Key-words: Adult - cerebellopontine angle - medulloblastoma - radiomics
Introduction
Adult medulloblastomas (MBs) are relatively rare and constitute 0.4%–1.0% of adult
primary brain tumors.[[1 ]] Cerebellopontine angle (CPA) is an atypical site for MB with 42 cases reported
in both pediatric and adult population, intra- and extraaxial locations combined.[[2 ]] Extraaxially, the tumors have been localized in two areas: the tentorial region
and the CPA.[[3 ]] In adults, only 12 cases were found in extra-axial CPA location.[[2 ]],[[4 ]],[[5 ]],[[6 ]],[[7 ]],[[8 ]],[[9 ]],[[10 ]],[[11 ]],[[12 ]],[[13 ]]Here, we report the 13th case of an extra-axial lesion in the CPA of a 27-year-old
male, which was presumed to be meningioma on preoperative imaging but turned out to
be desmoplastic MB.
As molecular subtyping is not available in resource-limited setups in low income countries,
we utilized ''for the first time'' the MRI based nomogram as an alternative to expensive
genetic profiling and propose its use in selective cases.
Case Report
A 27-year-old male patient presented with headache, vomiting, left-sided ataxia, and
left-sided mild hearing loss of 1-month duration. The vision was normal. Fundus examination
showed papilledema. He had left-sided House-Brackman Grade 1 facial paresis, left-sided
mild sensorineural hearing loss, and left cerebellar signs on physical examination.
Imaging
A computed tomography scan of the brain showed a left CPA mixed-density nonenhancing
lesion (5.4 cm × 2.8 cm) with broad-based tentorial attachment and displacement of
the fourth ventricle but not causing obstructive hydrocephalus. No evidence of calcification
was noted. Magnetic resonance imaging (MRI) showed T1 hypo/T2 heterointense, extra-axial
lesion in left CPA, which had a wide tentorial attachment and showed moderate heterogeneous
enhancement with IV gadolinium. There was no meatal extension [[Figure 1 ]]a, [[Figure 1 ]]b, [[Figure 1 ]]c. Adjacent white matter edema in the left cerebellar hemisphere was noted [[Figure 1 ]]d.
Figure 1: (a) Axial T1 showing hypointense and axial T2 showing heterointense mass in left
cerebellopontine angle (b) T2 Sagittal showing central hyperintensity (c) Coronal
contrast images showing heterogeneous moderate enhancement and wide tentorial attachment
(d) T2 FLuid Attenuated Inversion Recovery (FLAIR) showing adjacent cerebellar edema
Treatment
He underwent a left retromastoid craniectomy. A moderately vascular, soft suckable,
grayish pink lesion was found densely adherent to petro-tentorial junction laterally
and to the inferior surface of left hemi tentorium till hiatus medially. This attachment
was heavily coagulated [[Figure 2 ]]a. The tumor-brainstem interface was well defined, but the tumor-cerebellar interface
was ill-defined. All cranial nerves from 4 to 11 were preserved [[Figure 2 ]]b. Gross total excision achieved in a piecemeal fashion [[Figure 2 ]]c. The central necrotic part corresponded to T2 hyperintensity on imaging.
Figure 2: (a) Perioperative image showing heavy coagulation of tentorial attachment (b) Cranial
nerve 5 (yellow arrow), 7, 8 (green arrow), 9, 10 (Blue arrow), and 11 (Red arrow)
seen intact following tumor resection (c) tumor specimen
Histopathology showed a nodular arrangement of round-to-oval cells having hyperchromatic
nuclei and scanty cytoplasm, with brisk mitotic activity suggestive of Desmoplastic
MB–WHO Grade IV [[Figure 3 ]]a. Immunohistochemical staining showed patchy cytoplasmic staining of beta-catenin,
wild type p53, and synaptophysin positive with background GFAP activity [[Figure 3 ]]b, [[Figure 3 ]]c, [[Figure 3 ]]d. Molecular subtyping is currently not available in our country.
Figure 3: (a) H and E stain illustrating the peculiar nodular arrangement (b) cytoplasmic reactivity
of Beta catenin (c) wild type P53 (d) Synaptophysin positive (e) Axial T1 without
contrast at 9.month FU (f) Coronal Contrast at 9.month FU showing no evidence of recurrence
or mets
Postoperatively, his headache and ataxia improved but his facial paresis transiently
worsened to Grade 2 which improved to Grade 1 over 2 weeks. His craniospinal MRI with
contrast was done after 4 weeks of surgery. Cerebrospinal fluid analysis from lumbar
tap turned out to be negative for malignant cells. He was labelled as “standard risk”
based on the size of residual <1.5 cm2, no metastasis, nonlarge cell histology, and
age >3 years. The patient underwent craniospinal irradiation as well as chemotherapy
with vincristine, cisplatin, and cyclophosphamide. Postadjuvant therapy, craniospinal
imaging was clear for any metastasis or recurrence at 9 months [[Figure 3 ]]e and [[Figure 3 ]]f.
Discussion
Adult MB is known to be lateral hemispheric. CPA is a rarely reported location, challenging
to diagnose both clinically and radiologically. There have been only 12 reported cases
of extra-axial MB in the adult literature [[Table 1 ]].[[2 ]],[[4 ]],[[5 ]],[[6 ]],[[7 ]],[[8 ]],[[9 ]],[[10 ]],[[11 ]],[[12 ]],[[13 ]]
Table 1: Summary of Previously Reported Extra-axial Medulloblastoma in Adult Cerebellopontine
Angle
A review of all reported cases shows that CPA MB is twice more common in males in
their 2nd–3rd decade, with the classic variety seen in more than half of the cases.
Ten out of 12 reports so far published considered meningioma as a provisional diagnosis
based on broad tentorial attachment, despite heterogeneous enhancement. Meatal extension
which is considered pathognomonic for vestibular schwannoma was only seen in one case.[[13 ]]
Our preoperative diagnosis was based on a broad base extra-axial lesion, without extension
into the internal auditory meatus favoring meningioma as well as disapproving schwannoma.
Although the lack of dural tail and heterogeneous enhancement was not classical of
a meningioma, the short clinical course and white matter edema in the adjacent cerebellar
lobe favored atypical meningioma preoperatively. Classical radiological features of
adult MB are well delineated in the literature but not for atypical locations. All
12 cases reported so far were diagnosed after histopathology leaving MRI features
to be focused more than ever before.
Molecular subtyping has not yet been reported for adult CPA MBs. We lack the facility
in our setup but tried to speculate on the basis of location, histology, and immunohistochemistry
(IHC) retrospectively. Moreover, in the era of molecular genetics, radiogenomics have
provided sufficient data to define certain molecular subtypes noninvasively, although
it is quite clear that no single imaging feature is pathognomonic of any particular
subgroup, certain imaging characteristics are much more prevalent in one subgroup
compared to others and may even be highly specific for an individual molecular subgroup.[[14 ]]
Here, we intend to emphasize the use of an integrated approach utilizing location,
histology, IHC, and radiomics to predict molecular subgroup in adult cases of MB,
in developing countries that lack the facility of genetic subtyping. A brief review
of the correlation of molecular subgroup to location, histology, IHC, and radiomics
is presented individually, which can help predict the molecular subgroup with a fair
degree of accuracy when combined.
Location-molecular subtype correlation
Adult MBs unlike pediatric are reported to be of two molecular subtypes, instead of
four, with Group 3 occurring as an exception and Group 4 restricted to midline.[[15 ]],[[16 ]] Lateral location is reserved for WNT and SHH tumors, with the former mostly seen
in CPA and the latter in the cerebellar hemisphere. Furthermore, superior location
abutting and/or reaching the tentorium is a specific imaging feature of SHH-subgroup
MB seen in 48% of patients as opposed to 6% in other groups.[[14 ]]
Histology-molecular subgroup correlation
There is a consensus among researchers that all desmoplastic histology in adults belongs
to the SHH subgroup, but the opposite is not true: Not all SHH tumors in adults are
of desmoplastic histology.[[15 ]],[[17 ]],[[18 ]]Moreover, another study reports that 9% of adult SHH-MBs are located in CP/CPA.[[19 ]]
Immunohistochemistry-molecular subgroup correlation
Beta-catenin localization to cytoplasm instead of nucleus excludes WNT.[[18 ]] In addition, some markers such as GAP 1 and YAP1 are considered pathognomonic for
SHH and WNT type, respectively.[[20 ]]
Radiomics-molecular subgroup correlation
Nomograms based on multiparametric MRI have suggested 95% predictive accuracy for
SHH subgroup,[[21 ]] best among all other subgroups. We applied the SHH subgroup-specific nomogram to
our case [[Figure 4 ]] and the probability of the tumor to be SHH type turned out to be 98%. As only two
molecular subgroups can be expected in adult CPA, we propose to suggest WNT type in
cases where SHH type seems less probable on the nomogram, combined with nuclear reactivity
for beta-catenin.[[20 ]]
Figure 4: Nomogram for SHH used to predict the molecular subgroup (used with permission of
Dasgupta et al.[21]
In our case, CPA location was the only factor favoring WNT type, while tentorial attachment,
desmoplastic histology, and lack of nuclear reactivity of beta-catenin were favoring
SHH type. Hence, we used SHH specific nomogram in retrospect to predict the molecular
subgroup with fair degree of accuracy. Desmoplastic histology, p53 wild type and SHH
subgroup all confer good to intermediate prognosis in this case.
Conclusion
The use of an integrated approach, based on location, appearance, histology, and demographics
not only predicts molecular subgroup but also help defining prognosis. MRI-based nomograms
in selected cases can substitute expensive genetic profiling in resource-limited setups.
Declaration of patient consent
The authors certify that they have obtained all appropriate patient consent forms.
In the form, the legal guardian has given his consent for images and other clinical
information to be reported in the journal. The guardian understands that names and
initials will not be published and due efforts will be made to conceal identity, but
anonymity cannot be guaranteed.
