Child's Nervous System

https://doi.org/10.1007/s00381-021-05207-7

REVIEW ARTICLE

Pediatric ependymoma: an overview of a complex disease

Stephanie Theresa Jünger 1,2 & Beate Timmermann 3 & Torsten Pietsch 1

Received: 20 December 2020 / Accepted: 5 May 2021

# The Author(s) 2021

Abstract
Pediatric ependymomas comprise biologically distinct tumor entities with different (epi)genetics, age distribution and localiza-
tion, as well as a different prognosis. Regarding risk stratification within these biologically defined entities, histopathological
features still seem to be relevant. The mainstay of treatment is gross total resection (GTR) if possible, achieved with intraoper-
ative monitoring and neuronavigation—and if necessary second surgery—followed by adjuvant radiation therapy. However,
there is growing evidence that some ependymal tumors may be cured by surgery alone, while others relapse despite adjuvant
treatment. To date, the role of chemotherapy is not clear. Current therapy achieves reasonable survival rates for the majority of
ependymoma patients. The next challenge is to go beyond initial tumor control and use risk-adapted therapy to reduce secondary
effect and therapy-induced morbidity for low-risk patients and to intensify treatment for high-risk patients. With identification of
specific alterations, targeted therapy may represent an option for individualized treatment modalities in the future.

Keywords Ependymoma . Pediatric . Genetics . Chemotherapy . Radiation . Neuropathology

Introduction stratification and hence for treatment decisions. However, the

utility of grading for risk stratification has remained contro-
Ependymomas represent the third most common malignant versial, especially because of its lack of reproducibility due to
intracranial neoplasm in children. Pediatric ependymomas inter-observer variability [4, 5].
can occur in all compartments of the CNS; the most frequent Nevertheless, to date, histological diagnosis remains im-
location is the posterior fossa followed by supratentorial sites, portant since it can be performed immediately and almost
while spinal tumors are rather rare [1, 2]. A slight overall male everywhere. However, genetic and epigenetic studies in
predominance is reported [1–3]. Based on their histological ependymomas have led to the identification of biologically
resemblance, ependymomas across all locations were histori- distinct subtypes, suggesting more adequate means for risk
cally considered as one entity but assigned different tumor stratification [6–14].
grades. Assuming that histopathological indicators of anapla- Taylor et al. were able to show that tumors of different
sia, namely mitotic activity, vascular proliferation, and necro- locations in the CNS represent distinct tumor entities derived
sis, were adequate predictors for outcome independent of pa- from spatially restricted progenitor cell populations of radial
tient age and tumor location, WHO grading was used for risk glia cells, indicating that spatial heterogeneity may be associ-
ated with distinct tumor biology [13]. These findings indicate
that tumors in different locations constitute genetically and
* Stephanie Theresa Jünger biologically distinct diseases despite their histopathological
stephanie.juenger@uk-koeln.de
resemblance. Pajtler et al. presented an epigenetic classifica-
tion of ependymomas which divides supra-, infratentorial and
1
Department of Neuropathology, DGNN Brain Tumor Reference spinal ependymomas into three different, partly age-related,
Centre, University of Bonn Medical Centre, Bonn, Germany an d pr ogn ostica lly re leva nt gro ups eac h: sp inal
2
Centre for Neurosurgery, Department of General Neurosurgery, subependymoma (SP-SE), spinal myxopapillary
Faculty of Medicine and University Hospital Cologne, University of ependymoma (SP-MPE), spinal ependymoma (SP-EPN), pos-
Cologne, Cologne, Germany
3
terior fossa subependymoma (PF-SE), posterior fossa
Department of Particle Therapy, University Hospital Essen, West ependymoma-A (PF-EPN-A), posterior fossa ependymoma-
German Proton Therapy Centre Essen (WPE), West German Cancer
Center (WTZ), German Cancer Consortium (DKTK), B (PF-EPN-B), supratentorial subependymoma (ST-SE),
Essen, Germany supratentorial ependymoma YAP1-fused (ST-EPN-YAP),
 Childs Nerv Syst

and supratentorial ependymoma RELA-fused (ST-EPN- cystic areas as well as calcifications and more rarely hemor-
RELA) [11]. rhage and necrosis [21]. Supratentorial ependymomas of
In this review, the subdivision is partly orientated on the childhood are considered to have a better prognosis than those
classification by Pajtler et al. However, pediatric ependymoma located in the posterior fossa, possibly because gross-total
is distinct from adult tumors, as indicated in the section below; resection is more easily achieved in the former group.
therefore, we aimed to focus on ependymomas occurring in Supratentorial ependymal tumors can be divided into three
childhood. Current treatment modalities will be discussed sep- epigenetic groups: ST-EPN-RELA, ST-EPN-YAP, and ST-
arately, since most trials published to date have used histopath- SE, with supratentorial subependymoma being irrelevant in
ological grade, age, and extent of resection for stratification pediatric patients [11]. Furthermore, there is evidence of
rather than localization or epigenetically defined subgroups. supratentorial ependymal tumors different from ST-EPN-
Information regarding studies identifying prognostic parame- RELA nor ST-EPN-YAP [22–24].
ters and results from prospective clinical trials provided in the
text are summarized in Tables 1 and 2. Supratentorial ependymoma with C11orf95–RELA
fusions

Distinction between pediatric and adult Approximately 70 % of supratentorial ependymomas harbor

ependymoma one of several alternative C11orf95–RELA fusions resulting
from chromosomal instability/microchromothripsis involving
Approximately 30% of all ependymomas affect children and the chromosomal region 11q13.1 [11, 12, 20]. C11orf95–
adolescent patients [1, 2]. With several differences between RELA fusion proteins translocate into the nucleus, activating
pediatric and adult ependymoma, a differentiated approach is the canonical NF-κB signaling pathway. This pathway con-
warranted. Regarding clinical parameters, intracranial location trols cell proliferation, apoptosis, and vascularization. In addi-
is frequent in children, while spinal tumors are rather rare tion, it is involved in inflammatory response [12, 20].
compared to the adult population, where this location repre- C11orf95–RELA fusion-positive ependymomas have a dense-
sents the most frequent manifestation site [1, 2]. Regarding ly capillarized phenotype [20, 25], presumably due to the
histological classification, anaplastic ependymoma is more vasogenic action of the NF-κB signaling pathway, as well as
frequent in pediatric patients, while myxopapillary a clear cell morphology [26]. Regardless of the specific fusion
ependymomas are mostly found in the adult population [1]; gene, elevated nuclear protein expression of RelA protein can
subependymomas are extremely rare in childhood and will be identified by immunohistochemistry, offering an accessible
therefore not be discussed in this review. cost-effective tool for detection [20, 27]. Furthermore, com-
Regarding treatment, complete resection and radiation ther- pared to fusion detection by RT-PCR, immunohistochemistry
apy are administered in a higher frequency in pediatric pa- can identify all different fusions at once using antibodies
tients [1]; however, some bias may be owed to the fact that against p65/RELA (with higher specificity compared to
most controlled studies for ependymoma are exclusively in- L1CAM) [27], while RT-PCR-based analysis requires specif-
cluding pediatric patients. ic primers for each individual fusion transcript.
With respect to genetic differences, infants with posterior Regarding prognosis, lower survival rates for ST-EPN-
fossa tumors more frequently seem to have balanced genetic RELA ependymoma compared to other ependymal tumors
profiles compared to polyploid ones in older patients [7, 15]. in this location were reported in a retrospective analysis, in-
Furthermore, the gain of chromosome 1q seems to be only cluding adult and pediatric patients [11]. However, these find-
slightly more frequent in the pediatric population [16], while ings are challenged by data derived from exclusively pediatric
PF-EPN-A seem to be merely exclusive to the pediatric pop- cohorts enrolled in three recent prospective trials, showing no
ulation [17]. RELA-fusion as well as CDKN2A deletion may prognostic disadvantage for ST_EPN-RELA when compared
be detected in pediatric as well as adult patients; however, to other ependymoma subtypes [28–30]. In the revised WHO
differences in frequency have not been analyzed in detail so classification of CNS tumors of 2016, supratentorial
far [9, 11, 12, 18–20]. ependymomas with C11orf95–RELA fusions were introduced
as a novel specific tumor entity [31], initiating the era of a
genetically based ependymoma classification.
Supratentorial ependymoma
Supratentorial ependymoma with YAP1-MAMLD1
Supratentorial ependymomas are more frequently diagnosed in fusion
young patients and show a decreasing incidence with age [1].
On MRI supratentorial ependymomas appear as large, A small subset of pediatric supratentorial ependymomas harbor a
inhomogeneously contrast-enhancing tumors with frequent YAP1-MAMLD1 fusion. This fusion almost exclusively occurs in
Childs Nerv Syst

Table 1 Identification of prognostic parameters in pediatric ependymoma (review of the literature)

Study Location No. of patients (n), Results

pediatric (p), adult
(a), mixed (m)

Scheil et al. ST, PF, n=22; m Chromosome 1q gain only present in pediatric EP, N-MYC amplification present in adult spinal EP
(2001) SP
Dyer et al. ST, PF, n=53; p Characterization of 3 genomic EP types with possible prognostic relevance: numerical, structural
(2002) SP associated with chromosome 1q gain and an inferior prognosis, balanced predominant in infancy
Taylor et al. ST, PF, n=103; m Spatially defined radial glia cells are the cell of origin of ependymomas, with different genetic
(2005) SP characteristics despite histological resemblance
Benesch et al. SP n=29; p Extent of resection was the strongest predictor of outcome mainly PFS, the role of adjuvant
(2010) radio-chemotherapy still needs to be defined, due to risk of recurrence close surveillance is
warranted
Korshunov et al. ST, PF, n=122; m EPN with chromosome 1q gain and homozygous CDKN2A deletion showed an inferior prognosis
(2010) SP
Witt et al. (2011) ST, PF, n=177; m PF-EPN may be divided into two groups A and B with A showing a lateral growth pattern, cerebellar
SP invasion, and a poor prognosis event-free and overall survival, while B may be predominant in
older patients, representing with a less invasive/aggressive growth and a favorable prognosis
Kilday et al. PF n=48; p Pediatric PF-EPN with chromosome 1q gain and incomplete resection showed an inferior prognosis
(2012)
Godfraind et al. ST, PF n=146; p Residual tumor, gain of chromosome 1q, and high mitotic activity were predictors for an inferior
(2012) prognosis in pediatric PF-EPN, alike were brain invasion and homozygous CDKN2A deletion in
ST-EPN
Parker et al. ST n=77; m Characterization of ST EPN with RELA and YAP fusions
(2014)
Pietsch et al. ST n=17; p Characterization of ST EPN with RELA fusion in pediatric patients
(2014)
Mack et al. PF n=47; m PF-EPN may be divided into two epigenetically defined groups according to CpG island methylator
(2014) phenotype, with the methylated one depicting a poor prognosis PF-EPN-A
Pajtler et al. ST, PF, n=498; m Methylation-based classification, identifying 9 subgroups stratified according to location: SP-SE,
(2015) SP SP-MPE, SP-EPN, PF-SE, PF-EPN-A, PF-EPN-B, ST-SE, ST-EPN-YAP, ST-EPN-RELA;
ST-EPN-RELA and PF-EPN-A with poor and ST-EPN-YAP, PF-EPN-B, and SE with favorable
prognosis
Bayliss et al. ST, PF, n=20; m Epigenetically driven H3K27me3 loss may be detected by immunohistochemistry to identify
(2016) SP pediatric PF-EPN-A, which showed an inferior prognosis
Ramaswamy PF n=820; m PF-EPN-A displayed a dismal prognosis compared to PF-EPN-B, which may be cured by surgery
et al. (2016) alone
Panwalkar et al. PF n=460; m H3K27me3 immunohistochemistry reliably identified PF-EPN-A
(2017)
Cavalli et al. PF n=212; m Characterization of 5 PF-EPN-B subgroups with individual methylation profiles associated with an
(2018) overall favorable prognosis but possible late relapses across all groups; extent of resection rep-
resents the strongest prognostic parameter across all PF-EPN-B groups
Fukuoka et al. ST, PF n=107; m Characterization of ST-EPN without RELA and YAP fusions but alternative fusion-transcripts; val-
(2018) idation of poor prognosis in PF-EPN-A
Pajtler et al. PF n=675; m Methylation-based subclassification of EPN-PF-A with distinct prognosis: PFA-1c enriched by
(2018) chromosome 1q gain showed a poor prognosis, while PFA-2c with OTX2 expression displayed a
favorable prognosis, and PFA-1f may harbor H3K27me3 mutations
Witt et al. (2018) ST, PF, n=122; a PF-EPN-A are almost absent in adult patients, EPN in adulthood showed a favorable prognosis,
SP molecular classification may provide a more precise classification beyond histology
Andreiuolo et al. ST n=15; p Characterization of pediatric ST-EPN with YAP fusion with a favorable prognosis
(2019)
Benesch et al. ST, PF n=10 (primarily Primarily metastatic EPN were rare in pediatric patients, ST-EPN-RELA with radio-chemotherapy
(2019) metastatic); p showed a better prognosis compared to PF-EPN- A
Ghasemi et al. SP n=13; m N-MYC amplification occurred in spinal EP II and III and depicts a negative predictor of outcome
(2019)
Jünger et al. PF n=134; p Integrated risk stratification for pediatric PF-EPN with chromosome 1q gain, incomplete resection,
(2019) and high mitotic activity as negative prognostic parameters
Swanson et al. SP n=4; m N-MYC amplification occurred in spinal EP III and was an indicator for poor PFS and OS
(2019) disregarding administered therapy
 Childs Nerv Syst

Table 1 (continued)

Study Location No. of patients (n), Results

pediatric (p), adult
(a), mixed (m)

Jünger et al. ST n=54; p Identification of CDKN2A deletion as adverse prognostic factor in pediatric ST-EPN-RELA
(2020a)
Jünger et al. ST, PF n=28; p (<3 years) Infant PF-EPN-A showed completely balanced genomes, CDKN2A deletion were absent in very
(2020b) young children with ST-EPN
Neumann et al. ST, PF, n=48; m Integrated diagnostic methods in EP are warranted since histology alone may be insufficient;
(2020) SP ependymoma with clear cell and papillary morphology may represent distinct tumors with respect
to morphology, location, and methylation profile, while tanycytic ependymoma may not do so
Pagès et al. ST n=40; p Characterization of ST-EPN with(out) RELA and YAP fusions or SE-features;
(2020) Ependymal/subependymal mixed tumors showed a favorable prognosis
Ritzmann et al. ST, PF n=302; p No significant differences in PFS rates between ST and PF, most relapses occurred within the first 5
(2020) years, late relapses, however, did occur. Chromosome 1q gain and PF-EPN-A were indicators for
relapse. Current therapy regimes (resection+ irradiation) insufficiently prolonged survival after
relapse
Zschernack et al. ST n=18; p Characterization of ST-EPN without RELA and YAP fusions with RELA- like or tanycytic phenotype
(2021) and alternative fusion-transcripts

young children and shows an excellent prognosis, possibly even tumors may be sub-stratified, according to their predominant
without further therapy after resection [6, 11, 12, 30]. The in- histological appearance and biology as RELA-like, tanycytic,
volved fusion partners YES-associated protein 1 (Yap1) and and astroblastoma-like variants. Novel fusion transcripts may
mastermind-like domain containing 1 (Mamld1) are involved be encountered among these tumors as well [24].
in several pathways, e.g., Hippo signaling pathway and Wnt/β-
catenin pathway, active in various cancers [6].
Posterior fossa ependymoma
CDKN2A deletion in supratentorial ependymoma
Pediatric ependymomas are frequently located in the posterior
Besides the two aforementioned gene fusions, a frequently fossa [1, 2]. In addition to histological grading distinguishing
reported genetic alteration associated with poorer prognosis between classic (WHO grade II) and anaplastic ependymomas
in (pediatric) supratentorial ependymomas is homozygous de- (WHO grade III), there is an association between midline and
letion of CDKN2A [9, 11, 18, 19]. The CDKN2A locus on lateral tumor localization and prognosis.
chromosome 9p21 encodes the tumor suppressor proteins On MRI, posterior fossa ependymoma may appear as ho-
p14ARF and p16INK4A, which control the cell cycle via mogenous and well-demarcated tumors with hemorrhage and
Mdm2/p53 or CDK4/6 and the retinoblastoma (RB) family possible calcification spots showing variable contrast en-
of proteins [32]. Consequently, a deletion/inactivation of hancement due to necrosis and cyst formation. Tumors may
CDKN2A may result in uncontrolled cell growth. However, be located inside the fourth ventricle with possible lateral ex-
only a limited number of cases have been reported so far, and pansion through the foramina of Luschka or the foramen of
therefore general conclusions should be drawn with caution. Magendie.
In childhood ependymomas, CDKN2A deletion is restricted to With GTR being the strongest predictor of outcome [8, 29,
ependymomas with RELA fusion [19]. 33–35], tumors arising from the floor and the lateral aspect of
the fourth ventricle have a worse prognosis than those arising
Supratentorial ependymoma without C11orf95–RELA from the roof. One explanation is that achieving GTR without
or YAP1-MAMLD1 fusions (non-RELA/non-YAP postoperative deficits is more difficult [36] in the former
ependymomas) group. Another explanation may be that the location is asso-
ciated with different (epi-)genetically driven growth patterns
Finally, supratentorial ependymomas harboring neither [14, 37].
C11orf95–RELA nor YAP1-MAMLD1 fusions have been re- Analysis of genomic copy number profiles revealed dis-
ported. Some of these tumors may represent histological tinct cytogenetic patterns with different prognostic impact.
mimics of ependymoma misdiagnosed as ependymomas. Tumors displaying partial chromosomal alterations (structural
However, the remaining ones may represent truly novel enti- alterations) had the worst prognosis, especially those harbor-
ties and need to be characterized in more detail [22–24]. These ing a gain of the q-arm of chromosome 1 [7–9, 33]. Tumors
 Childs Nerv Syst

Table 2 Published pediatric ependymoma patient cohorts enrolled in clinical trials (review of the literature)

Study No. of patients (n) EFS/OS Prognostic factors Implications

Grill et al. (2001) Multi-center n=73 (<5 years) 4-year EFS/OS OS: Chemotherapy only or to delay radiation therapy may be
22%/59% EOR, St tumors suitable for a subset of tumors; however, results were
not competitive when compared to other studies
Timmermann et al. (2005) Multi-center n=34 (<3 years) 3-year EFS/OS EFS: Radiation therapy (in individual cases including the
27.3%/55.9% EOR neuroaxis) should be administered
Merchant et al. (2009) Single center n=153 5-year EFS/OS EFS: Maximal safe surgery and high dose (54–59.4Gy) at an
74%/85.0% gender, age, EOR early age (12 months) achieved good results with low
OS: risk for 2nd malignancies and brainstem necrosis; age
EOR, WHO grade, and sex may be used as risk stratification in future trials
ethnic group
Upadhyaya et al. (2019) Multi-center n=54 (<3 years) 4-year EFS/OS EFS: Radiation therapy may be feasible in young children (54Gy)
75.1%/92.6% EOR (incl. re-resection),
PFA + chromosome 1q gain
OS:
-
Merchant et al. (2019) Multi-center n=356 5-year EFS/OS EFS: Some supratentorial tumors may be cured by complete
62.7%/83.8% EOR, chromosome 1q gain resection and observation alone, early postoperative
in PF-EPN, WHO grade, Gender radiation therapy (54–59.4Gy) is beneficial, also for
patients younger than 3 years
Massimino et al. Multi-center n=160 5-year EFS/OS EFS: Re-resection is warranted, in case of residual tumor a boost
(2016 + 2020) 65.4%/81.1% EOR (incl. re-resection), WHO of 8Gy (additional to 59.4Gy) may be beneficial and
10-year EFS/OS grade, gender; PF-EPN-A, CDKN2A feasible, future trials should include molecular
58%/73% deletion, chromosome 1q gain classifications for risk stratification
OS:
EOR, WHO grade, VP-shunt, gender,
age, PF-EPN-A, CDKN2A deletion,
chromosome 1q gain chromosome 1q
gain and CDKN2A deletion may be
more frequent in children >3 years and are
associated with a higher risk of dissemination
 Childs Nerv Syst

with a balanced genetic profile without any chromosomal al- these tumors, avoiding the long-term sequelae of
terations were associated with a slightly better prognosis and radiotherapy.
occur predominantly in very young children [7, 15, 16]. In
contrast, whole chromosomal alterations (numerical genetic
profile) were found in tumors of older patients and adult— Spinal ependymoma
mostly spinal—tumors and associated with a better prognosis
[7, 16]. In contrast to the adult population, spinal ependymomas are
Epigenetically, ependymomas of the posterior fossa can be rare in children [1, 2]. Furthermore, pediatric patients suffer-
divided into two defined groups [10, 11, 14], which are ing from ependymomas in this location seem to be generally
discussed in detail in the following paragraphs. older than those with supratentorial or infratentorial tumors [1,
2]. Spinal ependymomas can be histologically classified into
myxopapillary ependymoma WHO grade I, as well as classic
PF-A ependymoma
(WHO grade II) and the less frequent anaplastic ependymoma
(WHO grade III) [31, 42, 43]. Additionally, a rare and aggres-
PF-A ependymomas occur predominantly in younger children
sive genetic variant with MYCN amplification has been de-
with a decreasing frequency in adolescence and near absence
scribed [44–46].
in adulthood. There is a male predominance, and male sex is
associated with poorer prognosis [11, 14, 37, 38]. PF-A tu-
Classic and anaplastic spinal ependymoma
mors show a predominantly balanced genetic profile without
frequent recurrent somatic mutations, but gain of the long arm
Classic and anaplastic ependymomas arise predominantly in
of chromosome 1 may be noted in 17–25 % [9, 11, 14, 33, 37].
the cervicothoracic cord and are located intramedullary [42].
However, in infants below the age of 18 months, chromosom-
On MRI, they appear as T1-hypo- and T2-hyperintense
al aberrations may be absent [15]. Furthermore, CpG-island
intramedullary, contrast-enhancing lesions with frequent cys-
hypermethylation (compared to PF-B tumors) is present in
tic, hemorrhagic, and necrotic components as well as possible
PF-A tumors [10, 11, 14], caused by overexpression of
calcifications. About 60 % are associated with syringomyelia
EZHIP (CXorf67; enhancer of Zeste inhibitory protein) [37].
[47]. Classic and anaplastic spinal ependymomas have a pre-
By binding EZH2, EZHIP silences targets of the polycomb
dominantly numerical cytogenetic profile, similar to PF-B
repressive complex 2 which represses the expression of dif-
ependymomas and are associated with a better prognosis [7,
ferentiation genes through trimethylation of H3K27 [10].
16]. The most frequent chromosomal alteration in spinal
Indeed, reduced or absent H3K27me3 expression is character-
ependymomas is the loss of the long arm of chromosome 22
istic of PF-A ependymomas [39, 40]. The loss of
[11], and the most frequent somatically mutated gene in clas-
trimethylation of H3K27 can be evaluated by immunohisto-
sic and anaplastic spinal ependymoma is NF2, located on
chemistry, which offers a cost-effective and readily accessible
chromosome 22q [48], while NF2 is neither mutated in intra-
tool to characterize posterior fossa ependymomas. Eventually,
cranial nor myxopapillary ependymomas. Epigenetically,
DNA methylation analysis revealed further possible subdivi-
based on a mixed and predominantly adult cohort, spinal
sion of posterior fossa ependymomas [37, 41]. Pajtler et al.
ependymomas were subdivided into three groups SP-SE,
published a subdivision of PF-A ependymomas into 2 main
SP-EPN, and SP-MPE [11]. Spinal ependymomas show a
and 9 epigenetically defined subgroups, characterized by dif-
methylation signature distinct from subependymomas,
ferent frequencies of H3K27M mutations, gain of chromo-
myxopapillary ependymomas, and MYCN-amplified
some 1q, and levels of CXorf67 or OTX2 expression [37].
(anaplastic) ependymomas [11, 46] (see below). Overall, pa-
tients with spinal ependymomas have a favorable prognosis
PF-B ependymoma regardless of age with progression-free and overall survival
rates of 70–90 and 90–100 %, respectively [49] as well as
Compared to PF-A, PF-B tumors are more frequently found in limited evidence of a worse prognosis for anaplastic tumors.
older children and adult patients and frequently show relative-
ly hypomethylated genomes and a polyploid chromosomal Myxopapillary ependymoma
profile with gains of complete chromosomes, retained
H3K27me3, and slight female predominance [10, 11, 14, 17, The name is derived from their histopathological features
37–41]. Cavalli et al. were able to identify five different epi- which include perivascular “sleeves” of myxoid ground sub-
genetically defined PF-B subgroups distinct from stance, microcysts, and radial (pseudopapillary) arrangements
subependymomas, PF-A, and spinal ependymomas [41]. of the tumor cells around vessels. In contrast to classic and
The relatively favorable prognosis of these tumors constitutes anaplastic tumors, myxopapillary ependymomas are mostly
a rationale for reduction of adjuvant therapy after surgery for found in the conus-cauda region in intradural but
Childs Nerv Syst

extramedullary location [42, 50] and are relatively rare in pe- 30, 34]. The mainstay of approved adjuvant treatment has
diatric patients [51]. Multifocal growth as well as locations been postoperative radiation therapy regardless of location
other than the conus-cauda region [52, 53] and extra spinal and extend of resection [34, 35]. Weighing possible long-
manifestation in the sacrococcygeal soft tissue [54] have been term sequelae against tumor recurrence, the cut-off age for
described. Furthermore, with a given tendency to radiation therapy is currently debated; it is performed as early
leptomeningeal dissemination, craniospinal MRI is warranted as 12 months in recently published studies [29, 30, 34, 35].
[55]. The most favorable single center results so far reported for a
The diagnosis of myxopapillary ependymoma is straight- cohort of pediatric ependymoma with supra- and infratentorial
forward if typical histological features are present. However, location were achieved by Merchant and colleagues [35] pub-
even if these are absent, their distinct methylation profile will lished in 2009 with 5-year EFS and OS of 74 % and 85.0 %,
establish the diagnosis, since it differentiates them from respectively, while promising multi-center results were pub-
classic/anaplastic ependymomas [11, 17, 26]. Of note, rare lished by Massimino et al. in 2016 [34] with 5-year EFS and
anaplastic variants with an increased mitotic index as well as OS of 65.4 % and 81.1 % and consecutive 10-year EFS and
vascular proliferation and necrosis have been described [54]; OS rates of 58 % and 73 % [28, 34], respectively; the
these features do not appear to influence survival, though [52]. largest cohort published to date (n=356) also achieved
Overall survival of myxopapillary ependymoma is favorable good 5-year EFS and OS rates of 62.7 % and 83.8 %,
with >90 % at 10 years [51, 56, 57]. Progression-free survival respectively [29]. Since local recurrence is reported
is less favorable, with frequent persistent and recurrent disease more frequently than distant failure [34, 62, 63],
as well as dissemination sometimes already present at diagno- Massimino et al. advocated local intensification of RT
sis [56, 57]. In incompletely resected cases, adjuvant radio- with an extra boost on the tumor region [34].
therapy improves the prognosis [58].
Genetic parameters, prognostic, and possible
Spinal ependymoma with MYCN amplification therapeutic implications

A rare and highly aggressive subtype of anaplastic spinal cord Regarding the different emerging biologically defined sub-
ependymoma with MYCN amplification predominantly af- types, Merchant et al. recommended an observation-only ap-
fecting young adults but older children as well has been iden- proach for completely resected WHO grade II supratentorial
tified recently, further supporting the prognostic relevance of tumors with promising results [29]. A similar approach for PF-
biological classification. Their site of predilection is the B-type tumors was suggested by Ramaswamy et al. [38]. The
cervico-thoracic spine [44–46, 59], and tumors are often large option of observation only for certain subentities is currently
at diagnosis, involving multiple spinal segments and investigated in the SIOP-Ependymoma II trial and in the
displaying an exophytic growth pattern as well as frequent ACNS0831 trial. In contrast to PF-B tumors, PF-A-type
dissemination [44, 46, 59]. Histologically, ependymomas with ependymoma with gain of the q-arm of chromosome 1
MYCN amplification present with characteristic ependymal showed a dismal prognosis and needed maximal safe surgery
features and signs of anaplasia [44, 46, 59]. MYCN protein as well as radiation- and chemotherapy, especially since local
overexpression can be demonstrated by immunohistochemis- and distant failure and the tendency for dissemination are
try [44, 59]. Genetically, all cases show high-level MYCN more frequently reported compared to other ependymoma
amplification, which remains stable throughout the course of subtypes [28, 29]. For supratentorial ependymoma, the nega-
disease [44–46, 59]. This subtype also shows a distinct meth- tive prognostic impact of RELA-fusion-positive ependymoma
ylation profile [44, 59]. The cases reported so far showed that was initially reported [11] could not be validated by others
unfavorable progression-free and overall survival. Therefore, [29, 30]. On the other hand, recent data by Jünger et al. and
their distinction from other spinal ependymomas is essential. Massimino et al. suggest that supratentorial ependymoma
with CDKN2A deletion represents a more aggressive variant,
possibly requiring intensified treatment [19, 28]. Given
Treatment strategies according to the results Palbociclib, an oral inhibitor of CDK4/CDK6, is currently
of latest trial cohorts being tested for clinical efficacy in high-grade gliomas with
amplification of CDK4/CDK6 or homozygous deletion of
Published trial cohorts CDKN2A [64], CDKN2A deletion in pediatric supratentorial
ependymomas should be the subject of further investigation
The most important factor in the treatment of pediatric since it might represent a therapeutic target. Based on the
ependymoma, regardless of age and location, is maximal safe advancing characterization of the biological features of
surgery [28–30, 33–35, 60, 61], ideally GTR, and may include ependymoma, targeted therapies may be identified and proven
second surgery, whenever needed and considered feasible [28, in a clinical setting in the future [65]. However, for now,
 Childs Nerv Syst

maximal safe surgery remains the upfront goal and postoper- have a higher rate of local recurrence than classic grade II
ative radiation therapy the most effective treatment. tumors, GTR seems to be the only predictor of progression-
Furthermore, despite ependymoma being divided into biolog- free and overall survival. Relapse rates for pediatric patients
ically distinct sub-entities, WHO grading was used for risk seem to be generally higher when compared to adult patients
stratification in most trials and confirmed to be a significant [66].
predictor of prognosis [28, 29, 34, 35]. Gender and age might
be prognostically relevant as well [29, 34, 35]. However, the
prognostic impact of age may be due to the fact that all treat- Treatment of primarily metastatic ependymoma and
ment modalities are not necessarily used for very young chil- recurrent ependymoma
dren in all departments.
CNS metastases in ependymoma are rare at presentation [31].
Radio- and chemotherapy Benesch et al. recently published data of ten such pediatric
patients (2.4 % of 402) enrolled in the multi-center HIT-
Regarding proton versus photon therapy, the two modalities 2000 ependymoma trial [67]. The four patients with ST-
are considered of equal efficacy. Regarding sequelae, proton EPN-RELA were alive in first or second remission 7.5–12.3
beam therapy, however, may be more suitable for years after initial diagnosis, while all 4 with PF-EPN-A had
ependymoma typically arising at very young age in order to died, indicating that chemotherapy may be required for these
spare the highly vulnerable surrounding developing brain [62, patients and that genetic subgrouping may adequately predict
63]. For clinical validation, further evaluation in prospective prognosis [67]. With regard to radiotherapy, in disseminated
trial cohorts with longer follow-up is needed. ependymomas, the full craniospinal volume will be targeted
To date, chemotherapy according to different protocols by the radiotherapy, whereas in localized tumor, only the treat-
was often administered in children with incompletely ment of tumor bed is standard of care [55].
resected or anaplastic (WHO grade III) tumors in addition For recurrent ependymomas, the EANO suggests reopera-
to radiotherapy [29, 30, 34]. In very young children, chemo- tion and reirradiation if possible as well as chemotherapy. If
therapy was applied with the aim to delay radiation therapy. no local treatment is feasible, chemotherapy can be adminis-
However, efficacy of chemotherapy could not be proven tered alone [55]. However, prognosis of children with recur-
[60, 61]. The value of chemotherapy in ependymoma thera- rent ependymoma remains poor [68]. A recently published
py is currently being re-evaluated in the ongoing SIOP retrospective analysis reported an increased risk for recurrence
Ependymoma II trial. Furthermore, the ACNS 0831 trial, in patients with tumors with gain of chromosome 1q and/or
also investigating the possible benefit of chemotherapy, has PFA-methylation profile, further demonstrating only little
shown first promising results, indicating a possible benefit of benefit from resurgery and reirradiation in case of relapse [69].
maintenance chemotherapy regarding EFS (presented at
ISPNO 2020).

Therapy of spinal ependymoma Conclusion

Data on treatment protocols specific for pediatric patients with In conclusion, pediatric ependymomas comprise biologi-
spinal ependymomas are sparse, since pediatric trials focused cally distinct tumor entities with different (epi)genetics,
on intracranial ependymoma and the majority of larger retro- age distribution and localization, as well as different
spective analyses described mixed or adult cohorts. Data from prognosis. However, histological classification within
the prospective HIT-trial on 29 spinal ependymomas have these biologically defined entities is still relevant for
shown gross-total resection as the main predictor of favorable risk stratification. Furthermore, molecular classification
outcome [49]. However, uniform adjuvant treatment protocols gains importance and will be addressed in the upcoming
for pediatric spinal ependymomas do not exist [49] and the WHO classification of CNS tumors.
relevance of postoperative radio- and chemotherapy is not To date the mainstay of treatment remains maximal
proven. Nevertheless, the European Association of Neuro- safe surgery (ideally GTR), achieved with intraoperative
Oncology (EANO) recommended postoperative radiotherapy monitoring and neuronavigation and if necessary second
for all anaplastic (WHO grade III) tumors and incompletely surgery. Adjuvant radiation therapy as soon as justifi-
resected classic (WHO grade II) tumors [55]. For able, given potential harm to the surrounding nervous
myxopapillary ependymomas, radiotherapy was recommend- tissue, is recommended for most tumors; however, there
ed even after complete resection as well as in relapsed cases is substantial evidence that it might be spared in defined
after—if feasible—second surgery in combination with che- subentities, while the role of chemotherapy still needs to
motherapy [55]. For myxopapillary ependymomas which be established.
Childs Nerv Syst

Remaining open questions DiLuna M (2020) Comparison of epidemiology, treatments, and

outcomes in pediatric versus adult ependymoma. Neurooncol Adv
2(vdaa019). https://doi.org/10.1093/noajnl/vdaa019
Since pediatric ependymomas depict rare entities, all children 2. McGuire CS, Sainani KL, Fisher PG (2009) Incidence patterns for
should be treated in clinical trials, analyzing demographic, ependymoma: a surveillance, epidemiology, and end results study.
clinical, histological, molecular, and treatment-derived param- J Neurosurg 110:725–729. https://doi.org/10.3171/2008.9.
JNS08117
eter in order to answer the questions remaining to date:
3. Ostrom QT, Gittleman H, Liao P, Vecchione-Koval T, Wolinsky
Y, Kruchko C, Barnholtz-Sloan JS (2017) CBTRUS statistical re-
i) Is a mere molecular risk stratification superior to classic port: primary brain and other central nervous system tumors diag-
histopathological classification to differentiate between nosed in the United States in 2010–2014. Neuro-Oncology 19:v1–
v88. https://doi.org/10.1093/neuonc/nox158
low- and high-risk tumors?
4. Ellison DW, Kocak M, Figarella-Branger D, Felice G, Catherine G,
ii) Does surgery alone depict sufficient treatment for low- Pietsch T, Frappaz D, Massimino M, Grill J, Boyett JM, Grundy
risk tumors? RG (2011) Histopathological grading of pediatric ependymoma:
iii) What time, dose, and extent are suitable for which type reproducibility and clinical relevance in European trial cohorts.
Journal of Negative Results in BioMedicine 10. https://doi.org/10.
of tumor in terms of photon and proton radiation; and
1186/1477-5751-10-7
which modality may be more suitable? 5. Tihan T, Zhou T, Holmes E, Burger PC, Ozuysal S, Rushing EJ
iv) What is the role of classic chemotherapy in pediatric (2008) The prognostic value of histological grading of posterior
ependymoma? fossa ependymomas in children: a Children’s Oncology Group
v) With potential therapeutic targets identified, what will study and a review of prognostic factors. Modern Pathology 21:
165–177. https://doi.org/10.1038/modpathol.3800999
be the role of targeted therapy? 6. Andreiuolo F, Varlet P, Tauziède-Espariat A, Jünger ST, Dörner E,
vi) Which is the optimal treatment for recurrent and pri- Dreschmann V, Kuchelmeister K, Waha A, Haberler C, Slavc I,
marily metastatic pediatric ependymoma; do we need a Corbacioglu S, Riemenschneider MJ, Leipold A, Rüdiger T,
more aggressive treatment regime? Körholz D, Acker T, Russo A, Faber J, Sommer C, Armbrust S,
Rose M, Erdlenbruch B, Hans VH, Bernbeck B, Schneider D,
Lorenzen J, Ebinger M, Handgretinger R, Neumann M, van
Buiren M, Prinz M, Roganovic J, Jakovcevic A, Park SH, Grill J,
Acknowledgements We thank Dr. Steffen Albrecht, Montreal, for criti- Puget S, Messing-Jünger M, Reinhard H, Bergmann M, Hattingen
cal reading of the manuscript. E, Pietsch T (2019) Childhood supratentorial ependymomas with
YAP1-MAMLD1 fusion: an entity with characteristic clinical, ra-
Data availability statement The data that support the findings of this diological, cytogenetic and histopathological features. Brain Pathol
study are available on reasonable request from the corresponding author. 29:205–216. https://doi.org/10.1111/bpa.12659
7. Dyer S, Prebble E, Davison V, Davies P, Ramani P, Ellison D,
Funding Open Access funding enabled and organized by Projekt DEAL. Grundy R (2002) Genomic imbalances in pediatric intracranial
The study was supported by grants from the German Children Cancer ependymomas define clinically relevant groups. The American
Foundation DKS 2006.03, 2009.19, 2011.01 and 2014.17 to TP. Journal of Pathology 161:2133–2141. https://doi.org/10.1016/
S0002-9440(10)64491-4
8. Kilday J-P, Mitra B, Domerg C, Ward J, Andreiuolo F, Osteso-
Declarations Ibanez T, Mauguen A, Varlet P, le Deley MC, Lowe J, Ellison
DW, Gilbertson RJ, Coyle B, Grill J, Grundy RG (2012) Copy
Conflict of interest On behalf of all authors, the corresponding author number gain of 1q25 predicts poor progression-free survival for
states that there is no conflict of interest. pediatric intracranial ependymomas and enables patient risk strati-
fication: a prospective European clinical trial cohort analysis on
Open Access This article is licensed under a Creative Commons behalf of the Children’s Cancer Leukaemia Group (CCLG),
Attribution 4.0 International License, which permits use, sharing, adap- Societe Francaise d’Oncologie Pediatrique (SFOP), and
tation, distribution and reproduction in any medium or format, as long as International Society for Pediatric Oncology (SIOP). Clin Cancer
you give appropriate credit to the original author(s) and the source, pro- Res 18:2001–2011. https://doi.org/10.1158/1078-0432.CCR-11-
vide a link to the Creative Commons licence, and indicate if changes were 2489
made. The images or other third party material in this article are included 9. Korshunov A, Witt H, Hielscher T, Benner A, Remke M, Ryzhova
in the article's Creative Commons licence, unless indicated otherwise in a M, Milde T, Bender S, Wittmann A, Schöttler A, Kulozik AE, Witt
credit line to the material. If material is not included in the article's O, von Deimling A, Lichter P, Pfister S (2010) Molecular staging of
Creative Commons licence and your intended use is not permitted by intracranial ependymoma in children and adults. Journal of Clinical
statutory regulation or exceeds the permitted use, you will need to obtain Oncology 28:3182–3190. https://doi.org/10.1200/JCO.2009.27.
permission directly from the copyright holder. To view a copy of this 3359
licence, visit http://creativecommons.org/licenses/by/4.0/. 10. Mack SC, Witt H, Piro RM, Gu L, Zuyderduyn S, Stütz AM, Wang
X, Gallo M, Garzia L, Zayne K, Zhang X, Ramaswamy V, Jäger N,
Jones DTW, Sill M, Pugh TJ, Ryzhova M, Wani KM, Shih DJH,
Head R, Remke M, Bailey SD, Zichner T, Faria CC, Barszczyk M,
Stark S, Seker-Cin H, Hutter S, Johann P, Bender S, Hovestadt V,
References Tzaridis T, Dubuc AM, Northcott PA, Peacock J, Bertrand KC,
Agnihotri S, Cavalli FMG, Clarke I, Nethery-Brokx K, Creasy
1. Elsamadicy AA, Koo AB, David WB, Lee V, Zogg CK, CL, Verma SK, Koster J, Wu X, Yao Y, Milde T, Sin-Chan P,
Kundishora AJ, Hong CS, DeSpenza T, Reeves BC, Kahle KT, Zuccaro J, Lau L, Pereira S, Castelo-Branco P, Hirst M, Marra
 Childs Nerv Syst

MA, Roberts SS, Fults D, Massimi L, Cho YJ, van Meter T, Network (2018) DNA methylation-based classification of
Grajkowska W, Lach B, Kulozik AE, von Deimling A, Witt O, ependymomas in adulthood: implications for diagnosis and treat-
Scherer SW, Fan X, Muraszko KM, Kool M, Pomeroy SL, Gupta ment. Neuro-oncology 20:1616–1624. https://doi.org/10.1093/
N, Phillips J, Huang A, Tabori U, Hawkins C, Malkin D, neuonc/noy118
Kongkham PN, Weiss WA, Jabado N, Rutka JT, Bouffet E, 18. Godfraind C, Kaczmarska JM, Kocak M, Dalton J, Wright KD,
Korbel JO, Lupien M, Aldape KD, Bader GD, Eils R, Lichter P, Sanford RA, Boop FA, Gajjar A, Merchant TE, Ellison DW
Dirks PB, Pfister SM, Korshunov A, Taylor MD (2014) (2012) Distinct disease-risk groups in pediatric supratentorial and
Epigenomic alterations define lethal CIMP-positive ependymomas posterior fossa ependymomas. Acta Neuropathol 124:247–257.
of infancy. Nature 506:445–450. https://doi.org/10.1038/ https://doi.org/10.1007/s00401-012-0981-9
nature13108 19. Jünger ST, Andreiuolo F, Mynarek M, Wohlers I, Rahmann S,
11. Pajtler KW, Witt H, Sill M, Jones DTW, Hovestadt V, Kratochwil Klein-Hitpass L, Dörner E, zur Mühlen A, Velez-Char N, von
F, Wani K, Tatevossian R, Punchihewa C, Johann P, Reimand J, Hoff K, Warmuth-Metz M, Kortmann RD, Timmermann B, von
Warnatz HJ, Ryzhova M, Mack S, Ramaswamy V, Capper D, Bueren A, Rutkowski S, Pietsch T (2020) CDKN2A deletion in
Schweizer L, Sieber L, Wittmann A, Huang Z, van Sluis P, supratentorial ependymoma with RELA alteration indicates a dis-
Volckmann R, Koster J, Versteeg R, Fults D, Toledano H, mal prognosis: a retrospective analysis of the HIT ependymoma
Avigad S, Hoffman LM, Donson AM, Foreman N, Hewer E, trial cohort. Acta Neuropathol 140:405–407. https://doi.org/10.
Zitterbart K, Gilbert M, Armstrong TS, Gupta N, Allen JC, 1007/s00401-020-02169-z
Karajannis MA, Zagzag D, Hasselblatt M, Kulozik AE, Witt O, 20. Pietsch T, Wohlers I, Goschzik T, Dreschmann V, Denkhaus D,
Collins VP, von Hoff K, Rutkowski S, Pietsch T, Bader G, Yaspo Dörner E, Rahmann S, Klein-Hitpass L (2014) Supratentorial
ML, von Deimling A, Lichter P, Taylor MD, Gilbertson R, Ellison ependymomas of childhood carry C11orf95-RELA fusions leading
DW, Aldape K, Korshunov A, Kool M, Pfister SM (2015) to pathological activation of the NF-κB signaling pathway. Acta
Molecular classification of ependymal tumors across all CNS com- Neuropathol 127:609–611. https://doi.org/10.1007/s00401-014-
partments, histopathological grades, and age groups. Cancer Cell 1264-4
27:728–743. https://doi.org/10.1016/j.ccell.2015.04.002 21. Nowak J, Jünger ST, Huflage H, Seidel C, Hohm A, Vandergrift
12. Parker M, Mohankumar KM, Punchihewa C, Weinlich R, Dalton LA, von Hoff K, Rutkowski S, Pietsch T, Warmuth-Metz M (2019)
JD, Li Y, Lee R, Tatevossian RG, Phoenix TN, Thiruvenkatam R, MRI phenotype of RELA-fused pediatric supratentorial
White E, Tang B, Orisme W, Gupta K, Rusch M, Chen X, Li Y, ependymoma. Clin Neuroradiol 29:595–604. https://doi.org/10.
Nagahawhatte P, Hedlund E, Finkelstein D, Wu G, Shurtleff S, 1007/s00062-018-0704-2
Easton J, Boggs K, Yergeau D, Vadodaria B, Mulder HL, 22. Fukuoka K, Kanemura Y, Shofuda T et al (2018) Significance of
Becksfort J, Gupta P, Huether R, Ma J, Song G, Gajjar A, molecular classification of ependymomas: C11orf95-RELA fusion-
Merchant T, Boop F, Smith AA, Ding L, Lu C, Ochoa K, Zhao negative supratentorial ependymomas are a heterogeneous group of
D, Fulton RS, Fulton LL, Mardis ER, Wilson RK, Downing JR, tumors. Acta Neuropathol Commun 6:134. https://doi.org/10.1186/
Green DR, Zhang J, Ellison DW, Gilbertson RJ (2014) C11orf95– s40478-018-0630-1
RELA fusions drive oncogenic NF-κB signalling in ependymoma. 23. Pagès M, Pajtler KW, Puget S, Castel D, Boddaert N, Tauziède-
Nature 506:451–455. https://doi.org/10.1038/nature13109 Espariat A, Picot S, Debily MA, Kool M, Capper D, Sainte-Rose C,
13. Taylor MD, Poppleton H, Fuller C, Su X, Liu Y, Jensen P, Chrétien F, Pfister SM, Pietsch T, Grill J, Varlet P, Andreiuolo F
Magdaleno S, Dalton J, Calabrese C, Board J, MacDonald T, (2019) Diagnostics of pediatric supratentorial RELA
Rutka J, Guha A, Gajjar A, Curran T, Gilbertson RJ (2005) ependymomas: integration of information from histopathology, ge-
Radial glia cells are candidate stem cells of ependymoma. Cancer netics, DNA methylation and imaging. Brain Pathol 29:325–335.
Cell 8:323–335. https://doi.org/10.1016/j.ccr.2005.09.001 https://doi.org/10.1111/bpa.12664
14. Witt H, Mack SC, Ryzhova M, Bender S, Sill M, Isserlin R, Benner 24. Zschernack V, Jünger ST, Mynarek M, Rutkowski S, Garre ML,
A, Hielscher T, Milde T, Remke M, Jones DT, Northcott PA, Ebinger M, Neu M, Faber J, Erdlenbruch B, Claviez A, Bielack S,
Garzia L, Bertrand KC, Wittmann A, Yao Y, Roberts SS, Brozou T, Frühwald MC, Dörner E, Dreschmann V, Stock A,
Massimi L, van Meter T, Weiss WA, Gupta N, Grajkowska W, Solymosi L, Hench J, Frank S, Vokuhl C, Waha A, Andreiuolo F,
Lach B, Cho YJ, von Deimling A, Kulozik AE, Witt O, Bader Pietsch T (2021) Supratentorial ependymoma in childhood: more
GD, Hawkins CE, Tabori U, Guha A, Rutka JT, Lichter P, than just RELA or YAP. Acta Neuropathol 141:455–466. https://
Korshunov A, Taylor MD, Pfister SM (2011) Delineation of two doi.org/10.1007/s00401-020-02260-5
clinically and molecularly distinct subgroups of posterior fossa 25. Figarella-Branger D, Lechapt-Zalcman E, Tabouret E, Jünger S, de
ependymoma. Cancer Cell 20:143–157. https://doi.org/10.1016/j. Paula AM, Bouvier C, Colin C, Jouvet A, Forest F, Andreiuolo F,
ccr.2011.07.007 Quintin-Roue I, Machet MC, Heitzmann A, Milin S, Sevestre H,
15. Jünger ST, Andreiuolo F, Mynarek M, Dörner E, zur Mühlen A, Godfraind C, Labrousse F, Metellus P, Scavarda D, Pietsch T
Rutkowski S, von Bueren AO, Pietsch T (2020) Ependymomas in (2016) Supratentorial clear cell ependymomas with branching cap-
infancy: underlying genetic alterations, histological features, and illaries demonstrate characteristic clinicopathological features and
clinical outcome. Childs Nerv Syst 36:2693–2700. https://doi.org/ pathological activation of nuclear factor-kappaB signaling. Neuro-
10.1007/s00381-020-04655-x Oncology 18:919–927. https://doi.org/10.1093/neuonc/now025
16. Kilday J-P, Rahman R, Dyer S, Ridley L, Lowe J, Coyle B, Grundy 26. Neumann JE, Spohn M, Obrecht D, Mynarek M, Thomas C,
R (2009) Pediatric ependymoma: biological perspectives. Mol Hasselblatt M, Dorostkar MM, Wefers AK, Frank S, Monoranu
Cancer Res 7:765–786. https://doi.org/10.1158/1541-7786.MCR- CM, Koch A, Witt H, Kool M, Pajtler KW, Rutkowski S, Glatzel
08-0584 M, Schüller U (2020) Molecular characterization of histopatholog-
17. Witt H, Gramatzki D, Hentschel B, Pajtler KW, Felsberg J, ical ependymoma variants. Acta Neuropathol 139:305–318. https://
Schackert G, Löffler M, Capper D, Sahm F, Sill M, von doi.org/10.1007/s00401-019-02090-0
Deimling A, Kool M, Herrlinger U, Westphal M, Pietsch T, 27. Gessi M, Giagnacovo M, Modena P, Elefante G, Gianno F,
Reifenberger G, Pfister SM, Tonn JC, Weller M, German Glioma Buttarelli FR, Arcella A, Donofrio V, Diomedi Camassei F,
Childs Nerv Syst

Nozza P, Morra I, Massimino M, Pollo B, Giangaspero F, Antonelli Gandola L (2016) Final results of the second prospective AIEOP
M (2019) Role of immunohistochemistry in the identification of protocol for pediatric intracranial ependymoma. Neuro-oncology
supratentorial C11ORF95-RELA fused ependymoma in routine 18:1451–1460. https://doi.org/10.1093/neuonc/now108
neuropathology. Am J Surg Pathol 43:56–63. https://doi.org/10. 35. Merchant TE, Li C, Xiong X, Kun LE, Boop FA, Sanford RA
1097/PAS.0000000000000979 (2009) Conformal radiotherapy after surgery for paediatric
28. Massimino M, Barretta F, Modena P, Witt H, Minasi S, Pfister SM, ependymoma: a prospective study. The Lancet Oncology 10:258–
Pajtler KW, Antonelli M, Gandola L, Luisa Garrè M, Bertin D, 266. https://doi.org/10.1016/S1470-2045(08)70342-5
Mastronuzzi A, Mascarin M, Quaglietta L, Viscardi E, Sardi I, 36. Merchant TE (2002) Current management of childhood
Ruggiero A, Pollo B, Buccoliero A, Boschetti L, Schiavello E, ependymoma. Oncology (Williston Park, NY) 16:629–642, 644;
Chiapparini L, Erbetta A, Morra I, Gessi M, Donofrio V, discussion 645-646, 648
Patriarca C, Giangaspero F, Johann P, Buttarelli FR (2020) The 37. Pajtler KW, Wen J, Sill M, Lin T, Orisme W, Tang B, Hübner JM,
AIEOP 2nd series of children and adolescents intracranial Ramaswamy V, Jia S, Dalton JD, Haupfear K, Rogers HA,
ependymoma. An integrated molecular and clinical characterization Punchihewa C, Lee R, Easton J, Wu G, Ritzmann TA, Chapman
with a long-term follow-up. Neuro Oncol 23:848–857. https://doi. R, Chavez L, Boop FA, Klimo P, Sabin ND, Ogg R, Mack SC,
org/10.1093/neuonc/noaa257 Freibaum BD, Kim HJ, Witt H, Jones DTW, Vo B, Gajjar A,
29. Merchant TE, Bendel AE, Sabin ND, Burger PC, Shaw DW, Chang Pounds S, Onar-Thomas A, Roussel MF, Zhang J, Taylor JP,
E, Wu S, Zhou T, Eisenstat DD, Foreman NK, Fuller CE, Anderson Merchant TE, Grundy R, Tatevossian RG, Taylor MD, Pfister
ET, Hukin J, Lau CC, Pollack IF, Laningham FH, Lustig RH, SM, Korshunov A, Kool M, Ellison DW (2018) Molecular hetero-
Armstrong FD, Handler MH, Williams-Hughes C, Kessel S, geneity and CXorf67 alterations in posterior fossa group A (PFA)
Kocak M, Ellison DW, Ramaswamy V (2019) Conformal radiation ependymomas. Acta Neuropathol 136:211–226. https://doi.org/10.
therapy for pediatric ependymoma, chemotherapy for incompletely 1007/s00401-018-1877-0
resected ependymoma, and observation for completely resected, 38. Ramaswamy V, Hielscher T, Mack SC, Lassaletta A, Lin T, Pajtler
supratentorial ependymoma. Journal of Clinical Oncology 37: KW, Jones DTW, Luu B, Cavalli FMG, Aldape K, Remke M,
974–983. https://doi.org/10.1200/JCO.18.01765 Mynarek M, Rutkowski S, Gururangan S, McLendon RE, Lipp
30. Upadhyaya SA, Robinson GW, Onar-Thomas A, Orr BA, Billups ES, Dunham C, Hukin J, Eisenstat DD, Fulton D, van
CA, Bowers DC, Bendel AE, Hassall T, Crawford JR, Partap S, Landeghem FKH, Santi M, van Veelen MLC, van Meir EG,
Fisher PG, Tatevossian RG, Seah T, Qaddoumi IA, Vinitsky A, Osuka S, Fan X, Muraszko KM, Tirapelli DPC, Oba-Shinjo SM,
Armstrong GT, Sabin ND, Tinkle CL, Klimo P, Indelicato DJ, Marie SKN, Carlotti CG, Lee JY, Rao AAN, Giannini C, Faria CC,
Boop FA, Merchant TE, Ellison DW, Gajjar A (2019) Molecular Nunes S, Mora J, Hamilton RL, Hauser P, Jabado N, Petrecca K,
grouping and outcomes of young children with newly diagnosed Jung S, Massimi L, Zollo M, Cinalli G, Bognár L, Klekner A,
ependymoma treated on the multi-institutional SJYC07 trial. Hortobágyi T, Leary S, Ermoian RP, Olson JM, Leonard JR,
Neuro-oncology. 21:1319–1330. https://doi.org/10.1093/neuonc/ Gardner C, Grajkowska WA, Chambless LB, Cain J, Eberhart
noz069 CG, Ahsan S, Massimino M, Giangaspero F, Buttarelli FR,
31. Louis et al (2016) WHO classification of tumours of the central Packer RJ, Emery L, Yong WH, Soto H, Liau LM, Everson R,
nervous system, Revised, 4th edn. International Agency for Grossbach A, Shalaby T, Grotzer M, Karajannis MA, Zagzag D,
Research on Cancer, Lyon Wheeler H, von Hoff K, Alonso MM, Tuñon T, Schüller U,
32. Shirahata M, Ono T, Stichel D, Schrimpf D, Reuss DE, Sahm F, Zitterbart K, Sterba J, Chan JA, Guzman M, Elbabaa SK, Colman
Koelsche C, Wefers A, Reinhardt A, Huang K, Sievers P, Shimizu H, Dhall G, Fisher PG, Fouladi M, Gajjar A, Goldman S, Hwang E,
H, Nanjo H, Kobayashi Y, Miyake Y, Suzuki T, Adachi JI, Kool M, Ladha H, Vera-Bolanos E, Wani K, Lieberman F,
Mishima K, Sasaki A, Nishikawa R, Bewerunge-Hudler M, Mikkelsen T, Omuro AM, Pollack IF, Prados M, Robins HI,
Ryzhova M, Absalyamova O, Golanov A, Sinn P, Platten M, Soffietti R, Wu J, Metellus P, Tabori U, Bartels U, Bouffet E,
Jungk C, Winkler F, Wick A, Hänggi D, Unterberg A, Pfister Hawkins CE, Rutka JT, Dirks P, Pfister SM, Merchant TE,
SM, Jones DTW, van den Bent M, Hegi M, French P, Baumert Gilbert MR, Armstrong TS, Korshunov A, Ellison DW, Taylor
BG, Stupp R, Gorlia T, Weller M, Capper D, Korshunov A, Herold- MD (2016) Therapeutic impact of cytoreductive surgery and irra-
Mende C, Wick W, Louis DN, von Deimling A (2018) Novel, diation of posterior fossa ependymoma in the molecular era: a ret-
improved grading system(s) for IDH-mutant astrocytic gliomas. rospective multicohort analysis. J Clin Oncol 34:2468–2477.
Acta Neuropathologica 136:153–166. https://doi.org/10.1007/ https://doi.org/10.1200/JCO.2015.65.7825
s00401-018-1849-4 39. Bayliss J, Mukherjee P, Lu C, Jain SU, Chung C, Martinez D,
33. Jünger ST, Mynarek M, Wohlers I, Dörner E, Mühlen A, Velez- Sabari B, Margol AS, Panwalkar P, Parolia A, Pekmezci M,
Char N, von Hoff K, Rutkowski S, Warmuth-Metz M, Kortmann McEachin RC, Cieslik M, Tamrazi B, Garcia BA, la Rocca G,
RD, Timmermann B, Rahmann S, Klein-Hitpass L, von Bueren Santi M, Lewis PW, Hawkins C, Melnick A, David Allis C,
AO, Pietsch T (2019) Improved risk-stratification for posterior fos- Thompson CB, Chinnaiyan AM, Judkins AR, Venneti S (2016)
sa ependymoma of childhood considering clinical, histological and Lowered H3K27me3 and DNA hypomethylation define poorly
genetic features - a retrospective analysis of the HIT ependymoma prognostic pediatric posterior fossa ependymomas. Sci Transl
trial cohort. Acta Neuropathol Commun 7:181. https://doi.org/10. Med 8:366ra161–366ra161. https://doi.org/10.1126/scitranslmed.
1186/s40478-019-0820-5 aah6904
34. Massimino M, Miceli R, Giangaspero F, Boschetti L, Modena P, 40. Panwalkar P, Clark J, Ramaswamy V, Hawes D, Yang F, Dunham
Antonelli M, Ferroli P, Bertin D, Pecori E, Valentini L, Biassoni V, C, Yip S, Hukin J, Sun Y, Schipper MJ, Chavez L, Margol A,
Garrè ML, Schiavello E, Sardi I, Cama A, Viscardi E, Scarzello G, Pekmezci M, Chung C, Banda A, Bayliss JM, Curry SJ, Santi M,
Scoccianti S, Mascarin M, Quaglietta L, Cinalli G, Diletto B, Rodriguez FJ, Snuderl M, Karajannis MA, Saratsis AM, Horbinski
Genitori L, Peretta P, Mussano A, Buccoliero A, Calareso G, CM, Carret AS, Wilson B, Johnston D, Lafay-Cousin L, Zelcer S,
Barra S, Mastronuzzi A, Giussani C, Marras CE, Balter R, Eisenstat D, Silva M, Scheinemann K, Jabado N, McNeely PD,
Bertolini P, Giombelli E, la Spina M, Buttarelli FR, Pollo B, Kool M, Pfister SM, Taylor MD, Hawkins C, Korshunov A,
 Childs Nerv Syst

Judkins AR, Venneti S (2017) Immunohistochemical analysis of 50. Cervoni L, Celli P, Caruso R, Gagliardi FM, Cantore GP (1997)
H3K27me3 demonstrates global reduction in group-A childhood Neurinomas and ependymomas of the cauda equina. A review of
posterior fossa ependymoma and is a powerful predictor of out- the clinical characteristics. Minerva Chir 52:629–633
come. Acta Neuropathol 134:705–714. https://doi.org/10.1007/ 51. Bates JE, Choi G, Milano MT (2016) Myxopapillary ependymoma:
s00401-017-1752-4 a SEER analysis of epidemiology and outcomes. J Neurooncol 129:
41. Cavalli FMG, Hübner J-M, Sharma T, Luu B, Sill M, Zapotocky 251–258. https://doi.org/10.1007/s11060-016-2167-0
M, Mack SC, Witt H, Lin T, Shih DJH, Ho B, Santi M, Emery L, 52. Sonneland PR, Scheithauer BW, Onofrio BM (1985)
Hukin J, Dunham C, McLendon RE, Lipp ES, Gururangan S, Myxopapillary ependymoma. A clinicopathologic and immunocy-
Grossbach A, French P, Kros JM, van Veelen MLC, Rao AAN, tochemical study of 77 cases. Cancer 56:883–893. https://doi.org/
Giannini C, Leary S, Jung S, Faria CC, Mora J, Schüller U, Alonso 10.1002/1097-0142
MM, Chan JA, Klekner A, Chambless LB, Hwang EI, Massimino 53. Yust Katz S, Cachia D, Kamiya-Matsuoka C, Olar A, Theeler B,
M, Eberhart CG, Karajannis MA, Lu B, Liau LM, Zollo M, Penas Prado M, Gilbert MR, Armstrong T (2018) Ependymomas
Ferrucci V, Carlotti C, Tirapelli DPC, Tabori U, Bouffet E, arising outside of the central nervous system: a case series and
Ryzhova M, Ellison DW, Merchant TE, Gilbert MR, Armstrong literature review. J Clin Neurosci 47:202–207. https://doi.org/10.
TS, Korshunov A, Pfister SM, Taylor MD, Aldape K, Pajtler KW, 1016/j.jocn.2017.10.026
Kool M, Ramaswamy V (2018) Heterogeneity within the PF-EPN- 54. Lee JC, Sharifai N, Dahiya S, Kleinschmidt-DeMasters BK,
B ependymoma subgroup. Acta Neuropathol 136:227–237. https:// Rosenblum MK, Reis GF, Samuel D, Siongco AM, Santi M,
doi.org/10.1007/s00401-018-1888-x Storm PB, Ferris SP, Bollen AW, Pekmezci M, Solomon DA,
42. Celano E, Salehani A, Malcolm JG, Reinertsen E, Hadjipanayis CG Tihan T, Perry A (2019) Clinicopathologic features of anaplastic
(2016) Spinal cord ependymoma: a review of the literature and case myxopapillary ependymomas. Brain Pathol 29:75–84. https://doi.
series of ten patients. J Neurooncol 128:377–386. https://doi.org/ org/10.1111/bpa.12673
10.1007/s11060-016-2135-8 55. Rudà R, Reifenberger G, Frappaz D, Pfister SM, Laprie A,
43. Engelhard HH, Villano JL, Porter KR, Stewart AK, Barua M, Santarius T, Roth P, Tonn JC, Soffietti R, Weller M, Moyal ECJ
Barker FG, Newton HB (2010) Clinical presentation, histology, (2018) EANO guidelines for the diagnosis and treatment of
and treatment in 430 patients with primary tumors of the spinal ependymal tumors. Neuro-oncology 20:445–456. https://doi.org/
cord, spinal meninges, or cauda equina. J Neurosurg Spine 13: 10.1093/neuonc/nox166
67–77. https://doi.org/10.3171/2010.3.SPINE09430 56. Abdallah A, Emel E, Gündüz HB, Sofuoğlu ÖE, Asiltürk M,
44. Ghasemi DR, Sill M, Okonechnikov K, Korshunov A, Yip S, Abdallah BG (2020) Long-term surgical resection outcomes of
Schutz PW, Scheie D, Kruse A, Harter PN, Kastelan M, Wagner pediatric myxopapillary ependymoma: experience of two centers
M, Hartmann C, Benzel J, Maass KK, Khasraw M, Sträter R, and brief literature review. World Neurosurg 136:e245–e261.
Thomas C, Paulus W, Kratz CP, Witt H, Kawauchi D, Herold- https://doi.org/10.1016/j.wneu.2019.12.128
Mende C, Sahm F, Brandner S, Kool M, Jones DTW, von 57. Bandopadhayay P, Silvera VM, Ciarlini PDSC, Malkin H, Bi WL,
Deimling A, Pfister SM, Reuss DE, Pajtler KW (2019) MYCN Bergthold G, Faisal AM, Ullrich NJ, Marcus K, Scott RM,
amplification drives an aggressive form of spinal ependymoma. Beroukhim R, Manley PE, Chi SN, Ligon KL, Goumnerova LC,
Acta Neuropathol 138:1075–1089. https://doi.org/10.1007/ Kieran MW (2016) Myxopapillary ependymomas in children: im-
s00401-019-02056-2 aging, treatment and outcomes. J Neurooncol 126:165–174. https://
45. Scheil S, Brüderlein S, Eicker M, Herms J, Herold-Mende C, doi.org/10.1007/s11060-015-1955-2
Steiner HH, Barth TFE, Möller P (2001) Low frequency of chro- 58. Weber DC, Wang Y, Miller R, Villa S, Zaucha R, Pica A,
mosomal imbalances in anaplastic ependymomas as detected by Poortmans P, Anacak Y, Ozygit G, Baumert B, Haller G,
comparative genomic hybridization. Brain Pathol 11:133–143. Preusser M, Li J (2015) Long-term outcome of patients with spinal
https://doi.org/10.1111/j.1750-3639.2001.tb00386.x myxopapillary ependymoma: treatment results from the MD
46. Swanson AA, Raghunathan A, Jenkins RB, Messing-Jünger M, Anderson Cancer Center and institutions from the Rare Cancer
Pietsch T, Clarke MJ, Kaufmann TJ, Giannini C (2019) Spinal cord Network. Neuro-oncology 17:588–595. https://doi.org/10.1093/
ependymomas with MYCN amplification show aggressive clinical neuonc/nou293
behavior. J Neuropathol Exp Neurol 78:791–797. https://doi.org/ 59. Raffeld M, Abdullaev Z, Pack SD, Xi L, Nagaraj S, Briceno N,
10.1093/jnen/nlz064 Vera E, Pittaluga S, Lopes Abath Neto O, Quezado M, Aldape K,
47. Koeller KK, Rosenblum RS, Morrison AL (2000) Neoplasms of the Armstrong TS, Gilbert MR (2020) High level MYCN amplification
spinal cord and filum terminale: radiologic-pathologic correlation. and distinct methylation signature define an aggressive subtype of
Radiographics 20:1721–1749. https://doi.org/10.1148/ spinal cord ependymoma. Acta Neuropathol Commun 8:101.
radiographics.20.6.g00nv151721 https://doi.org/10.1186/s40478-020-00973-y
48. Ebert C, von Haken M, Meyer-Puttlitz B, Wiestler OD, 60. Grill J, Le Deley M-C, Gambarelli D et al (2001) Postoperative
Reifenberger G, Pietsch T, von Deimling A (1999) Molecular ge- chemotherapy without irradiation for ependymoma in children un-
netic analysis of ependymal tumors. NF2 mutations and chromo- der 5 years of age: a multicenter trial of the French Society of
some 22q loss occur preferentially in intramedullary spinal Pediatric Oncology. Journal of Clinical Oncology 19:1288–1296.
ependymomas. Am J Pathol 155:627–632. https://doi.org/10. https://doi.org/10.1200/JCO.2001.19.5.1288
1016/S0002-9440(10)65158-9 61. Timmermann B, Kortmann R-D, Kühl J, Rutkowski S, Dieckmann
49. Benesch M, Weber-Mzell D, Gerber NU, von Hoff K, Deinlein F, K, Meisner C, Bamberg M (2005) Role of radiotherapy in anaplas-
Krauss J, Warmuth-Metz M, Kortmann RD, Pietsch T, Driever PH, tic ependymoma in children under age of 3 years: results of the
Quehenberger F, Urban C, Rutkowski S (2010) Ependymoma of prospective German brain tumor trials HIT-SKK 87 and 92.
the spinal cord in children and adolescents: a retrospective series Radiother Oncol 77:278–285. https://doi.org/10.1016/j.radonc.
from the HIT database. J Neurosurg Pediatr 6:137–144. https://doi. 2005.10.016
org/10.3171/2010.5.PEDS09553 62. MacDonald SM, Safai S, Trofimov A et al (2008) Proton radiother-
apy for childhood ependymoma: initial clinical outcomes and dose
Childs Nerv Syst

comparisons. Int J Radiat Oncol, Biol, Phys 71:979–986. https:// targeting of ependymoma as informed by oncogenic enhancer pro-
doi.org/10.1016/j.ijrobp.2007.11.065 filing. Nature 553:101–105. https://doi.org/10.1038/nature25169
63. Sato M, Gunther JR, Mahajan A, Jo E, Paulino AC, Adesina AM, 66. Feldman WB, Clark AJ, Safaee M, Ames CP, Parsa AT (2013)
Jones JY, Ketonen LM, Su JM, Okcu MF, Khatua S, Dauser RC, Tumor control after surgery for spinal myxopapillary
Whitehead WE, Weinberg J, Chintagumpala MM (2017) ependymomas: distinct outcomes in adults versus children: a sys-
Progression-free survival of children with localized ependymoma tematic review. J Neurosurg Spine 19:471–476. https://doi.org/10.
treated with intensity-modulated radiation therapy or proton-beam 3171/2013.6.SPINE12927
radiation therapy. Cancer 123:2570–2578. https://doi.org/10.1002/ 67. Benesch M, Mynarek M, Witt H, Warmuth-Metz M, Pietsch T,
cncr.30623 Bison B, Pfister SM, Pajtler KW, Kool M, Schüller U,
64. Pfaff E, Kessler T, Balasubramanian GP, Berberich A, Schrimpf D, Pietschmann K, Juhnke BO, Tippelt S, Fleischhack G, Schmid I,
Wick A, Debus J, Unterberg A, Bendszus M, Herold-Mende C, Kramm CM, Vorwerk P, Beilken A, Classen CF, Hernáiz Driever
Capper D, Schenkel I, Eisenmenger A, Dettmer S, Brors B, P, Kropshofer G, Imschweiler T, Lemmer A, Kortmann RD,
Platten M, Pfister SM, von Deimling A, Jones DTW, Wick W, Rutkowski S, Hoff K (2019) Newly diagnosed metastatic intracra-
Sahm F (2018) Feasibility of real-time molecular profiling for pa- nial ependymoma in children: frequency, molecular characteristics,
tients with newly diagnosed glioblastoma without MGMT promot- treatment, and outcome in the prospective HIT series. Oncologist
er hypermethylation-the NCT Neuro Master Match (N2M2) pilot 24:e921–e929. https://doi.org/10.1634/theoncologist.2018-0489
study. Neuro-oncology 20:826–837. https://doi.org/10.1093/ 68. Byer L, Kline CN, Coleman C, Allen IE, Whitaker E, Mueller S
neuonc/nox216 (2019) A systematic review and meta-analysis of outcomes in pe-
65. Mack SC, Pajtler KW, Chavez L, Okonechnikov K, Bertrand KC, diatric, recurrent ependymoma. J Neurooncol 144:445–452. https://
Wang X, Erkek S, Federation A, Song A, Lee C, Wang X, doi.org/10.1007/s11060-019-03255-3
McDonald L, Morrow JJ, Saiakhova A, Sin-Chan P, Wu Q, 69. Ritzmann TA, Rogers HA, Paine SML, Storer LCD, Jacques TS,
Michaelraj KA, Miller TE, Hubert CG, Ryzhova M, Garzia L, Chapman RJ, Ellison D, Donson AM, Foreman NK, Grundy RG
Donovan L, Dombrowski S, Factor DC, Luu B, Valentim CLL, (2020) A retrospective analysis of recurrent pediatric ependymoma
Gimple RC, Morton A, Kim L, Prager BC, Lee JJY, Wu X, reveals extremely poor survival and ineffectiveness of current treat-
Zuccaro J, Thompson Y, Holgado BL, Reimand J, Ke SQ, ments across central nervous system locations and molecular sub-
Tropper A, Lai S, Vijayarajah S, Doan S, Mahadev V, Miñan AF, groups. Pediatr Blood Cancer 67:e28426. https://doi.org/10.1002/
Gröbner SN, Lienhard M, Zapatka M, Huang Z, Aldape KD, pbc.28426
Carcaboso AM, Houghton PJ, Keir ST, Milde T, Witt H, Li Y, Li
CJ, Bian XW, Jones DTW, Scott I, Singh SK, Huang A, Dirks PB,
Bouffet E, Bradner JE, Ramaswamy V, Jabado N, Rutka JT, Publisher’s note Springer Nature remains neutral with regard to jurisdic-
Northcott PA, Lupien M, Lichter P, Korshunov A, Scacheri PC, tional claims in published maps and institutional affiliations.
Pfister SM, Kool M, Taylor MD, Rich JN (2018) Therapeutic