Article Text


Cellular schwannoma of the posterior fossa
  1. A L Green1,
  2. J S Yeh1,
  3. H L Brydon1,
  4. M P Carey2
  1. 1Department of Neurosurgery, North Staffordshire Royal Infirmary, Stoke on Trent ST4 6TA, UK
  2. 2Queen Elizabeth Hospital, Birmingham, UK
  1. Correspondence to:
 Mr Green;{at}

Statistics from

Schwannomas are slowly growing, non-invasive neoplasms derived from Schwann cells and usually arise from peripheral nerves.1 They may also arise from cranial nerves, most commonly the vestibular part of the VIIIth nerve. In this situation, they are sometimes associated with neurofibromatosis type 2 (NF2). The cellular variety of schwannoma has been described as a distinct “pseudosarcomatous” entity, composed of hypercellular areas of spindle shaped cells that can easily be mistaken for a malignant tumour.2 There are fewer than 60 reported cases of intraparenchymal schwannoma3 in which the tumour is not associated with any cranial or peripheral nerves.

In 1991, a 37 year old woman presented with left sided facial pain and numbness of several weeks' duration, accompanied by blurring of vision in her left eye. Computed tomography revealed a left trigeminal schwannoma. This was resected and, macroscopically, the tumour appeared to have been completely removed. Histology showed a schwannoma of normal cellularity with some areas composed of compact spindle cells arranged in short bundles and other areas with cells set in a loosely textured matrix containing some large, irregular vessels. There were only occasional mitotic figures and the tumour showed diffuse positivity for S100. She developed a recurrence in 1994 and underwent a further, presumed complete, resection. However, a remnant was discovered in 1995 and treated with stereotactic radiosurgery. She was followed up with regular cranial magnetic resonance (MR) scanning and a scan in March 1998 appeared satisfactory, with minimal further tumour growth and no tumour elsewhere in her brain (fig 1A).

She presented eight months later (now aged 44 years) with a two month history of increasing posterior and right sided neck pain. She also complained of headaches but had no other features of raised intracranial pressure. Examination was unremarkable except for the longstanding left trigeminal nerve palsy. A further MR scan was obtained and this revealed a 20 mm diameter tumour behind the brain stem, which was extending through the foramen magnum to the posterior arch of C1 (fig 1B).

The tumour was removed using a posterior approach. Macroscopically it originated from the posterior surface of the brain stem and was remote from any peripheral nerve. Histology showed tumour composed exclusively of compact spindle cells arranged in short bundles and with focal nuclear palisading. It was more cellular than previously and showed moderate nuclear polymorphism and slightly more frequent mitotic figures than the previous specimens. There was a large central area of necrosis and it was diffusely positive for S100. The appearance was that of a cellular schwannoma. The proliferative index was measured (two years later) using MIB1 (Ki67) antibody and counting automatically using the Kontron 3000 system. The result, counting 1000 nuclei, was 22%.

In early 2000, the patient underwent genetic testing which revealed no alterations in the NF2 gene.

In March 2000 she presented again with neck pain and headaches. An MR scan showed recurrence of the tumour at the foramen magnum. This was resected, and the histology indicated recurrences of the cellular schwannoma. Her postoperative recovery was complicated by a breakdown of her wound, following which she developed pseudomonas meningitis. This led to the development of hydrocephalus, which required external ventricular drainage. An MR scan two months after the operation showed several lesions in the mid-thoracic spinal cord, which were suggestive of metastases. There was also dural thickening at the level of S2 downwards. These findings suggested extensive meningeal spread of the disease. She died soon afterwards from aspiration pneumonia.

In this patient, the cervicomedullary schwannoma arose from the pia of the brainstem, and was well away from the lower cranial or upper cervical nerves. These intraparenchymal schwannomas are generally indolent in nature and present in a variety of ways. Younger patients tend to present with a longer duration of illness (often seizures or headaches), whereas older patients tend to have a more rapid clinical course, with marked neurological deficits.3 Comparison between the two MR scans (fig 1A and B) shows that the cellular schwannoma had grown to 20 mm diameter in eight months (that is, a growth rate of 30 mm a year). This is faster than in any previously published report.

However, it is possible that the cervicomedullary tumour was a metastasis from the previous trigeminal schwannoma. This cervicomedullary tumour was originally considered to be benign on the basis of the histological findings, but was later found to have a proliferation index of 22%—surprisingly high considering the relative sparsity of identifiable mitotic figures.4 As our patient developed spinal lesions that were suggestive of metastases (unfortunately these were not sectioned at necropsy), the possibility that the original foramen magnum tumour was a metastasis from the previous trigeminal schwannoma is more likely. The difficulty in determining whether or not these lesions are malignant has important implications for the surgeon when considering how aggressive to be with treatment.

Another possibility is that our patient may have had NF2 or schwannomatosis. Tumours in both of these conditions behave differently from solitary cases, with faster growth rates and a more fulminant clinical course.5 In our patient, genetic studies showed no alteration in the NF2 gene (although these are only 60–70% sensitive). Also, she had a trigeminal rather than a vestibular schwannoma. These two factors suggest that NF2 is less likely but not impossible. Schwannomatosis is characterised by multiple non-vestibular schwannomas, in the absence of meningiomas, intraspinal ependymomas, and other clinical signs of NF2.5 Although this is consistent with our case, the aggressive behaviour of the tumour, as well as the histological findings (in particular, a lack of Verocay bodies, which are common in schwannomatosis5), suggest that this diagnosis was unlikely.

In summary, our unusual case of a cellular schwannoma of the posterior fossa underlines the difficulty in determining the exact nature of these lesions, both histologically and clinically. Despite benign histological appearances, this posterior fossa tumour behaved as a malignant peripheral nerve sheath tumour (MPNST). This is the first time that growth rate has been reported for this particular type of tumour.

Figure 1

(A) Axial T1 weighted magnetic resonance imaging (MRI) (without contrast), eight months before discovery of a cervicomedullary lesion, showing no evidence of tumour. (B) Axial T1 weighted MRI (without contrast) at presentation showing a large tumour at the level of the foramen magnum.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.