:Acta Neurochirurgica

Acta Neurochir (Wien) (1990) 107:37-43

9 by Springer-Verlag 1990

Metastatic Spinal Cord Compression Occurrence, Symptoms, Clinical Presentations and Prognosis in 398 Patients with Spinal Cord Compression F. Bach l B. H. Larsen 2, K. Rohde 3, S. E. Bargescn 4, F. Gjerris 4, T. Bage-Rasmussen s, N. Agerlin 4, B. Rasmusson 3, P. Stjernholm 6, and P. S. S~rensen 2 Department of Oncology, University Hospital Herlev and 3 The Finscn Institute, University Department of 2 Neurology and 4 Neurosurgery, Righospitalet and Departments of Neurosurgery, University Hospitals 5 Hvidovre and 6 Glostrup, Copenhagen, Denmark

Summary We reviewed aI1 medical records concerning patients suffering from spinal cord or cauda equina compression (SCC) secondary to cancer, in the eastern part of Denmark, from 1979 through 1985. During the period the incidence of SCC in cancer patients went up from 4.4% to 6%. However, this increase was not significant. The series comprised 398 cases, with carcinoma of the prostate (19%), lung (18%), breast (14%) and kidney (10%) accounting for 61%. The symptoms were evaluated in accordance with the patients rating of pain, motor deficits, sphincter control and paraesthesia, whereas the clinical manifestations were classified on the basis of motor deficit and bladder dysfunction. During the period preceding the diagnosis of SCC, 83% of the patients suffered from back pain, 67% from deteriorating gait and 48% had retention of tile urine. In 35% of the patients there was no sphincter disturbance and 10% had normal sensory function. The outcome of treatment was estimated by changes in motor deficits and sphincter function, and depended primarily on the patients condition at the time of the diagnosis. Of the patients who were able to walk before treatment, 79% remained ambulatory, wheras only 18% of the non-ambulatory patients regained walking ability. Patients treated by decompressive laminectomy followed by radiotherapy apparantly had a better response than patients treated with surgery or irradiation alone, but when the patients pro-treatment motor function was taken into account, no significant difference was observed. The study may call for a properly randomized trial with careful stratification of tumour biology, performance status and neurological deficits.

Keywords: Cancer; spinal cord compression; occurrence; symptoms; diagnosis; treatment and prognosis.

Introduction C o m p r e s s i o n o f the spinal c o r d or c a u d a e q u i n a (SCC) is a d i s a b l i n g c o m p l i c a t i o n o f cancer. S y m p t o m -

p r o d u c i n g e p i d u r a l m e t a s t a s e s are r e p o r t e d to occur in 5% o f cancer patients, m o s t frequently as m e t a s t a s e s following lung, breast, prostate and kidney c a r c i n o m a s 1, 2, 8, 11, 19, 23. O c c a s i o n a l l y h a e m a t o l o g i c a l s p r e a d to the spinal c o r d occurs, b u t m o s t i n t r a s p i n a l m e t a s t a s e s are l o c a t e d e p i d u r a l l y a n d derived f r o m vert e b r a l or p a r a v e r t e b r a l metastases. The c o m p r e s s i o n o f the spinal c o r d is usually a c c o m p a n i e d b y severe b a c k pain, d i s t u r b a n c e o f sensation, loss o f sphincter functions a n d p a r a p l e g i a . A s y m p t o m a t i c i n t r a s p i n a l metastases, nevertheless, are s u p p o s e d to be even m o r e c o m m o n t h a n the s y m p t o m a t i c ones ;~. T h e initial s y m p t o m s m a y be sparse b u t w i t h o u t t r e a t m e n t the s y m p t o m s will p r o g r e s s to c o m p l e t e loss o f spinal function b e l o w the site o f the SCC. In a large p r o p o r t i o n o f p a t i e n t s the o u t c o m e o f t r e a t m e n t is unsastisfactory r e g a r d i n g quality o f life a n d survival, a n d p r e s u m a b l y d e p e n d i n g o n the n e u r o l o g i c a l status at the time o f the diagnosis a n d the origin o f the p r i m a r y m a l i g n a n t disease. H o w e v e r , due to the lack o f clinically c o n t r o l l e d investigations, it is still d e b a t e d whether decompressive laminectomy, radiotherapy or a c o m b i n a t i o n is the p r e f e r a b l e choice o f t r e a t m e n t 9' 12, 16, 17, 23

W e a n a l y z e d 398 cases o f SCC, all referred for treatm e n t in a well defined g e o g r a p h i c a l a r e a in the eastern p a r t o f D e n m a r k d u r i n g a seven-year p e r i o d f r o m 1979 to 1985. The p u r p o s e was to elucidate the occurrence, the initial s y m p t o m s , s y m p t o m d e v e l o p m e n t , t r e a t m e n t m o d a l i t i e s a n d the prognosis.

38

F. Bach et al. : Metastatic Spinal Cord Compression

Material and Methods The analysis comprised all patients admitted with SCC secondary to malignant disease in the departments of oncology or neurosurgery in the eastern part of Denmark from January 1979 through December 1985. The departments serve a population of 2.4 million, in whom 398 patients with spinal cord compression were detected. The patients were followed up during 1988 and the survival status of each patient was ascertained. The diagnosis of SCC was usually confirmed by lumbar metrizamide or omnipaque myelography combined with cervical or cisternal myelography (in order to locate the upper limit of the tumour in case of a complete block), or CT scan with subarachnoid contrast injection. When decompressive laminectomy was carried out, the procedure took place immediately after the myelography. Additional radiotherapy was usually started within 5-8 days after the operation. In patients treated with radiotherapy alone, the first dose was given as soon as possible after the myelography as a single posterior field directed against the involved part of the spine, and including at least one adjacent vertebra above and below the epidural block. Beam sources were photon rays, using mega-voltage X-rays, 1.2-6 MeV. Tumour dose was 20-45 Gy in 5-20 fractions during 6-28 days with single fraction of 2-6 Gy. The majority of the patients received 5 Gy* 6. Surgery was regularly undertaken if the patient had no known primary turnout, or if radiotherapy had been applied previously against epidural metastases at the same spinal level. Usually patients who had been paraplegic without residual motor function for more than 24 hours were not offered a laminectomy, but a few received radiotherapy. Finally, a few patients did not receive any specific treatment. Corticosteroids were not used consistently, but were administrated in dosages that varied from department to department. The symptoms were evaluated on the basis of recorded information about the patients rating of pain, motor symptoms, sphincter control and paraesthesia; the clinical findings were classified according to motor deficits and dysfunction of the bladder.

able to make a crude epidemiological evaluation of the incidence of SCC during the period, based on data from the Danish Cancer Registry. Statistical evaluation was performed by SPSS-PC using life-table analysis and analysis of variance or contingency table analysis.

Results

Population Characteristics' During the period spanned by the investigation no increase of cancer in this particular part of Denmark was recorded (unpublished data from the Danish Cancer Registry, 1989). The incidence of SCC in cancer patients went up from 4.4% to 6%. However, this increase is not significant and is apparent in all types of cancer (Fig. 1). The series consisted of 398 patients, 260 male (65%) and 138 female (35%), with an average age of 63 years (range 14-84 years). The age and sex distribution appears from Fig. 2 which shows the highest incidence

/ 4.0 ~ 3.6 3.2 2.8 2.4 2.0

~y 0.40.0 "

Pain was categorized into: 1) no pain at all, 2) localized back pain, 3) radicular or root pain. The motor function was divided into four grades: 1) no motor deficit, 2) moderate deficit, walking ability preserved, 3) paraparesis, unable to walk, 4) paraplegia. The spincter function was classified into three grades: 1) normal function, 2) sphincter dysfunction, 3) incontinence, bladder catheter required.

,~

0.8

~ - ~ 79 80

81

82

83

84

85

Year

Fig. 1. Incidence of spinal cord compression in the eastern part of Denmark in carcinoma of the kidney, prostate breast and lung, from January 1979 to December 1985 No. of patients 1~0-

~

Male(2~)

90-

~

Female(138)

i~ilili~iii

807060,50-

...............

4030-

The sensory function was classified into: 1) no sensory disturbance, 2) disturbance but level of sensation not established, 3) sensation level established. As the investigation was retrospective, it was not possible to evaluate all the parameters adopted in all patients. Since practically all patients treated for SCC were registered in one of the departments participating in the investigation, we were

20100

10-19 Years

L

20-29 Years

i

30-39 Years

40-49 50-59 Years Years Age groups

i

60-69 Years

70-79 Years

i

>8O Years

i

Fig. 2. Age and sex distribution of patients with spinal cord compression secondary to cancer in the eastern part of Denmark, from January 1979 to December 1985

F. Bach et al. : Metastatic Spinal Cord Compression

39

Table 1. Primary Malignancy in 398 Cancer Patients with Spinal Cord

Table 2. Time of the Primary Malignant DiagnosisAccording to Time

Compression

of Spinal Cord Compression in 398 Cancer Patients

Prostate Lung Breast Kidney Lymphoma Myeloma Gastrointestinal tract Melanoma Leukemia Others

Number

%

74 72 54 38 31 16 15 10 3 85

19 18 14 10 8 4 4 3 1 20

No. Mean time

Prostate Lung Breast Kidney

74 72 54 38

1.7 0.5 4.6 2.2

years years years years

Occurrence Before SCC

Sire. SCC

82% 65% 98% 79%

11% 31% 2% 11%

Following SCC 7% 4% 11%

Symptoms and Clinical Findings in the sixth decade. Table 1 shows the number of different malignancies causing SCC. The most frequent primary sites in males were the prostate and the lungs, in females the breast. It should however be noted that in the present study prostate outnumbered lungs. Carcinoma of the prostate (19%), lung (18%), breast (14%) and kidney (10%) accounted for 61%.

Time of SCC Development Altogether, 80% of the patients had had their malignant diagnosis established at the time of SCC, although the mean period from the primary malignant diagnosis to the SCC varied considerably according to the type of the primary turnout: e.g. lung: 0.5 years, prostate: 1.7 years, kidney: 2.2 years and breast: 4.6 years (Table 2). There were only few attestations of cancer being diagnosed subsequently to the SCC, although 31% of the patients with lung cancer had their diagnosis established simultaneously with SCC. In contrast, in breast cancer patients the primary malignant diagnosis had been established previously in nearly all.

The patients symptoms at the time of the diagnosis of SCC were evaluated by rating of pain, motor deficit, sphincter control and paresthesiae. The majority of patients, 83%, had pain at the time of SCC, either root pain or local back pain (Table 3). The initial motor symptoms were unsteadiness of gait (ataxia) (67%). Twenty-seven % of the patients had a feeling of heaviness or weakness in the legs, 6% had no motor symptoms at all. As regards sphincter control, 35% reported no symptoms, 18% moderate disturbance (l% urgency, 10% incontinence and 7% changed sensitivity during urination). The clinical evaluation included motor function, sphincter function and sensitivity. Nearly 65% of the patients had severe motor deficits and were unable to walk; Thirty-two % of the patients had moderate motor deficits with gait function still preserved, whereas 4% only had no motor deficits at all. Thirty-one % of the patients had normal control of bladder and bowel function, 20% had mild dysfunction (such as urgency of normal bladder sensation), and 48% had severe dysfunction (incontinence or retention requiring catheterization). As concerns sensitivity 29% had some, but

Table 3. Neurological Symptoms in 398 Cancer Patients with Spinal Cord Compression

Symptoms Pain Motor signs Sphincter disturbance

Radicular 47% Failing control 67% Severe (Catheter) 47%

Local back 36% Limpness, weakness 27% Moderate 18%

No pain 17% No symptoms 6% No symptoms 35%

Sensation level established 61% Paralytie/paraparetic 64% Retention 47%

Sensation level not established 29% Mild deficit 32% Moderate 22%

Normal sensitivity 10% No deficit 4% No symptoms 31%

Clinical Findings Sensitivity Motor functions Sphincter disturbance

40

F, Bach et al. : Metastatic Spinal Cord Compression

61% of the patients had sensory loss below the level of the spinal compression. Several patients had symptoms long before the SCC was diagnosed. The mean duration from the first symptom until the diagnosis was established was 58 days (average 30 days, range 0-420 days). No significant difference between primary malignancy was demonstrated, although lung cancers seem to progress more rapidly (65% within the first seven days). In 259 patients was were able to determine the time interval between the first visit to a doctor with symptoms indicating incipient SCC, and the diagnosis of SCC. This ,,doctor's delay" had a mean value of 23 days (average 4 days, range 0-360 days). Radiological Studies

The majority of the epidural metastases (71%) were located in the thoracic region. Apart from breast canc e r - mostly located in the lower thoracic r e g i o n - there was no obvious connection between the primary turnour and the location of the spinal metastases (Table 4).

Plain X-ray of the spine was performed in 89% of the patients, lumbar metrizamide or omnipaque myelography combined with cervical or cisternal myelography to locate the upper limit of the tumour in case of complete block in 80%, and/or CT scan with subarachnoid contrast injection in 6%. Plain X-ray examination was normal in 17%; metastases at the relevant spinal level in 48 % ; multiple vertebral metastases in 34%; whereas metastases of vertebrae unrelated to the location of the SCC were found in only 1% of the patients. In patients examined by myelography a complete block was found in 77%, a partial block in 18%, separate blocks caused by epidural deposits in 3% and normal findings in 2%. Only 3% of the patients had symptom producing epidural metastases at two or more separate levels of the spinal canal (Table 5).

Results of Treatment Of the 365 treated patients, the 345 patients eligible for response received treatment in the form of radiotherapy; 149pts. (43%); laminectomy: 105pts. (31%);

Table 4. Anatomical Localization of Spinal Cord Compression in 398 Cancer Patients No. Prostate Lung Breast Kidney Others Total

74 72 54 38 160

Cervical N (%)

U. Thorax N (%)

L. Thorax N (%)

Lumbar N (%)

Several N (%)

2 (3) 4 (6) 2 (4) 3 (9) 7 (4)

28 (38) 31 (43) 8 (15) 8 (21) 51 (32)

27 (36) 21 (29) 30 (55) 19 (50) 59 (37)

11 (15) 14 (19) 12 (22) 7 (18) 40 (25)

6 (8) 2 (3) 2 (4) 1 (2) 3 (2)

18 (5)

126 (32)

156 (39)

84 (21)

14 (4)

Cervical= Cervical. Upper thoracic=Th 1-6. Lower thoracic= Th 7-12. Lumbar = Lumbar.

Table 5. Radiologieal Investigations in 398 Cancer Patients with Spinal Cord Compression Number of vertebral bone metastases Plain X-ray

Normal

Single m e t .

Multiplemet.

Otherirrev, met.

No. 353.

60 (17%)

169 (48%)

120 (34%)

4 (1%)

Number of complete, partial and multiple blocks at myelography Myelography

Complete

Partial

Multiple

Normal

No. 319.

245 (77%)

59 (18%)

9 (3%)

6 (2%)

F,

Bach et al.: Metastatic Spinal Cord Compression

41

Table 6. Effect o f Motor Function on Resonse to Treatment in 245 Cancer Patients with Spinal Cord Compression Pretreatment motor function

Motor function after treatment Able to walk No deficit

Unable to walk Mild deficit Paraparetic Paraplegic

No deficit Mild deficit able to walk) Paraparetic unable to walk) Paraparetic

16 115 165 49

ll 18 5 0

1 73 30 3

4 19 98 15

0 5 32 31

Total

345

34

107

136

68

Table 7. Ef/ect on Bladder Function on Response to Treatment in 301 Cancer Patients with Spinal Cord Compression

Before treatment After treatment

4

2

~

43% Symptoms:

No.

Normal Dysfunction Catheter

[]

Total series

Normal 99 Dysfunction 61 Catheter 141 Total

301

74 16 14

9 22 16

16 23 111

104

47

150

47%

improved

[ ] Unchanged

1

[ ] Worse

~

440/o~

9'/0

Laminectomy and Radiotherapy (n = 91)

Laminectomy (n = 105)

laminectomy followed by radiotherapy: 91 pts. (26%). Fourty-two % of the patients had corticosteroids in doses varying from department to department. As to motor and bladder functions the response appears from Tables 6 and 7. Before treatment 38% were ambulatory, after treatment 41%. No major changes were seen in the number of patients with normal sphincter function and in the number with need of a catheter. The pre-treatment motor deficits seem to be the most important determinant for the motor function after treatment. A total of 79% of the patients who were able to walk before treatment remained ambulatory, whereas only 21% of the non-ambulatory paraparetic patients and 6% of the paralytic patients regained walking ability. Patients treated with laminectomy followed by radiotherapy seemd to respond better than those treated with radiotherapy or laminectomy alone (Fig. 3), but taking the patients pre-treatment motor function into account, no significant difference was observed. Additionally, in the group treated with laminectomy and radiotherapy, a longer survival (average values: radiotherapy 2.3 months, laminectomy 2.8 months, radiotherapy plus laminectomy 5.4 months) presumably reflects that these patients were in a lower stage of disease and had a better performance status.

Radiotherapy (n = 149)

Fig. 3. Immediate treatment respouse in patients with spinal cord compression on motor and sphincter function, to different treatment modalities

~,~

Probability (%) 100. . . . s 8070-

60-

.

Lung Breast ~ B Prostate ~ , - - Kidney

~.

5040-

~

~,

3020-

"",~..

,.

ll~" "~..o,,qK" ~" ~ , ~ ' ~

---%

-,..-,~_~

"'-..,.~"t-_ 0 0

_,..

. . . . . . . .

=u~q=-.-.-.-.-.-.-.~ _ :

i

i

i

~

r

r

I

I

2

4

6

8

10

t2

14

16

: "tl 18 20 Months I

"T

Fig. 4. Probability o f survivat in patients treated for spinal cord

compression secondary to carcinoma of the lung, breast, prostate and kidney. [Lung cancer differs significantly from the other malignancies. (Lee-Desu test, p < 0.01)]

The results of treatment have been given in details by Sorensen et al. 24. The average survival of the patients was 3.1 months (range 0.4-70.8 months). Ten patients were still alive

42 in December 1988 and had an average survival of 38.6 months. The patients with breast, prostate and kidney cancer survived longer than patients with lung cancer (Fig. 4). The number of long-term survivors (> 12 months after diagnosis of epidural metastases) amounted to 20 patients with breast cancer (45% of all), 23 with prostate cancer (38%), 8 with lymphoma/ myeloma (17%), whereas only 3 (5%) of the patients with lung cancer survived for more than one year. Discussion

Spinal cord compression constitutes one of the major problems in neuro-oncology. The treatment is often debated, mostly due to the lack of evidence of the best treatment modality. The present series is in principle unselected, and includes all patients referred for treatment for SCC in a geographically defined area during a seven-year period from January 1979 to December 1985. No increase in the occurrence of cancer in the area during the study period was observed, although the authors were able to demonstrate an increase in the rate of incidence in SCC from 4.4 to 6%. This increase is not significant. However, it is to the best of our knowledge the first study attempting to evaluate the incidence of SCC in different primary malignancies on the basis of epidemiological data. The overall incidence of SCC in general was similar to the rate of 5% usually reported 1, 2, 8, 11, 19, 23, although for various reasons obviously not all patients with SCC were referred for treatment (e.g. bad physical condition, heavy extra CNS tumour burden and terminal disease). Hence, the incidence obtained is probably an underestimation of the real incidence. The distribution of the primary malignancies does not differ from most other larger series 9, ~3, 27, nor did the neurological status of the patients before treatment (38% ambulatory) 7' 9, 11, 16, 21, 28 It is supposed, that in general a patient who has been paralytic for more than 24 hours, is unable to benefit from any treatment. This presumption may be wrong, since Tarlov has shown that recovery may be extended and last for several months 2s, findings supported by others ~5. These considerations seem important, especially when evaluating treatment results, since the period of observation ought to be prolonged as well. However, this point is unclear in most reported studies but preceeding paresis or paralysis for 24-48 hours, consequently, may not be an acceptable argument for rejecting a patient from treatment for SCC. During the period preceding the SCC 83% of the patients had suffered from back pain, either from ra-

F. Bach et al. : MetastaticSpinal Cord Compression diating pain (47%) or from local back pain (36%). Pain in the back, however, occurs frequently in cancer patients (83% had vertebral bone metastases, 34% had multiple symptoms producing vertebral bone metastases) and for a considerable period it may be the only symptom. It is generally accepted that early symptoms preceeding SCC include pain TM23, and Rodichock e t al. strongly recommend myelography in patients with onset of back pain when metastatic disease is suspected, since spinal epidural metastases were detected in 75 out of 140 cancer patients by myelography. It is noteworthy, however, that 25 of these had no evidence of myelopathy at the time of SCC. In addition, 90% of the patients diagnosed and treated while still ambulant remained so 2~ The initial symptoms that ought to lead to diagnostic procedures (such as myelography, X-rays or CT scans) may be sparse; but it is of interest that the ,,patient's and doctor's delays" observed, were as long as 58 days and 23 days, respectively (mean values). In a study concerning patients with small cell lung cancer nearly half of the cases with SCC were diagnosed simultaneously with the malignant disease. These patients suffered mainly from pain, pain preceding motor dysfunction and bone destruction with total block at myelography (symptom duration 65 days), while patients developing SCC during therapy mainly suffered from disturbances of sphincter function (symptom duration 11 days)18. The findings are in agreement with our results, however, it should be added that it seems to be primary lung cancer patients who develop SCC simultaneously with (31%) or prior to the primary malignant diagnosis. By contrast, 98 % of the breast cancer patients had their diagnosis established at iche time of the SCC, and cases of a duration of several years (1015 years) before developing SCC are reported. Treatment of SCC has usually been and still is laminectomy with or without postoperative irradiation s, 9, 13, 16, 22, 23, 26, 27 although a number of authors are claiming that radiotherapy alone is as effective 7' 9, 11,12. No controlled clinical studies of a sufficiently large number of patients are available. The overall results of treatment, without regard to the different treatment modalities, were preservation of gait function in 79% of those patients who were ambulatory before treatment, and restoration of walking abilities in 18% of the patients who before treatment were unable to walk. These results are equal to or better than the treatment response in previous series monitoring effect of therapy in a similar way 7' 8, 11, 12, a~, 28, and confirm that as to walking ability the results depend on the pretreatment ambulatory status.

F. Bach et al.: Metastatic Spinal Cord Compression

The significance of the primary tumour type as an indicator of survival appears clearly from the patients included in the present study. Except for small cell bronchogenic carcinoma, metastasis from pulmonary carcinomas respond poorly to all forms of therapy. Conversely, in spinal epidural metastases from lymphoma and carcinomas from the breast and prostate the response is considerably more favourable 3' 4, 6, 9, i1, ~_2.The choice of therapy (surgery, radiotherapy or the combination), cannot be assessed from the present retrospective study, but should be elucidated through a clinically controlled randomized tiral with distinct stratification of primary tumour, performance status and neurologic deficits. At present, the only way to improve the prognosis seems to be early detection. Myelography is safe and free of serious complications, but another major procedure for a patient who has already undergone many. Computer tomography of the spinal canal does not adequately detect epidural tumours in the absence of contrast material1~ Magnetic resonance imaging may permit visualization of the epidural space because of the unique characteristics of lipid-rich material 14, but conclusive studies with this technique have not yet been completed. Meanwhile, one must rely on careful clinical examination, especially in cancer patients with back pain, and bear in mind that appropriate diagnostic procedures and treatment may preserve the gait function in almost 80% of the ambulant patients developing spinal cord compression. References 1. Bansal S~ Brady LW, Olsen A eta[ (1967) The treatment of metastatic spinal cord tumors. JAMA 202:686-688 2. Barron KD, Hirano A, Araki S etal (1959) Experience with metastatic neoplasm involving the spinal cord. Neurology 9: 91106 3. Black P (1979) Spinal metastases: current status and recommended guidelines for management. J Neurosurg 5:726-746 4. Brady LW, Antoniades J, Prasavinichai Set al (1975) The treatment of metastatic disease of the nervous system by radiation therapy. In: Seydl HG (ed) Tumors of the nervous system. John Whiley and Sons, New York, pp 176-189 5. Brice J, McKissock WS (1965) Surgical treatment of malignant extradural spinal tumours. Br Med J 1:1341-1344 6. Chade HO (1976) Metastatic tumors of the spine and spinal cord. In: Vinken P J, Bruyn GW (eds) Handbook of clinical neurology. 20. North-Holland publishing Co, Amsterdam, pp415M33 7. Cobb CA, Leavens ME, Eckles N (1977) Indications for nonoperative treatment of spinai cord compression due to breast cancer. J Neurosurg 47:653-658 8. Constans JP, Divitis ED, Donzelli R etal (1983) Spinal metas-

43 tases with neurological manifestations. Review of 600 cases. J Neurosurg 59:111-118 9. Dunn RC, Kelly WA, Wohns RNW etal (1980) Spinal epidural neoplasia. A 15-year review of tile results of surgical therapy. J Neurosurg 52: 47-5t 10. Eldervik OP, Dugstad G, Orrison WW etal (1982) The effect of clinical bias of the interpretation of myelography and spinal computed tomography. Neuroradial 145:85-89 11. Gilbert RW, Kim JH, Posner J (1978) Epidural spinal cord compression from metastatic tumour: diagnosis and treatment. Ann Neurol 3:40-51 12. Greenberg HS, Kim JH, Posner JB (1980) Epidural cord compression from epidural turnout: results with a new treatment protocol. Ann NeuroI 8:361 366 13. Hall AJ, Mackay NNS (1973) The results of laminectomy for compression of cord and cauda equina by extradural malignant tumour. J Bone Joint Surg 55:497-595 14. Hutchison JMS, Smith FW (I983) NMR clinical results: Aberdeen. In: Partain, James, Rollo, Price (eds) Nuclear magnetic resonance imaging. Saunders, Philadelphia, pp 231-249 15. Larsen SH, Rasmusson B, Sorensen PS (1989) Recovery of gait by radiotherapy in patients with spinal cord compression secondary to metastatic cancer. European Conference of Cancer and Clinical Oncology V. Proceedings, P-0482 16. Livingston KE, Perrin RG (1978) The neurosurgical management of spinal metastases causing cord and cauda equina compression. J Neurosurg 3:839 843 17. Nather A, Base K (1982) The results of decompression from metastatic extradural tumors. Clin Orthop 169:103-108 18. Pedersen AG, Bach F, Melgaard B (1985) Frequency, diagnosis and prognosis of spinal cord compression in small cell lung bronchogenic carcinoma. Cancer 55: 1818-1822 19. Posner J (1971) Neurological complications of systemic cancer. Med Clin North Am 1971 55:625-646 20. Rodichock LD, Ruckdeschel JC, Harper GR etal (1986) Early detectio~'~ and treatment of spinal epidural metastases: the role of myelography. Ann Neuroi 20:696-702 21. Shaw MDM, Rose JE, Paterson A (1980) Metastatic extradural malignancy of the spine. Acta Neurochir (Wien) 52:113-120 22. Smith R (1965) An evaluation of surgical treatment for spinal cord compression due to metastatic carcinoma. J Neurol Neurosurg Psychiatry 28:152-158 23. Stark R J, Henson RA, Evans SJW (1982) Spinal metastases. A retrospective survey from a general hospital. Brain 105: 189213 24. Sorensen PS, Borgesen SE, Rohde K etal (1990) Metastatic epidural spinal cord compression: results of treatment and survival. Cancer. 65:1502-1508 25. Tarlov M and Herz E (1954) Spinal cord compression studies. IV. Outlook with complete paralysis in man. A.M.A. Archives of Neurology and Psychiatry 72:43-51 26. Wild WO, Porter RW (1963) Metastatic tumour of the spine: a study of 45 cases. Arch Surg 87:825-830 27. Wright RL (1963) Malignant tumors in the spinal extradural space. Results of surgical treatment. Ann Surg 157:227-231 28. Young RF, Post EM, King GA (1980) Treatment of epidural metastases. Randomized prospective comparison of laminectomy and radiotherapy. J Neurosurg 53:741-748 Correspondence and Reprints: FIemming Bach, M.D., Department of Oncology, 54C3, University Hospital Herlev, 2730 Copenhagen, Denmark.