CLINICAL STUDY - PATIENT STUDY

Survival after rehabilitation for spinal cord injury due to tumor:a 12-year retrospective study

Michael Tan • Peter New

Received: 17 July 2010 / Accepted: 8 November 2010 / Published online: 23 November 2010

� Springer Science+Business Media, LLC. 2010

Abstract Metastatic spinal cord compression occurs in

2.5% of cancer patients in the last 5 years of life. Spinal

cord injury (SCI) due to tumors accounts for 26–45% of

non-traumatic SCI. With the aging population and

improving survival, the number of patients with SCI due to

tumor who have the capacity to benefit from rehabilitation

is increasing. This study planned to document survival in

patients with primary and secondary tumors causing SCI

admitted to a spinal rehabilitation unit over a 12-year

period. This is a retrospective open-cohort case series of

patients admitted to a Spinal Rehabilitation Unit between 1

July 1996 and 30 June 2008 with a diagnosis of tumor

causing SCI. Linkage to the Registry of Births, Deaths and

Marriages (Victoria) was used to calculate survival in

months following discharge from inpatient rehabilitation.

108 patients were included in the study cohort of which

62% (n = 67) were male. The median age was 61.5 years

(IQR 53.6–74). The majority of patients had paraplegia

(n = 97, 89.8%) and a metastatic tumor (n = 71, 65.7%)

causing their SCI. Most (n = 78, 72%) patients died during

the study period. The median survival for primary tumors

was significantly greater (Cox regression HR = 1.8,

P = 0.03) 9.5 months (IQR 2.8–34.4 months) than that for

metastatic tumors (median 2.8, IQR 1.2–9.0 months).

Conclusion: Patients with longer survival are appropriate

for rehabilitation programs targeting longer-term goals

whilst patients with shorter survival are more suitable for

short, directed rehabilitation programs.

Keywords Spinal cord compression � Spinal neoplasms �Rehabilitation � Survival analysis � Non-traumatic spinal

cord injury

Introduction

The burden of having two devastating severe diseases can

have overwhelming consequences for the life of patients

with spinal cord injury (SCI) due to tumor. The overall rate

of confirmed spinal metastases at autopsy in patients whose

cause of death was malignancy has been reported to be as

high as 36% [1]. In a Canadian population based cohort

study, symptomatic metastatic spinal cord compression

(MSCC) has been reported to occur in 2.5% of cancer

patients in the last 5 years of life [2], and more commonly

results in paraplegia than tetraplegia [3]. The incidence of

MSCC varies by primary disease site and age [2]. In con-

trast, primary spinal cord tumors are rare, accounting for

4–8% of tumors arising from the central nervous system,

with a reported incidence rate of 0.74 per 100,000 [4].

SCI resulting from tumors account for 26–45% of non-

traumatic spinal cord injuries [5, 6]. They are associated with

short survival durations [7], and have a peak incidence

between 50 and 70 years of age [8]. With ageing of the pop-

ulation and improvements in early cancer diagnosis and

treatment leading to increased survival, the number of patients

with SCI due to tumor who have the capacity to benefit from

rehabilitation is likely to increase in coming decades.

M. Tan � P. New

Spinal Rehabilitation Unit, Caulfield Hospital, Caulfield, VIC,

Australia

P. New

Epworth Monash Rehabilitation Unit, Monash University,

Melbourne, VIC, Australia

M. Tan (&)

Department of Medicine, St Vincent’s Hospital Melbourne,

P.O. Box 2900, Fitzroy, VIC 3065, Australia

e-mail: Michael.tan@svhm.org.au

123

J Neurooncol (2011) 104:233–238

DOI 10.1007/s11060-010-0464-6

The overall life expectancy and oncologic prognosis is

important when determining treatment options for newly

diagnosed patients with MSCC. Prognostic predictors for

survival have been described in recent years that aid in the

decision-making process [9–11].

Although there have been studies of the rehabilitation

outcomes for patients with SCI due to tumor, including

MSCC [6, 7, 12, 13], studies reporting on survival of these

patients have been limited by small sample sizes [6, 14]

and short follow-up [14]. This information, however,

would be very helpful to justify participation in specialist

SCI rehabilitation to patients and their families, as well as

health care management and rehabilitation gatekeepers. To

address this knowledge gap, a project was planned to

perform a retrospective open-cohort consecutive case ser-

ies to determine the outcomes of patients admitted to a

specialized SCI rehabilitation unit over a 12-year period

with primary and secondary spinal cord tumors. The

objective of this component to the project was to document

the survival patterns in these patients.

Methods

Setting

The Spinal Rehabilitation Unit at Caulfield Hospital, Vic-

toria, Australia is a 10-bed inpatient unit. It is located in a

public hospital funded through the State health department.

The unit specializes in non-traumatic SCI patients, and is

one of two SCI rehabilitation services in the State. Patients

are referred to the Spinal Rehabilitation Unit from both

private and public acute hospitals, mainly in Melbourne,

but also sometimes from elsewhere in the State of Victoria.

The rehabilitation physician or advanced trainee in reha-

bilitation medicine attached to the Unit initially assesses

patients referred by acute hospitals to determine suitability

for admission. The selection process excludes patients who

are perceived to have a prognosis of less than 3 months.

These patients are directed to the palliative care services.

Patients

The health information department at Caulfield Hospital

used the International Classification of Diseases and

Related Health Problems, 10th edition, Australian modifi-

cation (ICD-10-AM) [15] coding to identify patients with

SCI due to tumor who were discharged from the spinal

rehabilitation unit with a diagnosis of tumor between 1 July

1996 and 30 June 2008. The medical records of these

patients were then reviewed by one of us (MT) to confirm

the etiology of SCI and that patients met the inclusion

criteria. Only patients with a new onset of a SCI as a result

of primary or secondary tumor were included. Patients who

were not admitted for their initial admission after the onset

of SCI but were re-admitted for late complications were

excluded. A total of 116 patients were identified that ful-

filled the study criteria. Eight records were found to be

missing from the list of eligible patients (five records had

been destroyed as part of a previous hospital policy and

three records were not able to be located).

Data collection

The medical records of eligible patients were reviewed by

one of the authors (MT) and relevant details were extracted.

This included demographic information (date of birth and

gender), tumor diagnosis and type (primary or secondary),

neurological level, and whether spinal surgery, radiotherapy

and chemotherapy prior to or during the rehabilitation

admission where documented. The date of acute hospital

admission, date of admission to the Spinal Rehabilitation

Unit and subsequent discharge were also recorded.

Information collected was de-identified and recorded

electronically into a secured database. Database entry was

performed using FileMaker Pro software (version 10.0;

FileMaker, Inc, Clara, CA). Information identifying indi-

viduals was also collected in a separate database for sub-

mission to the Registry of Births, Deaths and Marriages in

order to determine survival. Patient’s details were provided

to the Registry of Births, Deaths and Marriages (Victoria)

in September 2008 to identify the date of death of any

patient until 30th June 2008. The date of death received

from the Registry of Births, Deaths and Marriages was

added to the main database. Survivorship was calculated as

survival in months after discharge from the Spinal Reha-

bilitation Unit.

Following completion of the data collection, a random

sample of 10% of medical files were rechecked and com-

pared with the database for errors. If an error rate greater

than 5% was found, the study protocol required rechecking

another random sample of 40% of patient records and

correcting any errors identified.

Data analysis

Statistical analysis was performed using Stata software

(version 9; Statacorp, Texas). Descriptive statistics were

calculated for demographic characteristics and clinical

features. Statistical significance was defined as a P-

value \ 0.05. Survival analysis for primary and secondary

tumors was performed using the Kaplan–Meier method.

The Cox regression procedure was performed to detect

significant differences between the two groups.

Approval for this project was obtained from the

Research and Ethics Unit at Alfred Health.

234 J Neurooncol (2011) 104:233–238

123

Results

A total of 108 patients were included in the study cohort.

The median age of patients was 61.5 years (Interquartile

range (IQR) 53.6–74.0), with males comprising 62.0%

(n = 67) and females 38.0% (n = 41) of the sample. The

majority of patients (n = 97, 89.8%) had paraplegia and

9.3% (n = 10) tetraplegia (n = 1, 1% level not listed).

Tumor treatments undertaken prior to or during rehabili-

tation admission are shown in Table 1. The majority of

cases were due to metastatic disease (n = 71, 65.7%) and

34.3% (n = 37) had a primary tumor (Table 2).

The length of stay (LOS) in acute hospital preceding

admission to the Spinal Rehabilitation Unit was a median

of 25.5 days (IQR 16.0–47.5). There was no significant

difference (Z = 0.3, P = 0.76) between patients with a

primary tumor (median 25.0 days, IQR 17.0–39.0) com-

pared to those with a secondary tumor (median 26.0 days,

IQR 14.0–51.0). Patients spent a median of 47.5 days (IQR

25.0–80.5) in the Spinal Rehabilitation Unit. Although

there was no significant difference (Z = 1.5, P = 0.12)

between the LOS in rehabilitation for patients with a pri-

mary tumor (median 64.0, IQR 28.0–91.0) compared to

those with a secondary tumor (median 46.0, IQR

22.0–73.0), patients with a primary tumor had a trend

towards shorter admissions. The majority of patients in this

study were discharged home (Table 3) however there

appeared to be a trend with more patients with secondary

tumors being discharged to a nursing home and palliative

care (P = 0.16).

Survival

A total of 72.2% (n = 78) patients died during the study

period. Three patients died during their admission to the

Spinal Rehabilitation Unit and 12 patients were transferred

to an acute hospital facility due to medical instability, and

did not return.

The median survival following discharge differed

between groups, with patients who had a primary tumor

having a significantly greater (Cox regression HR = 1.8,

P = 0.03) median survival of 9.5 months (IQR 2.8–34.4),

and secondary tumors a median of 2.8 months (IQR

1.2–9.0 months) (Fig. 1). Patients with a secondary tumor

were 80% more likely to die at any given time than those

with a primary tumor. Survival at 1 year was 47.4% (95%

CI (Confidence Interval) 24.4–67.3%) in the primary tumor

group and 21.4% (95% CI 11.9–32.9%) in the secondary

tumor group. At 5 years, survival was 10.5% (95% CI

1.8–28.4%) in the primary tumor group and 3.6% (95% CI

0.7–10.9%) in the secondary tumor group. The survival of

patients in each tumor category is shown in Table 4. There

was no relationship between survival and age (P = 0.08)

or gender (P = 0.25).

Of those who remained alive at the end of the study

period, patients with a primary tumor (n = 18, 16.7%) had

survived a median of 78.1 months (IQR 52.2–128.0) post

discharge from the Spinal Rehabilitation Unit, while

patients with a secondary tumor (n = 12, 11.1%) survived

a median of 46.4 months (IQR 24.3–104.5) after discharge.

Table 1 Treatments provided to patients with SCI due to tumor

Surgery Radiotherapy Chemotherapy Total subjects

n (%) n (%) n (%)

Primary 31 (84) 8 (22) 3 (8) 37

Secondary 36 (51) 46 (65) 16 (23) 71

Total 67 (62) 54 (50) 19 (18) 108

Table 2 Etiology of tumors causing SCI

Primary n (%) Secondary n (%)

Ependymoma 7 (19) Prostate 20 (28)

Meningioma 7 (19) Lymphoma 11 (15)

Myeloma 7 (19) Lung 9 (13)

Othera 7 (19) Breast 8 (11)

Haemangioma 3 (8) Renal 8 (11)

Astrocytoma 2 (5) Otherb 4 (6)

Chloroma 1 (3) Melanoma 3 (4)

Giant cell tumour 1 (3) Endometrial 2 (3)

Plasmacytoma 1 (3) Colorectal 1 (1)

Osteosarcoma 1 (3) Esophagus 1 (1)

Total 37 Pancreatic 1 (1)

Gastric 1 (1)

Liver 1 (1)

Mesothelioma 1 (1)

Total 71

a Epidermal cyst; benign conus tumor; primitive neuroectodermal

tumor; Ganglioblastoma; schwanoma; lipoma; neurolemmomab Carcinoid; Dumbell tumor; Cervical SCC; Nasopharyngeal

Table 3 Discharge destination

Discharge destination Total Primary Secondary

n (%) n (%) n (%)

Private residence 67 (62) 29 (78) 38 (54)

Hospital 12 (11) 4 (11) 8 (11)

Nursing home 11 (10) 2 (5) 9 (13)*

Assisted living residence 5 (5) 1 (3) 4 (6)

Deceased 2 (2) 0 (0) 2 (3)

Palliative care 11 (10) 1 (3) 10 (14)

Total 108 37 71

* P = 0.16 (Fisher’s exact)

J Neurooncol (2011) 104:233–238 235

123

Ten medical records were randomly checked for data

integrity and validated to be correct, so no further random

samples were taken of the study population.

Discussion

We have reported the survival of patients with SCI due to

tumor who were discharged from a specialist SCI reha-

bilitation unit. Not surprisingly, patients with MSCC had a

reduced change of survival when compared to those with a

primary tumor.

It is important to emphasize that patients were accepted

into our service with a strong appreciation of their esti-

mated survival. This is a crucial aspect of our Units’ phi-

losophy towards patients with SCI due to tumor. The

treating Oncology unit is requested to provided an esti-

mation of the patient’s prognosis to assist with our decision

regarding whether to admit or not, and how our team

focuses its’ efforts and prioritizes goals. Patients who are

felt to have a very poor prognosis (generally less than

3 months) are excluded and referred to palliative care. We

accept that on occasions patients may deteriorate quicker

than expected and die during their inpatient rehabilitation

admission, or require transfer back to an acute hospital or

to palliative care. We also accept that some patients may

have reduced ability to participate fully in therapy com-

pared to patients with SCI without tumor.

The team in the Spinal Rehabilitation Unit modify their

goals and the focus of the admission according to the

patients’ prognosis and ability to participate. There is a

trade-off between functional outcomes and length of stay

that must be taken into account to allow those patients with

limited life expectancy to have optimal time with their

loved ones. This may involve trading off independent

function with equipment prescription and carer training to

ensure the patient may be managed at home with appro-

priate services.

More recently, a conceptual framework that guides

rehabilitative decision-making has been published. The

‘‘NOMPS’’ criteria includes assessment of the neurologic

(N), oncologic (O), medical (M), pain (P), and support (S)

status of the patient [16, 17], and may be useful for guiding

the assessment by rehabilitation teams.

When comparing our results of survival with other

studies, it needs to be emphasized that we calculated sur-

vival for our study cohort from the time of discharge from

the Spinal Rehabilitation Unit, not from diagnosis. We

used this approach for two reasons. Firstly, we did not

always have the exact date of diagnosis. More importantly,

we feel that this approach provides a more patient-centered

measure of life expectancy after patients’ are discharged

from hospital. Adding the LOS in acute hospital and

Rehabilitation to the survival results for our patients pro-

vides a more appropriate comparison to previous studies.

The median survival of 2.8 months for patients with

secondary tumors in our series is consistent with that

reported in other studies on these patients that have

reported a median survival from diagnosis of 4.6 months

[18], 4.1 months [7], 3.6 months [19] and 82 days [20].

Other studies, however, have reported both worse survival

of 59 days [21] and also better results of 10 months [14]

and 12.7 months after the onset of paraplegia [6]. The

differences in survival reported may be due to a number of

factors. The presence of visceral metastases has been

Fig. 1 Kaplan–Meier survival estimate by tumor group. Cox regres-

sion HR = 1.8, P = 0.03

Table 4 Survival of SCI patients by tumor type

n Median (IQR) (months)

Primary tumorsa

Otherb 7 46.1 (1.3–61.0)

Myeloma 7 6.3 (1.1–23.2)

Secondary tumorsc

Otherd 4 5.3 (2.1–13.9)

Breast 8 4.0 (2.2–28.5)

Prostate 20 3.2 (1.2–9.0)

Renal 8 2.8 (1.3–9.2)

Lymphoma 11 2.0 (0.7–51.1)

Lung 9 0.9 (0.2–5.8)

a Plasmacytoma (13 days); Osteosarcoma (121 days); Meningioma

(369 days); Haemangioma (all 3 alive); Giant cell tumor (alive);

Chloroma 379 daysb Epidermal cyst (alive); benign conus tumor (alive); primitive neu-

roectodermal tumor (1384 days); Ganglioblastoma (1829 days);

schwanoma (916 days); lipoma (2095 days); neurolemmoma

(40 days)c Colorectal (48 days); Esophagus: died in hospital; Pancreatic

(2 days); Gastric (15 days); Liver (130 days); Mesothelioma

(42 days)d Carcinoid (418 days); Dumbell tumor (alive); Cervical SCC

(64 days); Nasopharyngeal (160 days)

236 J Neurooncol (2011) 104:233–238

123

demonstrated to adversely affect overall survival whilst

bone metastases was not found to affect overall survival

[18]. Overall survival was 4.5 months in the visceral

metastases group as opposed to 8.1 months in the group

with no visceral metastases, and not surprisingly, patients

with Karnofsky performance scores between 80 and 100

had significantly prolonged survival compared with

patients who had lower performance scores [18].

With respect to median survivals by tumor type, other

studies have demonstrated major differences in survival

according to primary cancer diagnosis [22]. Breast cancer

patients had a significantly longer survival than both lung

cancer and prostate cancer, with a median survival of

13 months from diagnosis, compared to 7 months for

prostate cancer and 2 months for lung cancer [22]. This

differs from our results and may be due to a combination of

small sample size and advances in the effectiveness of

current therapeutics.

The majority of patients (62%) in our cohort were dis-

charged home. In Australia, nursing home accommodation

payments are determined according to a resident’s assets,

with a minimum payment of 85% of the aged or disability

pension. It would be useful from a health economics per-

spective to have data on the cost/benefit for patients with

SCI due to tumor receiving specialist SCI rehabilitation,

with an aim of returning home in a functional state that

they and their supports can cope with, compared with

remaining in an acute hospital or transferring to a nursing

home. However, given that this was a retrospective study

conducted over 12 years it is not possible to provide the

data accurately. SCI rehabilitation in Victoria is funded via

a casemix system on a per diem basis that varies from year

to year. Calculating the cost of care based on either the

length of stay or cost of a nursing home bed would be

inaccurate given the need to make several assumptions that

will not necessarily apply to all patients. Given that both

nursing home and acute hospital beds are limited resources

that are in high demand, it is undesirable for the SCI

patients admitted to our unit to remain in acute care once

the diagnosis and management has been established and

implemented or to go to a nursing home if this can be

avoided. Consequently, the role of discharge planning is

essential in ensuring a smooth transition to home, subacute

or palliative care services based on the current functional

status of the patient. We believe that empirically it is

possible to claim that specialty SCI inpatient rehabilitation

offers the cohort of patients reported upon here the most

cost-effective and patient-centered setting of care after

acute hospitalization.

There are a number of limitations to this study. The

sample size is relatively small, however, this present report

is still one of the largest studies on the survival of SCI

tumor patients admitted to rehabilitation. The study was

conducted over 12 years and it would be very difficult to

prospectively recruit larger numbers of SCI cancer patients

in a rehabilitation population within a shorter timeframe

and a more robust study design without considerable

resources and the involvement from multiple centers.

Another limitation of this study is the shortcoming inherent

with retrospective studies having potential inaccuracies in

medical record documentation. This study also reflects a

single rehabilitation institution.

There are a number of implications of this study.

Given the varying survival patterns of patients with

differing types of tumor etiology, rehabilitation teams

must be careful in trying to select only those patients

with potential to benefit from an inpatient rehabilitation

program and recommend palliative care where prognosis

is poor. The rehabilitation team must also balance opti-

mizing the functional goals and outcomes with an

appropriate length of stay to allow more time for patients

to be with family. The consequences of tumor beyond

the location of the primary site are well described. The

main consequences of SCI included reduced mobility and

sensation, increased risk of pressure ulcer, and inconti-

nence. These patients, therefore, have a need ideally for

joint involvement of both oncology and SCI rehabilita-

tion teams, and also palliative care teams in many cases.

It is important that oncology teams are aware that most

patients with SCI due to tumor should be referred for an

opinion from a SCI rehabilitation unit regarding suit-

ability for a specialized program focused on the patients’

needs and circumstances.

Conclusion

This retrospective study of patients who were selected for

admission into a spinal rehabilitation unit with a primary or

secondary tumor demonstrates that most were discharged

alive and had a survival that is acceptable and consistent

with previous studies. Patients with a likelihood of a longer

survival (primary tumors and those with SCI due to breast,

prostate, renal and lymphoma) are appropriate for reha-

bilitation programs targeting longer term goals. Those with

shorter survival are suitable for short, directed rehabilita-

tion programs aiming at ensuring a timely transition to

home.

Acknowledgments The authors would like to acknowledge the

assistance provided by Jane Kaye, Lil Powell, and the other staff in

the Health Information Department at Caulfield Hospital for their

assistance with the retrieval of the medical files used in this project.

Sharon Perera, Manager, Legal and Policy, Registry of Births, Deaths

and Marriages (Department of Justice, Victoria) is thanked for her

assistance with providing the information regarding the date of death.

The authors declare no financial sources of support.

J Neurooncol (2011) 104:233–238 237

123

References

1. Wong DA, Fornasier VL, MacNab I (1990) Spinal metastases:

the obvious, the occult, and the impostors. Spine 15(1):1–4

2. Loblaw DA, Laperriere NJ, Mackillop WJ (2003) A population-

based study of malignant spinal cord compression in ontario. Clin

Oncol (R Coll Radiol) 15(4):211–217. doi:10.1016/S0936-6555

(02)00400-4

3. Prasad D, Schiff D (2005) Malignant spinal-cord compression.

Lancet Oncol 6(1):15–24. doi:10.1016/S1470-2045(04)01709-7

4. Schellinger KA, Propp JM, Villano JL, McCarthy BJ (2008)

Descriptive epidemiology of primary spinal cord tumors. J Neu-

rooncol 87(2):173–179. doi:10.1007/s11060-007-9507-z

5. New PW (2005) Functional outcomes and disability after non-

traumatic spinal cord injury rehabilitation: results from a retro-

spective study. Arch Phys Med Rehabil 86(2):250–261. doi:

10.1016/j.apmr.2004.04.028

6. Fattal C, Gault D, Leblond C, Gossens D, Schindler F, Rouays-

Mabit H, Fabro M, Bauchet L (2008) Metastatic paraplegia: care

management characteristics within a rehabilitation center. Spinal

Cord. doi:10.1038/sc.2008.75

7. Guo Y, Young B, Palmer JL, Mun Y, Bruera E (2003) Prognostic

factors for survival in metastatic spinal cord compression: a ret-

rospective study in a rehabilitation setting. Am J Phys Med

Rehabil 82(9):665–668

8. McKinley WO, Huang ME, Brunsvold KT (1999) Neoplastic

versus traumatic spinal cord injury: an outcome comparison after

inpatient rehabilitation. Arch Phys Med Rehabil 80(10):

1253–1257

9. Rades D, Rudat V, Veninga T, Stalpers LJ, Hoskin PJ, Schild SE

(2008) Prognostic factors for functional outcome and survival

after reirradiation for in-field recurrences of metastatic spinal

cord compression. Cancer 113(5):1090–1096. doi:10.1002/cncr.

23702

10. Rades D, Hoskin PJ, Karstens JH, Rudat V, Veninga T, Stalpers

LJ, Schild SE, Dunst J (2007) Radiotherapy of metastatic spinal

cord compression in very elderly patients. Int J Radiat Oncol Biol

Phys 67(1):256–263. doi:10.1016/j.ijrobp.2006.08.011

11. Tokuhashi Y, Matsuzaki H, Oda H, Oshima M, Ryu J (2005) A

revised scoring system for preoperative evaluation of metastatic

spine tumor prognosis. Spine 30(19):2186–2191

12. McKinley WO, Conti-Wyneken AR, Vokac CW, Cifu DX (1996)

Rehabilitative functional outcome of patients with neoplastic

spinal cord compressions. Arch Phys Med Rehabil 77(9):892–895

13. Hacking HG, Van As HH, Lankhorst GJ (1993) Factors related to

the outcome of inpatient rehabilitation in patients with neoplastic

epidural spinal cord compression. Paraplegia 31(6):367–374

14. Tang V, Harvey D, Park Dorsay J, Jiang S, Rathbone MP (2007)

Prognostic indicators in metastatic spinal cord compression:

using functional independence measure and Tokuhashi scale to

optimize rehabilitation planning. Spinal Cord 45(10):671–677

15. National Centre for Classification in Health (2000) The interna-

tional statistical classification of diseases and related health

problems, 10th revision, Australian modification (ICD-10-AM),

2nd edn. National Centre for Classification in Health, Sydney

16. Stubblefield MD (2008) Barriers to rehabilitation of the malig-

nant spine. Top Spinal Cord Inj Rehabil 14(2):19–30

17. Stubblefield MD, Bilsky MH (2007) Barriers to rehabilitation of the

neurosurgical spine cancer patient. J Surg Oncol 95(5):419–426

18. van der Linden YM, Dijkstra SPDS, Vonk EJA, Marijnen CAM,

Leer JWH, Group DBMS (2005) Prediction of survival in patients

with metastases in the spinal column: results based on a ran-

domized trial of radiotherapy. Cancer 103(2):320–328. doi:

10.1002/cncr.20756

19. Helweg-Larsen S, Sørensen PS, Kreiner S (2000) Prognostic

factors in metastatic spinal cord compression: a prospective study

using multivariate analysis of variables influencing survival and

gait function in 153 patients. Int J Radiat Oncol Biol Phys

46(5):1163–1169

20. Cowap J, Hardy JR, A’Hern R (2000) Outcome of malignant

spinal cord compression at a cancer center: implications for

palliative care services. J Pain Symptom Manage 19(4):257–264

21. Conway R, Graham J, Kidd J, Levack P, Scottish Cord Com-

pression Group (2007) What happens to people after malignant

cord compression? Survival, function, quality of life, emotional

well-being and place of care 1 month after diagnosis. Clin Oncol

(R Coll Radiol) 19(1):56–62

22. Sande TA, Ruenes R, Lund JA, Bruland OS, Hornslien K,

Bremnes R, Kaasa S (2009) Long-term follow-up of cancer

patients receiving radiotherapy for bone metastases: results from

a randomised multicentre trial. Radiother Oncol 91(2):261–266.

doi:10.1016/j.radonc.2009.02.014

238 J Neurooncol (2011) 104:233–238

123