Rapid growth of primary cerebral fibrosarcoma with conversion to glioblastoma at second recurrence

  • Amos Olufemi Adeleye
  • Yakov Fellig
  • Felix Umansky
  • Yigal Shoshan
Case Report


We present a case of de novo fibrosarcoma in a 43-year-old male, with MRI documented evolution from a 5 mm hyperintense area to 5 cm tumor mass in a 12-month period. The diagnosis of low-grade fibrosarcoma was established by three experienced neuropathologists. The patient underwent gross total resection with adjuvant fractionated conformal radiotherapy. Following first recurrence 3 months later, the patient was reoperated and stereotactic radiosurgery of a residual tumor was performed thereafter. The pathological diagnosis was similar, but with additional extensive radiation effects. Six months later the patient underwent aggressive surgical resection for second recurrence. The pathological diagnosis was WHO grade IV glioblastoma. The etiology of this highly unusual progression from primary mesenchymal neoplasm to high-grade glioma is discussed.


Gliomatous differentiation Gliosarcoma Primary cerebral fibrosarcoma Rapid growth 



Computerized tomography


Fluid attenuated inversion recovery


Magnetic resonance imaging


Isodose line




Primary parenchymal fibrosarcomas are extremely rare tumors of the brain, with reported incidence of 0.2–0.7% [1]. Parenchymal and meningeal fibrosarcoma account for 1.5% for all intracranial neoplasms [2, 3]. They are characterized by very aggressive clinical behavior with early local recurrence despite oncologic treatment at the current best standard of care [3, 4]. This clinical aggressiveness is evident in a number of reports documenting a short latency period before tumor growth [5, 6], however these are cases of tumor recurrence following surgery and radiotherapy for treatment of a primary fibrosarcoma or other intracranial tumor.

We present here a case of de novo primary cerebral fibrosarcoma that is unique due to its very unusual cellular dedifferentiation to glioblastoma at the second recurrence. Also of interest is the documented relatively rapid growth of fibrosarcoma from an early stage to a full blown clinical picture during 12 months, demonstrating a possible natural history not previously reported.

Case report

A 43-year-old man presented in our department in December 2006 with history of headache and dysphasia of a few weeks duration. Prior history of primary brain tumor in other family members was negative. Clinical examination revealed slow mentation and dysphasia, with no other focal neurologic or general physical abnormality.

Cranial computerized tomography (CT) and magnetic resonance imaging (MRI) showed a left frontal pure intraparenchymal 5 cm mass lesion with nonhomogenous contrast enhancement and significant perilesional edema (Fig. 1a, b). CT of the chest, abdomen, and pelvis did not reveal any other lesion. One year earlier, he had undergone cranial CT and MRI for evaluation of vertigo. These studies were interpreted as normal; however, retrospective evaluation of that MRI revealed a left frontal cortico-subcortical 5 mm area of hyperintensity on the FLAIR protocol (Fig. 1c, d).
Fig. 1

a T1-weighted, gadolinium-enhanced. b FLAIR MRI studies at first presentation, showing a 5 cm nonhomogenously enhancing left frontal parenchymal mass lesion with marked perilesional edema. c, d FLAIR MRI study performed 1 year earlier. The only positive finding was a discrete 5 mm left cortico-subcortical hyperintense lesion (d)

A preoperative diagnosis of high-grade glioma was made based on these imaging studies, and the patient underwent an intraoperative MRI (iMRI)-guided craniotomy with gross total excision of the mass. The pathological diagnosis was low-grade fibrosarcoma, based on a consensus between our Hospital’s neuropathologists, as well as two experienced outside neuropathologists. Following surgery, the patient was treated by conformal radiotherapy, with 54 Gy in 30 fractions. Radiation was then followed with two cycles of themozolamide.

Clinical and radiologic follow up revealed local tumor recurrence 3 months later in March 2007. The patient had repeated craniotomy and subtotal excision of the 3.5 cm recurrent tumor. Pathology revealed mainly radiation effects and small focus of a neoplasm similar to the fibrosarcoma seen at first surgery. One month later, the patient received cyberknife-based single-fraction stereotactic radiosurgery with marginal dose of 17 Gy (70% IDL) for a 7.2 cc volume of lesion.

In September 2007, the patient presented with repeated seizures. Immediate MRI showed a second tumor recurrence, consisting of an irregular ring-enhancing lesion in the previous tumor bed (Fig. 2a) and a peripherally located solid nodule (Fig. 2b).
Fig. 2

T1-weighted, gadolinium-enhanced MRI study showing: a Cystic/necrotic. b Solid portions of the tumor at second recurrence, 9 months following resection of the primary Fibrosarcoma. c, d Gross-total excision of both the cystic and solid tumor portions. Histopathology at this secondary recurrence was glioblastoma

Resection of this secondary recurrence was performed with frameless stereotactic navigation-guided awake craniotomy. Pathology of this lesion was compatible with glioblastoma together with radiation necrosis, but with no evidence of fibrosarcoma. Immediate postoperative MRI study revealed gross total excision of this secondary tumor (Fig. 2c, d). The patient’s postoperative recovery was uneventful. Five months after surgery the patient is asymptomatic, repeat MRI 3 month following last surgery did not showed tumor recurrence at the resection site.

Pathology profiles

Formalin-fixed, paraffin-embedded sections were stained with hematoxylin and eosin (HE) using automatic devices (Leica autostainer XL, Leica slide coverer CV5000). Immunohistochemical stains were also performed using an automatic device (NexES®, Ventana Medical Systems, Tucson AZ) for glial fibrillary acidic protein (GFAP-DAKO, M0761, 1:600), S-100 protein (DAKO, Z0311, 1:5000), epithelial membrane antigen (EMA-DAKO, M0613, 1:40), CD34 (DAKO, M7165, 1:50), p53 (DAKO, M7001, 1:2000) and the MIB-1 (Ki-67-DAKO, M7240, 1:100) proliferation marker, as well as silver impregnation for reticulin fibers (Ventana). Stained sections were screened with a light microscope (Axioplan2, Carl Zeiss, Oberkochen, Germany), and photographs were taken using a digital camera (AxioCam, Zeiss).

The primary neoplasm was a low-grade fibrosarcoma (Fig. 3a–e) composed of interlacing bundles of spindle cells focally disposed in a herringbone-like pattern, infiltrating the brain parenchyma. The neoplasm showed moderate cellularity, and neoplastic cells showed moderate atypia. There was low-to-moderate mitotic activity with small necrotic focus. Loose, myxoid-like areas imparted a partial myxofibrosarcoma-like appearance. The neoplasm was rich in reticulin fibers surrounding individual cells. Neoplastic cells were immunonegative for GFAP, S-100 protein, EMA, and CD34. The MIB-1 labeling index was about 7%. p53 labeling index was relatively low, about 3–5% and stain intensity was weak to moderate.
Fig. 3

Pathological findings: The primary neoplasm is shown in the upper panel (ae); the second recurrence is shown in the lower panel (fj). The primary neoplasm was a spindle cell sarcoma consistent with low grade fibrosarcoma (a), “rich” in reticulin fibers (b), and immunonegative for GFAP (c), which highlights scattered entrapped astrocytes. The MIB-1 labeling index is low-to-moderate, up to 7% (d). p53 labeling index is relatively low, about 3–5% (e) and stain intensity is week to moderate. The second recurrence is a glioblastoma, WHO grade IV, with numerous mitotic figures, robust microvascular proliferation and pseudopalisading necrosis (f). This neoplasm is “poor” in reticulin fibers, which highlight blood vessels only (g). Neoplastic cells are strongly and diffusely immunoreactive for GFAP (h). The MIB-1 labeling index is very high, up to 30–40% (i). Original magnifications are ×200 (c, e), p53 labeling index is relatively low, about 3–5% in most areas and focally higher, up to 7–10% (j). Stain intensity is week to moderate with only few cells showing strong staining intensity. Original magnification are ×200 (c, f, g, h), ×400 (a, d, e, i, j), and ×100 (b)

At the first recurrence, a similar neoplasm with extensive areas showing radiation effect (necrosis, hyalinized vessels) was seen.

The second recurrence was a glioblastoma, WHO grade IV (Fig. 3f–j) infiltrating the brain parenchyma and showing high cellularity, prominent nuclear atypia, numerous mitotic figures, robust microvascular proliferation, and pseudopalisading necrosis. The neoplasm was poor in reticulin fibers. Extensive areas of radiation effect were noted. Neoplastic cells were strongly and diffusely immunoreactive for GFAP. The MIB-1 labeling index was about 30–40%. p53 labeling index was yet relatively low, about 3–5% in most areas and focally higher, about 7–10%. Stain intensity was strong in a few cells only. Certainly, the p53 labeling index was not extremely higher in the second recurrence as one would be expected in a secondary glioblastoma or a typical gliosarcoma. Thus, the morphoimmunophenotype of the second recurrence appears to be entirely different from that of the primary neoplasm.


We have presented here a case of primary cerebral parenchymal fibrosarcoma that progressed to glioblastoma at the second recurrence, 10 months after initial diagnosis.

Primary cerebral fibrosarcoma is a rare neoplasm. Some reports have estimated incidence at about 1.5% [2, 3, 6], however, many of these estimates are based on series predating the era of immunohistochemistry. Pathological studies in this earlier period did not distinguish between meningeal or parenchymal brain and spinal cord lesions [1], thus actual incidence may be somewhat less than these already low estimates.

Intracranial fibrosarcomas are considered to be intrinsically aggressive lesions [4, 6] that should be treated forcefully with multimodality measures, hence the decision for three surgical interventions and adjuvant treatment with two forms of conformal radiation techniques [6].

As compared to the literature, there are some unique characteristics in this case. First, it is firmly documented that a parenchymal brain fibrosarcoma developed de novo (Fig. 1c, d), and progressed from a 5 mm hyperintense nodule on T2-weighted MRI to a 5 cm mass within a 12-month period (Fig. 1a, b). The authors are not aware of another case documenting the early natural history of a de novo fibrosarcoma.

Several cases in the literature describe a short interval to tumor recurrence following surgical resection and irradiation of fibrosarcoma [5, 6]. One could presume that an intrinsically aggressive tumor biology was perhaps influenced or aggravated by the effects of ionizing radiation on tumor differentiation [7].

In this patient the fibrosarcoma recurred 3 months after gross total resection with adjuvant radiotherapy and recurred a second time only 6 months after stereotactic radiosurgery that was adjuvant to surgery for the first recurrence. The patient had no history of irradiation prior to the first surgery.

A phenotype of a glioblastoma at the second recurrence is most unusual, and runs contrary to another rare, but more well-documented process of tumor progression, in which cerebral fibrosarcoma manifests following radiation therapy for glioblastoma, resulting in a diagnosis of gliosarcoma [7, 8, 9]. Instead, our patient’s tumor progressed along a less well-defined course, transforming from fibrosarcoma to glioblastoma at the second recurrence [10].

It is possible to speculate about several interesting alternative explanations for the histogenesis of such pathological rarities as mixed glial and mesenchymal intracranial tumors, or mixed glioma–sarcoma [11].

Historically, such mixed tumors were classified into three types: collision tumors—juxtaposed independent primary neoplasms that intermingle only at their common borders; composite tumors—lesions with synchronous participation of both stromal and parenchymal cells in the neoplastic process; and dependent or consequential tumours—cases in which one tumor element is secondarily induced by a primary lesion [9, 10, 11].

Our case may therefore be considered a dependent or consequential tumor, or a reactive glioma, on the basis of this classification. It is hypothesized that there could be transition from reactive astrocytes to neoplastic cell forms in glial islands that are incorporated into the advancing edges of the primary sarcoma [8, 10].

Another possible explanation for post-irradiation malignant glial transformation of a primary mesenchymal tumor may be the putative indirect effect that radiation is suspected to have on the tissue microenvironment [7]. This microenvironment is thought to facilitate the preferential growth of preexisting malignant cells. Irradiation may also induce primitive pluripotential mesenchymal cells in the tumor bed to undergo malignant change [7, 8, 12], giving rise to a mixed mesenchymal-glial lesion, a gliosarcoma, which might be the overall diagnosis in our case. There may, however, be some yet-to-be elucidated explanation for this rare clinico-pathologic occurrence [11, 13].

Whatever the etiology of this tumor is, our case has further confirmed the very aggressive biological behavior of primary cerebral fibrosarcomas, and the continued need to apply equally aggressive multimodality treatment from the outset in each new lesion.


We present a case of de novo cerebral fibrosarcoma with documented evolution from a 5 mm hyperintense area to 5 cm tumor mass in a 12-month period. At the second recurrence within 9 months from initial diagnosis, the tumor had transformed into a glioblastoma. The etiology of this highly unusual post-irradiation progression from primary mesenchymal neoplasm to high-grade glioma is unknown. Our experience with this patient concurs with other reports in the literature of the aggressive nature of intracranial fibrosarcoma. Aggressive multimodality therapeutic approaches to maintain tumor control should be considered.



The authors wish to thank Shifra Fraifeld, a research associate in the Department of Neurosurgery, for her editorial assistance with the preparation of this manuscript.


  1. 1.
    Oliveira AM, Scheithauer BW, Salomao DR et al (2002) Primary sarcomas of the brain and spinal cord: a study of 18 cases. Am J Surg Pathol 26(8):1056–1063. doi:10.1097/00000478-200208000-00011 PubMedCrossRefGoogle Scholar
  2. 2.
    Bisogno G, Roganovic J, Carli M et al (2002) Primary intracranial fibrosarcoma. Childs Nerv Syst 18(11):648–651. doi:10.1007/s00381-002-0628-6 PubMedCrossRefGoogle Scholar
  3. 3.
    Cai N, Kahn LB (2004) A report of primary brain fibrosarcoma with literature review. J Neurooncol 68(2):161–167. doi:10.1023/B:NEON.0000027740.70380.5c PubMedCrossRefGoogle Scholar
  4. 4.
    Gaspar LE, Mackenzie IR, Gilbert JJ et al (1993) Primary cerebral fibrosarcomas. Clinicopathologic study and review of the literature. Cancer 72(11):3277–3281. doi:10.1002/1097-0142(19931201)72:11<3277::AID-CNCR2820721124>3.0.CO;2-CPubMedCrossRefGoogle Scholar
  5. 5.
    Couldwell WT, Scheithauer BW, Rice SG et al (1997) Osteosarcoma of the meninges in association with glioblastoma. Acta Neurochir (Wien) 139(7):684–689. doi:10.1007/BF01412005 CrossRefGoogle Scholar
  6. 6.
    Fraser RW, Limentani SA, Dollar JD et al (2003) Recurrent primary fibrosarcoma of the brain treated with the gliasite brachytherapy system: case report. Surg Neurol 60(6):579–583. doi:10.1016/S0090-3019(03)00456-7 (discussion 583–584)PubMedCrossRefGoogle Scholar
  7. 7.
    Chang SM, Barker FGII, Larson DA et al (1995) Sarcomas subsequent to cranial irradiation. Neurosurgery 36(4):685–690. doi:10.1097/00006123-199504000-00007 PubMedCrossRefGoogle Scholar
  8. 8.
    Perry JR, Ang LC, Bilbao JM et al (1995) Clinicopathologic features of primary and postirradiation cerebral gliosarcoma. Cancer 75(12):2910–2918. doi:10.1002/1097-0142(19950615)75:12<2910::AID-CNCR2820751219>3.0.CO;2-APubMedCrossRefGoogle Scholar
  9. 9.
    Smith DR, Hardman JM, Earle KM (1969) Contiguous glioblastoma multiforme and fibrosarcoma with extracranial metastasis. Cancer 24(2):270–276. doi:10.1002/1097-0142(196908)24:2<270::AID-CNCR2820240210>3.0.CO;2-5PubMedCrossRefGoogle Scholar
  10. 10.
    Lalitha VS, Rubinstein LJ (1979) Reactive glioma in intracranial sarcoma: a form of mixed sarcoma and glioma (“Sarcoglioma”): report of eight cases. Cancer 43(1):246–257. doi:10.1002/1097-0142(197901)43:1<246::AID-CNCR2820430136>3.0.CO;2-JPubMedCrossRefGoogle Scholar
  11. 11.
    Henry JM, Leestma JE (1973) Astrocytoma arising in meningeal fibrosarcoma. Acta Neuropathol 23(4):334–337. doi:10.1007/BF00687463 PubMedCrossRefGoogle Scholar
  12. 12.
    Shapiro S, Mealey J Jr, Sartorius C (1989) Radiation-induced intracranial malignant gliomas. J Neurosurg 71(1):77–82PubMedCrossRefGoogle Scholar
  13. 13.
    Kaminski JM, Yang CC, Yagmai F et al (2000) Intracranial fibrosarcoma arising 5 years after chemotherapy alone for glioblastoma multiforme in a child. Pediatr Neurosurg 33(5):257–260. doi:10.1159/000055965 PubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC. 2008

Authors and Affiliations

  • Amos Olufemi Adeleye
    • 1
    • 2
  • Yakov Fellig
    • 3
  • Felix Umansky
    • 1
  • Yigal Shoshan
    • 1
  1. 1.Department of NeurosurgeryHadassah – Hebrew University Medical CenterJerusalemIsrael
  2. 2.Division of Neurological Surgery, Department of SurgeryCollege of Medicine – University of Ibadan and University College HospitalIbadanNigeria
  3. 3.Department of PathologyHadassah – Hebrew University Medical CenterJerusalemIsrael

Personalised recommendations