (3.0%)0.331 NF1 0/37 (0.0%)1/14 (7.1%)4/67 (6.0%)0.298 PDGFRA 0/37 (0.0%)1/14 (7.1%)4/67 (6.0%)0.298 MET 0/37 (0.0%)0/14 (0.0%)1/67 (1.5%)0.681 Clinicogenomic comparison across operational grade groups in BRAF p.V600E‐positive gliomas and glioneuronal tumors. Grade‐stratified genomic landscape of BRAF p.V600E‐positive gliomas and glioneuronal tumors. Oncoplot of 118 operationally categorized BRAF p.V600E‐positive gliomas and glioneuronal tumors stratified into Low‐grade, High‐grade, and Indeterminate groups. Columns represent individual tumors and rows represent genomic alterations or clinicopathological annotations. Low‐grade tumors generally demonstrated relatively quiet genomic profiles, whereas High‐grade tumors exhibited accumulation of co‐occurring genomic alterations involving CDKN2A/B, TERT promoter, TP53, and ATRX. Indeterminate tumors, predominantly composed of pleomorphic xanthoastrocytoma, demonstrated particularly frequent CDKN2A/B alterations. Genomic alteration types are distinguished by color as SNV/indel, amplification, homozygous deletion/loss, and rearrangement/fusion. Cells containing multiple alteration types within the same gene are subdivided accordingly. Gray cells indicate alterations not routinely reported by the assay used. Tumor mutational burden (TMB) values are shown in the upper annotation panel with a numerical y‐axis scale. DLGG‐MAPK, diffuse low‐grade glioma with MAPK pathway alterations; DNT, dysembryoplastic neuroepithelial tumor; GG, ganglioglioma; HGG, NOS, high‐grade glioma, not otherwise specified; LGG, NOS, low‐grade glioma, not otherwise specified; PA, pilocytic astrocytoma; LGG, pediatric‐type low‐grade glioma; PLNTY, polymorphous low‐grade neuroepithelial tumor of the young; PXA, pleomorphic xanthoastrocytoma; TMB, tumor mutational burden. To determine whether the co‐occurring alterations in high‐grade BRAF p.V600E‐positive tumors were specific to BRAF‐altered disease, we compared BRAF p.V600E‐positive (n = 20) and BRAF‐wild‐type (n = 510) tumors registered as GBM, IDH‐wildtype. Co‐occurring alterations were common in both groups, but their spectra differed. CDKN2A and CDKN2B alterations were more frequent in BRAF p.V600E‐positive GBMs (45.0% each) than in BRAF‐wild‐type GBMs (24.5% and 20.4%), reaching significance for CDKN2B (p = 0.021). EGFR alterations were less frequent in BRAF p.V600E‐positive GBMs (5.0% vs. 38.8%, p = 0.0015), and TP53 alterations tended to be less frequent in BRAF p.V600E‐positive GBMs than in BRAF‐wild‐type GBMs (15.0% vs. 37.3%, p = 0.056). Thus, co‐occurring alterations are common in GBM irrespective of BRAF status, whereas their spectrum differs by BRAF p.V600E status (Table 4). TABLE 4Comparison of secondary genomic alterations between BRAF p.V600E‐positive and BRAF‐wild‐type glioblastoma, IDH‐wildtype.Feature BRAF p.V600E‐positive (n = 20) BRAF‐wild‐type (n = 510) p CDKN2A HD9/20 (45.0%)125/510 (24.5%)0.0620 CDKN2B HD9/20 (45.0%)104/510 (20.4%)0.0210 TERTp hotspot8/20 (40.0%)124/510 (24.3%)0.1190 TP53 3/20 (15.0%)190/510 (37.3%)0.0560 ATRX 1/20 (5.0%)45/510 (8.8%)1.0000 PTEN 4/20 (20.0%)176/510 (34.5%)0.2320 EGFR 1/20 (5.0%)198/510 (38.8%)0.0015 NF1 1/20 (5.0%)110/510 (21.6%)0.0920 PDGFRA 2/20 (10.0%)104/510 (20.4%)0.3930 MET 1/20 (5.0%)30/510 (5.9%)1.0000 Comparison of secondary genomic alterations between BRAF p.V600E‐positive and BRAF‐wild‐type glioblastoma, IDH‐wildtype. KIAA1549::BRAF fusion‐positive tumors were predominantly associated with low‐grade glioma contexts, particularly PA and pediatric low‐grade glioma tumors, although the alteration was not entirely restricted to canonical PA. Among the harmonized cohort of 344 gliomas and glioneuronal tumors, 15 tumors met the definition of BRAF amplification‐only tumors. In contrast, amplification‐only tumors demonstrated a distinct genomic profile enriched for receptor tyrosine kinase (RTK)‐associated and copy‐number‐driven alterations. Notably, MET alterations were identified in 10 of 15 amplification‐only tumors (66.7%), compared with 25 of 329 non‐amplification‐only BRAF‐altered tumors (7.6%) (Fisher's exact test, p = 6.91 × 10−8). All MET alterations in amplification‐only tumors represented MET amplification. EGFR alterations were also significantly enriched in amplification‐only tumors (5/15, 33.3%) compared with non‐amplification‐only tumors (21/329, 6.4%; p = 0.00298), as were TP53 alterations (9/15, 60.0% vs. 76/329, 23.1%; p = 0.00313). PTEN alterations were also more frequent in amplification‐only tumors (4/15, 26.7% vs. 34/329, 10.3%), although this difference did not reach statistical significance (p = 0.0707). Collectively, these findings support the concept that BRAF amplification may arise within an RTK‐amplified, copy‐number–rich genomic context rather than representing a canonical MAPK‐driven glioma subgroup (Figure 5). Excluding grade‐unclassifiable tumors, 11 of 13 grade‐evaluable BRAF amplification‐only tumors (84.6%) were categorized as high‐grade tumors. Among tumors harboring concurrent MET amplification, 8 of 9 evaluable cases (88.9%) were classified as high‐grade tumors. Genomic landscape of BRAF amplification‐only gliomas and glioneuronal tumors. Oncoplot summarizing genomic alterations in 15 BRAF amplification‐only gliomas and glioneuronal tumors. Columns represent individual tumors and rows represent genomic alterations or clinicopathological annotations. Amplification‐only tumors demonstrated frequent receptor tyrosine kinase‐associated and copy‐number–driven alterations, particularly MET amplification, together with frequent co‐alterations involving EGFR, PTEN, and TP53. These findings support the concept that BRAF amplification may arise within an RTK‐amplified, copy‐number–rich genomic context distinct from canonical MAPK‐driven glioma subgroups. Genomic alteration types are distinguished by color as SNV/indel, amplification, homozygous deletion/loss, and rearrangement/fusion. Cells containing multiple alteration types within the same gene are subdivided accordingly. Gray cells indicate alterations not routinely reported by the assay used. Tumor mutational burden (TMB) values are shown in the upper annotation panel with a numerical y‐axis scale; an axis break is used to display the single tumor with an exceptionally high TMB value while preserving visualization of the remaining tumors. Glioma, NOS, glioma, not otherwise specified; HGG, NOS, high‐grade glioma, not otherwise specified; N/A, not available; PA, pilocytic astrocytoma; PMA, pilomyxoid astrocytoma; TMB, tumor mutational burden. Detailed characterization of the other BRAF alteration group (n = 109) identified 92 tumors harboring non‐p.V600E SNV/indels and 18 harboring non‐KIAA1549::BRAF rearrangements or SVs, including one tumor harboring both alteration types. Three tumors additionally harbored BRAF amplification. Among the 18 rearrangement/SV‐positive tumors, 10 harbored partner‐resolved events, seven had BRAF rearrangements for which the partner was not specified in the CGP annotation, and one harbored a BRAF duplication. Partner‐resolved events included TRIM24::BRAF in two tumors, both classified as High‐grade glioma, NOS, and single cases of BRAF::WDFY2, PLEKHA8::BRAF, CTTNBP2::BRAF, OSBPL11::BRAF, TTYH3::BRAF, BCAS1::BRAF, GIT2::BRAF, and DENND2A::BRAF rearrangement. Thus, the non‐KIAA1549 rearrangement spectrum was highly heterogeneous, with TRIM24::BRAF representing the only recurrent partner‐resolved event. Discussion In this nationwide real‐world analysis of BRAF‐altered gliomas and glioneuronal tumors, we identified distinct clinicogenomic contexts across BRAF alteration classes and grade‐associated genomic patterns within BRAF p.V600E‐positive tumors. Rather than forming a uniform biological group, the major BRAF alteration classes were associated with different genomic and clinicopathological contexts. Age distributions also differed across BRAF alteration classes, driven largely by the pediatric predominance of KIAA1549::BRAF fusion‐positive tumors, whereas the remaining classes showed broader overlap. KIAA1549::BRAF fusion‐positive tumors were predominantly associated with low‐grade contexts, particularly PA, consistent with previous reports [16]. Their younger age distribution was therefore partly attributable to PA. However, fusion‐positive tumors remained significantly younger than BRAF p.V600E‐positive tumors after exclusion of PA, although residual confounding by histology cannot be excluded because centralized pathological review and comprehensive molecular classification were unavailable for all C‐CAT cases. Most prior studies focused on canonical KIAA1549::BRAF fusion‐positive pediatric low‐grade gliomas and BRAF p.V600E‐positive gliomas [6, 7], leaving less common alterations such as BRAF amplification and non‐p.V600E variants relatively underexplored. Ali et al. [17] similarly reported clinicogenomic heterogeneity among MAPK pathway‐altered gliomas, with BRAF p.V600E across diverse ages and diagnoses, KIAA1549::BRAF fusion restricted to cerebellar pediatric low‐grade gliomas, and non‐p.V600E variants or BRAF amplification in high‐grade tumors with additional alterations. These observations are broadly consistent with our findings, which extend them in a nationwide cohort. Detailed analysis of the other BRAF alteration group revealed diverse noncanonical rearrangements. Previously reported partners included GIT2::BRAF in PA [18], CTTNBP2::BRAF in pediatric low‐grade glioma [19], TTYH3::BRAF in glioblastoma [20], BCAS1::BRAF in pediatric brain tumors [21], and DENND2A::BRAF among rare CNS BRAF fusions [22]. TRIM24::BRAF recurred in two HGG, NOS cases; although reported in extracranial malignancies and shown to activate BRAF signaling, it has not, to our knowledge, been described in glioma [23]. BRAF::WDFY2 and PLEKHA8::BRAF may likewise represent very rare or underrecognized events. OSBPL11::BRAF has been reported in CNS Rosai‐Dorfman disease but not glioma [24]. Because these events were identified from clinical CGP annotations without systematic orthogonal validation, their novelty and functional significance require caution. Amplification‐only tumors showed a distinct RTK‐associated, copy‐number‐rich profile, particularly MET amplification. MET alterations occurred in more than two‐thirds of amplification‐only tumors and all represented amplification; EGFR and TP53 were also enriched, whereas PTEN showed a nonsignificant trend. Our previous institutional cohort likewise showed concurrent MET amplification in all evaluable BRAF amplification‐positive tumors [10]. Together, these findings suggest that BRAF amplification may arise within a broader RTK‐amplified context rather than as an isolated MAPK alteration. Most grade‐evaluable BRAF amplification‐only tumors were high‐grade, especially those with concurrent MET amplification, paralleling our institutional series [10]. Although BRAF amplification is not a definitive marker of high‐grade biology, it may preferentially occur in aggressive, copy‐number‐complex glioma contexts. MET co‐amplification may also be therapeutically relevant: recent clinical experience with tepotinib showed responses in selected MET‐altered high‐grade gliomas, including MET‐amplified tumors [25], and EANO guidelines support considering MET‐targeted therapy in clinical trials for MET amplification, fusion, or exon‐skipping alterations [26]. However, the predictive value of MET amplification itself remains to be established. Within BRAF p.V600E‐positive gliomas and glioneuronal tumors, low‐grade tumors generally had quiet genomic profiles, whereas high‐grade tumors accumulated co‐occurring alterations involving CDKN2A/B, TERT, TP53, and ATRX. However, our within‐tumor‐type comparison showed that such alterations were also common in BRAF‐wild‐type GBMs; the spectrum differed, with more frequent CDKN2A/B and fewer EGFR alterations in BRAF p.V600E‐positive GBMs. Thus, high‐grade BRAF p.V600E‐positive tumors may represent a distinct genomic context rather than simply a greater burden of alterations. Because the C‐CAT dataset is cross‐sectional, the temporal order of BRAF p.V600E and accompanying alterations cannot be established; CDKN2A/B, TERT, TP53, and ATRX should therefore be regarded as cooperating rather than necessarily secondary events. Longitudinal paired sampling and clonality analyses would be required to distinguish initiating from progression‐associated events. Resistance to RAF inhibition in BRAF p.V600E‐mutant gliomas is likewise heterogeneous and can involve MAPK reactivation, MAP2K1 alterations, loss of NF1, PTEN, or CBL, and increased EGFR signaling [27]. Our data cannot determine whether the co‐alterations observed here contribute to intrinsic or acquired resistance because paired pre/post‐treatment samples and detailed treatment histories were unavailable. CDKN2A/B alterations were particularly frequent in the Indeterminate group, largely PXA‐spectrum tumors, consistent with the established association between CDKN2A/B homozygous deletion and PXA [9]. In contrast, TERT promoter hotspot alterations tended to be enriched in high‐grade tumors [28], while TP53 and ATRX also accumulated in higher‐grade groups. These findings are compatible with partially distinct CDKN2A/B‐rich circumscribed/PXA‐like [9, 29] and TERT/TP53/ATRX‐associated diffuse high‐grade‐like trajectories, although substantial overlap likely exists. TMB was higher in high‐grade tumors but generally remained modest, suggesting that aggressive behavior may relate more to structural and tumor suppressor alterations than to hypermutation. Nationwide real‐world CGP datasets provide access to rare molecular subsets but introduce diagnostic heterogeneity from institutional practice, evolving WHO criteria, historical terminology, and incomplete molecular information. We therefore used biologically informed operational harmonization rather than rigid reclassification, avoiding reassignment of entities requiring essential molecular criteria without adequate support [30]. Because BRAF alterations were generally identified by CGP, most tumors were unlikely to have received molecularly guided BRAF‐targeted therapy before testing; therefore, the observed patterns may partly reflect their natural clinicogenomic architecture. These considerations illustrate both the value and limitations of nationwide CGP datasets. Several limitations should be acknowledged. First, centralized pathological review and comprehensive DNA methylation profiling were not uniformly available. Second, CGP platforms differed in gene coverage and structural‐variant detection; GenMineTOP uses broader RNA‐based fusion detection, whereas F1CDx and NOP assess more restricted rearrangement targets. Platform‐specific reporting also affects CDKN2A/B homozygous deletion, which GenMineTOP does not routinely report. Nevertheless, enrichment of CDKN2A/B loss in high‐grade BRAF p.V600E‐positive tumors persisted after excluding GenMineTOP and restricting the analysis to copy‐number loss events. Third, operational grade harmonization introduced uncertainty, particularly for PXA. Despite these limitations, this nationwide cohort delineates distinct BRAF alteration contexts and supports the utility of real‐world CGP data for studying rare CNS tumor subsets. In summary, the graphical summary of this study is shown in Figure 6. Graphical table of contents. Nationwide real‐world profiling of BRAF‐altered brain tumors identified distinct clinicogenomic contexts across four major BRAF alteration classes in harmonized gliomas and glioneuronal tumors. BRAF p.V600E‐positive tumors showed grade‐associated genomic differences; KIAA1549::BRAF fusion‐positive tumors were associated with younger age and predominantly low‐grade histology; BRAF amplification‐only tumors were characterized by frequent MET amplification and a copy‐number‐rich high‐grade context; and other BRAF alterations formed a heterogeneous group. Author Contributions Seiji Yamada: writing – original draft, writing – review and editing, conceptualization, methodology, data curation, investigation, formal analysis, visualization, project administration, validation. Tetsuya Takimoto: software, supervision, formal analysis, data curation. Sayaka Ueno: supervision. Sachiko Minamiguchi: resources. Shigeo Ohba: supervision, resources. Yuichi Hirose: supervision, resources. Hideyuki Saya: supervision. Tamotsu Sudo: supervision. Funding The authors have nothing to report. Ethics Statement Approval of the research protocol: This study was approved by the institutional review board of Fujita Health University (approval No. HM26‐021) and was conducted in accordance with the Declaration of Helsinki. Data were obtained through the C‐CAT data utilization portal under the approved research project number CDU2026‐009N. Informed consent: This study used de‐identified data from the C‐CAT database that had been registered with patient consent for information provision to C‐CAT and secondary use. The requirement for additional informed consent for this retrospective database study was waived by the institutional review board. Conflicts of Interest Hideyuki Saya is an Editorial Board Member of Cancer Science. The other authors have no conflicts of interest to declare. Acknowledgments The authors thank Takuki Sawada and Miki Shimmei of the Department of Genomic Medicine, Fujita Health University, for their administrative support and assistance with C‐CAT data retrieval and management. Data Availability Statement The data that support the findings of this study are available from C‐CAT. Restrictions apply to the availability of these data. Access is subject to C‐CAT approval and applicable data‐use procedures. References Associated Data Data Availability Statement The data that support the findings of this study are available from C‐CAT. Restrictions apply to the availability of these data. Access is subject to C‐CAT approval and applicable data‐use procedures.