adult intradural primary spinal cord tumors · and are considered part of the differential...

Original Article

© JNCCN–Journal of the National Comprehensive Cancer Network | Volume 9 Number 4 | April 2011

434

Key WordsPrimary spinal cord tumors, intradural intramedullary, intradural extramedullary

AbstractPrimary spinal cord tumors represent 4.5% of all central nervous system neoplasms. They are either intradural intramedullary or intradural extramedullary. Intramedullary tumors are predomi-nantly intrinsic gliomas (astrocytomas and ependymomas). Spinal ependymomas can usually be completely removed by separating the tumor from the spinal cord and, when complete, no further therapy is required. Astrocytomas, by contrast, infiltrate the my-elon, and therefore surgery is frequently incomplete. Intradural extramedullary tumors are mostly benign (WHO grade 1) and comprise either peripheral nerve sheath tumors (neurofibromas and schwannomas) or meningiomas. Complete resection can be performed on both lesions and is often curative. Radiotherapy is indicated for primary malignant tumors (WHO grade 3 and higher) and for patients in whom surgery is contraindicated. For grade 1 and 2 tumors, the role of radiotherapy is controversial. Chemo-therapy is reserved for recurrent primary spinal cord tumors with no other options. However, the lack of clinical trials for these tu-mors is problematic. Consequently, treatment is similar to that for intracranial histologies. Early recognition of the signs and symp-toms of primary spinal cord tumors facilitates early treatment, po-tentially minimizes neurologic morbidity, and improves outcome. Primary treatment for almost all spinal cord tumors is surgery, with predictors of outcome being preoperative functional status, grade of tumor, and extent of resection. (JNCCN 2011;9:434–447)

Primary spinal cord tumors account for 5% to 10% of all adult spinal tumors1 and 4.5% of primary central ner-vous system (CNS) tumors. Approximately 850 to 1700 new adult cases of primary CNS spinal cord tumors are diagnosed each year in the United States. Primary spi-nal cord tumors unlike intracranial tumors do not show an association between grade and age at diagnosis.

Because tumor origin varies by anatomic site in the spinal cord, primary spinal cord tumors are classified by anatomic sublocation as either intradural intramedullary, intradural extramedullary, or extradural. Intradural intra-medullary spinal cord tumors constitute 20% to 30% of all primary spinal cord tumors; the remaining 70% to 80% are intradural extramedullary. Most (90%) intradural in-tramedullary tumors are either ependymomas (60%–70%) or astrocytomas (30%–40%). For the remaining 10%, he-mangioblastomas account for 3% to 8%, of which 15% to 25% are associated with von Hippel-Lindau (VHL) disease. Another 2% of intradural intramedullary spinal cord tumors are metastatic. Intradural extramedullary spi-nal cord tumors predominantly comprise either menin-giomas (30%) or peripheral nerve sheath tumors (30%). Overall, and including the lumbar cistern, the most com-mon intradural extramedullary tumor types are meningio-mas (24.4%), ependymomas (23.7%), and schwannomas (21.2%). Anatomically, the sites of primary spinal cord tumors are the spinal cord (70.5%), spinal meninges (24.2%), and cauda equina (5.3%).2

The clinical presentation of primary spinal cord tu-mors is determined partially by the location (Tables 1 and 2). Pain is the most common presenting symptom regardless of location,1,3 and is manifested as back pain (27%), radicular pain (25%), or central pain (20%). Neurologic deficit is the second most common symp-tom and can occur in the form of motor (72%), sensory (39%), or sphincter disturbance (15%).3 Intramedul-lary tumors usually affect central gray matter and cause

From aH. Lee Moffitt Cancer Center, Department of Neuro-Oncology, Tampa, Florida, and bUniversity of Washington, Department of Neurology/Division of Neuro-Oncology, Seattle, Washington.*Made equal contributions to the manuscript.Submitted November 5, 2010; accepted for publication January 20, 2011.The authors have disclosed that they have no financial interests, arrangements, or affiliations with the manufacturers of any products discussed in this article or their competitors. Correspondence: Frank Vrionis, MD, PhD, H. Lee Moffitt Cancer Center, Department of Neuro-Oncology, 12902 Magnolia Drive, Tampa, FL 33612. E-mail: [email protected]

Adult Intradural Primary Spinal Cord Tumors

Kamran Aghayev, MDa; Frank Vrionis, MD, PhDa*; and Marc C. Chamberlain, MDb*; Tampa, Florida, and Seattle, Washington

Original Article



435

a syringomyelitic syndrome characterized by the disassociation between pain/temperature (loss) and tactile (preserved) sensations at affected levels, low-er motor nerve dysfunction at the affected level, and upper motor nerve dysfunction caudal to the lesion. No symptoms are pathognomic for spinal cord tu-mors, although sacral sparing (i.e., maintenance of sensation in the sacral dermatomes) and upper mo-tor deficits are common in intramedullary tumors, whereas a Brown-Séquard–type syndrome (hemi-cord dysfunction) is characteristic of extramedullary tumors. Extramedullary tumors mostly present clini-cally with pain and myelopathic symptoms, such as paraparesis. Schwannomas, in contrast to other tu-mors, are more prone to cause radicular symptoms.4

General Considerations

ImagingMRI (with and without contrast) plays an essential role in the diagnosis of primary spinal cord tumors. Currently, no other imaging modality can be used alone to establish a diagnosis. Plain radiographs may show scalloping, foraminal enlargement, bony ero-sion, and calcification. CT myelography may show spinal cord enlargement in cases of intramedullary tumors, or a filling defect in extramedullary tumors. However, the internal structure of the spinal cord and tumor and the tumor/cord interface cannot be visualized.

The presence of mass effect in the form of spi-nal cord segmental enlargement, cyst formation, contrast enhancement, and peritumoral edema favors the diagnosis of a neoplastic process. Non-neoplastic lesions that can be interpreted as tumors and are considered part of the differential diagnosis are multiple sclerosis, transverse myelitis, spinal cord ischemia, sarcoid, and CNS angiitis.5,6 Acute demy-

elination, such as in multiple sclerosis and transverse myelitis, may be associated with spinal cord edema and segmental enlargement, mimicking a tumor.1 Additional studies, such as brain MRI, cerebrospi-nal fluid (CSF) analysis, and angiography, should be performed in suspected cases to establish the correct diagnosis. In rare cases, when diagnosis cannot be established based on correlative data, patient obser-vation with follow-up imaging in 2 to 3 months is reasonable. Spinal cord edema in non-neoplastic le-sions usually regresses on interval MRI.

SurgeryMicrosurgery is the cornerstone of spinal cord tumor treatment. Tumor type and grade have been shown to be the most important factors affecting outcome (Table 3). Surgery allows tissue sampling, and con-sequently a pathologic diagnosis with correspond-ing prognosis, and removes bulk tumor (i.e., cyto-reduction). In most well-defined (circumscribed) low-grade tumors, resective surgery can be curative. With infiltrative tumors, maximal safe resection or biopsy helps provide a diagnosis and define further treatment. Diffusion tension tractography has been shown to be a useful tool that can demonstrate the passage of fibers around or though the tumor, and therefore can be used in preoperative planning.7

Class 3 evidence shows that the extent of re-section positively correlates with patient outcome. However, this approach is counterbalanced by the risk of neurologic injury from aggressive surgery. Most postsurgery deficits, however, are transient and improve with time and rehabilitation. The McCormick grading score that is based on patient functional status and limited longitudinal extent are the most important factors for a favorable outcome.8 Delayed postoperative neurologic deterioration may occur from tumor growth (most frequent) or opera-tive complications, such as spinal cord tethering to the dura, spinal instability, and resulting deformity. Spinal instability may be present preoperatively or postoperatively due to neurologic deficits, erosion of bony/ligamentous elements by the tumor, postlami-nectomy kyphosis, or radiation injury.

The role of laminoplasty versus laminectomy has been debated, with some studies showing an advantage with laminoplasty in preventing a future deformity, at least in the pediatric population. This effect seems to be less prevalent in the adult group, although laminoplasty facilitates surgical exposure

Table 1 Intramedullary Spinal Cord Tumor: Topography

Cervical 30%

Cervicothoracic 25%

Thoracic 29%

Conus 15%

Data from Raco A, Esposito V, Lenzi J, et al. Long-term follow-up of intramedullary spinal cord tumors: a series of 202 cases. Neurosurgery 2005;56:972–981

Original Article

Aghayev et al.


436

in recurrent cases.9 Yasargil et al.10 reported on a se-ries of 100 patients with intraspinal arteriovenous malformations and tumors treated with hemilami-nectomy. Although technically challenging, this approach ultimately eliminates the risk of surgery-induced instability.

Radiation TherapyRadiotherapy is an important adjuvant therapy for the treatment of spinal tumors. Radiotherapy is primar-ily administered for high-grade gliomas and recurrent or residual tumors with confirmed progression when

surgical resection is not possible. Radiotherapy-as-sociated side effects are characterized as acute, early delayed, and late delayed. Acute reactions usually reflect secondary inflammatory and transient effects on nearby tissues (in-field effects), especially skin and gastrointestinal tract. Early delayed side effects most often manifest as transient demyelination and posteri-or column dysfunction (i.e., Lhermitte’s sign) that can be seen on T2-weighted images. Surprisingly, a recent series of benign intradural spinal tumors treated with radiosurgery showed that all patients developed radi-

Table 2 Intradural Intramedullary TumorsTumor Type Incidence Presentation MRI Appearance Treatment Outcome

Ependymoma Most common intramedullary spinal cord tumor in adults

Mixed sensorimotor syndrome is most common

Syringomyelic syndrome may be present

Hyperintense on T2-weighted and hypo- or isointense on T1-weighted images with heterogenous contrast enhancement

Surgery is the most effective treatment

Radiotherapy is reserved for recurrent or malignant tumors

Chemotherapy is limited to recurrence after radiotherapy; agents include etoposide and carboplatin

Local control rate 90%–100% after complete resection

Astrocytoma Second most common intramedullary spinal cord tumor

Mixed sensorimotor syndrome is most common

Syringomyelic syndrome may be present

Fusiform expansion with irregular margins

Often with a cystic component, associated edema, or syrinx

Hypo- to isointense on T1, hyperintense on T2, with variable contrast enhancement

Maximal safe surgical resection followed by observation or radiotherapy

Total resection is accomplished infrequently

Radiotherapy only indicated for clinical or radiographic progression, not for pilocytic astrocytoma

Grade is the most important prognostic factor. 5-year survival exceeds 70%

Hemangioblastoma Rare, although common in patients with von Hippel-Lindau

Usually sensory, especially slowly worsening proprioception

Rarely, patients present with acute hemorrhage

Homogenously enhancing, hypervascular nodule with associated cyst or syrinx

Angiography shows enlarged feeding arteries, intense nodular stains, and early draining veins

Maximal safe surgical resection followed by observation

Preoperative embolization may be considered

Radiotherapy has no role and experience with chemotherapy is limited

Stereotactic radiosurgery is an option for recurrent or unresectable tumors

Local control rate is almost 100% after complete resection

Cavernous malformation

Rare Mixed sensorimotor, acute onset in 38% of cases

Mixed signal intensity with hypointense rim on T2

Maximal safe surgical resection followed by observation

Radiotherapy and chemotherapy have no role

Local control rate is almost 100% after complete resection

Original Article



437

ation-induced myelitis despite having a radiographic margin between the tumor and the spinal cord.11 De-layed late injuries include secondary malignancies, particularly in the pediatric population and in pa-tients with genetic tumor predisposition disorders.12,13

Intradural Intramedullary Tumors

EpendymomaEpendymoma is the most common intradural intra-medullary tumor type in adults.14,15 Myxopapillary ependymoma (WHO grade 1) is a distinct type, predominantly extramedullary and located in the lumbar cistern, and is considered in the intradural extramedullary section. Cellular ependymomas arise from the ependymal lining of the central canal and are classified as WHO grade 2 tumors. Anaplastic ep-endymomas are rare and considered WHO grade 3 tumors. The most common presentation is pain, fol-lowed by neurologic deficit. Rare cases of acute neu-rologic compromise from tumor hemorrhage have been reported.16

Histologically, ependymoma cells are character-ized by round to oval nuclei containing finely dis-persed chromatin with perivascular and ependymal rosettes.17 The association between neurofibroma-tosis type 2 (NF2) and spinal ependymoma is well known. This is an autosomal dominant disease caused by mutation of the merlin or schwannomin gene on chromosome 22, which is a member of the protein 4.1 family.18

The most common location of ependymomas is the cervical spine, followed by the cervicothoracic and thoracic spine. Ependymomas are hypointense on T1- and hyperintense on T2-weighted MR im-ages (Figure 1A). Contrast enhancement is usually

present. Ependymomas mostly reside centrally in the spinal cord, attributed to the origin of these tumors from the ependymal cells lining the central canal. Reactive syrinx usually occurs at the superior or in-ferior poles of the tumor, and does not require surgi-cal removal. Cyst presence does not seem to have an effect on overall prognosis.19,20 The main factor affecting postoperative morbidity is tumor width in relation to spinal cord19 and preoperative neurologic status.19,21 Tumor length does not have an effect on preoperative neurologic status or morbidity, but lon-ger tumors are associated with postoperative dyses-thesias,19 possibly because of longer myelotomy. A recent study of very long intramedullary spinal cord tumors showed that none of 13 holocord tumors were ependymomas.22

Ependymomas mostly follow a benign course, except for malignant histologic subtypes (anaplastic ependymoma; WHO grade 3). Surgery is the most effective treatment for all grades, and the same rate of complete resections can be achieved even for re-current tumors.23 Local control rates of 90% to 100% have been reported with total surgical resection, al-though in a minority of cases complete surgical re-section is not achieved.24–26 An inverse relationship exists between preoperative symptom duration and tumor resectability. This effect may be explained by the possible development of microadhesions be-tween tumor and normal tissue in longstanding cases. Overall survival increases with complete resection.27

Chemotherapy plays a limited role in the man-agement of ependymomas. Data regarding the effect of chemotherapy for patients with recurrent spinal cord ependymoma is extremely sparse. Etoposide and platinoids are generally considered the preferred agents commonly used for recurrent intracranial ependymomas.28 No evidence shows that temozolo-mide is effective in ependymomas, notwithstanding its widespread use in other gliomas.29 Recently, epi-dermal growth factor inhibitors (particularly lapa-tinib) and antiangiogenic inhibitors (i.e., bevaci-zumab) have been shown to be useful in recurrent intracranial ependymoma and may be applicable for recurrent spinal cord ependymoma.

Radiotherapy is an option for incomplete resec-tion and is indicated for malignant grade30 (external beam radiotherapy at 45 to 54 Gy).25,31–33 However, several studies have shown no benefit in survival or time to recurrence for its use in incompletely resected

Table 3 Primary Spinal Cord Tumors: 5-Year Survival

Astrocytoma

Juvenile pilocytic > 90%

Low-grade > 70%

High-grade 30%

Ependymoma

Low-grade 85%

Anaplastic 30%

Myxopapillary > 95%

Original Article

Aghayev et al.


438

low-grade ependymoma.23,27,34,35 Therefore, because of the slow growth of low-grade ependymomas, ob-servation with sequential MRI can be performed and radiotherapy administered at disease progression.36 Patients undergoing total resection should also be followed up with serial neurologic examinations and MRI. If no recurrence is present, time between MRIs can be increased and the patient can be followed up neurologically.

AstrocytomaAstrocytomas are a diverse group of gliomas with differing biological behavior. Histologically, gliomas can be differentiated as pilocytic (WHO grade 1), fibrillary (WHO grade 2), anaplastic (WHO grade 3), and glioblastoma (WHO grade 4). Grade of the glioma is the major prognostic factor with regard to outcome,27,37 yet studies have shown that histo-pathologic distinction between anaplastic astrocy-tomas and glioblastomas is not of prognostic signifi-cance.38 Astrocytomas are associated with NF1,39,40 although the exact mechanism of tumor formation is unknown.

Radiologically, astrocytomas are character-ized by heterogenous contrast-enhancement; show isointensity or hypointensity on T1-weighted and hyperintensity on T2-weighted images; and are chal-lenging to differentiate from ependymomas. Con-trast enhancement does not, however, predict tumor grade or behavior, because pilocytic astrocytomas commonly have intense and homogenous contrast enhancement but rarely behave aggressively. The likelihood of total surgical resection for pilocytic as-trocytomas is higher than for WHO grade 2 through 4 gliomas.8 Astrocytomas typically occupy an eccen-tric position in the spinal cord and may be associated with cysts and may not contrast-enhance as much as ependymomas (Figure 1B). WHO grade 2 through 4 astrocytomas lack distinct borders within normal neuronal tissue, reflecting their infiltrative nature.

Microsurgery is an important part of treat-ment of astrocytomas. Total excision rates are lower than in ependymoma, with a partial resec-tion or biopsy performed in most patients. Some authors have shown that complete resection sig-nificantly reduces the risk of disease progression,27 whereas others suggest no benefit to extensive resection.37,41Although rare, adult pilocytic astro-cytomas have a relatively benign course, and total resection can be attempted. However, if intraoper-

ative biopsy confirms the diagnosis of astrocytoma grade 2 or higher, and no clear surgical plane exists between tumor and neuronal tissue, further tumor resection is not recommended.42

As with other primary spinal cord tumors, lim-ited data address the role of chemotherapy. Temo-zolomide has been used43,44 with some success in the recurrent setting, but no trials establish a role for its adjuvant use (i.e., up-front). Nonetheless, and as in the management of intracranial gliomas, spinal cord high-grade tumors are commonly treated with con-comitant (chemoradiation) and postradiotherapy temozolomide. A recent small case series also sug-gested that salvage therapy with bevacizumab is of palliative benefit for recurrent spinal cord glioblas-toma.45 Lomustine, carboplatin, and vincristine, and the 8-in-1 chemotherapy protocol have also been used in children.46–49

Adjuvant radiotherapy plays an important role in the management of spinal cord astrocytomas.50 Because biopsy or partial resection is the surgical result in most patients, radiotherapy is usually ad-ministered. Radiotherapy does not confer a survival benefit in patients with WHO grade 2 tumors,37 but it reduces the risk of disease progression.27 In high-grade gliomas, the survival benefit of radiotherapy is established (or at least accepted as usual and cus-tomary) and commonly applied. In addition, radio-therapy is the preferred treatment for recurrent glio-mas not already irradiated. Serial neurologic status assessment and MRI is the usual follow-up strategy to monitor tumor recurrence.

Pilocytic astrocytomas differ by manifesting a low potential for malignant transformation, and therefore radiotherapy is usually deferred notwith-standing incomplete resection.51 The usual approach involves close follow-up with serial MRI and repeat surgery for recurrence.

HemangioblastomaHemangioblastomas are the third most common pri-mary spinal cord tumor in adults, representing 3% to 8% of all intramedullary tumors. They are seen predominately in men and presentation usually oc-curs in the fourth decade.52,53 Among patients with hemangioblastoma of the spinal cord, 10% to 30% have VHL syndrome, an autosomal dominant dis-order caused by deletion of chromosome 3p25. De-creased VHL protein levels induce cell prolifera-tion and increase expression of vascular endothelial

Original Article



439

growth factor (VEGF), platelet-derived growth fac-tor, and erythropoietin.54 Therefore, chemothera-peutic agents targeting VEGF have been shown to be effective in hemangioblastoma.

The origin of hemangioblastomas is unknown. Anatomically, most hemangioblastomas arise from the dorsal portion of the cervical spinal cord, close to the dorsal root entry zone.52,53,55,56 and can be in-tramedullary (30%), intramedullary and extramed-ullary (50%), or extramedullary (20%).56

Radiographically, the tumor presents as a ho-mogenously enhancing hypervascular nodule with associated cyst or syrinx and peritumoral edema on MRI.52,56 Spinal angiography shows enlarged feed-ing arteries, intense nodular stains, and early drain-ing veins. These tumors are highly vascular and can present with intramedullary or subarachnoid bleed-ing.57–62 Endovascular embolization can be performed

to decrease vascularity of the tumor and facilitate resection.52,63–66 However, usually neither diagnostic angiography nor endovascular embolization is re-quired in managing these tumors.56

Understanding the natural history of hemangio-blastomas in the setting of VHL is an essential part of treatment, and experts typically recommend that resection of hemangioblastomas be reserved until the onset of the symptoms.67

Surgical resection is the main treatment modal-ity. Well-defined margins are usually present, allow-ing for total resection and cure. Excessive intraop-erative bleeding that obscures the operative field can be the limiting factor in subtotal resection.52 This complication can be avoided through adherence to the technique of circumferential dissection, feeder coagulation, and preservation of draining vein up to end of the operation.55,56

A B

C D

➡

Figure 1. Radiologic appearance of common primary spinal cord tumors. (A) Ependymoma; note segmental enlargement of spinal cord at the level of tumor. (B) Recurrent pilocytic astrocytoma; note extensive cyst formation with spinal cord edema. (C) Schwannoma; note posteriolateral localization according to typical origin from sensory root. (D) Meningioma; note “dural tail” sign (arrow).

Original Article

Aghayev et al.


440

Radiotherapy has no role in the management of hemangioblastomas, and experience with chemother-apy is limited. Stereotactic radiosurgery is an option for patients with recurrent or unresectable tumors.68 Individual patients and small case series have shown clinical and radiographic response to SU5416 (semax-anib, a VEGF receptor inhibitor) and bevacizumab (monoclonal antibody against VEGF-A).69 Follow-up of patients with sporadic hemangioblastoma usually includes serial MRI with increasing time span. If no recurrence is present, the patient may be switched to clinical observation only. However, patients with VHL syndrome should undergo lifelong follow-up with MRI.

Intramedullary Cavernous MalformationCavernous malformations are considered a rare dis-order representing approximately 5% of all spinal lesions.70 Familial cases have been reported mostly related to KRIT1/CCM1 gene mutations.71,72 CCM2 and CCM3 mutations have been reported as a cause of multiple cerebral but not spinal cavernous mal-formations. Spinal cavernous malformations have a slight female preponderance, and the peak of inci-dence is in the fourth decade.73,74 Grossly, they have a mulberry-like appearance and are characterized by sinusoidal channels under magnification. Unlike hemangioblastomas, they are not highly vascular le-sions because of the lack of feeding arteries.

Despite the relatively small size and low growth potential of cavernous malformations, neurologic deficits are not infrequent. In the largest retrospec-tive series from the French Study Group of Spinal Cord Cavernomas, initial neurological symptoms were acute in 38% of patients, mostly because of he-matomyelia (70%), and were progressive in 32% of patients mostly because of compression (75%).73

On MRI, cavernous malformations typically have mixed signal intensity with a hemosiderin rim around the lesion on T2-weighted images. The ab-sence of contrast enhancement is a distinguishing feature of hemangioblastomas.

Microsurgery is the mainstay of cavernous mal-formation treatment. Usually these lesions are ap-proached posteriorly through hemilaminectomy/laminectomy.70,75 Lesions are usually well circum-scribed and sometimes associated with hematoma cavity, allowing total resection in most cases. Hemo-siderin rim of the tumor should be left intact to pre-vent additional deficit; however, transient new post-operative deficits are seen in half of the patients.73

RT and chemotherapy have no role in the man-agement of cavernous malformations. Postoperative MRI should be obtained to confirm resection ex-tent. The presence of a hemosiderin rim after sur-gery is not a sign of incomplete resection and does not require radiologic follow-up. Follow-up usually requires serial MRI with increasing time span, and if no recurrence is seen, patients can then be followed up with clinical observation.

Intramedullary Spinal Cord MetastasisWith current treatment modalities patients with can-cer have longer survival time, and thus the incidence of CNS metastasis is increasing. However, intramed-ullary spinal cord metastases (ISCMs) are rare and represent fewer than 2% of primary intramedullary spinal cord tumors.76 Although any malignant tumor can spread into the spinal cord, the most common types are lung cancer, breast cancer, and melanoma. The presence of ISCM should be suspected in patients with known malignancy that have an intramedullary mass. Clinically, these tumors often grow more rap-idly than primary spinal cord tumors. Therefore, pro-gression of neurologic symptoms is also rapid, with higher incidence of complete deficits.5 Consequently, these tumors have worse outcomes than primary spi-nal cord tumors.5,77–79

Radiologically, ISCM may be indistinguishable from primary spinal cord tumors, showing contrast enhancement and peritumoral edema formation on T2-weighted images.78

Treatment for patients with ISCM should be individualized based on tumor localization and size, neurologic deficit, cancer type, extension of meta-static disease, and other medical conditions. A high rate of CSF dissemination is seen with widespread leptomeningeal disease. Therefore, complete brain and whole-spine MRI should be obtained before treatment. Most cases do not require biopsy because these tumors are managed with radiotherapy and steroids, especially in the presence of radiosensitive tumors. However, given the aggressive behavior of these tumors, with 70% to 75% of patients develop-ing complete neurologic deficit in 1 month,5,78 surgi-cal intervention may be warranted. Additionally, in select cases, surgery is necessary, as in cancer of un-known origin. Patients treated with surgery may have a better prognosis than those treated only with radio-therapy,78 although the results are at least partially due to selection bias. Nonetheless, when an ISCM is

Original Article



441

amenable to total removal, vascularity and adhesive-ness can impede complete resection,5,78 making cyto-reduction an alternative approach.79–81Additionally, aggressive surgical treatment does not affect overall survival and is often associated with poor functional outcome.77 Intraoperative biopsy is important in se-lected patients, especially in those with no known history of cancer. Systemic chemotherapy has a limited role because these tumors are in a relatively secluded site behind a pharmacologic barrier much like brain parenchymal metastasis. Consequently, ra-diotherapy is the primary palliative therapy for most patients. Intra–CSF chemotherapy has no role in the treatment of ISCM because it has limited diffusion into spine parenchyma.

Miscellaneous Other rare primary spinal cord intramedullary tu-mors include ganglioglioma, lymphoma, germinoma, primitive neuroectodermal tumor, teratoma, dermoid cyst, epidermoid cyst, lipoma, and hamartoma.82–84

Intradural Extramedullary Tumors

Myxopapillary EpendymomaMyxopapillary ependymoma is a WHO grade 1 ep-endymal tumor usually found in the conus or filum region from which it arises. Radiologic features are similar to those of cellular ependymomas, with pro-nounced contrast enhancement. Grossly, these tu-mors are well circumscribed and encapsulated, but invasive tumors with nerve root encasement and even extravertebral invasion have been reported. Histologically, myxopapillary ependymomas are characterized by myxoid changes in the stroma and papillary organization of tumor cells.85

Microsurgery plays an essential role in the treat-ment of these tumors, given their benign nature and resectability. In cases of filum tumors, the tumor can be removed in en bloc fashion.86 However, some authors reported that complete resection was more difficult than with cellular ependymomas.87 Local recurrence and CSF dissemination can occur with piecemeal removal, and even with tumor capsule violation, during surgery.85

Similar to WHO grade 2 or 3 spinal ependy-momas, sparse literature suggests a role for chemo-therapy. Platinum salts and etoposide are the most common agents used for recurrent radiation- and

surgery-refractory ependymomas in the intracranial compartment, and a similar strategy has been applied to spinal ependymomas.

Radiotherapy plays a minor role in the treatment of these tumors, because given the benign nature of the tumor, radiotherapy is typically not given in in-stances of incomplete removal. However, one study showed that adjuvant radiotherapy resulted in better local tumor control in patients who had either gross total or subtotal resections.88 Remote metastasis can occur in the central nervous system, including the brain,88 but no compelling evidence currently sug-gests a benefit from wider neuraxis irradiation.85

Follow-up usually includes neurologic monitor-ing and serial MRI with an increasing time span. Surgery should be considered for recurrence of re-sidual tumor growth.

SchwannomasSchwannomas are common tumors that usually arise from the posterior (i.e., dorsal) nerve root, al-though sometimes they can begin in the anterior root. Schwannomas originate from a single fascicle, and the remainder of the nerve root is intact. Be-cause the most common origin is the dorsal nerve root, they usually reside posteriolateral to the spinal cord (Figure 1C). Anatomically, schwannomas are classified as intradural (group 1; 54%), intra- and ex-tradural (group 2; 8%), extradural (group 3; 14%), extradural with extension to the intervertebral fo-ramen (group 4; 17%), and intra- and extradural with extension to the intervertebral foramen (group 5; 6%).89 Intramedullary schwannomas are rare and believed to arise from the penetrating nerve fibers accompanying the anterior spinal artery,90 although different theories exist.91 Most schwannomas are be-nign encapsulated tumors referred to as WHO grade 1, although malignant subtypes exist (malignant schwannoma). Schwannomas usually present in the fourth through sixth decades.

The association between schwannomas and NF2 is well-known. NF2 is caused by mutation of the merlin gene on chromosome 22q12. This gene is involved in cell membrane stability and motility, and intercellular adhesion.54,92 Mutation of the sec-ond allele is necessary for tumor formation, and the phenotype correlates with the type of mutation. All spontaneous schwannomas also have merlin gene

Original Article

Aghayev et al.


442

mutations.93 Carney syndrome is another entity char-acterized by cardiac myxomas, spotty skin pigmenta-tion pituitary tumors, endocrine hyperfunction, and melanocytic schwannomas.94

Histologically, schwannomas are characterized by a biphasic pattern consisting of highly cellular re-gions with spindle-shaped cells in columns (Antoni type A) and hypocellular regions with large, vacu-olated, pleomorphic, stellate cells (Antoni type B).

On MRI, schwannomas appear as solid tumors in the dorsal sensory root region, with displacement of the spinal cord, conus medullaris, or filum termi-nale.95 They are iso- or hypointense on T1-weighted images and hyperintense on T2-weighted images.96

Contrast enhancement varies from intense homoge-nous to faint enhancement. If contrast enhancement is faint or absent, differential diagnosis from perineu-ral cyst may be difficult.95 A cyst may be present and is believed to arise from degeneration of Antoni B por-tion of the tumor.97 Anatomically, spinal cord nerve sheath tumors are evenly distributed throughout the spinal axis (upper cervical, 16%; cervical, 31%; tho-racic, 22%; conus medullaris, 7%; and cauda equina, 24%).5,89 Usually only neural foramina are enlarged because of tumor erosion; however, cases of extensive bony destruction have been reported.98 Hemorrhage is rare, but intratumoral, subdural, subarachnoid, and even intramedullary hemorrhage can be pres-ent.97 When large, these tumors may extend beyond neural foramina and have an hourglass appearance. Schwannoma and neurofibroma have no differenti-ating features on MRI. Melanocytic schwannomas are distinct and hyperintense on T1-weighted imag-es, and hypointense on T2-weighted images, because of the paramagnetic effect of melanin.97

Microsurgery is the primary treatment for schwannomas. The uninvolved nerve root of ori-gin can usually be spared. Because most schwanno-mas originate from dorsal sensory nerve roots, even amputation of the involved root usually does not cause postoperative deficits. Grossly, schwannomas tumors are tan-colored, globoid masses attached to nerve roots,5 in contrast to melanocytic schwanno-mas, which appear as dark masses and may be mis-interpreted as melanomas. Surgery is usually cura-tive, with improvement of symptoms and minimal morbidity.91 Adjuvant therapy is not recommended; incompletely resected tumors should be followed by serial MRI and resected if growth is documented.

Stereotactic radiosurgery can be considered for pa-tients who are poor surgical candidates.11,99 Malig-nant schwannomas are treated with radiotherapy, even if total resection was achieved. The role of adjuvant chemotherapy is unclear, but because ma-lignant schwannomas are considered soft tissue sar-comas, sarcoma anthracycline-based chemotherapy protocols are often used.

Follow-up usually requires serial postoperative MRI and neurologic monitoring. Because total re-section is achieved in most cases, patients can be followed up with only neurologic observation af-ter several postoperative MRIs. Patients with NF2, however, are usually followed up with MRI for longer times, and brain imaging should also be performed in these patients.

NeurofibromasNeurofibromas are WHO grade 1 tumors that arise from peripheral nerves. Two clinicopathologic types are recognized: solitary and plexiform. Solitary neuro-fibromas are often discretely localized, small, globular, or fusiform nodules originating from a single sensory nerve. Plexiform neurofibromas are characterized by redundant loops of nerve fiber bundles and tumor tis-sue intermixed in a disorganized pattern, involving multiple nerves. In contrast to schwannomas, neu-rofibromas encase nerve roots rather than displacing them, and tend to grow more longitudinally. Neurofi-bromas arising from brachial and lumbar plexuses can extend into the spinal canal via multiple nerve roots in retrograde fashion.100 Solitary neurofibromas are of-ten small, isolated, painless nodules that mostly reside in the cutaneous or subcutaneous layer of the skin and can cause cosmetic problems or pain.

Up to 60% of neurofibromas occur in patients with NF1.5 NF1 is one of the most common autoso-mal dominant disorders, caused by mutation of the neurofibromin gene on chromosome 17q11.2; how-ever, 50% of cases are caused by de novo mutations.54 Neurofibromin is a tumor suppressor gene involved in the regulation of multiple pathways, including cellular growth and division.101 Genetic testing is available, but sensitivity is approximately 70%,54 re-quiring clinical diagnosis that is based on National Institutes of Health criteria.102 Plexiform neurofibro-mas occur almost exclusively in NF1,103and 10% of them eventually progress to malignant peripheral

Original Article



443

nerve sheath tumors.54 Therefore, aggressive treat-ment should be attempted in patients with NF1 who have a rapidly enlarging mass and worsening pain.

On MRI, neurofibromas typically appear as rounded or fusiform tumors that are isointense on T1-weighted images and hyperintense on T2-weighted images, similar to schwannomas.95 Intense enhancement is seen postcontrast, and cyst forma-tion is rare, unlike with schwannomas.

The origin of neurofibromas is unknown; how-ever, most authors suggest a mesenchymal source. Histologically, neurofibromas are characterized by bundles of Schwann cells within an abundant ma-trix of collagen, with intermingling axons passing though the tumor.

Patients with spontaneous neurofibromas have a better prognosis than those with NF1. Surgical resec-tion is the primary treatment for patients with soli-tary neurofibromas. Complete resection is sometimes achieved through sacrificing the nerve root of origin. Neurologic sequelae are not usually significant. In most cases, the tumor is intradural and spinal stabili-zation is not needed. Resection of a plexiform neuro-fibroma is technically challenging, and complete re-section is rarely achieved. Malignant peripheral nerve sheath tumors usually have infiltrative character and extraspinal extension. Subtotal resections have a very high rate of recurrence, and therefore total en bloc resection should be the goal whenever possible. However, this technically challenging approach usu-ally requires a multidisciplinary surgical team. Radio-therapy and chemotherapy have no defined roles for benign neurofibromas. Patients with neurofibroma-tosis are at risk for malignant transformation either spontaneously or after radiotherapy.13 Chemotherapy is currently limited to malignant peripheral nerve sheath tumors and is similar to the adriamycin-based regimen used for other soft tissue sarcomas.

The follow-up strategy of these patients is close to that for schwannoma. Benign neurofibromas re-sected totally are monitored with serial neurologic examinations and radiologic follow-up. Malignant cases require more close observation, followed by ra-diotherapy for recurrence.

MeningiomasMeningiomas are dural-based tumors that arise from arachnoid cap cells. They are the most common in-

tradural extramedullary tumor, representing 25% of all primary spinal cord tumors. Similar to intracra-nial meningiomas, women are affected more often than men; more than 80% of all patients with spinal cord meningiomas are women, in whom the thoracic region is the most common location (80%). In men, meningiomas are equally encountered in the cervi-cal and thoracic segments.4 Overall, 12% of all me-ningiomas are located in the spinal canal, with 15% in the cervical, 81% in the thoracic, and 4% in the lumbar regions. Anatomically, cervical meningiomas frequently occupy ventral and ventrolateral posi-tions because of a higher concentration of arachnoid villi containing cap cells around nerve root sleeves. The same factor possibly contributes to the lower in-cidence of meningioma in the lumbosacral region. Thoracic meningiomas tend to occur more laterally.

Genetic predisposition (NF2) and prior expo-sure to ionizing radiation are the only known risk factors. Merlin gene inactivation contributes to fa-milial cases, and most spontaneous meningiomas harbor merlin gene mutations.93 Meningiomas in the setting of NF2 have more aggressive behavior and are more often multiple than spontaneous types.104

Most spinal cord meningiomas are slow-growing WHO grade 1 tumors. Histologically, they are a rather diverse group of neoplasms with numerous subtypes, including meningothelial, fibrocytic, transitional, se-cretory, psammomatous, and angiomatous variants corresponding to WHO grade 1. Common histologic findings are whorl and psammoma body formation. Higher-grade meningiomas (grades 2 or 3) can also occur in the spine, although incidence is much lower.

On MRI, meningiomas appear as solid, well-circumscribed lesions with broad attachment to the dura.95 The tumor is iso- or hypointense on T1-weighted and slightly hyperintense on T2-weighted images. Contrast enhancement usually is intense and homogenous. The dural “tail” sign can be pres-ent and highly suggestive of meningioma (Figure 1D).105 Among all spinal cord meningiomas, 94% are intradural and 6% are extradural.

Indications for treatment should be individual-ized, including age, presence of neurologic deficit, evidence of growth, and general medical condition. Observation might be appropriate in older and as-ymptomatic patients. If treatment is indicated, mi-crosurgery is the primary modality and can be cu-rative (with 5- and 10-year recurrence rates of 3%

Original Article

Aghayev et al.


444

and 6%, respectively).106–108 Even older patients with paraplegia or severe paraparesis can benefit from ag-gressive surgical resection.109 Significant preopera-tive neurologic deficit and pia/arachnoid invasion by the tumor are independent predictors of poor out-come.110 Sex, pregnancy, histologic type, and recur-rence do not influence outcome.111 Residual tumor should be followed up,111 because subtotally resected meningiomas have a recurrence rate of less than 15%.108 Repeat excision should be considered for re-currence in patients with incomplete resection.

The role of radiotherapy is controversial. It is in-dicated for malignant WHO grade 3 meningiomas (i.e., anaplastic meningiomas). For WHO grade 2 (atypical meningioma) and 1 tumors, radiotherapy can be considered for rare recurrent cases that can-not be resected completely. No established chemo-therapy regimen exists for meningiomas; however, small studies suggest a palliative benefit from treating recurrent intracranial meningioma with somatosta-tin analogues, hydroxyurea, interferon-α, PTK787, or sunitinib.112–123 However, no data exist on use of these agents for spinal meningiomas.

Follow-up usually requires serial postoperative MRI and neurologic monitoring. Patients with no evidence of recurrence can be followed up with neuro-logic observation after a few years of MRI monitoring.

Extradural Extramedullary TumorsMost extradural extramedullary tumors in adults are metastases, most often from lung, breast, or prostate cancers, or lymphoma, and may result in epidural spi-nal cord compression. Primary tumors of the verte-bral column can be separated into benign and malig-nant types. Benign tumors include osteoid osteoma, osteoblastoma, osteochondroma, hemangioma, aneu-rysmal bone cyst, giant cell tumor, and eosinophilic granuloma. Malignant varieties include chordoma, multiple myeloma, osteosarcoma, chondrosarcoma, Ewing’s sarcoma, lymphoma, soft tissue sarcomas, and plasmacytoma. A complete discussion of these tumor types is outside the scope of this article.

ConclusionsEarly recognition of the signs and symptoms of pri-mary spinal cord tumors facilitates early treatment, potentially minimizes neurologic morbidity, and

improves outcome. In most series, pain is the pre-dominant symptom and often persists after surgery. Primary treatment is surgery in most spinal cord tumors, and predictors of outcome include preop-erative functional status, histologic tumor grade, and extent of surgical resection. Postoperative care after resective spinal cord surgery often mandates a lengthy rehabilitation with delayed return of neuro-logic function, often requiring a year or more. Symp-tom management for spinal cord tumors can be chal-lenging, particularly for intramedullary tumors, with quality of life issues involving motor, sensory, and autonomic dysfunction.

References 1. Parsa AT, McCormick PC. Intramedullary spinal tumors. Phila-

delphia, PA: W.B. Saunders Co.; 2006. 2. Engelhard HH, Villano JL, Porter KR, et al. Clinical presentation,

histology, and treatment in 430 patients with primary tumors of the spinal cord, spinal meninges, or cauda equina. J Neurosurg Spine 2010;13:67–77.

3. Raco A, Esposito V, Lenzi J, et al. Long-term follow-up of intra-medullary spinal cord tumors: a series of 202 cases. Neurosurgery 2005;56:972–981.

4. Berger MS, Prados M. Textbook of neuro-oncology, 1st ed. Phila-delphia, PA: Elsevier Saunders; 2005.

5. Dickman CA, Fehlings M, Gokaslan ZL. Spinal cord and spinal column tumors: principles and practice. New York, NY: Thieme; 2006.

6. Baumer D, Flossmann E, Cudlip S, et al. Primary angiitis of the CNS mimicking a spinal cord tumour. J Neurol 2008;255:1970–1972.

7. Setzer M, Murtagh RD, Murtagh FR, et al. Diffusion tensor imag-ing tractography in patients with intramedullary tumors: compari-son with intraoperative findings and value for prediction of tumor resectability. J Neurosurg Spine 2010;13:371–380.

8. Eroes CA, Zausinger S, Kreth FW, et al. Intramedullary low grade astrocytoma and ependymoma. Surgical results and predicting fac-tors for clinical outcome. Acta Neurochir (Wien) 2010;152:611–618.

9. Taricco MA, Guirado VM, Fontes RB, et al. Surgical treatment of primary intramedullary spinal cord tumors in adult patients. Arq Neuropsiquiatr 2008;66:59–63.

10. Yasargil MG, Tranmer BI, Adamson TE, et al. Unilateral partial hemi-laminectomy for the removal of extra- and intramedullary tumours and AVMs. Adv Tech Stand Neurosurg 1991;18:113–132.

11. Gerszten PC, Burton SA, Ozhasoglu C, et al. Radiosurgery for benign intradural spinal tumors. Neurosurgery 2008;62:887–895; discussion 895–896.

12. Amirjamshidi A, Abbassioun K. Radiation-induced tumors of the central nervous system occurring in childhood and adolescence. Four unusual lesions in three patients and a review of the litera-ture. Childs Nerv Syst 2000;16:390–397.

13. Evans DG, Birch JM, Ramsden RT, et al. Malignant transforma-tion and new primary tumours after therapeutic radiation for be-

Original Article



445

nign disease: substantial risks in certain tumour prone syndromes. J Med Genet 2006;43:289–294.

14. Campello C, Le Floch A, Parker F, et al. Neuroepithelial intra-medullary spinal cord tumors in adults: study of 70 cases [abstract]. Presented at American Academy of Neurology Annual Meeting; April 25–May 2, 2009; Seattle, Washington.

15. Helseth A, Mork SJ. Primary intraspinal neoplasms in Norway, 1955 to 1986. A population-based survey of 467 patients. J Neu-rosurg 1989;71:842–845.

16. Heuer GG, Stiefel MF, Bailey RL, et al. Acute paraparesis from hemorrhagic spinal ependymoma: diagnostic dilemma and surgi-cal management. J Neurosurg Spine 2007;7:652–655.

17. Godfraind C. Classification and controversies in pathology of ep-endymomas. Childs Nerv Syst 2009;25:1185–1193.

18. Parsa AT, Chi JH, Acosta FL Jr, et al. Intramedullary spinal cord tumors: molecular insights and surgical innovation. Clin Neuro-Clin Neuro-surg 2005;52:76–84.

19. Peker S, Ozgen S, Ozek MM, et al. Surgical treatment of intra-medullary spinal cord ependymomas: can outcome be predicted by tumor parameters? J Spinal Disord Tech 2004;17:516–521.

20. Maesawa S, Takayasu M, Kajita Y, et al. Effects of opening peritu-moral syrinxes on motor evoked potentials during resection of an intramedullary spinal cord tumor—case report. Neurol Med Chir (Tokyo) 2004;44:258–262.

21. Alkhani A, Blooshi M, Hassounah M. Outcome of surgery for in-tramedullary spinal ependymoma. Ann Saudi Med 2008;28:109–113.

22. Tobias ME, McGirt MJ, Chaichana KL, et al. Surgical manage-ment of long intramedullary spinal cord tumors. Childs Nerv Syst 2008;24:219–223.

23. Hanbali F, Fourney DR, Marmor E, et al. Spinal cord ependy-moma: radical surgical resection and outcome. Neurosurgery 2002;51:1162–1172; discussion 1172–1174.

24. McCormick PC, Torres R, Post KD, et al. Intramedullary ependy-moma of the spinal cord. J Neurosurg 1990;72:523–532.

25. Volpp PB, Han K, Kagan AR, et al. Outcomes in treatment for intradural spinal cord ependymomas. Int J Radiat Oncol Biol Phys 2007;69:1199–1204.

26. Cooper PR, Epstein F. Radical resection of intramedullary spinal cord tumors in adults. Recent experience in 29 patients. J Neuro-surg 1985;63:492–499.

27. Abdel-Wahab M, Etuk B, Palermo J, et al. Spinal cord gliomas: a multi-institutional retrospective analysis. Int J Radiat Oncol Biol Phys 2006;64:1060–1071.

28. Chamberlain MC. Etoposide for recurrent spinal cord ependy-moma. Neurology 2002;58:1310–1311.

29. Chamberlain MC, Johnston SK. Temozolomide for recurrent in-tracranial supratentorial platinum-refractory ependymoma. Can-cer 2009;115:4775–4782.

30. Gomez DR, Missett BT, Wara WM, et al. High failure rate in spinal ependymomas with long-term follow-up. Neuro Oncol 2005;7:254–259.

31. Isaacson SR. Radiation therapy and the management of intramed-ullary spinal cord tumors. J Neurooncol 2000;47:231–238.

32. Linstadt DE, Wara WM, Leibel SA, et al. Postoperative radio-therapy of primary spinal cord tumors. Int J Radiat Oncol Biol Phys 1989;16:1397–1403.

33. Waldron JN, Laperriere NJ, Jaakkimainen L, et al. Spinal cord ependymomas: a retrospective analysis of 59 cases. Int J Radiat Oncol Biol Phys 1993;27:223–229.

34. Cervoni L, Celli P, Fortuna A, et al. Recurrence of spinal ependy-moma. Risk factors and long-term survival. Spine (Phila Pa 1976) 1994;19:2838–2841.

35. Sgouros S, Malluci CL, Jackowski A. Spinal ependymomas—the value of postoperative radiotherapy for residual disease control. Br J Neurosurg 1996;10:559–566.

36. Gilbert MR, Ruda R, Soffietti R. Ependymomas in adults. Curr Neurol Neurosci Rep 2010;10:240–247.

37. Kim MS, Chung CK, Choe G, et al. Intramedullary spinal cord astrocytoma in adults: postoperative outcome. J Neurooncol 2001;52:85–94.

38. Santi M, Mena H, Wong K, et al. Spinal cord malignant astrocyto-mas. Clinicopathologic features in 36 cases. Cancer 2003;98:554–561.

39. Yagi T, Ohata K, Haque M, et al. Intramedullary spinal cord tu-mour associated with neurofibromatosis type 1. Acta Neurochir (Wien) 1997;139:1055–1060.

40. Lee M, Rezai AR, Freed D, et al. Intramedullary spinal cord tu-mors in neurofibromatosis. Neurosurgery 1996;38:32–37.

41. Innocenzi G, Salvati M, Cervoni L, et al. Prognostic factors in in-tramedullary astrocytomas. Clin Neurol Neurosurg 1997;99:1–5.

42. Roonprapunt C, Houten JK. Spinal cord astrocytomas: pre-sentation, management, and outcome. Neurosurg Clin N Am 2006;17:29–36.

43. Kaley T, Abrey L. Temozolomide for spinal cord high-grade glio-mas [abstract]. Presented at American Academy of Neurology Annual Meeting; April 25–May 2, 2009; Seattle, Washington.

44. Chamberlain MC. Temozolomide for recurrent low-grade spinal cord gliomas in adults. Cancer 2008;113:1019–1024.

45. Chamberlain MC, Johnston SK. Recurrent spinal cord glioblas-toma: salvage therapy with bevacizumab. J Neurooncol, in press.

46. Allen JC, Aviner S, Yates AJ, et al. Treatment of high-grade spi-nal cord astrocytoma of childhood with “8-in-1” chemotherapy and radiotherapy: a pilot study of CCG-945. Children‘s Cancer Group. J Neurosurg 1998;88:215–220.

47. Lefkowitz IB, Packer RJ, Sutton LN, et al. Results of the treatment of children with recurrent gliomas with lomustine and vincristine. Cancer 1988;61:896–902.

48. Lowis SP, Pizer BL, Coakham H, et al. Chemotherapy for spinal cord astrocytoma: can natural history be modified? Childs Nerv Syst 1998;14:317–321.

49. Packer RJ, Lange B, Ater J, et al. Carboplatin and vincristine for recurrent and newly diagnosed low-grade gliomas of childhood. J Clin Oncol 1993;11:850–856.

50. Rodrigues GB, Waldron JN, Wong CS, et al. A retrospective analysis of 52 cases of spinal cord glioma managed with radiation therapy. Int J Radiat Oncol Biol Phys 2000;48:837–842.

51. Minehan KJ, Brown PD, Scheithauer BW, et al. Prognosis and treatment of spinal cord astrocytoma. Int J Radiat Oncol Biol Phys 2009;73:727–733.

52. Lee DK, Choe WJ, Chung CK, et al. Spinal cord hemangio-blastoma: surgical strategy and clinical outcome. J Neurooncol 2003;61:27–34.

53. Murota T, Symon L. Surgical management of hemangioblastoma of the spinal cord: a report of 18 cases. Neurosurgery 1989;25:699–707; discussion 708.

54. Hottinger AF, Khakoo Y. Neurooncology of familial cancer syn-dromes. J Child Neurol 2009;24:1526–1535.

55. Miller DJ, McCutcheon IE. Hemangioblastomas and other un-common intramedullary tumors. J Neurooncol 2000;47:253–270.

Original Article

Aghayev et al.


446

56. Lonser RR, Oldfield EH. Spinal cord hemangioblastomas. Neuro-surg Clin N Am 2006;17:37–44.

57. Irie K, Kuyama H, Nagao S. Spinal cord hemangioblastoma pre-senting with subarachnoid hemorrhage. Neurol Med Chir (To-kyo) 1998;38:355–358.

58. Cerejo A, Vaz R, Feyo PB, et al. Spinal cord hemangioblastoma with subarachnoid hemorrhage. Neurosurgery 1990;27:991–993.

59. Kormos RL, Tucker WS, Bilbao JM, et al. Subarachnoid hemor-rhage due to a spinal cord hemangioblastoma: case report. Neuro-surgery 1980;6:657–660.

60. Dijindjian M, Djindjian R, Houdart R, et al. Subarachnoid hem-orrhage due to intraspinal tumors. Surg Neurol 1978;9:223–229.

61. Minami M, Hanakita J, Suwa H, et al. Cervical hemangioblasto-ma with a past history of subarachnoid hemorrhage. Surg Neurol 1998;49:278–281.

62. Yu JS, Short MP, Schumacher J, et al. Intramedullary hemorrhage in spinal cord hemangioblastoma. Report of two cases. J Neuro-surg 1994;81:937–940.

63. Eskridge JM, McAuliffe W, Harris B, et al. Preoperative endo-vascular embolization of craniospinal hemangioblastomas. AJNR Am J Neuroradiol 1996;17:525–531.

64. Standard SC, Ahuja A, Livingston K, et al. Endovascular embo-lization and surgical excision for the treatment of cerebellar and brain stem hemangioblastomas. Surg Neurol 1994;41:405–410.

65. Tampieri D, Leblanc R, TerBrugge K. Preoperative embolization of brain and spinal hemangioblastomas. Neurosurgery 1993;33:502–505; discussion 505.

66. Vazquez-Anon V, Botella C, Beltran A, et al. Preoperative em-bolization of solid cervicomedullary junction hemangioblastomas: report of two cases. Neuroradiology 1997;39:86–89.

67. Wanebo JE, Lonser RR, Glenn GM, et al. The natural history of hemangioblastomas of the central nervous system in patients with von Hippel-Lindau disease. J Neurosurg 2003;98:82–94.

68. Ryu SI, Kim DH, Chang SD. Stereotactic radiosurgery for hem-angiomas and ependymomas of the spinal cord. Neurosurg Focus 2003;15:E10.

69. Schuch G, de Wit M, Holtje J, et al. Case 2. Hemangioblastomas: diagnosis of von Hippel-Lindau disease and antiangiogenic treat-ment with SU5416. J Clin Oncol 2005;23:3624–3626.

70. Lu DC, Lawton MT. Clinical presentation and surgical manage-ment of intramedullary spinal cord cavernous malformations. Neurosurgical Focus 2010;29:E12.

71. Kuhn J, Brummendorf TH, Brassat U, et al. Novel KRIT1 mu-tation and no molecular evidence of anticipation in a family with cerebral and spinal cavernous malformations. Eur Neurol 2009;61:154–158.

72. Lee YW, Lee ST, Cha JG, et al. A novel KRIT1 gene mutation in a patient with cerebral and multiple spinal cavernous malforma-tions. Ann Clin Lab Sci 2010;40:290–294.

73. Labauge P, Bouly S, Parker F, et al. Outcome in 53 patients with spinal cord cavernomas. Surg Neurol 2008;70:176–181.

74. Zevgaridis D, Medele RJ, Hamburger C, et al. Cavernous haeman-giomas of the spinal cord. A review of 117 cases. Acta Neurochir (Wien) 1999;141:237–245.

75. Bian LG, Bertalanffy H, Sun QF, et al. Intramedullary cavernous malformations: clinical features and surgical technique via hemi-laminectomy. Clin Neurol Neurosurg 2009;111:511–517.

76. Chi JH, Parsa AT. Intramedullary spinal cord metastasis: clinical management and surgical considerations. Neurosurg Clin N Am 2006;17:45–50.

77. Gasser T, Sandalcioglu IE, El Hamalawi B, et al. Surgical treat-Surgical treat-ment of intramedullary spinal cord metastases of systemic cancer: functional outcome and prognosis. J Neurooncol 2005;73:163–168.

78. Kalayci M, Cagavi F, Gul S, et al. Intramedullary spinal cord me-tastases: diagnosis and treatment—an illustrated review. Acta Neurochir (Wien) 2004;146:1347–1354; discussion 1354.

79. Sandalcioglu IE, Gasser T, Asgari S, et al. Functional outcome after surgical treatment of intramedullary spinal cord tumors: ex-perience with 78 patients. Spinal Cord 2005;43:34–41.

80. Aryan HE, Farin A, Nakaji P, et al. Intramedullary spinal cord metastasis of lung adenocarcinoma presenting as Brown-Sequard syndrome. Surg Neurol 2004;61:72–76.

81. Xu Q, Bao W, Mao R. Microsurgery of intramedullary cervical cord tumor. Chin Med J (Engl) 1996;109:756–761.

82. Sharma MC, Jain D, Sarkar C, et al. Spinal teratomas: a clin-ico-pathological study of 27 patients. Acta Neurochir (Wien) 2009;151:245–252; discussion 252.

83. Jallo GI, Freed D, Epstein FJ. Spinal cord gangliogliomas: a review of 56 patients. J Neurooncol 2004;68:71–77.

84. Biswas A, Puri T, Goyal S, et al. Spinal intradural primary germ cell tumour--review of literature and case report. Acta Neurochir (Wien) 2009;151:277–284.

85. Nakamura M, Ishii K, Watanabe K, et al. Long-term surgical out-comes for myxopapillary ependymomas of the cauda equina. Spi-Spi-ne (Phila Pa 1976) 2009;34:E756–760.

86. Meneses MS, Leal AG, Periotto LB, et al. Primary fi lum ter-Primary filum ter-minale ependymoma: a series of 16 cases. Arq Neuropsiquiatr 2008;66:529–533.

87. Lin YH, Huang CI, Wong TT, et al. Treatment of spinal cord ep-endymomas by surgery with or without postoperative radiothera-py. J Neurooncol 2005;71:205–210.

88. Akyurek S, Chang EL, Yu TK, et al. Spinal myxopapillary ependy-moma outcomes in patients treated with surgery and radiotherapy at M.D. Anderson Cancer Center. J Neurooncol 2006;80:177–183.

89. Jinnai T, Koyama T. Clinical characteristics of spinal nerve sheath tumors: analysis of 149 cases. Neurosurgery 2005;56:510–515.

90. Russell DS, Rubinstein LJ. Pathology of tumours of the nervous system, 5th ed. London, United Kingdom: Edward Arnold; 1989.

91. Conti P, Pansini G, Mouchaty H, et al. Spinal neurinomas: retro-spective analysis and long-term outcome of 179 consecutively op-erated cases and review of the literature. Surg Neurol 2004;61:34–43; discussion 44.

92. Houshmandi SS, Emnett RJ, Giovannini M, et al. The neuro-fibromatosis 2 protein, merlin, regulates glial cell growth in an ErbB2- and Src-dependent manner. Mol Cell Biol 2009;29:1472–1486.

93. Hanemann CO. Magic but treatable? Tumours due to loss of mer-lin. Brain 2008;131:606–615.

94. Carney JA. Psammomatous melanotic schwannoma. A distinc-tive, heritable tumor with special associations, including car-diac myxoma and the Cushing syndrome. Am J Surg Pathol 1990;14:206–222.

95. Abul-Kasim K, Thurnher MM, McKeever P, et al. Intradural spi-nal tumors: current classification and MRI features. Neuroradiol-ogy 2008;50:301–314.

96. Friedman DP, Tartaglino LM, Flanders AE. Intradural schwanno-mas of the spine: MR findings with emphasis on contrast-enhance-ment characteristics. AJR Am J Roentgenol 1992;158:1347–1350.

Original Article



447

97. Parmar HA, Ibrahim M, Castillo M, et al. Pictorial essay: diverse imaging features of spinal schwannomas. J Comput Assist Tomogr 2007;31:329–334.

98. Singrakhia MD, Parmar H, Maheshwari M, et al. Cervical schwan-noma presenting as an expansile vertebral body lesion: report of two cases with a technical note on the surgical management. Surg Neurol 2006;66:192–196; discussion 196.

99. Dodd RL, Ryu MR, Kamnerdsupaphon P, et al. CyberKnife radio-surgery for benign intradural extramedullary spinal tumors. Neu-rosurgery 2006;58:674–685.

100. Parsa AT, Lee J, Parney IF, et al. Spinal cord and intradural-extra-parenchymal spinal tumors: current best care practices and strate-gies. J Neurooncol 2004;69:291–318.

101. Gutmann DH. The neurofibromatoses: when less is more. Hum Mol Genet 2001;10:747–755.

102. Neurofibromatosis. Conference statement. National Institutes of Health Consensus Development Conference. Arch Neurol 1988;45:575–578.

103. Packer RJ, Gutmann DH, Rubenstein A, et al. Plexiform neu-Plexiform neu-rofibromas in NF1: toward biologic-based therapy. Neurology 2002;58:1461–1470.

104. Perry A, Giannini C, Raghavan R, et al. Aggressive phenotypic and genotypic features in pediatric and NF2-associated menin-giomas: a clinicopathologic study of 53 cases. J Neuropathol Exp Neurol 2001;60:994–1003.

105. Quekel LG, Versteege CW. The “dural tail sign” in MRI of spinal meningiomas. J Comput Assist Tomogr 1995;19:890–892.

106. Gezen F, Kahraman S, Canakci Z, et al. Review of 36 cases of spinal cord meningioma. Spine (Phila Pa 1976) 2000;25:727–731.

107. Peker S, Cerci A, Ozgen S, et al. Spinal meningiomas: evaluation of 41 patients. J Neurosurg Sci 2005;49:7–11.

108. Solero CL, Fornari M, Giombini S, et al. Spinal meningiomas: review of 174 operated cases. Neurosurgery 1989;25:153–160.

109. Sacko O, Haegelen C, Mendes V, et al. Spinal meningioma sur-gery in elderly patients with paraplegia or severe paraparesis: a multicenter study. Neurosurgery 2009;64:503–509; discussion 509–510.

110. Setzer M, Vatter H, Marquardt G, et al. Management of spinal meningiomas: surgical results and a review of the literature. Neu-

rosurg Focus 2007;23:E14. 111. Roux FX, Nataf F, Pinaudeau M, et al. Intraspinal meningiomas:

review of 54 cases with discussion of poor prognosis factors and modern therapeutic management. Surg Neurol 1996;46:458–463; discussion 463–464.

112. Chamberlain MC, Glantz MJ, Fadul CE. Recurrent meningioma: salvage therapy with long-acting somatostatin analogue. Neurol-ogy 2007;69:969–973.

113. Kaley T, Wen P, Karimi S, et al. Phase II trial of sunitinib (SU011248) in patients with recurrent or inoperable meningio-ma [abstract]. Presented at Society for Neuro-Oncology Annual Meeting; November 20–23, 2008; Las Vegas, Nevada.

114. DeBoer R, Grimm SA, Chandler J, et al. A phase II trial of PTK-787 (PTK/ZK) in recurrent or progressive meningiomas [abstract]. J Clin Oncol 2008;26(suppl 1):Abstract 2060.

115. Chamberlain MC, Blumenthal DT. Intracranial meningiomas: diagnosis and treatment. Expert Rev Neurother 2004;4:641–648.

116. Chamberlain MC, Glantz MJ. Interferon-alpha for recurrent World Health Organization grade 1 intracranial meningiomas. Cancer 2008;113:2146–2151.

117. Chamberlain MC, Tsao-Wei DD, Groshen S. Salvage chemo-therapy with CPT-11 for recurrent meningioma. J Neurooncol 2006;78:271–276.

118. Goodwin JW, Crowley J, Eyre HJ, et al. A phase II evaluation of tamoxifen in unresectable or refractory meningiomas: a South-west Oncology Group study. J Neurooncol 1993;15:75–77.

119. Grunberg SM, Weiss MH, Spitz IM, et al. Treatment of unresect-Treatment of unresect-able meningiomas with the antiprogesterone agent mifepristone. J Neurosurg 1991;74:861–866.

120. Grunberg SM. Role of antiprogestational therapy for meningio-mas. Hum Reprod 1994;9(Suppl 1):202–207.

121. Mason WP, Gentili F, Macdonald DR, et al. Stabilization of dis-ease progression by hydroxyurea in patients with recurrent or un-resectable meningioma. J Neurosurg 2002;97:341–346.

122. Kyritsis AP. Chemotherapy for meningiomas. J Neurooncol 1996;29:269–272.

123. Sioka C, Kyritsis AP. Chemotherapy, hormonal therapy, and immunotherapy for recurrent meningiomas. J Neurooncol 2009;92:1–6.

adult intradural primary spinal cord tumors · and are considered part of the differential...

Documents