Childs Nerv Syst (2016) 32:1895–1906

DOI 10.1007/s00381-016-3148-5

SPECIAL ANNUAL ISSUE

Clinical considerations and surgical approaches for low-grade

gliomas in deep hemispheric locations: thalamic lesions
Tai-Tong Wong 1,2 & Hsin-Hung Chen 2 & Muh-Lii Liang 2 & Kevin Li-Chun Hsieh 3,4 &
Yi-Shan Yang 2 & Donald Ming-Tak Ho 5 & Kai-Ping Chang 6 & Yi-Yen Lee 2 &
Shih-Chieh Lin 5 & Ting-Rong Hsu 6 & Yi-Wei Chen 7 & Sang-Hue Yen 7 & Feng-Chi Chang 8 &
Wan-You Guo 8 & Kuo-Wei Chen 9 & Wei-Kang Kwang 10 & Wu-Yu Hou 11 &
Chung-Yih Wang 12

Received: 4 June 2016 / Accepted: 12 June 2016

# Springer-Verlag Berlin Heidelberg 2016

Abstract 75, 2007; Albright, J Neurosurg 100(5 Suppl Pediatrics): 468–

Background Tumors with epicenter in the thalamus occur in 472, 2004; Kelly, Neurosurgery 25:185–195, 1989; Drake
about 4 % of pediatric brain tumors. The histological diagno- et al., Neurosurgery 29: 27–33, 1991).
sis is mainly gliomas. Among them, low-grade glioma (LGG) Materials and methods Sixty-nine cases of thalamic tumors in
constituted of a significant entity of the tumors (Cuccia et al., children were retrospectively reviewed. There were 25 cases
Childs Nerv Syst 13:514–521, 1997; Puget et al., J Neurosurg of LGGs. We analyzed our experience and correlated it with
106:354–362, 2007; Bernstein et al., J Neurosurg 61:649– reported series.
656, 1984; Bilginer et al., Childs Nerv Syst 30:1493–1498, Results Summing up of 4 reported series and the present se-
2014). Since Kelly’s report in 1989, >90 % resection of tha- ries, there were 267 cases of thalamic tumors in children.
lamic tumors were achieved in reported series (Ozek and Ture, Among these tumors, 107 (40.1 %) were LGGs and 91
Childs Nerv Syst 18:450–6, 2002; Villarejo et al., Childs Nerv (34.1 %) were low-grade astrocytomas (LGAs). In the present
Syst 10:111–114, 1994; Moshel et al., Neurosurgery 61:66– series, all of the 25 LGGs were LGAs that consisted of 11

Tai-Tong Wong, Hsin-Hung Chen, Muh-Lii Liang, and Kevin Li-Chun

Hsieh contributed equally to this work

* Tai-Tong Wong 6
Department of Pediatrics, Taipei Veterans General Hospital and
ttwong99@tmu.edu.tw; ttwong99@gmail.com National Yang Ming University, School of Medicine, Sec 2 Shi Pai
Rd, Taipei 11217, Taiwan
Hsin-Hung Chen 7
Department of Oncology, Taipei Veterans General Hospital and
roberthhchen3@gmail.com National Yang Ming University, School of Medicine, Sec 2 Shi Pai
Rd, Taipei 11217, Taiwan
1
Department of Neurosurgery, Taipei Medical University Hospital, 8
Department of Radiology, Taipei Veterans General Hospital and
Taipei Medical University, 252 Wuxing St, Taipei 11031, Taiwan
National Yang Ming University, School of Medicine, Sec 2 Shi Pai
2
Department of Neurosurgery, Neurological Institute, Taipei Veterans Rd, Taipei 11217, Taiwan
General Hospital and National Yang Ming University, School of 9
Medicine, Sec 2 Shi Pai Rd, Taipei 11217, Taiwan Department of Internal Medicine, Cheng Hsin General Hospital,
Taipei, Taiwan
3
Department of Medical Imaging, Taipei Medical University Hospital, 10
Taipei, Taiwan Department of Pathology, Cheng Hsin General Hospital,
4
Taipei, Taiwan
Translational Imaging Research Center, College of Medicine, Taipei
11
Medical University, Taipei, Taiwan Department of Radiology, Chen Hsin General Hospital,
5 Taipei, Taiwan
Department of Pathology and Laboratory Medicine, Taipei Veterans
12
General Hospital and National Yang Ming University, School of Department of Radiation Oncology, Cheng Hsin General Hospital,
Medicine, Sec 2 Shi Pai Rd, Taipei 11217, Taiwan Taipei, Taiwan
1896 Childs Nerv Syst (2016) 32:1895–1906

pilocytic astrocytomas (PAs) and 14 diffuse astrocytomas segment of the nucleus of tumor origination with distortion
(DAs). Six cases received biopsy sampling only. The remain- of surrounding structures, (2) the tumor expanded beyond the
ing 19 cases received different degrees of surgical resection thalamus into the surrounding white matter, and (3) the tumor
via several approaches. Radical (>90 %) resection was expanded into the lateral ventricle [5]. Moshel et al. classified
achieved better in PAs comparing with DAs. There was no thalamic tumors into three anatomical locations: anterior thal-
operative mortality. Two patients had increased neurological amus, posterior superior thalamus, and posterior inferior thal-
deficits. In a mean follow-up period of 11.9 years, three pa- amus in 2007 [7]. Rangel-Castilla and Spetzler in 2015 further
tients died of tumor progression and one patient died of ana- divided the thalamus into six anatomical regions and
plastic change. The 5- and 10-year overall survival (OS) was established six surgical procedures accordingly [21]. These
87.1 and 87.1 %, respectively. approaches can be applied for radical resection of thalamic
Conclusion Thalamic LGGs are mainly LGAs and are indo- tumor resection in specific locations. Except bilateral thalamic
lent. The rate of >90 % resection was relatively low in the tumors, in reports series since 1989, >90 % resection could be
present series. By applying contemporary diagnostic MRI achieved in the majority of selective thalamic gliomas [5–10].
studies, surgical facilities, and appropriate approaches in se- Biopsy sampling is for infiltrative tumors involving internal
lective cases, we may try maximum neuroprotective radical capsule and basal ganglia [5, 22]. However, based on
(>90 %) resection. neuronavigation, tractography of the corticospinal tract, and
IONM, maximum neuroprotective radical resection of infiltra-
Keywords Thalamic tumor . Low-grade glioma . Children . tive tumor can also be performed selectively [5, 17, 18].
Surgical approach . Outcome Hydrocephalus with or without tumor excision may be treated
by selective ventriculoperitoneal (VP) shunt or endoscopic
third ventriculostomy (ETV). The selection is based on wheth-
Introduction er the tumor obliterates CSF flow at the level of foramen
Monro or aqueduct [2, 22, 23]. In this paper, we reviewed
Thalamic tumors are deep-seated with the epicenter locating our experience on 25 low-grade thalamic gliomas in a hospital
in the unilateral or bilateral thalamus. The clinical presenta- series of thalamic tumors in children.
tions are mainly increased intracranial pressure due to tumor
mass, obstructive hydrocephalus, motor deficits, sensory dys-
functions, visual problem, and other symptoms consisting of Materials and methods
involuntary movement, spasticity, seizures, behavior problem,
and central pain [2, 3]. The duration of clinical symptoms is From 1971 to 2015, 1529 cases of primary intracranial tumors
usually short within few months [3]. Other than computed in children aged ≤18 years were treated in Taipei Veterans
tomography (CT), magnetic resonance imaging (MRI) is the General Hospital (Taipei VGH, 1917–2011), Cheng Hsin
most significant diagnostic neuroimaging modality to define General Hospital (CHGH, 2012–2014), and Taipei Medical
the anatomical location of the tumor in thalamus. The study University Hospital (TMUH, 2015). There were 74 (4.9 %)
may also provide information regarding whether the tumor is cases of tumors with the epicenter in the thalamus. Among
well-circumscribed or infiltrates to the neighboring internal them, 69 cases were primarily treated or received major treat-
capsule, basal ganglia, subthalamic nucleus, and midbrain ments in these hospitals. We retrospectively reviewed these 69
[11, 12]. In addition, modern MRI techniques integrating with cases for tumor location, presence of hydrocephalus at diag-
MRI white matter tractography (diffusion tensor imaging, nosis, biopsy sampling, degree of tumor resection, and histo-
DTI), MRI/iMRI (intraoperative MRI), neuronavigation, and logical diagnosis. These 69 tumors were categorized into uni-
intraoperative neurophysiological monitoring (IONM) help to lateral thalamic, thalamopeduncular, and bilateral thalamic tu-
define the tumor and the surrounding functional structures. mors according to the report of Puget et al. in 2007 [2]. The
Safe stereotactic biopsy and neuroprotective maximum radical degree of tumor resection was based on operation note and/or
resection [11, 13–19] can be achieved. Fluorescence-guided a comparison of preoperative and the first postoperative CT or
surgery with 5-aminolevulinic-acid has been used for maxi- MRI studies. Except biopsy sampling, the degree of tumor
mum resection of malignant thalamic glioma [20]. resections was defined as partial (<50 %), subtotal (>50–
Apart from contemporary surgical facilities, the selection 90 %), near total (>90 % with small residue tumor), or total
of surgical strategies for biopsy sampling or radical resection (no residue tumor). We further focused on LGGs for the pres-
of thalamic tumors are based on the preoperative CT or MRI ence of hydrocephalus at diagnosis, radiographic features, op-
features, such as tumor location, local expansion, direction of eration approaches, biopsy, or other extents of surgical resec-
extension, and infiltration to surrounding structures [11]. tion, clinical outcomes, long-term follow-up, and survival
Yasargil defined three modes of tumor extension and spread- evaluation. The results were correlated with representative
ing in thalamic tumors as (1) the tumor remained in the reported series.
Childs Nerv Syst (2016) 32:1895–1906 1897

Results [Fig. 4]. No tumor extension to mesial temporal region was

observed. Two tumors had cystic component. There was no
Sixty-nine cases of thalamic tumors in children were reviewed perilesional edema. Surgical strategies of these 25 histologi-
and analyzed for our experience on LGGs in thalamus. The cally verified LGAs were biopsy in 6 (24 %), <90 % resection
mean age at diagnosis was 7.8 years (range, 1 day to in 11 (44 %), >90 % resection with residue tumor in 7 (28 %),
17.8 years). The sex distribution was 30 males and 39 females. and complete resection in 1 (4 %) of tumors. The surgical
Thirty-eight (55.1 %) patients presented obstructive hydro- approaches included transfrontal in one, transcallosal in four
cephalus at diagnosis. Referring to Puget’s categorization of (anterior 3, posterior 1), transtemporal in nine (anterior 5, pos-
thalamic tumor in children, there were 59 unilateral thalamic, terior 4), transparietal (posterior superior) in two, transtemporal-
6 thalomopeduncular, and 4 bilateral thalamic tumors. Sixty- occipital (posterior inferior) in one, transtemporal-parietal in one,
one (88.4 %) patients received tumor operations including subtentorial in one, and no data in one tumors. Anterior
biopsy sampling in 13 (18.8 %), partial (<50 %) resection in transtemporal approach was used in five of six tumors with ex-
16 (23.2 %), subtotal (>50–90 %) resection in 16 (23.2 %), tension to mesial temporal region (Table 3) (Fig. 1). Reviewing
near total (>90 %) resection in 13 (18.8 %), and gross total of immediate postoperative status and morbidity, there was nor-
resection in 5 (7.3 %). Of the 61 patients with histological mal motor function in five patients. New occurrence or increased
verification, there were 50 astrocytic tumors (25 low-grade motor deficits developed in two patients. Visual field study be-
tumors and 25 high-grade tumors), 1 anaplastic ependymoma fore and after surgery was not regularly studied. Among 18
(AEP), 8 embryonal tumors, and 2 neuronal tumors. All of the LGAs after incomplete resection, in a mean follow-up of
25 cases of LGGs were LGAs that consisted of 11 PAs and 14 11.3 years (range 1–22.2 years), the residue tumors showed re-
DAs. Among the 26 cases of high-grade gliomas (HGGs), gression in two PA and progression in three tumors including
there were 17 anaplastic astrocytomas (AAs), 8 glioblastomas (one PA, two DAs). Fourteen patients received radiation and/or
(GBMs), and 1 AEP. The mean age of LGAs at diagnosis was chemotherapy for residue tumors after biopsy or surgical resec-
8.0 years (range, 4–15) compared to the mean age of 9.4 years tion. Anaplastic change occurred in two DAs at 0.3 year and
(range, 1.8 to 17.8) of HCGs (Table 1). 0.55 year after the first tumor resections without nonsurgical
Looking at the histological types in the three categories of treatment. Excluding one case of loss of follow-up after the first
thalamic tumors, the distributions of 25 LGAs were 18 unilat- admission, the mean follow-up period of the remaining 24 LGAs
eral thalamic tumors, 6 thalamopeduncular tumors, and 1 was 11.9 years (range, 1–25.9), and 4 patients died of the diseases
bithalamic tumors. In 11 Pas, there were 6 unilateral thalamic with mean follow-up period of 4.3 years (range, 1–11.1). Three
tumors and 5 thalamopeduncular tumors. In 14 DAs, there of them died of progressing residue tumors from 2.1, 2.8, to
were 12 unilateral thalamic, 1 thalamopeducular tumor, and 11.2 years. One of the two patients with anaplastic change died
1 bithalamic tumor (Table 2). Among these 25 thalamic 1 year after diagnosis. The 5-, 10-, and 15-year overall survival
LGAs, 13 (52 %) tumors presented obstructive hydrocephalus (OS) for LGAs was 87.1, 87.1, and 79.9 %, respectively. There
at diagnosis. Twelve patients required shunt procedures in- was no difference between PAs and DAs [Fig. 5].
cluding 11 VP shunt implantations and 3 ETV procedures.
Clinical manifestations in 24 patients with data were
hemiparesis in 14, increased intracranial pressure (IICP) in Discussion
10, visual problem in 4, seizures in 3, and involuntary move-
ment in 2 patients. Nine tumors were diagnosed by CT and 16 Thalamic tumor is an uncommon entity of brain tumors that
tumors were diagnosed by MRI. Reviewing of MRI studies of constitutes 4 % of brain tumors in children [1, 2]. In the pres-
8 PAs, the typical MRI features of monothlamic PAs were ent series, it was 4.9 %. We sum up four series of pediatric
homogeneous and well-circumscribed lesions that was thalamic tumors reported from 1980s to 2010s along with the
hypointense on T1WI, hyperintense on T2WI, and with strong present series, and there were a total of 267 patients [1–4]. The
enhancement in gadolinium-enhanced study [Fig. 1]. Four age at diagnosis of these five series varied from 7.7 to
tumors had cystic component and only one tumor had thin 11.6 years. Obstructive hydrocephalus presented from 46.7
rim of perilesional edema. Extending of the tumors to mesial to 65.2 % of patients at diagnosis in three of the five series.
temporal region in six of eight tumors, to amygdala area in Thalamic tumors in children consist of gliomas, neuronal tu-
five of eight tumors, to cerebral peduncle in five of eight mors, and embryonal tumors. Among tumors of these five
tumors, to ambient cistern in five of eight tumors, and to series, 82 % had histological diagnosis and 18 % were histo-
suprasellar cistern in two of eight was observed [Figs. 2 and logically unverified. LGAs (34.1 %) and HGAs (33.0 %) were
3]. Reviewing of the MRI studies in eight DAs, the typical the two main types of tumors (Table 1). Puget et al. reported
MRI features were homogenous to heterogeneous lesions that 69 cases of thalamic tumors in children and categorized these
showed mild hypointense on T1WI, hyperintense on T2WI, tumors into three groups as unilateral thalamic tumors,
and no or faint enhancement in contrast-enhanced study thalamopeduncular tumors, and bilateral thalamic tumors
 1898

Table 1 Comparison of demography, hydrocephalus at diagnosis, types of tumors in four reported series and the present series of thalamic tumors in children

Tumor type Bernstein et al., (1984) Cuccia et al., (1997) Puget, et al., (2007) Bilginer et al., Wong TT et al., (2016) Total/%
(2014)

No. of cases 58 26 69 45 69 267

Age (mean) 7.7 years 9.2 ± 4 years 9.5 ± 4.4 years (0–16years ) 11.06 ± 4.49 years (3–19 years) 7.8 years (0–17.8 years)
(5 months–18years)
Hydrocephalus at diagnosis ND ND 45 (65.2 %) 21 (46.7) 38 (55.9 %) 104 (56.8 %)
Tumor type
Low-grade 19 9 35 17 27 107 (40.1 %)
High-grade 20 17 25 16 34 112 (41.9 %)
No histological diagnosis 19 0 9 12 8 48 (18.0 %)
Astrocytic Tumor Low grade 17 9 26 14 25 91
PA 14 14 11 (34.1 %)
DA 12 14
High grade 18 13 19 13 25 88 (33.0 %)
AA 13 11 2 17
GBM 5 12 11 8
Oligodendroglioma OL 1 0 2 0 0 8 (3.0 %)
AOL 0 0 4 1 0
Mixed glioma OLA 1 0 5 0 0 8 (3.0 %)
AOLA 0 2 0 0 0
Ependymomal tumor EP 0 0 0 1 0 3 (1.1 %)
AEP 0 0 0 1 1
Neuronal mixed neuronal CNC 0 1 1 5 (1.9 %)
glial tumor DNET 0 0 0 0 1
GG 0 0 1 1 0
Embryonal tumor PNET 2 0 2 1 6 13 (4.9 %)
AT/RT 0 0 0 0 2

ND no data, PA pilocytic astrocytoma, DA diffuse astrocytoma, AA anaplastic astrocytoma, GBM glioblastoma, OL oligodendroglioma, AOL anaplastic oliogodendroglioma, OLA oligoastrocytoma, AOLA
anaplastic oligoastrocytoma, EP ependymoma, AEP anaplastic ependymoma, CNC central neurocytoma, DNET dysembryoplastic neuroepithelial tumor, GG ganglioglioma, PNET primitive
neuroectodermal tumor, AT/RT atypical teratoid/rhabdoid tumor
Childs Nerv Syst (2016) 32:1895–1906
Table 2 Distribution of histological types in three categories of thalamic tumors in children: a comparison between the Puget et al. series and the present series

Category of tumor Unilateral thalamic Thalamopeduncular Bithalamic Total

Series Puget S et al. (2007) Present series Puget S et al. (2007) Present series Puget S et al. (2007) Present series
Childs Nerv Syst (2016) 32:1895–1906

No. of cases 54 59 6 6 9 4 138

Hydrocephalus at diagnosis 34 (63 %) 34 (57.6 %) 5 (83 %) 2 (33.3 %) 6 (67 %) 2 (50 %) 83 (60.1 %)/
Histological diagnosis Unilateral thalamic Thalamopeduncular Bithalamic
Histologically verified 46 52 5 6 9 3 87.7 %
Astrocytic tumor LGA 16 18 4 6 6 1 40 %
PA 9 6 5 0
DA/GIIA 7 (DA) 12 (GIIA) 1 (DA) 1 (DA)
HGA 17 23 1 0 1 2 31.9 %
AA 15 2
GBM 8 0
Oligodendroglioma OL 2 0 0 0 0 0 1.5 %
AOL 2 0 0 0 1 0 2.2 %
Mixed tumor OLA 4 0 0 0 1 0 3.6 %
Ependymal tumor EP 0 0 0 0 0 0 0%
AEP 0 1 0 0 0 0 0.7 %
Neuronal mixed neuronal glial tumor DNET 0 1 0 0 0 0 0.7 %
GG 1 0 0 0 0 0 0.7 %
CNC 1 1 0 0 0 0 1.5 %
Embryonal tumor PNET 2 6 0 0 0 0 5.8 %
AT/RT 0 2 0 0 0 0 1.5 %
Histologically unverified 8 7 1 0 0 1 17 (12.3 %)

LGA low-grade astrocytoma, PA pilocytic astrocytoma, DA diffuse astrocytoma, GIIA grade II astrocytoma, low-grade astrocytoma other than pilocytic astrocytoma, HGA high grade astrocytoma,
malignant astrocytoma, AA anaplastic astrocytoma, GBM glioblastoma, OL oligoastrocytoma, AOL anaplastic oligodendroglioma, EP ependymoma, AEP anaplastic ependymoma, DNET
dysembryoplastic neuroepithelial tumor, GG ganglioglioma, CNC central neurocytoma, PNET primitive neuroectodermal tumor, AT/RT atypical teratoid/rhabdoid tumor
1899
1900 Childs Nerv Syst (2016) 32:1895–1906

Fig. 1 a–l Eight-year- and 2-

month-old boy with a left
thalamopeduncular PA. Axial
T1WI (A-C), axial T2WI (D-F),
and axial contrast enhanced T1WI
(G-I) revealed a well-
circumscribed and largely cystic
neoplastic lesion extending into
the ipsilateral amygdala region
and ambient cistern. Follow-up
contrast enhanced T1WI (j–l)
showed small residue after >90 %
tumor resection through a left
anterior transtemporal approach

[2]. Comparing with Puget’s series, we did not have >95 % to total resection in 21 (53.8 %) [24]. Albright in
oligodenglioma (OL) and mixed glioma. All of the LGGs in 2004 summed up five series of pediatric thalamic tumors in-
the present series were PAs (11 cases) and DAs (14 cases) volved in surgical resection. In a total of 65 tumors, there were
(Table 2). 46 low-grade tumors and 19 high-grade tumors. Among them,
The attitude toward the resection of thalamic tumors there were 19 PAs, 10 grade II astrocytomas, and 2 mixed
changed from conservative to more aggressive since the late gliomas. Radical (>90 %) resections were achieved in 59 of
1980s. Berstein et al. first reported a pediatric series of 60 65 (90.8 %) tumors with 3 % surgical-related death and
thalamic tumors in 1984 [3]. The series covered patients treat- 12.3 % surgical-related morbidity [1]. Regarding surgical ap-
ed between 1951 and 1983. Surgical procedures for tumors proaches for monothalamic tumor resection, based on the an-
were performed in 44 patients that consisted of 23 biopsies atomical locations of the tumors, a variety of surgical ap-
and 21 partial resections. In 2007, Puget et al. reported a series proaches was developed. Surgical anatomy of thalamic le-
of 69 thalamic tumors in children diagnosed in between 1989 sions (tumors and cavernoma) and related surgical approaches
and 2003 [2]. In the unilateral thalamic group of tumors, have been described in references [5, 9, 15, 21, 24, 25]. The
>90 % resection was achieved in 25 of 49 (51 %) patients most frequently used approaches were transfrontal (anterior),
who underwent surgical resections. In 2016, Steinbok et al. transparietal (posterior superior), transtemporal-occipital (pos-
reported a Canadian multicenter study series of 72 thalamic terior), transcallosal, and transtemporal. Other selected ap-
tumors in children. Resection surgery was performed in 39 proaches included parasplenial and subtemporal [1, 5, 6, 9,
patients. The degrees of tumor resection achieved were 10, 24]. In the present series, the 25 cases of LGAs were
Childs Nerv Syst (2016) 32:1895–1906 1901

Fig. 2 a–l A 10-year- and 2-

month-old boy with a
thalamopeduncular PA. Axial
T1WI revealed a well-
circumscribed multilobulated
tumor extends into the ipsilateral
mesial temporal and amygdala
regions, as well as suprasellar
ambient cisterns (a–d).
Heterogeneous high T2 signal
within the tumor is noted in T2WI
(e–h). After contrast medium
administration, the tumor shows
heterogeneous enhancement (i–l)

treated from 1971 to 2015. Nineteen tumors underwent surgi- high rate of >90 % radical resection, our low radical resection
cal excision, and >90 % resection was achieved in 8 (42.1 %) rate was due to the conservation attempts to radical resection
tumors only. In the group of PAs, >90 % excision was in early years. This decision was quite often before the era of
achieved in 60 % of patients. Compared with other series with MR image and neurophysiological-guided surgery for deep-

Fig. 3 a–l A 10-year- and 9-

month-old girl with a
thalamopeduncular PA. Axial
T1WI (A-D), axial T2WI (E-H),
and axial contrast enhanced T1WI
(i–l) revealed a heterogeneous
tumor lesion extending to the
ipsilateral mesial temporal and
amygdala regions, ambient
cistern, cerebral peduncle, and
midbrain-pontine junction. The
tumor was heterogeneous on
T1WI, slightly mixed high and
low signal on T2W1, and had
heterogeneous enhancement after
contrast medium administration
1902 Childs Nerv Syst (2016) 32:1895–1906

Fig. 4 a–i A 4-year- and 10-

month-old girl with a
thalamopeduncular DA. This
tumor has a well-defined irregular
margin extending into the
ipsilateral cerebral peduncle. The
signal of tumor is relatively
homogeneous on both T1WI
(a–c) and T2WI (d–f). No tumor
enhancement can be found after
contrast medium administration
(g–i)

seated brain tumor resection. In our experience, there was high recent Canadian multicenter review, unilateral low-grade tha-
association of thalamic PAs with extension to the mesial temporal lamic tumors had a 5-year overall survival rate of 84 ± 17 %
and amygdala region. The existence of thalamoamygdaloid con- compared to 6.9 ± 13.2 % of high-grade tumors [24]. In an-
nections may be a correlation [26]. The anterior transtemporal other recent multicenter study of 102 cases of thalamic LGG
approach was used more frequently for tumor located in such an in children, there were 87 monothalamic tumors and 15
anatomical location [Figs. 1, 6, and 7]. Several surgical ap- bithalamic tumors. Ninety patients received at least one surgi-
proaches according to tumor location and extension are shown cal procedure. The 10-year OS was 88 % (±3 %), but 10-year
in Fig. 6. event-free survival (EFS) was 28 % (±5 %). The 10-year OS
Long-term survival of LGGs in children is excellent with 5- was higher in monothalamic tumors (91 % (±3 %)) compared
year OS reaching 87 % at 5 years in a large cohort series [27]. to 65 % (±13 %) of bithalamic tumors. However, compared
In a long-term outcome study of SEER database, the 20-year with PAs, DAs did have an inferior 10-year OS (68 vs 93 %).
OS of 4040 children diagnosed with pediatric LGGs was Patients received total or radical resection with residue tumor
87 ± 0.8 %. The 20-year accumulation incidence of death <1.5 cm3 having a 10-year OS/EFS of 100/50 % as opposed to
due to glioma was 12 ± 0.8 % [39]. Compared with histolog- 84/20 % after biopsy only or >1.5-cm3 residue tumor. In the
ical grading, both grade I and grade II LGGs had excellent present series, excluding one case of loss of follow up, includ-
outcome. Compared with extent of surgical resection, there ing one case of bithalamic DA, the 10-year OS of 24 thalamic
was no significant difference of OS between patients with LGAs was 87.1 % compared to 4.9 % of HGAs [31].
gross total resection and patients with residue tumors after In LGGs, in the study of Boesten et al. 2016, tumor pro-
subtotal resection, biopsy, or no resection. Children who re- gression was observed in 11 (23 %) of 47 patients without
ceived radiation therapy had an increased risk of cancer- nonsurgical treatment [29]. Although spontaneous regression
specific death [28]. It is also true for thalamic LGGs. In a after incomplete resection of LGGs in children is rare in
Table 3 Comparison of disease characteristics with the Boesten el al. series by tumor location in children with thalamic low-grade glioma

Characteristics All patients Monohtalamic Bithalamic

Beosten et al. (2016) Present series Beosten et al. (2016) Present series Beosten et al. (2016) Present series

No. of patients 102 25 87 (85 %) 24 15 (15 %) 1 (4 %)

Childs Nerv Syst (2016) 32:1895–1906

Age at Dx (years)
Median 8.0 8.2 8.3 8.5 7.5 5.9
Range 0.4–17.5 1.8–15 0.4–17.5 1.9–15 2.0–15.1 5.9
Sex
Male 45 (44 %) 12 (48 %) 38 (44$) 11 (45.8 %) 7 (47 %) 1 (100 %)
Female 57 (56 %) 13 (52 %) 49 (56 %) 13 (54.2 %) 8 (53 %) 0
Histology
PA 50 (49 %) 11 (44 %) 48 (55 %) 11 (45.8 %) 2 (13 %) 0
DA 23 (23 %) 14 (56 %) 14 (16 %) 13 (54.2 %) 9 (60 %) 1 (100 %)
AG 1 (1 %) 0 1 (1 %) 0 0 0
GG 3 (3 %) 0 3 (3 %) 0 0 0
SEGA 1 (1 %) 0 1 (1 %) 0 0 0
OL 3 (3 %) 0 3 (3 %) 0 0 0
LGG, NOS 4 (4 %) 0 3 (3 %) 0 1 (7 %)` 0
Non-representative histology 4 (4 %) 0 3 (3 %) 0 1 (7 %)` 0
Only clinical/radiological Dx 12 (12 %) 0 10 (11 %) 0 2 (13 %) 0
Not documented 1 (1 %) 0 1 (%) 0 0 0
Extent of resection
Complete resection 8 (8 %) 1 (4 %) 8 (9 %) 1 (4.2 %) 0 0
<1.5 cm3 residual tumor (>90 % resection) 4 (4 %) 7 (28 %) 4 (5 %) 7 (29.2 %) 0 0
>1.5 cm3 residual tumor (<90 % resection) 23 (23 %) 11 (44 %) 22 (25 %) 10 (41.7 %) 1 (7 %) 1 (100 %)
Biopsy 55 (54 %) 6 (24 %) 43 (49 %) 6 (25 %) 12 (80 %) 0
No surgery 12 (12 %) 0 10 (11 %) 0 2 (13 %) 0

Dx diagnosis, PA pilocytic astrocytoma, DA diffuse astrocytoma, AG anaplastic ganglioglioma, GG ganglioglioma, SEGA subependymal giant cell astrocytoma, OL oligodendroglioma, LGG low-grade
glioma, NOS;
1903
1904 Childs Nerv Syst (2016) 32:1895–1906

change in two diffused astrocytomas at latency of 0.3 and

0.55 years after the first tumor resections. The first tumor with
malignant transformation was at initial diagnosis by endo-
scopic resection. The other tumor received >90 % resection
in the first operation. Compared with low MIB-1 index score
(insignificant to 1.8 %) of initial diagnosis, both tumors with
malignant transformation, the MIB-1 index score went up to
30.2 to 41.7 %, respectively.

Conclusion
Fig. 5 Outcome analysis of thalamic astrocytomas of the present series.
Overall survival curves split by low-grade astrocytoma and high grade Low-grade glioma is a significant entity of thalamic tumors in
astrocytomas: blue—low-grade tumors, red—high-grade tumors, children and the majority is low-grade astrocytoma. Pilocytic
purple—pilocytic astrocytomas, green—diffused astrocytomas
astrocytoma is usually well-circumscribed and diffuse astro-
cytoma is infiltrated. For low-grade tumor of thalamus in chil-
dren, radical (>90 %) to total resection through appropriate
patients with or without neurofibromatosis, it has been ob- approach is the goal of treatment and can be achieved in re-
served in PAs and grade II astrocytomas of visual pathway, ported series. The overall survival and EFS for tumors after
cerebellum, hypothalamus, and pons [30, 31]. Spontaneous radical resection had better OS and EFS as opposed to tumors
regression of a thalamic PA was reported in an adult [32]. In after biopsy or with large residues. By using contemporary
the present series, among 25 LGAs, 18 cases had residue neurosurgical facilities, neuroprotective maximum resection
tumors after surgical resection. In a mean follow-up of of tumor can be performed in selective patients. Since thalam-
11.6 years (range, 1to 22.2), there were three cases of tumor ic tumors, either benign or malignant, are deep-seated tumors,
progression. Slow regression of residue tumors was observed infiltration to important surrounding functional structures are
in two children after radical resection [Fig. 7]. Compared with very often. Not all of the thalamic tumors are suitable for
adults, anaplastic change of pediatric LGGs is uncommon. radical resection. Besides, low-grade thalamic gliomas are in-
Mistry et al. identified 26 cases of malignant transformation dolent tumors, even after incomplete resection, and although
from a large cohort series of 886 of pediatric LGGs. Five EFS is low, the 10-year and 20-year OS survival are still ex-
patients were treated with radiotherapy after diagnosis. Nine cellent. In long-term follow-up, the rate of tumor progression
were thalamic tumors. The latency of malignant transforma- is low, and spontaneous regression occurs. For low-grade tu-
tion was from 0.32 to 20.3 years [33]. We observed anaplastic mors after incomplete resection, we can observe by clinical and

Fig. 6 Selective surgical approaches and radical resections of for a tumor extends to the medial temporal and amygdala regions and
monothalamic tumors in children. Frameless stereotactic frontal ambient cistern (c, d). Transsylvian-transinsular approach for a tumor
transcortical approach for a tumor arises in the basal ganglia thalamic arises neighboring to the lateral thalamus (e, f). Occipital
region (not in the present series) (a, b). Anterior transtemporal approach interhemispheric transsplenial approach (g, h)
Childs Nerv Syst (2016) 32:1895–1906 1905

Fig. 7 a–h A 5-year- and 6-month-old boy with a thalamopeduncular radical resection via anterior temporal approach (e, f). Eight years after
PA. In axial contrast enhanced T1WI after contrast administration, the surgical excision without radiation or chemotherapy, follow-up images
tumor extends to the ipsilateral cerebral peduncle, mesial temporal region, showed regression of residue tumor (g, h)
and suprasellar cistern (a–d). Residue tumor was seen at 3.5 months after

radiographic follow-up initially. Nonsurgical managements such 5. Ozek MM, Ture U (2002) Surgical approach to thalamic tumors.
Childs Nerv Syst 18:450–456
as radiation and or chemotherapy are left for tumor progression.
6. Villarejo F, Amaya C, Perez Diaz C, Pascual A, Alvarez Sastre C,
Goyenechea F (1994) Radical surgery of thalamic tumors in chil-
Acknowledgment We thank Ms. Ching-Wen Tsai and Dr. Yen-Kuang dren. Childs Nerv Syst 10:111–114
Lin at the Biostatistics Center, Taipei Medical University, Taipei, Taiwan 7. Moshel YA, Link MJ, Kelly PJ (2007) Stereotactic volumetric re-
for their contribution to the data analysis of this work. section of thalamic pilocytic astrocytomas. Neurosurgery 61:66–75
8. Albright AL (2004) Feasibility and advisability of resections of
thalamic tumors in pediatric patients. J Neurosurg 100(5 Suppl
Compliance with ethical standards Pediatrics):468–472
9. Kelly PJ (1989) Stereotactic biopsy and resection of thalamic as-
Conflict of interest The authors have nothing to disclose. trocytomas. Neurosurgery 25:185–195
10. Drake JM, Joy M, Goldenberg A, Kreindler D (1991) Computer-
Source of funding This study was supported by the Aim for the Top and robot- assisted resection of thalamic astrocytomas in children.
University Project—Cancer Translational Center of Taipei Medical Neurosurgery 29:27–33
University, Ministry of Science, and Technology Grant MOST102– 11. Colosimo C, di Lella GM, Tartaglione T, Riccardi R (2002)
2314-B-038-060-MY3 and Health and Welfare Surcharge of Tobacco Neuroimaging of thalamic tumors in children. Childs Nerv Syst
Products MOHW105-TDU-B-212-134,001. 18:426–439
12. Lyons MK, Kelly PJ (1992) Computer assisted stereotactic biopsy
and volumetric resection of thalamic pilocytic astrocytomas. Report
of 23 cases. Stereotact Funct Neurosurg 59:100–104
13. Moshel YA, Elliot RE, Monoky DJ, Wisoff JH (2009) Role of
diffusion tensor imaging in resection of thalamic juvenile pilocytic
References astrocytoma. J Neurosurg Pediatr 4:495–505
14. Matsumoto K, Higashi H, Tomita S, Furuta T, Ohmoto T (1995)
1. Cuccia V, Monges J (1997) Thalamic tumors in children. Childs Resection of deep-seated gliomas using neuroimaging for stereo-
Nerv Syst 13:514–521 tactic placement of guidance catheters. Neurol Med Chir (Tokyo)
2. Puget S, Crimmins DW, Garnett MR, Grill J, Oliveira R, Boddaert 35:148–155
N, Wray A, Lelouch-Tubiana A, Roujeau T, Di Rocco F, Zerah M, 15. Steiger HJ, Götz C, Schmid-Elsaesser R, Stummer W (2000)
Sainte-Rose C (2007) Thalamic tumor in children: a reaprasial. J Surgical anatomy and results of a pilot series using maximum mi-
Neurosurg 106:354–362 crosurgical removal. Acta Neurochir 142:1327–1337
3. Bernstein M, Hoffman HJ, Halliday WC, Hendrick EB, 16. Laundre BJ, Jellison BJ, Badie B, Alexander AL, Field AS (2004)
Humphreys RP (1984) Thalamic tumors in children. Long-term Diffusion tensor imaging of the corticopspinal tract before and after
follow-up and treatment guidelines. J Neurosurg 61:649–656 mass resection as correlated with clinical motor findings: prelimi-
4. Bilginer B, Narin F, Isikay I, Oguz KK, Soylemezoglu F, Akalan N nary data. AJNR Am J Neuroradiol 26:791–796
(2014) Thalamic tumors in children. Childs Nerv Syst 30:1493– 17. Carrabba G, Bertani G, Ardolino G, Cogiamanian F, Zarino B,
1498 Locatelli M, Barbieri S, Caroli M, Rampini P (2014)
1906 Childs Nerv Syst (2016) 32:1895–1906

Neurophysiological monitoring in the management of thalamic 27. Fisher PG, Tihan T, Goldthwaite PT, Goldthwaite PT, Wharam
astrocytoms. Neuro-Oncology 16:ii1–ii112 MD, Repka MX, Cohen KJ, Burger PC (2008) Outcome analysis
18. Berman JI, Berger MS, Mukherjee P, Henry RG (2004) Diffusion- of childhood low-grade astrocytoma. Pediatr Blood Cancer 51(2):
tensor imaging-guided tracking of fibers of the pyramidal tract com- 245–250
bined with intraoperative cortical stimulation mapping in patients 28. Bandopadhayay P, Bergthold G, London WB, Goumnerova LC,
with gliomas. J Neurosurg 101:66–72 Morales La Madrid A, Marcus KJ, Guo D, Ullrich NJ, Robison
19. Zheng X, Xu X, Zhang H, Wang Q, Chen X, Sn G, Zhang J, Jiang J, NJ, Chi SN, Beroukhim R, Kieran MW, Manley PE (2014) Long-
Xu B, Zhang J (2016) A preliminary experience with use of intra- term outcome of 4,040 children diagnosed with pediatric low-grade
operative magnetic resonance imaging in thalamic glioma surgery: gliomas: an analysis of the Surveillance Epidemiology and End
a case series of 38 patients. World Neurosurg 89:434–441 Results (SEER) database. Pediatr Blood Cancer 61(7):1173–1179
20. Coburger J, Hagel V, Wirtz CR, Konig R (2015) Surgery for glio- 29. Beosten T, Gerber NU, Kandels D, Azizi AA, Schmidt R,
blastoma: impact of the combined use of 5-aminolevulinic acid and Warmuth-Metz MW, Pietsch T, Kortmann RD, Gnekow A,
intraoperative MRI on extent of resection and survival. PLoS One Grotzer MA (2016) Management of primary thalamic low-grade
10(6):e0131872 glioma in pediatric patients: results of the multicenter treatment
21. Rangel-Castilla L, Spetzler RF (2015) The six thalamic regions: studies HIT-LGG 1996 and SIOP-LGG 2004. Neuro Oncol Pract.
surgical approaches to thalamic cavernous malformations, opera- doi:10.1093/nop/npw007
tive results, and clinical outcomes. J Neurosurg 123:676–685 30. Rozen WM, Joseph S, Lo PA (2008) Spontaneous regression of
22. Baroncini M, Vinchon M, Minéo JF, Pichon F, Francke JP, low-grade gliomas in pediatric patients without neurofibromatosis.
Dhellemmes P (2007) Surgical resection of thalamic tumors in Pediatr Neurosurg 44(4):324–328
children: approaches and clinical results. Childs Nerv Syst 23: 31. Steinbok P, Poskitt K, Hendson G (2006) Spontaneous regression
753–760 of cerebellar astrocytoma after subtotal resection. 22(6):572–576
23. Wong TT, Liang ML, Chen HH, Chen FC (2011) Hydrocephalus 32. Balkhoyor KB, Bernstein M (2000) Involution of diencephalic
with brain tumors in children. Childs Nerv Syst 27(10):1723–1734 pilocytic astrocytoma after partial resection. Report of two cases
24. Steinbok P, Gopalakrishnan CV, Hengel AR, Vitali AM, Poskitt K, in adults. J Neurosurg 93:484–448
Hawkins C, Drake J, Lamberti-Pasculli M, Ajani O, Hader W, 33. Mistry M, Zhukova N, Merico D, Rakopoulos P, Krishnatry R,
Mehta V, McNeely PD, McDonald PJ, Ranger A, Vassilyadi M, Shago M, Stavropoulos J, Alon N, Pole JD, Ray PN, Navickiene
Atkinson J, Ryall S, Eisenstat DD, Hukin J (2016) Pediatric tha- V, Mangerel J, Remke M, Buczkowicz P, Ramaswamy V, Guerreiro
lamic tumors in the MRI era: a Canadian perpective. Childs Nerv Stucklin A, Li M, Young EJ, Zhang C, Castelo-Branco P, Bakry D,
Syst 32(2):269–280 Laughlin S, Shlien A, Chan J, Ligon KL, Rutka JT, Dirks PB,
25. Yasargil MG (1996) Microneurosurgery, vol 4B. Thieme, Stuttgart, Taylor MD, Greenberg M, Malkin D, Huang A, Bouffet E,
pp 29–91, 252–312, 291–342 Hawkins CE, Tabori U (2014) BRAF mutation and CDKN2A dele-
26. Nitecka L (1979) Connection of the posterior thalamus with the tion define a clinically distinct subgroup of childhood secondary
amygdaloid body of the rat. Acta Neurobiol Exp 39:49–55 high-grade glioma. J Clin Oncol. doi:10.1200/JCO.2014.58.3922