Neurol India Home 

Year : 1999  |  Volume : 47  |  Issue : 3  |  Page : 224--8

Management of intramedullary spinal cord tumours : review of 68 patients.

MJ Chandy, S Babu 
 Department of Neurological Sciences, Christian Medical College and Hospital, Vellore, 632004, India., India

Correspondence Address:
M J Chandy
Department of Neurological Sciences, Christian Medical College and Hospital, Vellore, 632004, India.


68 consecutive patients admitted with intramedullary spinal cord tumours and operated at Vellore during a six year period from January 1990 are discussed. 41 tumours were radically resected, 11 partially excised while 14 had only a biopsy. Radiation therapy was advised post operatively to those patients for whom a partial excision or biopsy was done. There was no postoperative mortality. Two patients developed wound infection and one developed postoperative hydrocephalus. Postoperative clinical assessment between four to eight weeks after surgery showed that 25 out of 68 patients improved, 29 remained unchanged, while 14 had worsening of deficits. Immediate post operative assessment, however, was less encouraging. Evaluation of these patients was done using a functional scoring system and Karnofsky rating. The follow up period ranged from 2 weeks to 64 months after discharge from hospital with a mean of 14.6 months. The indicators of radical excision were good tumour-cord interface, cranially located tumours, presence of syringomyelia and histology of ependymoma. Two patients had recurrence of tumour.

How to cite this article:
Chandy M J, Babu S. Management of intramedullary spinal cord tumours : review of 68 patients. Neurol India 1999;47:224-8

How to cite this URL:
Chandy M J, Babu S. Management of intramedullary spinal cord tumours : review of 68 patients. Neurol India [serial online] 1999 [cited 2021 Dec 3 ];47:224-8
Available from:

Full Text

  ::   IntroductionTop

Inspite of modern micro surgical technical advances, the surgery of intramedullary tumours remains challenging to every neurosurgeon. The contemporary approach to these lesions is one of aggressive surgery.[1],[2],[3] Total or radical resection with good functional outcome and acceptable morbidity has become the present day gold standard for ependymomas, haemangioblastomas and other benign lesions. Low grade infiltrating astrocytomas and malignant gliomas have been found to have a guarded prognosis inspite of radical excision. This report is a summary of our 6 year experience in the management of a spectrum of intramedullary mass lesions excluding vascular malformations. This report also includes the indicators of radical excision, retrospective analysis of the clinical presentation, imaging, surgical strategy, histology, postoperative radiation and the immediate and long term outcome of these patients.

  ::   Material and methodsTop

Patient population : A study of a consecutive cases of 68 patients with intramedullary mass lesions operated during a six year period from January 1990 at the Neurological surgery department of the Christian Medical College and Hospital, Vellore, was undertaken. There were 47 males and 21 females and their ages ranged from 1 year to 58 years. Eight of these patients had a biopsy elsewhere and were then referred for definitive surgery. 36 patients were followed up in our hospital at the outpatient clinic or as an inpatient. The follow-up period ranged from 2 weeks to 64 months with a mean of 14.6 months.

Clinical Assessment : The clinical features evaluated were broadly categorized into 4 groups. They were in relation to the sensory and motor systems, and gait, bladder and bowel disturbances. Pre and post operatively all patients were evaluated using the clinical scoring system developed by Klekamp and Sami.[4] This formed the basis for comparison of the pre and post operative clinical status. The functional status on admission and at follow-up was determined using the Karnofsky scale.

Diagnostic imaging : Patients with intramedullary tumours seen during the early part of the series were investigated with computed tomographic (CT) myelography and later with magnetic resonance(MR) imaging. 18 patients had CT myelography while 50 patients had MR imaging.

Surgery : Although the surgical goal was radical excision in all cases, this was not always possible. The extent of resection was therefore considered under the following groups : radical excision, partial excision and biopsy with decompression. This was based on the surgeon's operative assessment and postoperative contrast MR in the later half of the study. Radical excision was defined as excision of the entire lesion as seen through the surgical microscope, although on occasion a few millimeters of subpial tumour on the anterior median sulcus was left behind. Partial excision referred to cases where fifty percent or more of the tumour was excised. A standard bilateral laminectomy was done, one level above and below the lesion. This was followed by dural and arachnoidal opening. Following a dorsal myelotomy and subpial retraction sutures, surgical excision was started by first defining the tumour-cord plane and then continued till radical excision was achieved, usually following an internal decompression of the tumour. The exception to this strategy being haemangioblastomas, where internal decompression is contraindicated. If an infiltrating lesion, with a poor tumour cord plane was encountered, then either a biopsy or partial excision was done. Surgical strategy was not determined by the smear or frozen section histological report obtained during surgery but rather by the tumour-cord plane.

Radiotherapy : Radiation therapy was advised depending on the extent of the surgical excision and the nature of the lesion. Most patients who had radical excision of astrocytomas and had good tumour interface were not advised radiation. A total of 30 out of the 68 patients underwent radiation therapy.

Follow Up : All patients were assessed in the out patient clinic post-operatively at the end of periods ranging from one week to 6 weeks following radiation therapy. Long term follow up was obtained by clinical and functional evaluation in the hospital or by correspondence with the patient in a standard format.

  ::   ResultsTop

The age of patients in this series ranged from 1 year to 58 years with a mean of 31.28 years. There were 47 male and 21 female patients. The duration of symptoms and or signs ranged from 3 weeks to 15 years. It may be noted that there were 11 patients in the 1-16 year age group. The clinical picture although suggestive of myelopathy due to a mass was not always typical of an intramedullary lesion. 95.8% of the patients had motor weakness and or spasticity, 90.5% had sensory involvement, of which interestingly, only 82.5% had features of dissociated sensory loss. 86.2% of the patients had gait disturbance and only 61.7% of patients had bladder or bowel involvement.

The preoperative clinical status was assessed based on the clinical scoring for each of the chief symptoms on a scale of 0 to 5. The neurological status was divided into 3 groups based on the scores, viz. satisfactory (score 4-5), unsatisfactory (score 2-3), poor (score 0-1). Similarly, the functional status was grouped into 3 based on Karnofsky scoring on admission, viz. normal activity (100-80%), disabled (70-40%) and sick (30-0%).

Syringomyelia was diagnosed in only 2 out of 18 cases who had myelo CT against 19 out of 50 patients who had MR imaging, highlighting the importance of MRI in diagnosing syrinx. On analysing the MR images it was seen that associated syringomyelia was seen equally in astrocytomas and ependymomas. Astrocytomas were predominantly hypointense on TlW images and hyper intense on T2W images. Contrast enhancement was more marked in malignant gliomas. Ependymomas had similar signal characteristics but they were usually well defined, compact and enhanced uniformly and well with contrast. Benign lesions such as haemangioblastomas and schwannomas were well defined and seen separately from the surrounding cord. Haemangioblastomas had marked contrast uptake as well. Metastatic tumours were seen to have significant cord oedema. Tuberculomas also had associated oedema. Benign lesions in this series were confined to three segments of the cord while astrocytomas and ependymomas on the other hand were usually between three to six segments in length. Metastatic tumours were the exception and were confined to two segments. The site or location of the tumour is given in [Table I]. There were more tumours in the cervical cord followed by the cervico-thoracic and thoracic cord.

Surgery : Out of the 68 patients reported here, radical excision of the tumour was done in 41, partial excision in 11 and biopsy in 16 cases. [Table I]shows the relationship between the location of the tumour and the extent of excision. It may be noted that radical excision was accomplished in 50% of cases arising from the cervical and cervico-thoracic junction, whereas it was less in other areas. 14 of the 16 patients, where only a biopsy was possible, had astrocytoma. All the 16 patients had poor tumour cord plane impeding radical excision. Interestingly only one patient in this group had a high grade astrocytoma. [Table II] shows the relationship between the extent of tumour resection and the histological diagnosis. In the present series 75% of ependymomas were radically excised whereas only 25% of astrocytomas were amenable to radical excision.

Syringomyelia was encountered in 21 cases (28%). 7 patients had syringomyelia cranial to the tumour, 3 had caudal and 10 had cranial and caudal syringomyelia. In one patient there was cavity within the tumour. Syringomyelia was correlated with the tumour histology and its occurrence was found to be high in glial tumours. Contrary to reports in the literature syringomyelia was not found to be specific for ependymomas in this series. Syringomyelia was associated with ependymomas (8 cases) astrocytomas, (8 cases) haemangioblastomas (3 cases), schwannoma and ganglioglioma (1 case each). The presence of syringomyelia seemed to indicate the possibility of a good tumour cord interface. Consequently, radical excision was carried out in 17 of the 21 patients with syringomyelia. Arachnoid adhesions were seen at surgery in 11 patients. Six of these had been operated earlier. Radical excision was possible in only five of these eleven patients. Re-operation is a poor indicator of radical excision. The plane of cleavage was good in 36 patients, unsatisfactory in 30 while in 2 patients it was good in most parts except anteriorly in the midline. The plane of cleavage was the most important indicator of radical excision [Table III].

It is well known that surgery should be done before the onset of significant disabling neurological deficits. In our series, patients presenting with minimal symptoms and signs and who were operated early showed improvement or had no deterioration in their neurological status while those with major symptoms and signs and operated late in the course of the disease deteriorated post operatively. Two patients had wound infection. One of these had wound dehiscence and needed resuturing and daily lumbar punctures. One patient had post operative hydrocephalus needing shunt surgery. One patient had CSF collection around the posterior aspect of the cord which needed a cysto-peritoneal shunt.

More than 50% of the intramedullary tumours were glial in origin [Table II]. Vascular malformations were excluded from this study. In the paediatric age group, astrocytomas were the commonest tumours. In the adult group, astrocytomas while being commoner in the second and third decades were as common as ependymomas in the third and fourth decades.

Radiation therapy was advised depending on the extent of the surgical excision. A total of 30 out of the 68 patients were advised radiation although 12 of these had had radical excision, 4 had partial excision and 14 had only a biopsy. Among the radiated group 21 of the 30 patients had astrocytomas, 8 had ependymomas and 1 patient had a metastatic lesion.

Postoperatively 17 patients showed improvement, 25 had no change from the preoperative status, and 26 had immediate worsening of motor, sensory or bladder status. 5 patients had additional deficits in the immediate postoperative period. Of the 26 patients who worsened, 22 had worsening of the motor status (one MRC grade). Additional sensory deficit (mainly posterior column loss) was seen in 20 cases. 12 patients had worsening of bladder status. Of the 5 patients who developed additional deficit, 4 had both motor and sensory deficit and 2 had bladder dysfunction. Those patients who showed improvement post operatively did so during their stay in the ward during a period ranging from 1 week to 4 weeks. 17 patients showed improvement in their neurological status during their hospital stay. This was with particular reference to motor and autonomic function. Long term follow up was available in 36 patients [Table IV]. The follow up period ranged from 2 weeks to 5 years and 4 months, with a mean of 14.6 months. 14 patients had further improvement, 14 remained same and 8 had progression of the disease. Follow up until the present time showed that 2 patients had recurrence of tumour. One patient had had radical excision of a grade II astrocytoma which recurred 2 years after the first surgery. This patient underwent re-operation and the tumour was partially excised since there were adhesions present with poor tumour cord interface. The other patient who had an ependymoma, had recurrence after 18 months. He was advised radiotherapy as he was unwilling for a second operation. A longer period of follow up in this series would give a truer picture of recurrence and outcome.

  ::   DiscussionTop

Morbidity and mortality following surgery of intramedullary tumours made many surgeons wary of the aggressive approach. For several decades, radiation therapy with or without surgical intervention was the mainstay of the management of these lesions. Surgery was employed primarily for biopsy, aspiration of cyst or decompressive laminectomy. Greenwood in 1954 pointed out that those patients with the least preoperative disability and without prior surgery benefited the most from aggressive surgery. These observations set the pace for future advances in the microsurgical excision of these tumours. Astrocytomas and ependymomas have been reported to be the commonest tumors even though a wide variety of pathological processes which either arise primarily from the spinal cord or secondarily involve the cord have been reported. These include lipomas, neurofibromas, epidermoids, haemangioblastoma, granulomas, and metastatic tumors.[5],[6],[7],[8] Data regarding the age of presentation, duration of symptoms, preoperative clinical and functional status and location of the lesion seen in this series is comparable to those reported in the literature.[3],[6],[7],[8],[9] Evaluation of the pre and postoperative clinical and functional status in the present series have been described. Interestingly, this does not differ significantly from the earlier series.[3],[7],[12] The present report reiterates the fact that patients with minimal symptoms and signs operated early did better than those with major deficits and operated late in the course of the disease. This finding is in keeping with earlier workers who stress that early intervention in well preserved patients increases the possibility of a good outcome.[1],[7],[8],[12],[13],[14] The present day goal of most surgeons is aimed at radical surgery within the bounds of safety. Some earlier authors were of the opinion that there was a definite role for radiotherapy in the management of intramedullary glial tumours.[15],[16] Kopelson et al were of the opinion that only the minimum surgical procedure necessary to establish a diagnosis should be undertaken and that radiation therapy was the most effective treatment modality.[15] At the present time most authors feel that total or radical excision should be the goal for all intramedullary lesions particularly in the management of ependymomas.[1],[2],[3],[6],[7],[8],[9],[10],[12],[13] Radical excision is usually not necessary or recommended in the case of tuberculomas. This is because of the absence of a good plane, presence of cord oedema and the satisfactory outcome following appropriate antibiotics. In the present series 60.2% of lesions were radically excised. The histology of the tumour may influence surgical strategy since total excision of an intramedullary ependymoma is fairly straight forward because of the good tumour cord plane. 14 out of the 15 ependymomas in the present series were radically excised. It may be noted that the decision regarding radical excision should be governed by the tumour cord plane rather than the histology obtained during surgery. The treatment of astrocytomas is however more controversial.[17] In the present series 28 patients with astrocytomas were operated upon and a radical excision was possible only in 9 cases due to poor tumour-cord plane. The association of syringomyelia with intramedullary tumours is well recognized and occurs in about 45% of cases.[1],[8] The presence of a syrinx is considered a favourable indicator for radical resection and this is well demonstrated in the present series where 17 of the 21 patients who had an associated syrinx underwent radical excision. This is also in keeping with the experience of earlier authors.[6],[8] In the present series it was interesting to note that the incidence of syringomyelia was similar for astrocytomas and ependymomas. Stein in 1991, reaffirmed the concept that most intramedullary tumours are benign and therefore total excision should be the goal of surgery.[18] One third of the benign astrocytomas in his series showed enough pseudo-membrane around the tumour and, therefore, a good tumour cord plane to accomplish radical excision. They felt that radical resection of low grade astrocytomas significantly increases survival rates. There was, however, a note of caution in the outcome of malignant astrocytomas where biopsy followed by radiation and chemotherapy was advocated.[6] Haemangioblastomas warrant total excision since a cure could be obtained and was carried out successfully in the present series.[19],[20]

  ::   ConclusionsTop

Glial tumours form a majority of intramedullary tumours. This series includes both adult and childhood tumours and the latter account for a significant number of cases (11 out of 68). Patients presenting early had better immediate and long term outcome compared to those presenting late.[21] Radical resection should be the goal in all ependymomas and the majority of low grade astrocytomas. Radical excision in the cervical and cervico-thoracic region is probably easier accomplished compared to other areas. Majority of patients who had radical excision of tumour either improved or remained unchanged at long term follow up while most patients with either partial excision or biopsy of the tumour showed progression of the disease. The strategy for the management of tuberculomas has been outlined.[22] Presence of a good plane of cleavage between the lesion and the cord was the best indicator of radical excision regardless of the pathological nature of the lesion although ependymomas were more easily excised. Radiation as an adjuvant therapy may be indicated in partially excised astrocytomas or for malignant tumours keeping in mind the guarded prognosis.


1Cooper PR , Epstein F : Radical resection of intramedullary spinal cord tumours in adults. Recent experience in 29 patients. J Neurosurg 1985; 63 : 492-499.
2Fischer G, Mansuy L : Total removal of intramedullary ependymomas : follow up study of 16 cases. Surg Neurol 1980; 14 :243-249.
3Herrmann HD, Neuss M, Winkler D : Intramedullary spinal cord tumors resected with CO2 laser microsurgical technique : Recent experience with 15 patients. Neurosurgery 1988; 22 : 518-522.
4Citow JS, Ammirati M : Intramedullary tuberculoma of the spinal cord : case report. Neurosurgery 1994; 35(2) : 327-330.
5Malis LI : Intramedullary spinal cord tumors. Clin Neurosurg 1978; 25 : 512-539.
6Stein BM, McCormick PC : Intramedullary neoplasms and vascular malformations. Clin Neurosurg 1991; 39 : 361-387.
7Cristante L, Hermann HD : Surgical management of intramedullary spinal cord tumors. Functional outcome and sources of morbidity. Neurosurgery 1994; 35 : 69-76.
8Samii M, Kleklamp J : Surgical results of 100 intramedullary tumors in relation to accompanying syringomyelia. Neurosurgery 1994; 35 : 865-873.
9Malis LI : Intramedullary spinal cord tumors. Clin Neurosurgery 1978; 25 : 512-539.
10Guidetti B, Mercuri S, Vagbozzi R : Long term results of the surgical treatment of 129 intramedullary spinal gliomas. J Neurosurg 1981; 54 : 323-330.
11Guidetti B, Fortuna A : Differential diagnosis of intramedullary and extramedullary tumors. In : Handbook of Clinical Neurology, (Eds) Vinken PJ, Bruyn GW. Amsterdam, North Holland 1975; 19 : 51-75.
12Epstein FJ, Farmer JP, Freed D : Adult intramedullary astrocytomas of the spinal cord. J Neurosurg 1992; 77 : 355-359.
13Epstein F, Epstein N : Surgical treatment of spinal cord astrocytomas of childhood. A series of 19 patients. J Neurosurg 1982; 57 : 685-689.
14Kopelson G, Linggood RM, Kleinmann GM : Management of intramedullary spinal cord tumors. Radiology 1980; 135 : 473-479.
15Wood EH, Berne AS, Taveras JM : The value of radiation therapy in the management of intrinsic tumors of spinal cord. Radiology 1954; 63 : 11-24.
16Minehan KJ, Shaw EG, Scheithauer BW et al : Spinal cord astrocytoma : pathological and treatment considerations. J Neurosurg 1995; 83 : 590-595.
17McCormick PC, Torres R, Post KD et al : Intramedullary ependymoma of the spinal cord. J Neurosurg 1990; 72 : 523-532.
18Guidetti B, Fortuna A : Surgical treatment of intramedullary haemangioblastomas of the spinal cord. J Neurosurg 1967; 27 : 530-540.
19Neumann HPH, Eggert HR, Weigel K et al : Haemangioblastomas of the central nervous system : A 10 year study with special reference to von Hippel-Lindau syndrome. J Neurosurg 1989; 70 : 24-30.
20Kleklamp J, Samii M : Introduction of a score system for the clinical evaluation of patients with spinal processes. Acta Neurochir Suppl (Wein) 1993; 123 : 221-223.
21Bucy PC, Oberhill NR : Intradural spinal granulomas. J Neurosurg 1950; 7 : 1-12.