Int. J. Radiation Oncology Biol. Phys., Vol. 63, No. 1, pp.

91100, 2005
Copyright 2005 Elsevier Inc.
Printed in the USA. All rights reserved
0360-3016/05/$see front matter

doi:10.1016/j.ijrobp.2005.01.009

CLINICAL INVESTIGATION

Central Nervous System

LONG-TERM SURVIVAL AND FUNCTIONAL STATUS OF PATIENTS WITH

LOW-GRADE ASTROCYTOMA OF SPINAL CORD
CLIFFORD G. ROBINSON, M.D.,* RICHARD A. PRAYSON, M.D., JOSEPH F. HAHN, M.D.,
IAIN H. KALFAS, M.D., MELVIN D. WHITFIELD, M.D., SHIH-YUAN LEE, M.S.P.H.,*
AND JOHN H. SUH, M.D.*
Departments of *Radiation Oncology, Anatomic Pathology, and Neurosurgery, Brain Tumor Institute,
Cleveland Clinic Foundation, Cleveland, Ohio
Purpose: To determine survival and changes in neurologic function and Karnofsky performance status (KPS) in a
series of patients treated for low-grade astrocytoma of the spinal cord during the past two decades.
Methods: This study consisted of 14 patients with pathologically confirmed low-grade astrocytoma of the spinal cord
who were treated between 1980 and 2003. All patients underwent decompressive laminectomy followed by biopsy (n
7), subtotal resection (n 6), or gross total resection (n 1). Ten patients underwent postoperative radiotherapy
(median total dose 50 Gy in 28 fractions). The overall survival, progression-free survival, and changes in neurologic
function and KPS were measured.
Results: The overall survival rate at 5, 10, and 20 years was 100%, 75%, and 60%, respectively. The progression-free
survival rate at 5, 10, and 20 years was 93%, 80%, and 60%, respectively. Neither overall survival nor progressionfree survival was clearly correlated with any patient, tumor, or treatment factors. Neurologic function and KPS
worsened after surgery in 8 (57%) of 14 and 9 (69%) of 13 patients, respectively. At a mean follow-up of 10.2 years,
neurologic function had stabilized or improved in 8 (73%) of 11 remaining patients, but the KPS had worsened in 5
(50%) of 10. Most patients who were employed before surgery were working at last follow-up.
Conclusion: Patients who undergo gross total resection of their tumor may be followed closely. Patients who undergo
limited resection should continue to receive postoperative RT (50.4 Gy in 1.8-Gy fractions). The functional measures
should be routinely evaluated to appreciate the treatment outcomes. 2005 Elsevier Inc.
Astrocytoma, Surgery, Radiotherapy, Spinal cord.

INTRODUCTION
Tumors arising from the spinal canal are rare. Such tumors will
account for only 15% of the approximately 18,400 primary
central nervous system tumors diagnosed in 2004 (1). Of the
primary spinal tumors, roughly one-third are intramedullary,
with 85% of these having a glial origin in adults (2). In a recent
pathology review by Miller (3), astrocytomas of the spinal cord
were the most common spinal cord tumor in children (39%)
and the second most common in adults (24%). Most astrocytomas are low grade, with 25% of adult cases and 10% of
pediatric cases demonstrating malignant histologic features (4,
5). Low-grade astrocytoma of the spinal cord (LGASC) affects
males slightly more than females and may occur at any age,
although the tumor tends to present in the first 30 years of life
(2, 6).
No consensus has been reached in the literature regarding
the optimal treatment for LGASC. Series from the past two
decades reporting on the results of conservative surgery and
radiotherapy (RT) for LGASC have demonstrated overall 5-

and 10-year survival rates in the range of 58 91% and 43

91%, respectively (6 27). One recent single-institution series
reported a 5-year overall survival rate of 100% in patients
treated with extensive tumor resection and no postoperative
RT, with a low incidence of morbidity (17). Still, the predominant form of failure is local, with failure rates ranging from
14% to 52% (7, 9, 14).
In addition to looking at measures of survival, many authors
have also recognized the importance of analyzing functional
measures, such as neurologic function and Karnofsky performance status (KPS), with most relying on one or the other (7,
8, 10 14, 17, 19, 21, 25, 28 31). These measures are particularly important in the case of a tumor whose natural history is
so prolonged and for which treatment has many recognized
potential long-term sequelae. Historically, at least one-half of
the patients can expect to maintain their baseline functional
status, with the remainder equally likely to improve or worsen
(8, 10 14, 17, 19, 21, 25, 28 31).
This review was of our own series of 14 patients with

Reprint requests to: John H. Suh, M.D., Department of Radiation Oncology, Cleveland Clinic Foundation, T28, 9500 Euclid
Ave., Cleveland, OH 44195. Tel: (216) 444-5574; Fax: (216)

444-5331; E-mail: suhj@ccf.org

Received Nov 16, 2004, and in revised form Jan 8, 2005.
Accepted for publication Jan 11, 2005.
91

92

I. J. Radiation Oncology

Biology Physics

LGASC, treated predominantly with conservative surgery

and postoperative RT, with special consideration given to
careful long-term follow-up of neurologic function and
KPS. This series differed from most in that we included
only those patients with pathologically confirmed low-grade
astrocytoma.
METHODS AND MATERIALS
Patient population
The institutional review boardapproved low-grade glioma database at the Cleveland Clinic Foundation Brain Tumor Institute
was searched to identify all patients with LGASC treated between
December 1980 and January 2003. To be included in this study,
the patients must have been followed for at least 1 year from the
time of treatment. In total, 21 patients were identified who had
been treated at our institution. On additional review of the records
and repeated review of the pathologic materials, 3 patients were
found to have high-grade tumors and 4 did not have at least 1 year
of follow-up. Of the 4 patients with insufficient follow-up, 1 died
postoperatively of gram-negative sepsis after undergoing gross
total resection (GTR). The records of the remaining 14 patients
were retrospectively reviewed, and the information was entered
into a secure database. The identifying data was removed according to Health Insurance Portability and Accountability Act
(HIPAA) protocols. Information was collected on patient age,
gender, tumor grade, symptoms at diagnosis, symptom duration,
imaging type, disease location and extent, surgery technique, RT
technique, failure site, repeat operations, and neurologic function
and KPS at each treatment and subsequent follow-up examination.
(Table 1). Additionally, any complications arising from the medical treatment, surgery, or RT of the patient were recorded.

Assessment of neurologic function and KPS

Neurologic function and KPS were assessed pre- and postoperatively, before and after RT, and at each subsequent follow-up
visit. Additionally, all surviving patients were telephoned and
surveyed with regard to their neurologic function and KPS using
an institutional review boardapproved questionnaire. Neurologic
function was assessed in accordance with the scale developed by
McCormick et al. (32). The performance status was assessed using
the Karnofsky scale (33). No KPS values were recorded for one
pediatric patient. Patients who had been employed before surgery
were also asked about their current work history.

Pathologic features
All histologic sections available for evaluation were re-reviewed
by a single neuropathologist (R.A.P.) at the Cleveland Clinic. The
tumors were graded according to the most recent World Health
Organization classification for astrocytomas (34). Only Grade I or
Grade II tumors were studied.

Volume 63, Number 1, 2005

to the superior and inferior poles of the tumor, and the cord was
inspected for the presence of discrete surgical planes. In most
cases, no such plane was encountered, and a decision was then
made to either attempt to resect the tumor or terminate the procedure. Subtotal resection (STR) and GTR were performed entirely
with standard microsurgical techniques in most patients (see Table
2). After biopsy or attempted radical resection, watertight closure
of the dura was obtained, and the fascia, subcutaneous tissue, and
skin were closed in standard fashion.
Gross total resection was defined as complete removal of the
tumor without any evidence of residual disease according to the
operative notes and any available postoperative imaging. Anything
less than a GTR was considered STR. All STRs performed in this
analysis resulted in 75% removal, with the exception of one
resection that resulted in only 25% removal.

Radiotherapy
Radiotherapy was delivered to the tumor and margin based on
the imaging results and intraoperative findings. The field length
typically encompassed the tumor and 35 cm above and below the
tumor. The width was typically taken to be 2 cm on either side of
the most lateral aspect of the vertebral bodies. In 3 cases, a
cone-down field was used for the last 3.6, 6, and 14.4 Gy. The
radiation dose was delivered via PA fields for thoracic lesions,
prescribed to the spinal cord depth, and lateral fields for cervical
lesions, prescribed relative to the midline. In two cases (both
thoracic lesions), a three-field technique was used.

Chemotherapy
No patient in this study received chemotherapy.

Statistical analysis
Overall survival (OS) was measured from the date of surgery
until death or, for living patients, the last telephone interview,
clinical examination, or imaging study. Likewise, progression-free
survival (PFS) was measured from the date of surgery until tumor
progression; patients without progression were censored on the
date of their last clinical examination or imaging study. The
Kaplan-Meier method was used to summarize the 5-, 10-, and
20-year OS and PFS (35). The logrank test was used to determine
whether a statistically significant difference existed between gender, tumor extent, presence of syrinx, resection extent, symptom
duration, and the use of RT on OS and PFS. The correlation
between age and pre- and postoperative neurologic function and
KPS and survival was estimated using the Cox proportional hazard
model. The chi-square test was used to test for statistical significance; 95% confidence intervals (CIs) were also added to the
survival rate and hazard ratio to address the variance of our
estimation because of the small sample size. All tests were two
tailed, and p 0.05 was considered statistically significant. Statistical analyses were performed using the StatView software
package, version 5.0 (SAS Institute, Cary, NC).

Surgery
All patients underwent decompressive laminectomy at the level
indicated by previous imaging using a midline approach. After
opening the dura, a midline myelotomy was made, and the remainder of the surgery was performed with the assistance of an operating microscope in all but one case. In 13 cases, a biopsy was
taken and sent for frozen section analysis, with one sent for
permanent examination only. The myelotomy was then extended

RESULTS
Clinical findings on presentation
Of the 14 patients, 7 were males. The median age at
diagnosis was 40.5 years (range, 5.277.2 years). The most
common presenting symptoms were weakness (71%), sensory disturbances (43%), pain (43%), and autonomic dys-

Low-grade spinal cord astrocytoma

function (36%). The median duration of symptoms before

diagnosis was 8 months (range, 0.1120 months). The initial
imaging diagnosis was made with myelography (3 patients,
21%), CT (1 patient, 7%), and MRI (10 patients, 71%).
The most common tumor location was thoracic (11 patients, 79%) followed by cervical (2 patients, 14%) and
cervicomedullary (1 patient, 7%). The extent of the spinal
cord tumor was measured as a function of the number of
vertebral bodies it spanned. The median tumor extent was
two vertebral bodies (range one to six). No holocord
tumors were present in this series. Six patients (43%) had a
tumor-associated syrinx. A summary of patient characteristics is presented in Table 1.
Surgical results
All 14 patients in this series underwent surgery for their
tumor. The median time to surgery from the date of the
imaging diagnosis was 38.5 days (range, 0 595 days). Surgery was performed as previously described; 7 patients
(50%) underwent biopsy, 6 (43%) underwent STR, and 1
patient (7%) underwent GTR. In all 6 patients with a tumorassociated syrinx, the fluid was drained during surgical
exploration.
Pathologic analysis revealed all 14 tumor specimens to be
low-grade astrocytoma, with 11 clearly Grade II and 3
classified as low-grade astrocytoma, not otherwise specified
owing to limited tissue sampling. None of the tumors demonstrated identifiable mitotic activity, vascular proliferation
changes, or necrosis.
Two patients (Patients 2 and 4) required repeat surgery.
In Patient 2, neurologic function began to worsen 11 years

C. G. ROBINSON et al.

93

after biopsy and RT, and repeat MRI suggested tumor

progression. GTR, however, demonstrated an old hematoma
and necrosis. In Patient 4, neurologic function began to
worsen 15 years after STR and postoperative RT, and repeat
MRI suggested tumor progression. Biopsy demonstrated
blood clot and fibrous tissue.
RT results
Ten patients received postoperative RT as a part of their
initial therapy for the tumor. No patients received any
additional RT in their lifetime. RT was delivered as previously described. The median time to the start of RT from the
date of surgery was 32 days (range, 12 87 days). The
median dose delivered was 5020 cGy (range, 4500 5400
cGy) in 28 fractions (range, 2530), with a median dose per
fraction of 180 cGy (range, 170 200 cGy). Of the 10
patients who received postoperative RT, 3 (30%) were
treated with 60Co, 1 (10%) with 4-MV photons, 5 (50%)
with 6-MV photons, and 1 (10%) with 10-MV photons.
Four patients did not receive postoperative RT. Of these
patients, 2 underwent biopsy, 1 underwent STR, and 1
underwent GTR.
Complications
Surgery. In 8 (57%) of the 14 patients, previous neurologic deficits immediately worsened or new deficits developed postoperatively. The most common complaint was
worsening of lower extremity weakness (4 patients, 29%).
The symptoms were temporary in 5 (62.5%) of the 8 patients but became permanent in the remaining 3 (37.5%).
Neurologic function worsened immediately after the sec-

Table 1. Patient characteristics

Pt. No.

Age (y)

Gender

Year

Imaging
modality

1
2
3

32
38
34

F
M
F

1980
1982
1983

Myelography
Myelography
Myelography

4
5

47
57

M
F

1984
1987

CT
MRI

6
7
8
9
10

50
41
33
40
48

M
F
F
M
F

1988
1993
1994
1998
1999

MRI
MRI
MRI
MRI
MRI

11
12

5
43

F
M

1999
2000

MRI
MRI

13

77

2001

MRI

14

24

2002

MRI

Symptoms at diagnosis
Weak LLE, numb RLE, back pain
Weak BLE, spastic LLE
Weak BLE, numb LLE, urinary
retention
Weak BLE, back pain
Weak BLE, numb R foot, pain
LLE, bowel urgency
Numb RLE
Pain in neck shoulder, vertigo
Paralysis BLE, numb T10 level
Dysesthesia BUE/LE shoulders
Weak LLE, numb LLE,
paresthesias BLE, urinary
incontinence
Weak BLE, torticollis
Weak RLE, pain in thorax at T4,
erectile dysfuntion
Weak BUE, gait difficulty, urinary
frequency/urgency
Pain in back RLE

Symptom
duration
(mo)

Location

Extent

Syrinx

13
0.1
24

T9T10
T5T9
T9T11

2
5
3

No
Yes
No

1
36

T10T12
T11

3
1

No
Yes

18
7
1
3
120

T6T7
CM
T8T11
C3C4
T12L1

2
1
4
2
2

Yes
No
No
No
Yes

6
48

T1T6
T5T7

6
3

Yes
Yes

C4C5

No

T10

No

Abbreviations: Pt. No. patient number; B bilateral; L left; R right; UE upper extremity; LE lower extremity; CM
cervicomedullary.

Pt.
No.

Time to
surgery (d)

Extent

Field

Dose (cGy), fraction,

dose/Fx (cGy)

34

PA

5000, 25, 200

24

PA

5040, 28, 180

29

PA

4930, 29, 170

60

Co

30

ND

4915, 26, 189

60

Co

frozen,

53

PA

4500, 25, 180

10 MV

frozen,

13

PA

5400, 30, 180

6 MV

12
No RT

PA

5400, 30, 180

6 MV

87

RPO/LPO/PA

4500, 25, 180

6 MV

Lateral
LPO/RPO/AP

5400, 27, 200

5400, 30, 180

6 MV
6 MV

STR

Microscope

Bx

Frozen

STR

Microscope, frozen

71

Bx

Microscope, frozen

STR (25%)

57

Bx

595

Bx

8
9

40
44

Bx
Bx

Microscope,
laser.
Microscope,
SEP
Microscope,
SEP
Microscope,
Microscope,

10

23

STR

11

GTR

12

37

STR

13
14

112
155

STR
Bx

GTR at 11 y

frozen,
frozen
frozen

Microscope, frozen,
US
Microscope, frozen,
SEP, CUSA, US
Microscope, frozen,
SEP/MEP
Microscope, frozen
Microscope, frozen

Bx at 15 y

New dysesthesia BLE, loss

of proprioception LLE
1st surgery: Worsening
weakness BLE; 2nd
surgery: E. coli
bacteremia
Worsening weakness BLE

1st surgery: Worsening

weakness BLE; 2nd
surgery: Permanent
worsening of motor
function

New weakness BUE

Wound dehiscence, CSF
leak
Worsening numbness BLE
and paresthesias BLE

60

Complications

Co

Multiple rib fractures

4 MV

Worsening weakness
BLE
Multiple vertebral
body collapse in
surgical/radiation
site

No RT

Volume 63, Number 1, 2005

Energy

Biology Physics

Complications

No RT
New neurogenic bladder,
neurogenic bowel, and
LE DVT
Worsening weakness BLE

No RT
63
53

I. J. Radiation Oncology

Time to
RT (d)

Surgical technique

Second
surgery

94

Table 2. Treatment and complications summary

Decadron induced
AVN of both hips
and both ankles

Abbreviations: Bx biopsy; STR subtotal resection; GTR gross total resection; frozen frozen section analysis at surgery; SEP sensory evoked potential monitoring; MEP
motor evoked potential monitoring; CUSA Cavitron ultrasound aspirator; US intraoperative ultrasound guidance; ND not determined; RPO right posterior oblique; LPO left
posterior oblique; DVT deep venous thrombosis; other abbreviations as in Table 1.

Low-grade spinal cord astrocytoma

ond surgery in both patients who required repeat surgery for

putative tumor recurrence; in both cases, symptoms were
presented on last follow-up. Additional postoperative complications developed in 2 (14%) of the 14 patients that were
unrelated to their neurologic status and included wound
dehiscence, cerebrospinal fluid leak, lower extremity deep
venous thrombosis, and gram-negative sepsis.
Radiotherapy. All patients experienced some degree of
skin discoloration in the radiation field, although none experienced skin breakdown. In most patients, neurologic
function improved or stabilized shortly after the administration of RT. However, 1 patient developed complications
during RT, and 2 patients developed complications more
than one decade later. Patient 2 reported worsening bilateral
lower extremity weakness shortly after RT was initiated,
which gradually improved in the months after treatment.
The same patient later underwent GTR for worsening neurologic function and putative tumor progression on MRI, as
described previously. He had originally received 5040 cGy
in 28 fractions via a single PA field using 4-MV photons.
After the second surgery, his neurologic symptoms progressively worsened, and he died 5 years later of complications
related to paralysis. Patient 4 also underwent a second
surgery as a result of a similar presentation. However, he
was still alive 4 years later. He had originally received 4915
cGy in 26 fractions via an unknown field using 60Co.

C. G. ROBINSON et al.

95

available for follow-up at a mean of 10.2 years (range,

1.323.4 years) after surgery.
The OS rate at 5, 10, and 20 years was 100% (95% CI,
100 100%), 75% (95% CI, 45100%), and 60% (95% CI,
24 96%), respectively. The PFS rate at 5, 10, and 20 years
was 93% (95% CI, 79 100%), 80% (95% CI, 53100%),
and 60% (95% CI, 20 99%; Fig. 1a,b), respectively. No
statistically significant correlation was found with OS or
PFS for patient gender or age, presence of syrinx, tumor
extent, resection extent, symptom duration, RT use, or preor postoperative KPS or neurologic function. However, for
patients undergoing biopsy vs. more extensive resection, a
trend toward significance at 10 years for both OS (60% vs.
100%, p 0.0979) and PFS (60% vs. 100%, p 0.0746)
was found.
Failure patterns
Four patients developed MRI-diagnosed tumor progression at a median of 9.2 years (range, 0.9 15.7 years) after
surgery. Two of the four underwent repeat surgery at 11 and
15.7 years after their initial surgery for imaging-diagnosed

Other
Some complications were not obviously attributable to
either surgery or RT, but were likely exacerbated by both. In
Patient 3, multiple vertebral bodies within the operated and
radiated field started to collapse 18 years after treatment.
Likewise, not all complications came as a direct result of
surgery or RT. Patient 14 underwent biopsy and postoperative RT for a T10 tumor. His neurologic status remained
stable on steroids. However, several attempts to wean him
from the steroids resulted in worsening pain. Although he
was eventually tapered off the steroids, his hip and then
ankle pain worsened. He was later diagnosed with avascular
necrosis of both hips and ankles. Although at last follow-up,
he showed no signs of residual neurologic deficit, he had
chronic hip and ankle pain that had only recently come
under control with the use of multiple narcotics. Another
patient (Patient 1) recently began to have recurrent rib
fractures 23 years after receiving 5000 cGy in 25 fractions
via a single PA field using 60Co energy. A review of the
original port films indicated the affected ribs were not
directly in, or adjacent to, the radiation field.
A summary of the treatment and resulting complications
is given in Table 2.
Survival analysis
Three patients died during the period of analysis, with a
median time to death of 9.1 years (range, 6.116.0 years).
The cause of death was attributable to tumor progression in
Patients 6 and 8 and radiation necrosis in Patient 2. The
remaining 11 patients were alive at analysis, and all were

Fig. 1. Kaplan-Meier survival curves for (a) overall survival (OS)

and (b) progression-free survival (PFS) for all patients with lowgrade astrocytoma of the spinal cord. Survival data for all patients,
with 5-, 10-, and 20-year survival rates reported and 95% confidence intervals in parentheses.

96

I. J. Radiation Oncology

Biology Physics

recurrence. Neither patient showed evidence of recurrent

tumor in the second specimen. For the 2 remaining patients
who did not undergo pathologic confirmation of their progression, both died 1.7 and 5.2 years later of complications
related to worsening neurologic function, and thus they
were considered true failures. Patient 6 developed recurrence completely within the original tumor volume and
Patient 8 did so both locally and three vertebral levels above
the original site. Thus, the local control rate at 5 and 10
years for our series was 89% and 71%, respectively. A
summary of patient failure data is given in Table 3.
Neurologic function and performance status analysis
Of the original 14 patients, 11 were alive at analysis, and
all were interviewed; their neurologic function and KPS
were reviewed as described above. The median preoperative
neurologic function was two (range, one to four), and the
median preoperative KPS was 80 (range, 60 90). At follow-up, the median neurologic function was one (range, one
to four), and the median KPS was 75 (range, 60 90).
During the immediate postoperative period, when all 14
patients were alive, the neurologic function had worsened in
8 (57%) of 14 patients, and the KPS worsened in 9 (69%) of
13 patients. The most common neurologic complaint was
worsening of sensory deficits, and the most common complaint overall was loss of ability to perform normal activities
of daily living. In contrast, after RT, 10 (100%) of 10 and 9
(90%) of 10 patients experienced stable or improved neurologic function and KPS, respectively. Compared with the
immediate postoperative period, at last follow-up, 11

Volume 63, Number 1, 2005

(100%) of 11 and 8 (80%) of 10 patients experienced stable

or improved neurologic function and KPS, respectively.
These numbers are somewhat misleading, however, because
the preoperative neurologic function had stabilized or improved in 8 (67%) of 11 patients and KPS had worsened in
5 (50%) of 10 patients. This discrepancy resulted because
several patients developed significant worsening of neurologic function after surgery and never improved. For those
3 patients with persistent worsening of neurologic function,
2 ultimately worsened by 1 point on the McCormick scale,
and 1 by 3 points. Likewise, for the 5 patients with worsening KPS, the median change was 20 points, with only 1
patient worsening by 30 points. The most common reason
for worsening KPS was a declining ability to perform normal activities secondary to pain. In most cases, this pain was
similar in character to their presenting pain, but increased in
intensity and/or distribution. In a notable exception, the 1
patient who developed steroid-induced avascular necrosis of
the hips and ankles experienced a 30-point decline in KPS
in the face of stable neurologic function.
Of the 8 patients whom we queried about work history, 6
continued to work a median of 5 years (range, 116 years)
after treatment. Of the remaining 2 patients, 1 worked for 17
years before stopping secondary to worsening pain. Only 1
patient unable to work immediately after treatment secondary to pain and weakness. Of the 11 patients who were alive
at analysis, 7 were ambulating without assistance, 3 required a cane or walker, and 1 required a wheelchair. A
summary of the neurologic function and KPS measures is
presented in Table 4.

Table 3. Summary of survival and failure data

Pt. No.

Survival
(y)

1
2

16.0

Cause of
death

Time to
failure (y)

Time to death
from failure
(y)

Failure location

Comments

23.4

3
4

Necrosis

Progression by imaging at
11 y; GTR findings of
necrosis and old
hematoma

21.3
19.9

5
6

6.1

7
8

9.1

9
10
11
12
13
14

Follow-up
(y)

Progression by imaging at
15 y; Bx findings of
blood clot and fibrous
tissue

16.4
Tumor

0.9

5.2

Local

Progression by imaging
no biopsy

Tumor

7.4

1.7

Local plus 3 levels

above original level

Progression by imaging
no biopsy

9.5
5.7
4.6
4.4
3.9
2.3
1.3

Abbreviations: GTR gross total resection; Bx biopsy.

Table 4. Summary of neurologic function and performance status

KPS

Neurologic function*

Preoperative

Postoperative

Before
RT

After
RT

Follow-up
(mo)

Change
(preoperative
to follow-up)

Postoperative

Before
RT

After
RT

Follow-up
(mo)

1
2
3
4

80
80
80
90

50
50
60
70

60
70
70
80

70
80
70
80

80
Dead
60
70

0
Dead
20
20

2
3
3
1

4
4
4
4

3
3
3
3

3
3
3
2

2
Dead
3
4

0
Dead
0
3

80

80

90

90

90

10

6
7
8
9
10
11
12
13
14

80
90
60
70
70
NA
80
80
90

70
80
40
70
70
NA
70
50
90

70
None
50
None
70
None
None
50
90

70
None
50
None
70
None
None
80
70

Dead
90
Dead
90
90
NA
60
70
60

Dead
0
Dead
20
20
NA
20
10
30

2
1
4
2
1
4
2
2
1

2
2
4
2
2
3
3
3
1

2
None
4
None
2
None
None
4
1

2
None
4
None
1
None
None
3
1

Dead
1
Dead
1
1
1
3
3
1

Dead
0
Dead
1
0
3
1
1
0

NA (homemaker)
NA (dead)
No, 17 y, pain
No, 0 y, pain
weakness
Yes, 16 years to
date
NA (dead)
Yes, 10 y to date
NA (dead)
Yes, 6 y to date
Yes, 4 y to date
NA (child)
Yes, 4 y to date
NA (retired)
Yes, 1 y to date

C. G. ROBINSON et al.

Preoperative

Working now? How

many years after
surgery? Reason
stopped?

Low-grade spinal cord astrocytoma

Pt.
No.

Change
(Preoperative
to follow-up)

Abbreviations: KPS Karnofsky performance status; RT radiotherapy.

* McCormick scale.

97

98

I. J. Radiation Oncology

Biology Physics

DISCUSSION
We have reported on the survival and changes in functional status of patients with pathologically confirmed
LGASC who were treated predominantly with limited resection and postoperative RT. This series differs from most
in that we included only those patients with pathologically
confirmed low-grade astrocytomas. Our patient population
was otherwise similar to the others described in the literature with regard to patient, tumor, and treatment factors.
The use of adjuvant RT for spinal cord astrocytomas is
justified primarily because local recurrence remains the
primary pattern of treatment failure. Furthermore, considerable evidence has shown that patients undergoing STR for
intracranial astrocytoma benefit from adjuvant RT, thus
implying a certain radioresponsiveness for this tumor. This
implication has been further bolstered by evidence of a
doseresponse relationship; radiation doses 40 Gy have
been associated with improved local control (36). However,
no evidence has suggested additional local control or survival benefit with doses 50.4 Gy (15, 37). Thus, the
typical radiation doses given for spinal cord astrocytomas
are 4550.4 Gy in 1.8-Gy fractions.
In our study, the 5- and 10-year OS rate was 100% and 75%
and 5- and 10-year PFS rate was 93% and 80%, respectively,
for patients with LGASC. These results compare favorably
with other reported survival rates for patients undergoing limited resection and RT. When that body of the literature was
reviewed carefully, and the data relating only to LGASC were
extracted, we found OS rates of 6191% and 4391% at 5 and
10 years, respectively (6 13, 15, 16, 37). Thus, our results
confirm that for patients with low-grade astrocytoma, biopsy or
STR followed by RT can produce long-term OS and PFS.
Our OS and PFS rates were also equivalent to those
reported in a number of recent surgical series dealing with
LGASC. Because of the advances in technology that have
occurred during the past two decades within the neurosurgical field, surgeons can now perform more extensive spinal
tumor resections without the high morbidity previously
associated with such procedures. Several authors have recently reported 5-year OS rates of 72 89% in adults and
80 100% in children who underwent extensive resection
without adjuvant RT. These data are, overall, less mature
than those reported for limited resection and postoperative
RT, although the results are promising. Thus, many authors
now recommend that whenever possible, patients undergo
radical resection for LGASC without postoperative RT (10,
1719, 21, 22, 24, 25, 28, 29, 31, 38). These data must be
interpreted carefully, however, because anywhere from 4%
to 53% of patients even those in these surgical series
generally receive RT at some point in their treatment (17
19, 28 30, 39). In those patients who might safely undergo
complete tumor resection, a watchful waiting strategy may,
in fact, be appropriate.
Although both surgery and RT have recognized shortand long-term risks associated with their use, the reporting
of complication rates for LGASC has been inconsistent. In

Volume 63, Number 1, 2005

our series, 57% of the patients reported that their neurologic

function had worsened after surgery and 14% developed
other surgical complications. Of these patients, 38% had
persistent worsening of neurologic function at last followup. Our reported surgical complication rate was greater than
the range of 513% reported elsewhere in the literature.
However, many studies did not document worsening neurologic function as a complication (17, 19, 2325, 27, 28,
30, 31, 39).
In those patients who received postoperative RT in our
series, 2 patients underwent repeated surgery for suspected
tumor recurrence after prior surgery and postoperative RT.
Both had progressive decline in neurologic function and
imaging evidence of tumor progression. Neither had evidence of recurrent tumor on repeat surgery, but were instead
found to have clear evidence of necrosis in one and scarring
in the other. Both patients had received typical radiation
doses and fractions of 50 Gy in 28 fractions and 49 Gy in 26
fractions. The patient with necrosis on biopsy developed
progressively worsening neurologic function and ultimately
died of complications related to paralysis, and, thus, that
patients case was deemed true radiation myelopathy. The
patient with scarring on biopsy was alive at last follow-up,
with poor lower extremity motor function that developed
immediately after his second surgery, and has remained
stable. Therefore, our rate of radiation myelopathy was 7%
(1 of 14). The incidence of myelopathy as reported elsewhere is dose related, with rates reported using typical
fractions from 0.2% to 0.5% after 50 Gy to 1% to 5% after
60 Gy (24, 40, 41). Although the tolerance dose producing
5% of a specific radiotoxic effect within 5 years after
treatment (TD 5/5) using conventional fractionation has
been traditionally reported as 4550 Gy (42, 43), reviews of
human data have suggested that doses of 45 Gy with conventional fractionation results in almost no myelopathy (44,
45). A better estimate for a TD 5/5 risk of myelopathy with
conventional fractionation may be closer to 57 61 Gy (44,
45), suggesting that the spinal cord may be irradiated safely
to the accepted doses of radiation tolerance and even greater
(up to 55 Gy) without an increased complication risk (14,
15, 46). Although the general rule is that a dose reduction
should accompany increases in target volume, little clinical
data have supported this for the spinal cord (47, 48). No
second malignant tumors were reported in our series. One
possible method to reduce the complications due to RT
would be to implement more conformal techniques. Because astrocytomas, even low-grade varieties, are frequently
infiltrative to some extent, the radiation field has been
traditionally set to encompass the gross tumor volume and a
margin that inevitably includes normal cord. Thus, even
with conformal techniques, the normal cord immediately
surrounding the tumor will not be spared. However, conformal techniques will spare a larger proportion of other
surrounding normal tissue, thereby reducing the rate of
overall radiation toxicity to other systems such as the GI
tract, kidney, and soft tissue.
Most authors have reported at least some functional status

Low-grade spinal cord astrocytoma

measure in their analysis. The most commonly reported

functional measures are neurologic function, typically measured with the McCormick or Cooper and Epstein scales,
and performance status, using the Karnofsky scale (32, 33,
49). In a complete review of the RT and surgical literature
for spinal cord astrocytoma, we found that the median
reported rate of stable or improved functional status at last
follow-up was 73% (range, 2791%), with no observable
difference between the different treatment modalities (7, 8,
10 13, 17, 19, 21, 25, 28 31, 37). We reported comparable
rates from our series, with stable or improved neurologic
function in 73% and KPS in 50%. The reasons why a high
percentage of our patients had worsened KPS at last follow-up are unclear. Compared with earlier series with
greater mortality rates, our patients may have lived long
enough to develop long-term consequences of extensive
surgery with RT, which can result in vertebral body collapse, spinal deformity, and so forth. The vast majority of
studies, our own included, have reported on patients treated
with older surgical and RT techniques. With the use of
modern neurosurgical and conformal RT methods, many of
these long-term side effects may be avoided.
As an additional quality-of-life measure, we asked those
living patients who were employed before surgery whether
they were still able to work. Surprisingly, even in the face of
overall worsening KPS, fully 75% of the patients continued
to work, with 1 patient working 17 years before finally

C. G. ROBINSON et al.

99

retiring secondary to pain, and only 1 patient unable to work

postoperatively. Even with such issues as pain, motor or
sensory disturbances, and other chronic problems related to
their tumor or treatment, most patients in this series were
able to adapt and continue to function at a relatively high
level. Thus, only with the routine application of multiple
functional measurements can the effects of our treatments
truly be measured.
CONCLUSION
On the basis of the most recent literature and our own
experience, we believe it is reasonable to watch closely
patients with LGASC in whom GTR (confirmed with postoperative imaging) has been performed. In patients who
have undergoing anything less than GTR, sufficient evidence has shown that long-term OS, PFS, neurologic function, and KPS can be improved with the addition of RT. For
these patients, adjuvant RT to 50 Gy in 1.8-Gy fractions
should be delivered, preferably with conformal techniques
to minimize the dose to the normal surrounding structures.
Functional status measurements and complications should
be meticulously recorded to elucidate further the relative
risks and benefits of the various treatment options. To
determine the optimal treatment of these rare tumors, a
multi-institutional analysis of only those patients with documented LGASC, treated with modern neurosurgical and
RT techniques, should be undertaken.

REFERENCES
1. American Cancer Society. Cancer facts and figures 2004.
Atlanta: American Cancer Society; 2004.
2. Winn HR, Youmans JR. Youmans neurological surgery. 5th
ed. Philadelphia: WB Saunders; 2004.
3. Miller DC. Surgical pathology of intramedullary spinal cord
neoplasms. J Neurooncol 2000;47:189 194.
4. Malis LI. Intramedullary spinal cord tumors. Clin Neurosurg
1978;25:512539.
5. Epstein FJ, Farmer JP. Pediatric spinal cord tumor surgery.
Neurosurg Clin North Am 1990;1:569 590.
6. Minehan KJ, Shaw EG, Scheithauer BW, et al. Spinal cord
astrocytoma: Pathological and treatment considerations. J Neurosurg 1995;83:590 595.
7. Lee HK, Chang EL, Fuller GN, et al. The prognostic value of
neurologic function in astrocytic spinal cord glioma. Neurooncol 2003;5:208 213.
8. 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.
9. McLaughlin MP, Buatti JM, Marcus RB Jr, et al. Outcome
after radiotherapy of primary spinal cord glial tumors. Radiat
Oncol Invest 1998;6:276 280.
10. Nishio S, Morioka T, Fujii K, et al. Spinal cord gliomas:
Management and outcome with reference to adjuvant therapy.
J Clin Neurosci 2000;7:20 23.
11. Jyothirmayi R, Madhavan J, Nair MK, et al. Conservative
surgery and radiotherapy in the treatment of spinal cord astrocytoma. J Neurooncol 1997;33:205211.
12. Shirato H, Kamada T, Hida K, et al. The role of radiotherapy

13.

14.

15.

16.

17.

18.

19.
20.

21.

22.

in the management of spinal cord glioma. Int J Radiat Oncol

Biol Phys 1995;33:323328.
Hulshof MC, Menten J, Dito JJ, et al. Treatment results in
primary intraspinal gliomas. Radiother Oncol 1993;29:294
300.
Kopelson G, Linggood RM. Intramedullary spinal cord astrocytoma versus glioblastoma: The prognostic importance of
histologic grade. Cancer 1982;50:732735.
Linstadt DE, Wara WM, Leibel SA, et al. Postoperative radiotherapy of primary spinal cord tumors. Int J Radiat Oncol
Biol Phys 1989;16:13971403.
Cooper PR. Outcome after operative treatment of intramedullary spinal cord tumors in adults: Intermediate and long-term
results in 51 patients. Neurosurgery 1989;25:855 859.
Jallo GI, Danish S, Velasquez L, et al. Intramedullary lowgrade astrocytomas: Long-term outcome following radical
surgery. J Neurooncol 2001;53:61 66.
Innocenzi G, Salvati M, Cervoni L, et al. Prognostic factors in
intramedullary astrocytomas. Clin Neurol Neurosurg 1997;99:
15.
Epstein FJ, Farmer JP, Freed D. Adult intramedullary astrocytomas of the spinal cord. J Neurosurg 1992;77:355359.
Constantini S, Houten J, Miller DC, et al. Intramedullary
spinal cord tumors in children under the age of 3 years.
J Neurosurg 1996;85:1036 1043.
Constantini S, Miller DC, Allen JC, et al. Radical excision of
intramedullary spinal cord tumors: Surgical morbidity and
long-term follow-up evaluation in 164 children and young
adults. J Neurosurg 2000;93:183193.
Merchant TE, Kiehna EN, Thompson SJ, et al. Pediatric

100

23.
24.
25.
26.
27.
28.
29.

30.
31.
32.
33.
34.
35.
36.

I. J. Radiation Oncology

Biology Physics

low-grade and ependymal spinal cord tumors. Pediatr

Neurosurg 2000;32:30 36.
Bouffet E, Pierre-Kahn A, Marchal JC, et al. Prognostic
factors in pediatric spinal cord astrocytoma. Cancer 1998;83:
23912399.
OSullivan C, Jenkin RD, Doherty MA, et al. Spinal cord
tumors in children: Long-term results of combined surgical
and radiation treatment. J Neurosurg 1994;81:507512.
Przybylski GJ, Albright AL, Martinez AJ. Spinal cord astrocytomas: Long-term results comparing treatments in children.
Childs Nerv Syst 1997;13:375382.
Reimer R, Onofrio BM. Astrocytomas of the spinal cord in
children and adolescents. J Neurosurg 1985;63:669 675.
Rossitch E Jr, Zeidman SM, Burger PC, et al. Clinical and
pathological analysis of spinal cord astrocytomas in children.
Neurosurgery 1990;27:193196.
Cristante L, Herrmann HD. Surgical management of intramedullary spinal cord tumors: Functional outcome and sources of
morbidity. Neurosurgery 1994;35:69 74.
Brotchi J, Dewitte O, Levivier M, et al. A survey of 65 tumors
within the spinal cord: Surgical results and the importance of
preoperative magnetic resonance imaging. Neurosurgery 1991;
29:651 657.
Guidetti B, Mercuri S, Vagnozzi R. Long-term results of the
surgical treatment of 129 intramedullary spinal gliomas.
J Neurosurg 1981;54:323330.
Goh KY, Velasquez L, Epstein FJ. Pediatric intramedullary
spinal cord tumors: Is surgery alone enough? Pediatr Neurosurg 1997;27:34 39.
McCormick PC, Post KD, Stein BM. Intradural extramedullary tumors in adults. Neurosurg Clin N Am 1990;1:591 608.
Karnofsky D, Abelman W, Craver L, et al. The use of nitrogen
mustards in the palliative treatment of carcinoma. Cancer
1948;1:634 656.
Kleihues P, Cavenee WK, et al., for the International Agency
for Research on Cancer. Pathology and genetics of tumours of
the nervous system. Lyon: IARC Press; 2000.
Kaplan E, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457 448.
Garcia DM. Primary spinal cord tumors treated with surgery

Volume 63, Number 1, 2005

37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.

49.

and postoperative irradiation. Int J Radiat Oncol Biol Phys

1985;11:19331939.
Kopelson G, Linggood RM, Kleinman GM, et al. Management of intramedullary spinal cord tumors. Radiology 1980;
135:473 479.
Abdel-Wahab M, Corn B, Wolfson A, et al. Prognostic factors
and survival in patients with spinal cord gliomas after radiation therapy. Am J Clin Oncol 1999;22:344 351.
Samii M, Klekamp J. Surgical results of 100 intramedullary
tumors in relation to accompanying syringomyelia. Neurosurgery 1994;35:865 873.
Rampling R, Symonds P. Radiation myelopathy. Curr Opin
Neurol 1998;11:627 632.
Marcus RB Jr, Million RR. The incidence of myelitis after
irradiation of the cervical spinal cord. Int J Radiat Oncol Biol
Phys 1990;19:3 8.
Phillips TL, Buschke F. Radiation tolerance of the thoracic
spinal cord. Am J Roentgenol Radium Ther Nucl Med 1969;
105:659 664.
Wara WM, Phillips TL, Sheline GE, et al. Radiation tolerance
of the spinal cord. Cancer 1975;35:1558 1562.
Schultheiss TE, Stephens LC. The pathogenesis of radiation
myelopathy: Widening the circle. Int J Radiat Oncol Biol Phys
1992;23:1089 1093.
Schultheiss TE, Kun LE, Ang KK, et al. Radiation response of
the central nervous system. Int J Radiat Oncol Biol Phys
1995;31:10931112.
Chun HC, Schmidt-Ullrich RK, Wolfson A, et al. External
beam radiotherapy for primary spinal cord tumors. J Neurooncol 1990;9:211217.
Schultheiss TE, Stephens LC, Ang KK, et al. Volume effects
in rhesus monkey spinal cord. Int J Radiat Oncol Biol Phys
1994;29:6772.
van den Aardweg GJ, Hopewell JW, Whitehouse EM. The
radiation response of the cervical spinal cord of the pig:
Effects of changing the irradiated volume. Int J Radiat Oncol
Biol Phys 1995;31:5155.
Cooper PR, Epstein F. Radical resection of intramedullary
spinal cord tumors in adults: Recent experience in 29 patients.
J Neurosurg 1985;63:492 499.