Keywords
astrocytoma - extracranial - glioblastoma - IDH - metastases
Introduction
Glioblastomas are malignant brain tumors associated with rapid progression with unfavorable
prognosis. Isocitrate dehydrogenase (IDH) wild-type primary glioblastomas are usually
MGMT unmethylated tumors, with overall median survival less than 18 months. Secondary
glioblastomas (grade 4 astrocytomas, as per the World Health Organization [WHO] 2021
classification)[1] are usually IDH-mutant and most of them are MGMT methylated tumors. Despite being
a tumor with highly aggressive behavior, extracranial metastases are exceedingly rare.
We report a rare case of a young man with IDH-mutant glioblastomas presenting with
extracranial metastases.
Case Report
A 39-year-old male, evaluated in 2010 for history of headache, underwent magnetic
resonance imaging (MRI) of brain with contrast which showed a right parietal lesion
suggestive of a low-grade glioma. He underwent surgery for the same with histopathology
reported as diffuse astrocytoma, grade II, post which he received adjuvant radiotherapy
without concurrent chemotherapy in 2010. He was on follow-up till 2015, when he developed
recurrence on the right parietal region, for which he underwent right parietal craniotomy
and excision of the lesion; histopathology of the same was reported as diffuse astrocytoma,
grade II, IDH1 positive, MGMT methylated. He underwent adjuvant chemoradiation to
a total dose of 41.4 Gy in 23 fractions along with concurrent temozolomide followed
by adjuvant temozolomide in 2015 and was on regular follow-up.
In October 2020, he again developed seizures and an MRI of the brain showed recurrence
in the right parietal region at the same site ([Fig. 1A, B]). He underwent a redo right parietal awake craniotomy and gross total excision of
the lesion. Histopathology was reported as glioblastoma with primitive neuroectodermal
tumor (PNET) pattern, WHO grade IV with IDH1-mutant ATRX loss, p53 positive, Ki67
of 65%, brisk mitotic activity with a count of 30 to 32/10 high power field, and large
areas of necrosis. He then underwent reirradiation using image-guided intensity-modulated
proton beam therapy to a total dose of 55.8 CGE in 31 fractions during December 2020
along with 140 mg of concurrent temozolomide. His follow-up MRI showed excellent response
with no residue, and he completed eight cycles of adjuvant temozolomide till December
2021.
Fig. 1 (A) Presurgery contrast T1 magnetic resonance imaging (MRI) in October 2020—heterogeneously
enhancing mass lesion with necrotic areas seen involving the right parietal lobe region
favoring high-grade glioma likely grade 4 astrocytoma. (B) Postsurgery contrast T1 MRI November 2020—complete excision of the mass lesion with
postoperative changes seen. (C) Recent contrast T1 MRI—no significant residual or recurrent tumor mass seen. (D) (Short-tau inversion recovery [STIR] coronal sequence MRI) STIR hyperintense large
mass lesion seen at right proximal femur. (E) T1 postcontrast T1 fat saturated (FS) showing heterogeneously enhancing lesion at
right proximal femur. Small lesion at left femur shaft. (F) Multiple fluorodeoxyglucose (FDG) avid lesions in positron emission tomography-computed
tomography (PET-CT) at dorso-lumbar vertebra. (G) Multiple FDG avid lesions in PET-CT in bilateral proximal femur.
In October 2021, he started complaining of pain in the right hip region, local imaging
(X-ray) did not reveal any abnormality. However, MRI of the same picked up a thin
hairline fracture without any associated soft tissue component, and subsequent positron
emission tomography-computed tomography (PET-CT) scan showed localized uptake and
was managed conservatively. He was pain-free till December 2021 when he reported to
the emergency room with severe pain (pain score of 7/10), plain X-ray of local part
done did not reveal any abnormality, but MRI of both the hip joints showed altered
signal intensity in the subtrochanteric area of the right femur, with intramedullary
skip lesions noted 5 cm distal to the main lesion with avulsion fracture of lesser
trochanter ([Fig. 1D, E]). A CT-guided biopsy of the right femoral lesion done showed marrow infiltrating
lesion arranged in diffuse sheets interspersed with septae, positive for glial fibrillary
acidic protein, synaptophysin, CD56, IDH1 positive, and ATRX loss highly suggestive
of metastases from a high-grade glioma. PET-CT scan done subsequently showed intense
fluorodeoxyglucose (FDG) avid ill-defined lytic lesion with cortical erosion in proximal
shaft of the right femur (standard uptake value 31.3), with foci of abnormal intramedullary
skip lesions in the right femur, proximal shaft of the left femur, multiple dorsal
and lumbar vertebrae, iliac bones, and right third rib ([Fig. 1F, G]). MRI brain with contrast, MR spectroscopy, and perfusion studies along with spine
screening ([Fig. 1C]) did not reveal any abnormality and showed only postoperative cavity with treatment-related
changes. In view of severe pain and need for mobilization, he then underwent prosthetic
replacement of the right proximal femur in January 2022, and subsequent histopathology
from main sample also confirmed the immunohistochemistry findings of metastases from
high-grade glioma favoring glioblastoma ([Fig. 2]). In view of PET-CT showing multiple FDG avidity at bony areas, and patient being
asymptomatic for the same, palliative radiotherapy was deferred, and was discussed
extensively in our neuro-oncology multidisciplinary tumor board (MDT) as well as national
MDTs and was planned for single-agent lomustine (CCNU) and to consider palliative
radiotherapy later. He is at present asymptomatic, completed one cycle of single-agent
lomustine, and is able to take care of his daily day-to-day activities.
Fig. 2 Histopathology of right femur. (A) Hematoxylin and eosin (H&E) low power showing tumor tissue with normal bone marrow
cells. (B) H&E high power showing tumor tissue with normal bone marrow cells. (C) Glial fibrillary acidic protein (GFAP) positive suggesting glial lineage. (D) Isocitrate dehydrogenase (IDH) 1 low power (mutant) suggesting metastases from glial
neoplasm. (E) IDH1 high power (mutant) suggesting metastases from glial neoplasm. (F) Synaptophysin positive showing positivity for primitive neuroectodermal tumor (PNET)
pattern. (G) ATRX loss. (H) P53 positive. (I) Ki67 high indicating aggressiveness of the tumor.
Discussion
IDH-mutant glioblastomas (previously known as secondary glioblastomas and renamed
as grade 4 astrocytomas at present) are rare compared to de novo primary glioblastomas,
manifest in younger patients, and usually progress from low-grade diffuse astrocytomas.[2] Extracranial metastases from glioblastoma are extremely rare but can affect 0.4
to 0.5% of all patients with glioblastomas. This can be attributed to rapid intracranial
progression of the disease leading to poor overall survival, thereby not having sufficient
time for dissemination, lack of favorable cerebral environment for extracranial tumor
cell spreading, and absence of neural stroma for metastatic cells to adhere and multiply.
The most common sites of metastases so far noted are lymph nodes, lung, and bone;
liver, soft tissue, and skin can also be involved. Among bone metastases, spine (73%)
is the most common site, followed by ribs (23%), sternum (18%), skull (14%), and acetabulum
(9%).[2] Mean time between diagnoses of metastases and death is about 12 months, with better
prognosis to lymph nodes and worse to lungs and liver. Exact pathogenesis of extracranial
spread is not understood but is hypothesized to be invasion through venous system
or directly through dura and breakdown of blood–brain barrier favoring diffusion through
systemic circulation. Also, recent studies have demonstrated the existence of lymphatic
system in meninges, called “glymphatic system,” which can be attributed to disease
spread.[3] It is also postulated that predilection for bone metastases may come from both tumor-derived
and extracellular niche-derived cues, where glioblastoma cells express same hematopoietic
stem cell proteins that are critical for growth within the bone marrow, including
stromal cell-derived factor 1 alpha (SDF-1α), C-X-C chemokine receptor 4 (CXCR4),
osteopontin (OPN), and cathepsin K (CATK). Glioblastoma cells are also thought to
recruit bone marrow-derived progenitor cells that support tumor-associated angiogenesis
including hypoxia-inducible factor-1α and vascular endothelial growth factor, which
are known to increase glioma aggressiveness and invasion.[4]
Treatment of extracranial metastases varies widely, with no consensus due to the rarity
of the tumors and paucity of available literature. They are based on symptomatology
such as palliative radiotherapy and/or chemotherapy.[5] Progressive systemic involvement in the absence of intracranial progression, leads
to a conundrum of clinical diagnoses. Initial favorable histopathological and molecular
pattern can increase the likelihood of systemic spread from a glioblastoma. In our
patient, we reviewed the histopathology comparing with the previous histopathology
of the craniotomy specimen in 2020, where we did find the PNET pattern to be common
in both the samples thereby confirming the clinical diagnosis. Our initial working
diagnosis was toward a granulomatous/tuberculous etiology, with myelodysplasia also
being thought of; however, after careful histopathological, clinical, and radiological
correlation, extracranial metastases were confirmed and treated accordingly.
Conclusion
Long-term survivors of IDH-mutant glioblastomas without any intracranial disease presenting
with systemic symptoms should warrant the thought of extracranial metastases in the
differential diagnosis, as these tumors are known to cause systemic spread; however,
treatment paradigm remains futile and managed as per palliative care principles. Further
molecular analysis can throw more light on treatment of such tumors.
