Sievers et al. 
Rosette-forming glioneuronal tumors share a distinct DNA methylation profile and 
mutations in FGFR1, with recurrent co-mutation of PIK3CA and NF1 
Philipp Sievers1,2, Romain Appay3,4, Daniel Schrimpf1,2, Damian Stichel1,2, David E. Reuss1,2, 
Annika K. Wefers1,2, Annekathrin Reinhardt1,2, Roland Coras5, Viktoria Ruf6, Karin de 
Stricker7, Henning Boldt8, Bjarne Winther Kristensen8,9, Jeanette Krogh Petersen8, Benedicte 
P. Ulhøi10, Maria Gardberg11, Eleonora Aronica12,13, Martin Hasselblatt14, Wolfgang Brück15, 
Wolfgang Wick16,17, Christel Herold‑Mende18, Daniel Hänggi19, Sebastian Brandner20,21, 
Felice Giangaspero22,23, David Capper24,25,26, Elisabeth Rushing27, Pieter Wesseling28,29, 
Stefan M. Pfister30,31,32, Dominique Figarella-Branger3,4, Andreas von Deimling1,2, Felix 
Sahm1,2,30*, David T. W. Jones30,33* 
 
1Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany 
2Clinical Cooperation Unit Neuropathology, German Consortium for Translational Cancer Research (DKTK), 
German Cancer Research Center (DKFZ), Heidelberg, Germany 
3APHM, Hôpital de la Timone, Service d'Anatomie Pathologique et de Neuropathologie, Marseille, France 
4Aix-Marseille Univ, CNRS, INP, Inst Neurophysiopathol, Marseille, France 
5Department of Neuropathology, University of Erlangen-Nürnberg, Erlangen, Germany 
6Department of Neuropathology, Ludwig-Maximilian University, Munich, Germany 
7Department of Pathology, Rigshospitalet, Copenhagen, Denmark 
8Department of Pathology, Odense University Hospital, Odense, Denmark 
9Department of Clinical Research, University of Southern Denmark, Odense, Denmark 
10Department of Pathology, Aarhus University Hospital, Aarhus, Denmark 
11Department of Pathology University of Turku and Turku University Hospital, Turku, Finland 
12Department of (Neuro)Pathology, Amsterdam UMC, University of Amsterdam, Amsterdam, The Netherlands 
13Stichting Epilepsie Instellingen Nederland (SEIN), Zwolle, The Netherlands 
14Institute of Neuropathology, University Hospital Münster, Münster, Germany 
15Institute of Neuropathology, University Medical Center, Göttingen, Germany 
16Clinical Cooperation Unit Neurooncology, German Consortium for Translational Cancer Research (DKTK), 
German Cancer Research Center (DKFZ), Heidelberg, Germany 
17Department of Neurology and Neurooncology Program, National Center for Tumor Diseases, Heidelberg 
University Hospital, Heidelberg, Germany 
18Division of Experimental Neurosurgery, Department of Neurosurgery, University Hospital Heidelberg, 
Heidelberg, Germany 
19Department of Neurosurgery, University Medical Centre Mannheim, University of Heidelberg, Mannheim, 
Germany 
20Division of Neuropathology, National Hospital for Neurology and Neurosurgery, University College London 
Hospitals NHS Foundation Trust, Queen Square, London, UK 
21Department of Neurodegenerative Disease, UCL Queen Square Institute of Neurology, Queen Square, London, 
UK 
22Department of Radiological Sciences, Oncology and Anatomical Pathology, Sapienza University Rome, Rome, 
Italy 
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
23IRCCS Neuromed, Pozzilli, Italy 
24Charité-Universitätsmedizin Berlin, Freie Universität Berlin, Humboldt-Universität zu Berlin, Berlin, Germany 
25Institute of Neuropathology, Berlin Institute of Health, Berlin, Germany 
26German Cancer Consortium (DKTK), Partner Site Berlin, Berlin, Germany 
27Institute of Neuropathology, University Hospital Zurich, Zurich, Switzerland 
28Department of Pathology, VU University Medical Center/Brain Tumor Center Amsterdam, Amsterdam, The 
Netherlands 
29Department of Pathology, Princess Máxima Center for Pediatric Oncology and University Medical Center 
Utrecht, Utrecht, The Netherlands 
30Hopp Children’s Cancer Center at the NCT Heidelberg (KiTZ), Heidelberg, Germany 
31Division of Pediatric Neurooncology, German Cancer Consortium (DKTK), German Cancer Research Center 
(DKFZ), Heidelberg, Germany 
32Department of Pediatric Oncology, Hematology, Immunology and Pulmonology, University Hospital Heidelberg, 
Heidelberg, Germany 
33Pediatric Glioma Research Group, German Consortium for Translational Cancer Research (DKTK), German 
Cancer Research Center (DKFZ), Heidelberg, Germany 
 
* These authors share senior authorship 
 
 
Corresponding authors: 
Felix Sahm 
felix.sahm@med.uni-heidelberg.de  
 
David T. W. Jones 
david.jones@kitz-heidelberg.de 
  
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
Abstract 
Rosette-forming glioneuronal tumor (RGNT) is a rare brain neoplasm that primarily affects 
young adults. Although alterations affecting the mitogen-activated protein kinase (MAPK) and 
phosphoinositide 3-kinase (PI3K) signaling pathway have been associated with this low-
grade glioneuronal entity, comprehensive molecular investigations of RGNT in larger series 
have not been performed to date, and an integrated view of their genetic and epigenetic 
profiles is still lacking. Here we describe a genome-wide DNA methylation and targeted 
sequencing-based characterization of a molecularly distinct class of tumors (n = 30), initially 
identified through genome-wide DNA methylation screening amongst a cohort of > 30,000 
tumors, of which most were diagnosed histologically as RGNT. FGFR1 hotspot mutations 
were observed in all tumors analyzed, with co-occurrence of PIK3CA mutations in about two 
thirds of the cases (63%). Additional loss-of-function mutations in the tumor suppressor gene 
NF1 were detected in a subset of cases (33%). Notably, in contrast to most other low-grade 
gliomas, these tumors often displayed co-occurrence of two or even all three of these 
mutations. Our data highlight that molecularly defined RGNTs are characterized by combined 
genetic alterations affecting both MAPK and PI3K signaling pathways. Thus, these two 
pathways appear to synergistically interact in the formation of RGNT, and offer potential 
therapeutic targets for this disease.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Keywords: Rosette-forming glioneuronal tumor, RGNT, Brain tumor, DNA methylation 
profile, Molecular classification, MAPK, PI3K, FGFR1, PIK3CA, NF1  
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
Introduction 
Rosette-forming glioneuronal tumor (RGNT) is an uncommon central nervous system (CNS) 
neoplasm that primarily affects young adults [17]. It typically arises in the midline, usually 
occupying a substantial fraction of the fourth ventricle. In the World Health Organization 
(WHO) classification of brain tumors 2007, RGNT was therefore per definition associated 
with the fourth ventricle. However, more recent reports have shown that it can also affect 
other sites [1,2,17,25,29]. Thus, the extension “of the fourth ventricle” was abandoned in the 
current update of the classification [17]. 
Histologically, RGNT is characterized by a biphasic histologic architecture consisting of well-
differentiated neurocytic cells forming rosettes or perivascular pseudorosettes and a glial 
component resembling pilocytic astrocytoma [9,14,19,23]. While large-scale genomic and 
epigenomic analyses over the past decade have immensely contributed to our understanding 
of molecular mechanisms underlying many primary brain tumors, current knowledge on the 
molecular background of RGNT is based mainly on individual case reports or small series. 
Targeted molecular analyses revealed absence of KIAA1549-BRAF fusions or activating 
BRAF mutations [6] which are a hallmark of pilocytic astrocytoma [10,11] and can also be 
found in other low-grade glial and glioneuronal tumors [30]. In contrast, activating mutations 
in FGFR1 [7,16] and/or PIK3CA [3,5,16,28] have been described in a subset of RGNTs.  
We here performed a combined (epi)genomic analysis of genome-wide DNA methylation 
profiling and targeted next-generation DNA sequencing data from 30 tumors in order to 
evaluate the underlying molecular background and to identify new diagnostic biomarkers as 
well as potentially targetable alterations.  
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
Materials and methods 
Study population and sample collection 
Tumor samples and retrospectively determined clinical data from 30 patients were obtained 
from multiple international collaborating centers and collected at the Department of 
Neuropathology of the University Hospital Heidelberg (Heidelberg, Germany). Case selection 
was based on unsupervised hierarchical clustering of genome-wide DNA methylation data in 
a cohort of > 30,000 tumors that revealed a molecularly distinct group of tumors comprising 
30 samples, of which the majority was diagnosed histologically as RGNT. Tissue was 
available for 24 cases thereof. Additionally, DNA methylation data of numerous well 
characterized reference samples representing CNS tumors of known histological and/or 
molecular subtype were used for comparative analyses [4]. Detailed descriptions of the 
reference methylation classes are outlined under https://www.molecularneuropathology.org. 
Tissue sample collection and processing, data collection, and use were performed in 
accordance with local ethics regulations and approvals. Clinical patient details are listed in 
Fig. 2 and Supplementary Table 1. 
 
Histology and immunohistochemistry 
All samples with available tissue (n = 24/30) were histopathologically re-assessed according 
to the WHO 2016 classification of tumors of the central nervous system. Formalin-fixed, 
paraffin-embedded (FFPE) tissue samples were stained with hematoxylin and eosin (H&E) 
according to standard protocols. For all cases with sufficient material, immunohistochemistry 
was performed on a Ventana BenchMark ULTRA Immunostainer using either the OptiView 
DAB IHC Detection Kit or the ultraView Universal DAB Detection Kit (Ventana Medical 
Systems, Tucson, Arizona, USA). Antibodies were directed against: glial fibrillary acid protein 
(GFAP; Z0334, rabbit polyclonal, 1:1000 dilution, Dako Agilent, Santa Clara, CA, USA), Olig2 
(clone EPR2673, rabbit monoclonal, 1:100 dilution, Abcam, Cambridge, UK), Synaptophysin 
(clone MRQ-40, rabbit monoclonal, 1:160 dilution, Cell Marque Corp., Rocklin, CA, USA), 
NeuN (clone A60, mouse monoclonal, 1:100 dilution, Millipore, Burlington, MA, USA), CD34 
(clone QBEnd/10, mouse monoclonal, Ventana Medical Systems), Ki-67 (clone MIB-1, 
mouse monoclonal, 1:100 dilution, Dako Agilent). 
 
DNA extraction  
Representative tumor tissue with highest available tumor cell content (~50-90%) was 
histologically identified and selected for nucleic acid extraction. Genomic DNA was extracted 
from fresh-frozen or formalin-fixed and paraffin-embedded (FFPE) tissue samples using the 
automated Maxwell system with the Maxwell 16 Tissue DNA Purification Kit or the Maxwell 
16 FFPE Plus LEV DNA Purification Kit (Promega, Madison, WI, USA), according to the 
manufacturer’s instructions. 
 
DNA methylation array processing and copy number profiling 
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
The Infinium HumanMethylation450 (450k) BeadChip or Infinium MethylationEPIC (850k) 
BeadChip array (Illumina, San Diego, CA, USA) were used to obtain genome-wide DNA 
methylation profiles of tumor samples according to the manufacturer’s instructions at the 
Genomics and Proteomics Core Facility of the German Cancer Research Center (DKFZ; 
Heidelberg, Germany). DNA methylation data were generated from both fresh-frozen and 
FFPE tissue samples. On-chip quality metrics of all samples were carefully controlled. 
Processing of DNA methylation data was performed with custom approaches as previously 
described [8,27]. Copy number profiles were generated using the ‘conumee’ package for the 
“R” environment (http://bioconductor.org/packages/release/bioc/html/conumee.html). All 
samples were checked for duplicates by pairwise correlation of the genotyping probes on the 
450k/850k array. 
 
Targeted next-generation DNA sequencing and mutational analysis 
Targeted exon capture and next-generation sequencing covering the coding regions of 130 
genes of particular relevance in brain tumors was performed on a NextSeq 500 sequencer 
(Illumina) as previously described [24] for all tumor samples (n = 30). Fusion discovery was 
done based on panel sequencing data using deFuse [18] and arriba 
(https://github.com/suhrig/arriba/). 
 
Statistical analysis 
DNA methylation array data were processed with the R/Bioconductor package minfi (version 
1.20). For unsupervised hierarchical clustering of samples, the 20,000 most variably 
methylated probes by median absolute deviation across the dataset were selected. Samples 
were hierarchically clustered using Euclidean distance and Ward’s linkage method. DNA 
methylation probes were reordered using Euclidian distance and complete linkage. For 
unsupervised 2D representation of pairwise sample correlations, dimensionality reduction by 
t-distributed stochastic neighbor embedding (t-SNE) was performed using the 20,000 most 
variable CpG sites according to standard deviation, a perplexity value of 10 and 3000 
iterations. Survival data were analyzed by Kaplan-Meier analysis and compared by log-rank 
test using GraphPad Prism 8 (GraphPad Software, La Jolla, CA, USA). P values < 0.05 were 
considered significant.  
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
Results 
DNA methylation profiling highlights a distinct epigenetic signature of RGNT 
Based on an unsupervised analysis of genome-wide DNA methylation data in a cohort of 
>30,000 tumors, we identified a molecularly distinct group of tumors forming a cluster 
separate from other established entities. The majority of these were diagnosed histologically 
as RGNT (Supplementary Table 1 and Fig. 2). A subsequent focused unsupervised 
hierarchical clustering and t-distributed stochastic neighbor embedding (t-SNE) analysis of 
DNA methylation patterns of these cases, alongside 106 well-characterized CNS neoplasms 
encompassing other low-grade glial/glioneuronal tumor entities and control tissue (white 
matter), consistently confirmed the distinct nature of this class (Fig. 1). This unique pattern 
supports the consideration of RGNT as a distinct molecular tumor type. Analysis of copy 
number variations (CNVs) calculated from the DNA methylation array data revealed a flat 
(=balanced) profile in most of the cases. Only single cases showed a small number of copy 
number alterations, with no obvious recurrent patterns. 
 
RGNTs are characterized by alterations affecting MAPK and PI3K signaling pathways 
To investigate the mutational landscape in RGNT, we performed targeted next-generation 
sequencing on genomic DNA isolated from 30 tumors (Fig. 2 and Supplementary Table 1). In 
all analyzed tumors, a missense mutation within the kinase domain of FGFR1 was identified 
– resulting in either a p.N546K (n = 22) or p.K656E (n = 8) substitution (details are listed in 
Supplementary Table 2). Both result in activating alterations within the kinase domain of 
FGFR1 and were previously reported in other glioneuronal tumors [21]. Additionally, 19 of 30 
(63%) of the FGFR1-mutant tumors harbored a concomitant mutation in PIK3CA which acts 
as an integral part of the PI3K pathway, including 14 with a p.H1047R, two with a p.E545K, 
and one each with a p.E542K, p.H1047L or p.G1049R substitution. The mutant allele 
frequency for the FGFR1 and PIK3CA variants ranged from 23 to 58% / 23 to 48%, 
consistent with a heterozygous somatic mutation on the background of ~50-90% tumor purity 
(Supplementary Table 2). 
Besides activating FGFR1 and PIK3CA mutations, missense or damaging mutations in NF1 
were identified in 10 of the cases. Allele frequencies (13-55%) were consistent with being 
heterozygous somatic variants. No indication for a loss of heterozygosity was found by copy 
number analysis. Notably, seven of the tumors (23%) had a combined triple alteration of 
FGFR1, PIK3CA and NF1 - indicating a cooperative role in tumorigenesis.  
Two of the FGFR1-mutant tumors also harbored a PTPN11 variant, typically altered in 
patients with Noonan syndrome. The PTPN11 variants could not be verified as being in the 
germline because analysis was performed without matched normal tissue sample. In one of 
the cases, however, a Noonan syndrome was clinically known. 
 
Clinical characteristics and morphological features within the molecularly defined 
RGNT cohort 
All cases were located infratentorially, preferentially in the posterior fossa occupying the 4th 
ventricle and the cerebellum. However, in line with other studies, the molecularly-defined 
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
RGNTs described here also arose in mesencephalic or diencephalic regions [1,2,17,25,29]. 
Median age of the patients at the time of diagnosis was 32 years (range 10–69) and the sex 
distribution was balanced (male:female ratio 1.0). Basic clinical characteristics of the cases 
are summarized in Figure 2 and Supplementary Table 1. Outcome data were available for 
only ten patients. Analysis of overall survival (OS) of RGNT patients in comparison to 
reference glioma groups supports classification of the molecularly-rendered RGNT as WHO 
grade I, in line with the so far histologically-defined entity (Supplementary Fig. 1). 
Histologically, all tumors were characterized by a moderate cellularity of neuroepithelial 
tumor cells. While a glial component consisting of spindle to stellate-shaped astrocytic cells 
with oval or elongated nuclei in a dense fibrillary background were seen in all tumors, 
populations of uniform neurocytic cells arranged in rosettes and perivascular pseudorosettes 
could be observed in only 20 of the 24 cases (Fig. 3). Neurocytic cells typically showed round 
nuclei with fine stippled chromatin dispersed in a mucoid or fibrillary matrix. In seven of the 
cases an oligodendroglioma-like cytology was seen focally. Some of the tumors 
demonstrated eosinophilic granular bodies (n = 9) or Rosenthal fibers (n = 8). Calcification 
were seen in a small number of tumors (n = 5). Hyalinized vessels and focal reactive 
vascular proliferation were observed in nine cases. Necrosis was uniformly absent. Mitotic 
activity was very low to absent. Immunoreactivity for synaptophysin was present in the 
neurocytic component (Fig. 3), whereas the glial component exhibited strong positivity for 
GFAP. The tumor cells were Olig2 positive and NeuN and CD34 negative. Proliferation index 
(Ki67) ranged from 1 to 3% in 22 of the cases, only two of the cases showed a slightly 
elevated proliferation index of up to 7 or 10%. Notably, all cases were histologically 
compatible with the diagnosis RGNT, although some cases would more likely have been 
favored as a typical differential diagnosis of RGNT. 
  
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
Discussion 
Using genome-wide DNA methylation profiling, we have confirmed a highly distinct 
epigenetic signature of RGNT and shown that tumors within this group share recurrent 
alterations within the MAPK and PI3K/AKT/mTOR signaling pathways.  
As alterations within the kinase domain of FGFR1 were detected in all tumors analyzed, it 
seems that RGNTs are primarily driven by mutations leading to constitutive activation of 
FGFR signaling. Genetic alterations within the FGFR signaling pathway are also common in 
other low-grade glial and glioneuronal tumors, with missense mutations in FGFR1, internal 
tandem duplication of the kinase domain, and FGFR1-TACC1 fusions being observed 
[10,20,21,26,30]. In contrast to those other tumor types, however, RGNT showed further 
recurrent mutations in the oncogene PIK3CA, which acts as an integral part of the PI3K 
pathway, and in the NF1 tumor suppressor. Although these genes are each mutated in 
several other primary brain tumors, concomitant mutations (in such a high frequency as 
observed here) seem to be extremely uncommon, particularly in low-grade glioneuronal 
tumors. However, combined activation of MAPK and PI3K/AKT/mTOR signaling pathways 
has been previously reported in single cases of RGNT [7,16] and therefore appears to be 
highly characteristic of this molecular class. While FGFR alterations are also common in 
dysembryoplastic neuroepithelial tumor, another low-grade glioneuronal tumor with 
histological overlap to RGNT, no mutations of PIK3CA and only one concomitant NF1 
mutation were found in 32 analyzed cases. 
Although there is no certain association with hereditary syndromes, RGNT has been 
described in patients with neurofibromatosis type 1 [13], as well as Noonan syndrome 
[12,16]. None of the NF1 mutations in our cohort could be confirmed as being present in the 
germline, although only very few cases had germline material available. Two of the FGFR1-
mutant tumors also displayed a PTPN11 variant, typically altered in patients with Noonan 
syndrome. Although matched normal tissue samples were unfortunately not available for 
these cases, in one of them a Noonan syndrome had already been clinically diagnosed. 
In comparison to the vast majority of other low-grade glial and glioneuronal tumors that have 
alterations exclusively within a single molecular pathway [20,30], RGNT seem to be a 
multiple-pathway disease, a fact that is more characteristic for high-grade tumors. 
Furthermore, PI3K/AKT/mTOR pathway alterations have been associated with clinical and 
histologic aggressiveness in LGG [22] as well as poor outcome in patients with RGNT [7]. 
However, clinical follow-up data of our series appear in line with the current WHO grade I 
designation.  
Our findings also have important implications for treatment management of patients with 
subtotally resected or recurrent tumors. Highly potent inhibitors of FGFR1, several of which 
are currently in various phases of clinical development [15], may be valuable treatments for 
these patients. Similarly, mutations in PIK3CA and NF1 might also be targeted with specific 
inhibitors of the PI3K and MAPK pathways. However, this will need confirmation in clinical 
trials. 
In conclusion, our findings provide new insight into the molecular genetic background of 
RGNT and suggest that RGNTs are highly linked to combined MAPK and PI3K/AKT/mTOR 
signaling pathway activation via concomitant mutations in FGFR1 and PIK3CA, thereby 
offering options for targeted therapies. 
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
 
 
 
Acknowledgements  
We thank V. Zeller, U. Lass and J. Meyer for excellent technical support and the microarray unit of the DKFZ 
Genomics and Proteomics Core Facility for providing Illumina DNA methylation array-related services. D. Jones is 
supported by the Everest Centre for Low-grade Paediatric Brain Tumours (The Brain Tumour Charity, UK). 
  
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
References 
1. Allinson KS, O'Donovan DG, Jena R, Cross JJ, Santarius TS (2015) Rosette-forming glioneuronal 
tumor with dissemination throughout the ventricular system: a case report. Clin Neuropathol 
34:64-69. doi:10.5414/NP300682 
2. Anan M, Inoue R, Ishii K et al. (2009) A rosette-forming glioneuronal tumor of the spinal cord: the 
first case of a rosette-forming glioneuronal tumor originating from the spinal cord. Hum 
Pathol 40:898-901. doi:10.1016/j.humpath.2008.11.010 
3. Cachia D, Prado MP, Theeler B, Hamilton J, McCutcheon I, Fuller GN (2014) Synchronous rosette-
forming glioneuronal tumor and diffuse astrocytoma with molecular characterization: a case 
report. Clin Neuropathol 33:407-411. doi:10.5414/NP300767 
4. Capper D, Jones DTW, Sill M et al. (2018) DNA methylation-based classification of central nervous 
system tumours. Nature 555:469-474. doi:10.1038/nature26000 
5. Ellezam B, Theeler BJ, Luthra R, Adesina AM, Aldape KD, Gilbert MR (2012) Recurrent PIK3CA 
mutations in rosette-forming glioneuronal tumor. Acta Neuropathol 123:285-287. 
doi:10.1007/s00401-011-0886-z 
6. Gessi M, Lambert SR, Lauriola L, Waha A, Collins VP, Pietsch T (2012) Absence of KIAA1549-BRAF 
fusion in rosette-forming glioneuronal tumors of the fourth ventricle (RGNT). J Neurooncol 
110:21-25. doi:10.1007/s11060-012-0940-2 
7. Gessi M, Moneim YA, Hammes J et al. (2014) FGFR1 mutations in Rosette-forming glioneuronal 
tumors of the fourth ventricle. J Neuropathol Exp Neurol 73:580-584. 
doi:10.1097/NEN.0000000000000080 
8. Hovestadt V, Remke M, Kool M et al. (2013) Robust molecular subgrouping and copy-number 
profiling of medulloblastoma from small amounts of archival tumour material using high-
density DNA methylation arrays. Acta Neuropathol 125:913-916. doi:10.1007/s00401-013-
1126-5 
9. Jacques TS, Eldridge C, Patel A et al. (2006) Mixed glioneuronal tumour of the fourth ventricle with 
prominent rosette formation. Neuropathol Appl Neurobiol 32:217-220. doi:10.1111/j.1365-
2990.2005.00692.x 
10. Jones DT, Hutter B, Jager N et al. (2013) Recurrent somatic alterations of FGFR1 and NTRK2 in 
pilocytic astrocytoma. Nat Genet 45:927-932. doi:10.1038/ng.2682 
11. Jones DT, Kocialkowski S, Liu L et al. (2008) Tandem duplication producing a novel oncogenic 
BRAF fusion gene defines the majority of pilocytic astrocytomas. Cancer Res 68:8673-8677. 
doi:10.1158/0008-5472.CAN-08-2097 
12. Karafin M, Jallo GI, Ayars M, Eberhart CG, Rodriguez FJ (2011) Rosette forming glioneuronal 
tumor in association with Noonan syndrome: pathobiological implications. Clin Neuropathol 
30:297-300 
13. Kemp S, Achan A, Ng T, Dexter MA (2012) Rosette-forming glioneuronal tumour of the lateral 
ventricle in a patient with neurofibromatosis 1. J Clin Neurosci 19:1180-1181. 
doi:10.1016/j.jocn.2011.12.013 
14. Komori T, Scheithauer BW, Hirose T (2002) A rosette-forming glioneuronal tumor of the fourth 
ventricle: infratentorial form of dysembryoplastic neuroepithelial tumor? Am J Surg Pathol 
26:582-591 
15. Lasorella A, Sanson M, Iavarone A (2017) FGFR-TACC gene fusions in human glioma. Neuro Oncol 
19:475-483. doi:10.1093/neuonc/now240 
16. Lin FY, Bergstrom K, Person R et al. (2016) Integrated tumor and germline whole-exome 
sequencing identifies mutations in MAPK and PI3K pathway genes in an adolescent with 
rosette-forming glioneuronal tumor of the fourth ventricle. Cold Spring Harb Mol Case Stud 
2:a001057. doi:10.1101/mcs.a001057 
17. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK (2016) WHO Classification of Tumours of the 
Central Nervous System. Revised 4th edition edn. IARC, Lyon 
18. McPherson A, Hormozdiari F, Zayed A et al. (2011) deFuse: an algorithm for gene fusion discovery 
in tumor RNA-Seq data. PLoS Comput Biol 7:e1001138. doi:10.1371/journal.pcbi.1001138 
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
19. Preusser M, Dietrich W, Czech T, Prayer D, Budka H, Hainfellner JA (2003) Rosette-forming 
glioneuronal tumor of the fourth ventricle. Acta Neuropathol 106:506-508. 
doi:10.1007/s00401-003-0758-2 
20. Qaddoumi I, Orisme W, Wen J et al. (2016) Genetic alterations in uncommon low-grade 
neuroepithelial tumors: BRAF, FGFR1, and MYB mutations occur at high frequency and align 
with morphology. Acta Neuropathol 131:833-845. doi:10.1007/s00401-016-1539-z 
21. Rivera B, Gayden T, Carrot-Zhang J et al. (2016) Germline and somatic FGFR1 abnormalities in 
dysembryoplastic neuroepithelial tumors. Acta Neuropathol 131:847-863. 
doi:10.1007/s00401-016-1549-x 
22. Rodriguez EF, Scheithauer BW, Giannini C et al. (2011) PI3K/AKT pathway alterations are 
associated with clinically aggressive and histologically anaplastic subsets of pilocytic 
astrocytoma. Acta Neuropathol 121:407-420. doi:10.1007/s00401-010-0784-9 
23. Rosenblum MK (2007) The 2007 WHO Classification of Nervous System Tumors: newly recognized 
members of the mixed glioneuronal group. Brain Pathol 17:308-313. doi:10.1111/j.1750-
3639.2007.00079.x 
24. Sahm F, Schrimpf D, Jones DT et al. (2016) Next-generation sequencing in routine brain tumor 
diagnostics enables an integrated diagnosis and identifies actionable targets. Acta 
Neuropathol 131:903-910. doi:10.1007/s00401-015-1519-8 
25. Schlamann A, von Bueren AO, Hagel C et al. (2014) An individual patient data meta-analysis on 
characteristics and outcome of patients with papillary glioneuronal tumor, rosette 
glioneuronal tumor with neuropil-like islands and rosette forming glioneuronal tumor of the 
fourth ventricle. PLoS One 9:e101211. doi:10.1371/journal.pone.0101211 
26. Sievers P, Stichel D, Schrimpf D et al. (2018) FGFR1:TACC1 fusion is a frequent event in 
molecularly defined extraventricular neurocytoma. Acta Neuropathol 136:293-302. 
doi:10.1007/s00401-018-1882-3 
27. Sturm D, Witt H, Hovestadt V et al. (2012) Hotspot mutations in H3F3A and IDH1 define distinct 
epigenetic and biological subgroups of glioblastoma. Cancer Cell 22:425-437. 
doi:10.1016/j.ccr.2012.08.024 
28. Thommen F, Hewer E, Schafer SC, Vassella E, Kappeler A, Vajtai I (2013) Rosette-forming 
glioneuronal tumor of the cerebellum in statu nascendi: an incidentally detected diminutive 
example indicates derivation from the internal granule cell layer. Clin Neuropathol 32:370-
376. doi:10.5414/NP300612 
29. Xu J, Yang Y, Liu Y et al. (2012) Rosette-forming glioneuronal tumor in the pineal gland and the 
third ventricle: a case with radiological and clinical implications. Quant Imaging Med Surg 
2:227-231. doi:10.3978/j.issn.2223-4292.2012.09.03 
30. Zhang J, Wu G, Miller CP et al. (2013) Whole-genome sequencing identifies genetic alterations in 
pediatric low-grade gliomas. Nat Genet 45:602-612. doi:10.1038/ng.2611 
 
 
  
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Sievers et al. 
Figure legends 
Fig. 1 Rosette-forming glioneuronal tumor (RGNT) by DNA Methylation Profiling. 
Unsupervised hierarchical clustering of DNA methylation profiles of 30 rosette-forming 
glioneuronal tumors alongside 106 well-characterized CNS neoplasms encompassing other 
low-grade glial/glioneuronal tumor entities and control tissue. Shown in a two-dimensional 
representation of pairwise sample correlations using the 20,000 most variant probes by t-
distributed stochastic neighbor embedding (t-SNE) dimensionality reduction. Reference 
methylation classes: IDH-mutant astrocytoma (A IDH), IDH-mutant oligodendroglioma (O 
IDH), posterior fossa pilocytic astrocytoma (LGG, PA PF), ganglioglioma (LGG, GG), midline 
pilocytic astrocytoma (LGG, PA MID), supratentorial/hemispheric pilocytic astrocytoma and 
ganglioglioma (LGG, PA/GG), diffuse leptomeningeal glioneuronal tumor subgroup 1 
(DLGNT_1), diffuse leptomeningeal glioneuronal tumor subgroup 2 (DLGNT_2), 
extraventricular neurocytoma (EVN), dysembryoplastic neuroepithelial tumor (LGG, DNT), 
rosette-forming glioneuronal tumor (LGG, RGNT) and control tissue white matter 
(CONTROL). 
 
Fig 2 Clinicopathological characteristics and recurrent genetic alterations in rosette-forming 
glioneuronal tumors examined by DNA methylation profiling and targeted next-generation 
sequencing. M, male; F, female; RGNT, rosette-forming glioneuronal tumor; LGG/LGGNT, 
low-grade glial/glioneuronal tumor; PA, pilocytic astrocytoma 
 
Fig. 3 Morphological and immunohistochemical features of rosette-forming glioneuronal 
tumors. Typical histomorphological features of RGNT showing well-differentiated neurocytic 
cells forming rosettes and perivascular pseudorosettes (a) with immunoreactivity for 
synaptophysin (b). 
 
Supplementary Table 1 Summary of clinicopathological characteristics and key genetic 
alterations identified in the RGNT cohort. 
 
Supplementary Table 2 Single nucleotide variants (SNVs), small 
insertions/deletions/substitutions and splice site variants detected in the RGNT cohort. 
 
Supplementary Fig. 1 Kaplan–Meier curves for overall survival of methylation class RGNT 
patients in comparison to reference glioma group patients (pilocytic astrocytoma WHO grade 
I; diffuse astrocytoma, IDH-mutant, WHO grade II). 
 
Supplementary Fig. 2 Unsupervised hierarchical clustering of DNA methylation profiles of 
30 rosette-forming glioneuronal tumors alongside 106 well-characterized CNS neoplasms 
encompassing other low-grade glial/glioneuronal tumor entities and control tissue. 
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Supplementary Table 1 - Summary of clinicopathological characteristics and key genetic alterations identified in the RGNT cohort
Case # Age (years) Sex Tumor location Diagnosis FGFR1 PIK3CA NF1 PTPN11 Copy number variations
1 12 m 4th ventricle LGGNT p.N546K p.H1047R p.N58H balanced
2 23 m 4th ventricle RGNT p.N546K p.H1047R splicing balanced
3 25 f Cerebellum PA p.N546K p.H1047R p.M1149V balanced
4 43 m Pineal region RGNT p.K656E gain of chr 7, 12
5 39 m 4th ventricle LGG p.K656E loss of chr 4, 19, 21q
6 28 f Tectum PA p.N546K balanced
7 69 f 3th ventricle LGG/LGGNT p.K656E p.R1276X and p.N1259S balanced
8 67 f Cerebellum PA p.N546K balanced
9 20 f Cerebellum LGG p.N546K p.H1047R p.E1790X balanced
10 31 m 4th ventricle RGNT p.N546K p.H1047R balanced
11 34 m 4th ventricle RGNT p.N546K p.G1049R loss of chr 8p, 20p
12 36 f Thalamus, brainsteam RGNT p.N546K p.H1047R p.W1831X balanced
13 44 m 3th ventricle RGNT p.K656E balanced
14 55 f 4th ventricle RGNT p.N546K balanced
15 43 m Cerebellum PA p.N546K gain of chr 5, 6, 7, 8, 11, 12, 20
16 14 f 4th ventricle RGNT p.N546K p.H1047R p.P491H balanced
17 24 f Pineal region PA p.N546K p.H1047R p.R2637X balanced
18 21 m 4th ventricle RGNT p.N546K p.E545K balanced
19 59 m Cerebellum PA p.N546K p.R897W and p.H647R gain of chr 6p, segmnetal gain of 2q, 7q, 12q
20 40 m 4th ventricle RGNT p.N546K balanced
21 50 m Cerebellum RGNT p.K656E p.E542K balanced
22 10 f Cerebellum PA p.K656E p.H1047R balanced
23 20 m 4th ventricle RGNT p.K656E p.H1047R splicing balanced
24 38 f 4th ventricle RGNT p.N546K p.H1047R balanced
25 21 f Cerebellum PA p.N546K p.H1047R balanced
26 18 f 4th ventricle RGNT p.N546K p.H1047R p.R416X balanced
27 28 f 4th ventricle RGNT p.N546K p.H1047R balanced
28 41 f 4th ventricle RGNT p.K656E p.E1264G and p. L1161V balanced
29 25 m Cerebellum RGNT p.N546K p.H1047L balanced
30 21 m Cerebellum RGNT p.N546K p.E545K balanced
Abbreviations: m, male; f, female; RGNT I, rosette-forming glioneuronal tumor WHO grade I; PA I, pilocytic astrocytoma WHO grade I; LGG, low-grade glioma; LGGNT, low-grade glioneuronal tumor; chr, chromosome
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated version is available online at: http://dx.doi.org/10.1007/s00401-019-
02038-4”.
Suppl. Table 2
Supplementary Table 2 - Single nucleotide variants (SNVs), small insertions/deletions/substitutions and splice site variants detected in the RGNT cohort
Case # Chromosome Position (hg19) Reference Allele Alternative Allele Mutant Allele Fraction Gene Symbol Class RefSeq ID CDS Change Amino Acid Change Annovar Output Function
1 chr12 112888156 A C 0,51 PTPN11 nonsynonymous SNV NM_080601 c.A172C p.N58H PTPN11:NM_080601:exon3:c.A172C:p.N58H exonic
1 chr8 38274849 G T 0,29 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
1 chr4 55984939 G T 0,50 KDR nonsynonymous SNV NM_002253 c.C190A p.P64T KDR:NM_002253:exon3:c.C190A:p.P64T exonic
1 chr3 178952085 A G 0,23 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
2 chr3 178952085 A G 0,48 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
2 chr8 38274849 G T 0,43 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
2 chr17 29528504 - T NF1 splicing NA NA p.NA NA splicing
3 chr3 178952085 A G 0,24 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
3 chr17 29559848 A G 0,29 NF1 nonsynonymous SNV NM_001042492 c.A3445G p.M1149V NF1:NM_001042492:exon26:c.A3445G:p.M1149V exonic
3 chr8 38274849 G T 0,36 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
3 chr12 18719885 T G 0,51 PIK3C2G nonsynonymous SNV NM_001288772 c.T3905G p.F1302C PIK3C2G:NM_001288772:exon29:c.T3905G:p.F1302C exonic
4 chr8 38274857 T C 0,39 FGFR1 nonsynonymous SNV NM_023105 c.A1363G p.I455V FGFR1:NM_023105:exon11:c.A1363G:p.I455V exonic
4 chr8 38272308 T C 0,38 FGFR1 nonsynonymous SNV NM_023105 c.A1699G p.K567E FGFR1:NM_023105:exon13:c.A1699G:p.K567E exonic
5 chr8 38272308 T C 0,38 FGFR1 nonsynonymous SNV NM_023105 c.A1699G p.K567E FGFR1:NM_023105:exon13:c.A1699G:p.K567E exonic
5 chr6 36169331 A G 0,45 BRPF3 nonsynonymous SNV NM_015695 c.A1232G p.E411G BRPF3:NM_015695:exon2:c.A1232G:p.E411G exonic
6 chr2 211179766 T - 0,37 MYL1 frameshift deletion NM_079420 c.1delA p.M1fs MYL1:NM_079420:exon1:c.1delA:p.M1fs exonic
6 chr5 67589597 ACT - 0,40 PIK3R1 nonframeshift deletion NM_181504 c.550_552del p.184_184del PIK3R1:NM_181504:exon5:c.550_552del:p.184_184del exonic
6 chr17 41243509 T C 0,44 BRCA1 nonsynonymous SNV NM_007294 c.A4039G p.R1347G BRCA1:NM_007294:exon10:c.A4039G:p.R1347G exonic
6 chr9 5022130 G A 0,45 JAK2 nonsynonymous SNV NM_004972 c.G143A p.G48E JAK2:NM_004972:exon3:c.G143A:p.G48E exonic
6 chr5 1282733 C A 0,52 TERT nonsynonymous SNV NM_198253 c.G1580T p.G527V TERT:NM_198253:exon3:c.G1580T:p.G527V exonic
6 chr8 39872972 G T 0,47 IDO2 nonsynonymous SNV NM_194294 c.G1114T p.A372S IDO2:NM_194294:exon11:c.G1114T:p.A372S exonic
6 chr12 25398281 C T 0,36 KRAS nonsynonymous SNV NM_033360 c.G38A p.G13D KRAS:NM_033360:exon2:c.G38A:p.G13D exonic
6 chr8 38274849 G T 0,43 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
7 chr8 38272321 G C 0,42 FGFR1 nonsynonymous SNV NM_023105 c.C1686G p.I562M FGFR1:NM_023105:exon13:c.C1686G:p.I562M exonic
7 chr8 38272308 T C 0,40 FGFR1 nonsynonymous SNV NM_023105 c.A1699G p.K567E FGFR1:NM_023105:exon13:c.A1699G:p.K567E exonic
7 chr3 47162897 T C 0,52 SETD2 nonsynonymous SNV NM_014159 c.A3229G p.T1077A SETD2:NM_014159:exon3:c.A3229G:p.T1077A exonic
7 chr3 10183605 C T 0,51 VHL nonsynonymous SNV NM_198156 c.C74T p.P25L VHL:NM_198156:exon1:c.C74T:p.P25L exonic
7 chr1 120548095 C A 0,25 NOTCH2 nonsynonymous SNV NM_024408 c.G272T p.R91L NOTCH2:NM_024408:exon3:c.G272T:p.R91L exonic
7 chr17 29562696 A G 0,43 NF1 nonsynonymous SNV NM_001042492 c.A3776G p.N1259S NF1:NM_001042492:exon28:c.A3776G:p.N1259S exonic
7 chr11 108139154 G T 0,48 ATM nonsynonymous SNV NM_000051 c.G2656T p.A886S ATM:NM_000051:exon18:c.G2656T:p.A886S exonic
7 chr17 29562746 C T 0,43 NF1 stopgain SNV NM_001042492 c.C3826T p.R1276X NF1:NM_001042492:exon28:c.C3826T:p.R1276X exonic
8 chr8 38274849 G C 0,35 FGFR1 nonsynonymous SNV NM_023105 c.C1371G p.N457K FGFR1:NM_023105:exon11:c.C1371G:p.N457K exonic
9 chr2 29416250 T G 0,49 ALK nonsynonymous SNV NM_004304 c.A4703C p.E1568A ALK:NM_004304:exon29:c.A4703C:p.E1568A exonic
9 chr3 178952085 A G 0,36 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
9 chr5 67575523 C T 0,47 PIK3R1 nonsynonymous SNV NM_181523 c.C596T p.P199L PIK3R1:NM_181523:exon5:c.C596T:p.P199L exonic
9 chr7 55238037 C A 0,52 EGFR nonsynonymous SNV NM_201284 c.C1918A p.L640I EGFR:NM_201284:exon16:c.C1918A:p.L640I exonic
9 chr8 38274849 G C 0,35 FGFR1 nonsynonymous SNV NM_023105 c.C1371G p.N457K FGFR1:NM_023105:exon11:c.C1371G:p.N457K exonic
9 chr17 29654616 G T 0,36 NF1 stopgain SNV NM_001042492 c.G5368T p.E1790X NF1:NM_001042492:exon38:c.G5368T:p.E1790X exonic
9 chr17 41245466 GCTGTC ACTGTT 0,44 BRCA1 nonframeshift substitution NM_007294 c.2077_2082AACAGT p.NA BRCA1:NM_007294:exon10:c.2077_2082AACAGT exonic
10 chr3 178952085 A G 0,44 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
10 chr13 32915313 G T 0,43 BRCA2 nonsynonymous SNV NM_000059 c.G6821T p.G2274V BRCA2:NM_000059:exon11:c.G6821T:p.G2274V exonic
10 chr12 25368405 A T 0,44 KRAS stopgain SNV NM_033360 c.T540A p.C180X KRAS:NM_033360:exon5:c.T540A:p.C180X exonic
10 chr7 55259460 G A 0,49 EGFR nonsynonymous SNV NM_005228 c.G2518A p.A840T EGFR:NM_005228:exon21:c.G2518A:p.A840T exonic
10 chr3 47061329 G A 0,44 SETD2 nonsynonymous SNV NM_014159 c.C7352T p.A2451V SETD2:NM_014159:exon19:c.C7352T:p.A2451V exonic
10 chr1 27023483 G A 0,49 ARID1A nonsynonymous SNV NM_139135 c.G589A p.G197R ARID1A:NM_139135:exon1:c.G589A:p.G197R exonic
10 chr8 38274849 G T 0,37 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
11 chr9 139405209 C T 0,50 NOTCH1 nonsynonymous SNV NM_017617 c.G2636A p.R879Q NOTCH1:NM_017617:exon17:c.G2636A:p.R879Q exonic
11 chr9 5126715 A G 0,46 JAK2 nonsynonymous SNV NM_004972 c.A3323G p.N1108S JAK2:NM_004972:exon25:c.A3323G:p.N1108S exonic
11 chr8 38274849 G T 0,58 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
11 chr4 55980379 C A 0,43 KDR nonsynonymous SNV NM_002253 c.G712T p.V238F KDR:NM_002253:exon6:c.G712T:p.V238F exonic
11 chr3 178952090 G C 0,40 PIK3CA nonsynonymous SNV NM_006218 c.G3145C p.G1049R PIK3CA:NM_006218:exon21:c.G3145C:p.G1049R exonic
11 chr16 3824628 C A 0,50 CREBBP nonsynonymous SNV NM_004380 c.G2225T p.R742L CREBBP:NM_004380:exon12:c.G2225T:p.R742L exonic
11 chr20 57429775 C A 0,43 GNAS nonsynonymous SNV NM_001077490 c.C1268A p.P423H GNAS:NM_001077490:exon1:c.C1268A:p.P423H exonic
11 chr1 120479948 T C 0,50 NOTCH2 nonsynonymous SNV NM_024408 c.A3479G p.H1160R NOTCH2:NM_024408:exon21:c.A3479G:p.H1160R exonic
11 chr1 51439754 T G 0,48 CDKN2C nonsynonymous SNV NM_001262 c.T319G p.L107V CDKN2C:NM_001262:exon3:c.T319G:p.L107V exonic
12 chr8 38274849 G T 0,35 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
12 chr17 29654741 G A 0,13 NF1 stopgain SNV NM_001042492 c.G5493A p.W1831X NF1:NM_001042492:exon38:c.G5493A:p.W1831X exonic
12 chr3 178952085 A G 0,28 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
12 chr5 131915022 G A 0,48 RAD50 nonsynonymous SNV NM_005732 c.G379A p.V127I RAD50:NM_005732:exon4:c.G379A:p.V127I exonic
13 chr8 38272308 T C 0,39 FGFR1 nonsynonymous SNV NM_023105 c.A1699G p.K567E FGFR1:NM_023105:exon13:c.A1699G:p.K567E exonic
13 chr5 67589588 GAA - 0,34 PIK3R1 nonframeshift deletion NM_181504 c.541_543del p.181_181del PIK3R1:NM_181504:exon5:c.541_543del:p.181_181del exonic
14 chr8 38274849 G T 0,35 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
13 chr9 133759575 C A 0,23 ABL1 nonsynonymous SNV NM_007313 c.C1955A p.A652D ABL1:NM_007313:exon11:c.C1955A:p.A652D exonic
13 chr14 105242084 C A 0,52 AKT1 stopgain SNV NM_005163 c.G340T p.E114X AKT1:NM_005163:exon5:c.G340T:p.E114X exonic
13 chr19 40771132 G A 0,14 AKT2 nonsynonymous SNV NM_001626 c.C43T p.R15C AKT2:NM_001626:exon2:c.C43T:p.R15C exonic
13 chr2 29443583 G T 0,29 ALK nonsynonymous SNV NM_004304 c.C3634A p.R1212S ALK:NM_004304:exon23:c.C3634A:p.R1212S exonic
13 chr13 53418870 C T 0,19 PCDH8 nonsynonymous SNV NM_032949 c.G2747A p.G916E PCDH8:NM_032949:exon3:c.G2747A:p.G916E exonic
15 chr16 3789606 C A 0,29 CREBBP nonsynonymous SNV NM_004380 c.G4253T p.G1418V CREBBP:NM_004380:exon25:c.G4253T:p.G1418V exonic
15 chr17 7579717 G A 0,53 TP53 nonsynonymous SNV NM_001126112 c.C79T p.P27S TP53:NM_001126112:exon3:c.C79T:p.P27S exonic
15 chr11 94192659 T G 0,42 MRE11A nonsynonymous SNV NM_005591 c.A1415C p.E472A MRE11A:NM_005591:exon13:c.A1415C:p.E472A exonic
15 chr8 38274849 G T 0,26 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
15 chr3 142272098 A G 0,43 ATR nonsynonymous SNV NM_001184 c.T2776C p.F926L ATR:NM_001184:exon13:c.T2776C:p.F926L exonic
15 chr19 42797203 G A 0,46 CIC nonsynonymous SNV NM_015125 c.G3565A p.V1189M CIC:NM_015125:exon15:c.G3565A:p.V1189M exonic
16 chr5 1293676 TCC - 0,34 TERT nonframeshift deletion NM_198253 c.1323_1325del p.441_442del TERT:NM_198253:exon2:c.1323_1325del:p.441_442del exonic
16 chr3 142212087 T C 0,52 ATR nonsynonymous SNV NM_001184 c.A5965G p.T1989A ATR:NM_001184:exon35:c.A5965G:p.T1989A exonic
16 chr3 178952085 A G 0,39 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
16 chr6 33286923 G A 0,44 DAXX nonsynonymous SNV NM_001141969 c.C2014T p.P672S DAXX:NM_001141969:exon7:c.C2014T:p.P672S exonic
16 chr8 38274849 G T 0,47 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
16 chr12 112926852 C A 0,50 PTPN11 nonsynonymous SNV NM_002834 c.C1472A p.P491H PTPN11:NM_002834:exon13:c.C1472A:p.P491H exonic
16 chr13 53420063 C G 0,47 PCDH8 nonsynonymous SNV NM_002590 c.G2509C p.A837P PCDH8:NM_002590:exon1:c.G2509C:p.A837P exonic
16 chr14 21488110 G A 0,52 NDRG2 nonsynonymous SNV NM_001282216 c.C401T p.P134L NDRG2:NM_001282216:exon8:c.C401T:p.P134L exonic
Page 1
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
Suppl. Table 2
16 chr19 11096912 C G 0,49 SMARCA4 nonsynonymous SNV NM_001128846 c.C403G p.P135A SMARCA4:NM_001128846:exon3:c.C403G:p.P135A exonic
17 chr3 178952085 A G 0,26 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
17 chr8 38274849 G C 0,23 FGFR1 nonsynonymous SNV NM_023105 c.C1371G p.N457K FGFR1:NM_023105:exon11:c.C1371G:p.N457K exonic
17 chr13 53419568 A G 0,45 PCDH8 nonsynonymous SNV NM_032949 c.T2540C p.M847T PCDH8:NM_032949:exon2:c.T2540C:p.M847T exonic
17 chr17 29684326 C T 0,21 NF1 stopgain SNV NM_001042492 c.C7909T p.R2637X NF1:NM_001042492:exon54:c.C7909T:p.R2637X exonic
18 chr3 178936091 G A 0,31 PIK3CA nonsynonymous SNV NM_006218 c.G1633A p.E545K PIK3CA:NM_006218:exon10:c.G1633A:p.E545K exonic
18 chr3 9782480 A G 0,50 BRPF1 nonsynonymous SNV NM_004634 c.A1577G p.N526S BRPF1:NM_004634:exon4:c.A1577G:p.N526S exonic
18 chr2 29446388 C A 0,54 ALK nonsynonymous SNV NM_004304 c.G3179T p.R1060L ALK:NM_004304:exon20:c.G3179T:p.R1060L exonic
18 chr8 38274849 G C 0,31 FGFR1 nonsynonymous SNV NM_023105 c.C1371G p.N457K FGFR1:NM_023105:exon11:c.C1371G:p.N457K exonic
19 chr11 108203572 T A 0,53 ATM stopgain SNV NM_000051 c.T7872A p.C2624X ATM:NM_000051:exon53:c.T7872A:p.C2624X exonic
19 chr7 140624404 CGGCGC - BRAF nonframeshift deletion NM_004333 c.95_100del p.32_34del BRAF:NM_004333:exon1:c.95_100del:p.32_34del exonic
19 chr5 131924421 G A 0,47 RAD50 nonsynonymous SNV NM_005732 c.G1094A p.R365Q RAD50:NM_005732:exon8:c.G1094A:p.R365Q exonic
19 chr8 38274849 G T 0,39 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
19 chr17 29556322 C T 0,37 NF1 nonsynonymous SNV NM_001042492 c.C2689T p.R897W NF1:NM_001042492:exon21:c.C2689T:p.R897W exonic
19 chrX 76874309 G C 0,75 ATRX nonsynonymous SNV NM_000489 c.C5413G p.H1805D ATRX:NM_000489:exon21:c.C5413G:p.H1805D exonic
19 chr5 67589599 TCAGTT - PIK3R1 nonframeshift deletion NM_181504 c.552_557del p.184_186del PIK3R1:NM_181504:exon5:c.552_557del:p.184_186del exonic
19 chr17 29552207 A G 0,55 NF1 nonsynonymous SNV NM_001042492 c.A1940G p.H647R NF1:NM_001042492:exon17:c.A1940G:p.H647R exonic
20 chr6 36181882 G A 0,58 BRPF3 nonsynonymous SNV NM_015695 c.G2708A p.G903D BRPF3:NM_015695:exon8:c.G2708A:p.G903D exonic
20 chr8 38274849 G T 0,29 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
20 chr13 32968951 C T 0,42 BRCA2 stopgain SNV NM_000059 c.C9382T p.R3128X BRCA2:NM_000059:exon25:c.C9382T:p.R3128X exonic
21 chr8 38272310 T A 0,40 FGFR1 nonsynonymous SNV NM_023105 c.A1697T p.K566I FGFR1:NM_023105:exon13:c.A1697T:p.K566I exonic
21 chr8 38272308 T C 0,40 FGFR1 nonsynonymous SNV NM_023105 c.A1699G p.K567E FGFR1:NM_023105:exon13:c.A1699G:p.K567E exonic
21 chr3 178936082 G A 0,36 PIK3CA nonsynonymous SNV NM_006218 c.G1624A p.E542K PIK3CA:NM_006218:exon10:c.G1624A:p.E542K exonic
21 chr8 38272308 TTT CTA 0,42 FGFR1 nonframeshift substitution NM_023105 c.1697_1699TAG p.NA FGFR1:NM_023105:exon13:c.1697_1699TAG exonic
22 chr8 38272308 T C 0,34 FGFR1 nonsynonymous SNV NM_023105 c.A1699G p.K567E FGFR1:NM_023105:exon13:c.A1699G:p.K567E exonic
22 chr8 38272306 CTT GTC 0,34 FGFR1 nonframeshift substitution NM_023105 c.1699_1701GAC p.NA FGFR1:NM_023105:exon13:c.1699_1701GAC exonic
22 chr1 27099302 G C 0,47 ARID1A spiling NA NA p.NA NA splicing
22 chr3 178952085 A G 0,26 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
22 chr8 38272306 C G 0,33 FGFR1 nonsynonymous SNV NM_023105 c.G1701C p.K567N FGFR1:NM_023105:exon13:c.G1701C:p.K567N exonic
22 chr12 49443634 G A 0,53 KMT2D nonsynonymous SNV NM_003482 c.C3737T p.T1246M KMT2D:NM_003482:exon11:c.C3737T:p.T1246M exonic
22 chr16 2120559 G A 0,49 TSC2 nonsynonymous SNV NM_001077183 c.G1819A p.A607T TSC2:NM_001077183:exon17:c.G1819A:p.A607T exonic
22 chr16 68846048 C T 0,46 CDH1 nonsynonymous SNV NM_004360 c.C1019T p.T340M CDH1:NM_004360:exon8:c.C1019T:p.T340M exonic
22 chr4 55972049 G A 0,49 KDR nonsynonymous SNV NM_002253 c.C1595T p.A532V KDR:NM_002253:exon12:c.C1595T:p.A532V exonic
23 chr9 133760871 C T 0,49 ABL1 nonsynonymous SNV NM_007313 c.C3251T p.T1084M ABL1:NM_007313:exon11:c.C3251T:p.T1084M exonic
23 chr9 139413909 G A 0,44 NOTCH1 nonsynonymous SNV NM_017617 c.C851T p.P284L NOTCH1:NM_017617:exon5:c.C851T:p.P284L exonic
23 chr16 3779329 C T 0,43 CREBBP nonsynonymous SNV NM_004380 c.G5719A p.A1907T CREBBP:NM_004380:exon31:c.G5719A:p.A1907T exonic
23 chr17 29684388 GTAA - NF1 splicing NA NA p.NA NA splicing
23 chr8 38272308 T C 0,32 FGFR1 nonsynonymous SNV NM_023105 c.A1699G p.K567E FGFR1:NM_023105:exon13:c.A1699G:p.K567E exonic
23 chr3 178952085 A G 0,30 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
23 chr5 112174802 C T 0,47 APC nonsynonymous SNV NM_000038 c.C3511T p.R1171C APC:NM_000038:exon16:c.C3511T:p.R1171C exonic
23 chr8 38274849 G T 0,39 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
23 chr3 178952085 A G 0,30 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
23 chr5 67591132 GAC - PIK3R1 nonframeshift deletion NM_181504 c.915_917del p.305_306del PIK3R1:NM_181504:exon7:c.915_917del:p.305_306del exonic
25 chr13 53420536 T A 0,46 PCDH8 nonsynonymous SNV NM_032949 c.A2036T p.Q679L PCDH8:NM_032949:exon1:c.A2036T:p.Q679L exonic
25 chr11 108170506 A C 0,47 ATM nonsynonymous SNV NM_000051 c.A5071C p.S1691R ATM:NM_000051:exon34:c.A5071C:p.S1691R exonic
25 chr8 38274849 G T 0,23 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
25 chr5 112179179 G A 0,49 APC nonsynonymous SNV NM_000038 c.G7888A p.V2630I APC:NM_000038:exon16:c.G7888A:p.V2630I exonic
25 chr3 178952085 A G 0,26 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
25 chr3 10191545 A G 0,44 VHL nonsynonymous SNV NM_000551 c.A538G p.I180V VHL:NM_000551:exon3:c.A538G:p.I180V exonic
26 chr3 178952085 A G 0,41 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
26 chr7 128843395 C T 0,48 SMO nonsynonymous SNV NM_005631 c.C502T p.R168C SMO:NM_005631:exon2:c.C502T:p.R168C exonic
26 chr8 38274849 G C 0,30 FGFR1 nonsynonymous SNV NM_023105 c.C1371G p.N457K FGFR1:NM_023105:exon11:c.C1371G:p.N457K exonic
26 chr13 53420063 C G 0,49 PCDH8 nonsynonymous SNV NM_002590 c.G2509C p.A837P PCDH8:NM_002590:exon1:c.G2509C:p.A837P exonic
26 chr17 29528489 C T 0,31 NF1 stopgain SNV NM_001042492 c.C1246T p.R416X NF1:NM_001042492:exon11:c.C1246T:p.R416X exonic
26 chr9 2191388 G A 0,50 SMARCA2 nonsynonymous SNV NM_001289399 c.G775A p.D259N SMARCA2:NM_001289399:exon7:c.G775A:p.D259N exonic
27 chr12 18499770 GGTGAGTCGT - 0,91 PIK3C2G frameshift deletion NM_001288774 c.959_959del p.320_320del PIK3C2G:NM_001288774:exon11:c.959_959del:p.320_320del exonic
27 chr16 85695351 G A 0,52 GSE1 nonsynonymous SNV NM_001278184 c.G2021A p.R674Q GSE1:NM_001278184:exon8:c.G2021A:p.R674Q exonic
27 chr8 38274849 G C 0,30 FGFR1 nonsynonymous SNV NM_023105 c.C1371G p.N457K FGFR1:NM_023105:exon11:c.C1371G:p.N457K exonic
27 chr3 178952085 A G 0,41 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047R PIK3CA:NM_006218:exon21:c.A3140G:p.H1047R exonic
27 chr12 18499769 AGGTGAGTCGT AGGTGAGTGGT 0,91 PIK3C2G nonframeshift substitution NM_001288774 c.958_959AGGTGAGTGGT p.958_959AGGTGAGTGGT PIK3C2G:NM_001288774:exon11:c.958_959AGGTGAGTGGT exonic
28 chr8 38272308 T C 0,36 FGFR1 nonsynonymous SNV NM_023105 c.A1699G p.K567E FGFR1:NM_023105:exon13:c.A1699G:p.K567E exonic
28 chr6 114262250 G T 0,54 HDAC2 nonsynonymous SNV NM_001527 c.C1439A p.T480N HDAC2:NM_001527:exon14:c.C1439A:p.T480N exonic
28 chr17 29559883 TC GG 0,28 NF1 nonframeshift substitution NM_001042492 c.3480_3481GG NF1:NM_001042492:exon26:c.3480_3481GG exonic
28 chr17 29562711 A G 0,30 NF1 nonsynonymous SNV NM_001042492 c.A3791G p.E1264G NF1:NM_001042492:exon28:c.A3791G:p.E1264G exonic
28 chr17 29559884 C G 0,29 NF1 nonsynonymous SNV NM_001042492 c.C3481G p.L1161V NF1:NM_001042492:exon26:c.C3481G:p.L1161V exonic
29 chr17 41245466 GCTGTC ACTGTT 0,62 BRCA1 nonframeshift substitution NM_007294 c.2077_2082AACAGT BRCA1:NM_007294:exon10:c.2077_2082AACAGT exonic
29 chr8 38274849 G T 0,37 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
29 chr4 55561723 C T 0,42 KIT nonsynonymous SNV NM_001093772 c.C113T p.S38F KIT:NM_001093772:exon2:c.C113T:p.S38F exonic
29 chr3 178952085 A G 0,41 PIK3CA nonsynonymous SNV NM_006218 c.A3140G p.H1047L PIK3CA:NM_006218:exon21:c.A3140G:p.H1047L exonic
30 chr8 38274849 G T 0,37 FGFR1 nonsynonymous SNV NM_023105 c.C1371A p.N457K FGFR1:NM_023105:exon11:c.C1371A:p.N457K exonic
30 chr3 178936091 G A 0,31 PIK3CA nonsynonymous SNV NM_006218 c.G1633A p.E545K PIK3CA:NM_006218:exon10:c.G1633A:p.E545K exonic
Page 2
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●●●●
●
●
●
●
●
●
●
●
●
●
●●●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●● ●
●● ●
●
−25
0
25
50
−20 0 20
TSNE 1
TS
N
E 
2
A IDH
CONTROL
DLGNT_1
DLGNT_2
EVN
LGG, DNT
LGG, GG
LGG, PA MID
LGG, PA PF
LGG, PA/GG
O IDH
RGNT
136 samples
(20,000 most variant probes)
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
850k
DNA-seq
M
ol
ec
ul
ar
 
an
al
ys
is
FGFR1
PIK3CA
NF1
PTPN11A
lte
ra
tio
ns
10 - 20
21 - 40
>40
Age (years)
M
F
Sex
p.N546K
FGFR1
Completed
Not available
Molecular analysis
Age
Sex
LocationC
lin
ic
al
Histology
1 2 3 4 5 6 7 8 9 18 19 20 2111 12 13 14 15 1710 16 22 23 24 25 26 27 28 29 30 Case #
RGNT
LGG/LGGNT
PA
Histology
4th ventricle
Cerebellum
Location
p.H1047R
PIK3CA
p.N58H
PTPN11NF1
splice site
stopgain
p.K656E missensep.G1049R p.P491H
p.E545K
p.E542K
p.H1047L
Diencephalon
Mesencephalon
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated
version is available online at: http://dx.doi.org/10.1007/s00401-019-02038-4”.
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica
vol.138 issue 3, 2019. The final authenticated version is available online at: http://dx.doi.org/
10.1007/s00401-019-02038-4”.
0 50 100 150 200
0
50
100
Months since diagnosis
O
ve
ra
ll
su
rv
iv
al
(%
)
RGNT
PA I
A IDH-mut II
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated version is available online at: http://dx.doi.org/10.1007/s00401-019-
02038-4”.
A IDHO IDH LGG, PA PFLGG, GG LGG, PA MID LGG, PA/GG DLGNT_1 DLGNT_2 CONTROLEVN LGG, DNTLGG, RGNTgr
ou
p
20
,0
00
 m
os
t 
va
ri
ab
ly
 m
et
hy
la
te
d 
C
pG
s
(n
 =
 1
36
)
0
0.
5
1
be
ta
-v
al
ue
s
“This is a post-peer-review, pre-copyedit version of an article published in Acta Neuropathologica vol.138 issue 3, 2019. The final authenticated version is available online at: http://dx.doi.org/10.1007/s00401-019-
02038-4”.