0360-3016/82/050925-05$03.00/O Copyright 0 1982 Pcrgamon Press Ltd.

??Brief Communication

RADIATION TOLERANCE OF THE SPINAL CORD PREVIOUSLY-DAMAGED BY TUMOR AND OPERATION: LONG TERM

NEUROLOGICAL IMPROVEMENT AND TIME-DOSE-VOLUME RELATIONSHIPS AFTER IRRADIATION OF INTRASPINAL GLIOMAS

GENE KOPELSON, M.D. Department of Radiation Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA

021 I4

Of 26 patients with intramedullary spinal cord gliomas (9 astrocytomas, 5 glioblastomas, 12 ependymomas) seen at the Massachusetts General Hospital from 1962-1980, 24 were irradiated (21 initially and 3 after post-surgical recurrence). Those 19 patients who survived at least 1 year after completion of irradiation were evaluated for post-irradiation neurological changes. No patient developed radiation myelopathy. Return to a permanently and completely normal neurological status occurred for 33/51 (65%) of pre-irradiation neurological deficits. The major cause of post-irradiation neurological deterioration was tumor recurrence.

Although 18/19 patients had their thoracic or lumbar spinal cords irradiated, each with field sizes greater than 10 cm, spinal cord doses approaching, equalling, or occasionally exceeding various definitions of spinal cord tolerance were tolerated well without evidence of radiation myelopathy. Spinal cords of patients with intramedullary gliomas, often with major neurological deficits prior to irradiation, may be treated safely to doses approaching or equalling spinal cord tolerance levels. These doses are expected to locally control most ependymomas and astrocytomas without an increased risk of radiation myelopathy. Caution should be observed if doses higher than this are contemplated in an attempt to cure a glioblastoma, because the 5 % tolerance level of the damaged spinal cord remains to be defined.

Spinal cord neoplasms, Spinal cord-radiation effects.

INTRODUCTION

Most analyses of the radiation tolerance of the spinal cord

have centered on patients with non-spinal cord neoplasms in whom the spinal cord was irradiated because it was within the treated volume. Patients with intraspinal glio- mas have neurological defects as a result of their tumor, and the effects of surgery and/or irradiation. Some radiation therapists may be reluctant to treat these patients’ tumors to tumoricidal doses fearing excessive post-irradiation complications in a spinal cord possibly at increased radiosensitivity as a result of the prior damaged status of the cord.

The present series addresses the issue of long-term neurological functioning in patients with intraspinal glio- mas who received irradiation, and determines doses approaching or equalling spinal cord tolerance which may be safely given to these patients without an increased risk of radiation myelopathy.

METHODS AND MATERIALS

From 1962 to 1980 at the Tumor Registry of the Massachusetts General Hospital, 26 patients had intra- medullary spinal cord gliomas (9 astrocytomas, 5 glio- blastomas, 12 ependymomas). Of 24 who were irradiated (21 as part of initial management and three after post- operative recurrence), I9 patients who survived at least I year after completion of irradiation were selected for review. A minimum follow-up period of one year was selected because chronic radiation myelopathy usually begins at approximately I year post-irradiation.2.3.n,9 Of the five irradiated patients who were excluded because of less than one year post-irradiation follow-up, two returned abroad and were lost to follow-up at two and three months respectively, one died at IO months of arteriographically-documented cerebrovascular acci- dents, one died of intracranial subarachnoid tumor dis- semination at five months,’ and one is alive without

Reprint requests to: Dept. of Radiation Oncology, Salem Hospital, 81 Highland Ave., Salem, MA 01970. Acknowledgements-The author would like to thank Mindy

925

and Barry P. Kopelson for their help in the preparation of this paper.

Accepted for publication 4 December 198 1.

926 Radiation Oncology 0 Biology 0 Physics May 1982, Volume 8, Number 5

Table I. General clinical characteristics

No. males:females Age

Astrocytoma Glioblastoma Ependymoma

412 0:2 I:4

Range 6-69 18834 Median 39 26

Extent of prior surgery Biopsy 4 2 Subtotal resection 2 0 Total resection 0 0

Tumor location Cervical 0 I Cervico-thoracic 2 I Thoracic 0 0 Thoraco-lumbar 3 0 Lumbo-sacral 0 0 Diffuse I* 0

Irradiation volume Primary tumor + margins 5 2 Whole spinal cord

(2 boost) I 0 Craniospinal

(t boost) 0 0 Field length (cm.):

Range I2250 15-32 Median 23 24

*Cervico-thoraco-lumbar + conus medullaris. +Thoraco-lumbar + cauda equina in one patient, and cervico-thoraco-lumbar in another.

I l-61 26

I 9 I

0 I 3 0 5 2+

8

2

I

12-62 25

evidence of tumor at six months. Each of these five patients had improved or static neurological changes at last follow-up, and none had evidence of progressive early (less than I year post-irradiation) radiation myelopathy.

For the I9 patients analyzed in this report, detailed neurological examinations were reviewed immediately prior to initiation of irradiation, the subsequent neurologi- cal course, and most recent follow-up status. The general clinical characteristics and management appear in Table I.

For each patient the spinal cord dose (with appropriate corrections for Roentgen-to-rad and orthovoltage-to- megavoltage) was recorded along with the number of fractions and elapsed days. Because of the recent empha- sis on the importance of the number of fractions as it affects neural tissue tolerance,“.“.‘4 scattergrams were drawn using three recently-proposed modifications of time-dose-fractionation relationships for neurological tis- sue: Equivalent Dose (ED),14 Brain Tolerance Unit (btu)” and Neurological-NSD (Neuret),13 Fig. I. The standard spinal cord tolerance of 5000 rad/25 fractions/S weeks was converted into ED, btu, and Neuret for com- parison, as was the interpolated 5% tolerance level for ED from the data of Wara et ~1.‘~ (Fig. I). In addition, because most patients had tumors involving thoracic and lumbar segments which are felt to be more sensitive to irradiation,9.‘2 and because each patient had a field size greater than IO cm, two 10% dose reductions from the above standard tolerance level were made to yield a lower spinal cord tolerance level of 4000 rad/20 fractions/4

weeks, which also was converted into ED, btu, and Neuret (Fig. I).

RESULTS

Neurological functioning The majority of pre-irradiation neurological deficits

returned to normal permanently or improved perma- nently post-irradiation (Table 2). As explained above, post-irradiation neurological deterioration in the early post-irradiation period (ie, the five excluded patients) was not due to radiation myelopathy in any case. The major cause of post-irradiation neurological deterioration was tumor recurrence (Table 3). No patient demonstrated any evidence of chronic radiation myelopathy; each patient survived past the peak time-period incidence of radiation myelopathy (l-2 years post-irradiation2~3*8~9). Long-term normal neurological functioning was achieved for many patients (Table 4).

Time-dose-fractionation relationships When the doses for the patients are converted into

fractionation regimens which emphasize fraction number (Fig. I), even patients whose spinal cord doses equalled, approached, or exceeded spinal cord tolerance did not demonstrate radiation myelopathy. Specifically, when using the ED scheme, two patients who exceeded stan- dard tolerance (four using the level of Wara et ~1.‘~) and four patients who exceeded the lower tolerance level (corrected for field size and anatomic segment [see above]) did not develop myelopathy (Fig. 1). Similarly,

Radiation tolerance of the damaged spinal cord 0 G. KOPELSON 921

1100

iOO0

900

800

EQUIVALENT BRAIN TOLERANCE NEUROLOGICAL- NSD DOSE UNIT

D x ,,-0.3777 x T-O.056 Dx~ -0.45 x T-0.03 DxN-0.44 x T-0.06

I I r 1 I I

: - Wara et al

WO ED)

L

:

600 0

8 ??

i!firFz Y L

e 0

0

m PROPOSED TOLERANCE LEVELS (See Text)

A ONE PATIENT RE-IRRADIATED

Fig. I. Scattergrams depicting Equivalent Dose (ED), Brain Tolerance Unit (btu), and Neuret for all patients. Both standard spinal cord tolerance (5000 rad/25 fractions/5 weeks) and this dose corrected downward 10% for large field sizes and downward an additional 10% for thoracic/lumbar segments (4000 rad/20 fractions/4 weeks) have been converted into ED, btu, and Neuret values for comparison. In addition, the interpolated 5% tolerance level of the thoracic spinal cord using ED from the data of Wara et ~1.‘~ is shown.

for the btu scheme, two who exceeded standard tolerance and seven who exceeded the lower tolerance level did not develop myelopathy. For the Neuret scheme, three patients who exceeded the standard level and six who exceeded the lower tolerance level did not develop my- elopathy.

DISCUSSION The generally-accepted spinal cord tolerance level (5%

incidence of chronic myelopathy at five years) of 5000 rad/25 fractions/5 weeks can be modified by: the specific segment irradiated (ie, an increased sensitivity of the thoracicx,” and lumbar9 segments), the length of irra-

diated cord,3 alternate fractionation regimens,‘~2~“~‘4 hyperbaric oxygen,* and hyperthermia.4

Patients with intraspinal gliomas have had their spinal cords damaged by the tumor itself and by the biopsy/ resection procedures’ (Table 2). These pre-irradiation neurological deficits are a result of not only the direct tumor and/or the surgical destruction of normal and abnormal spinal cord tissue, but also the indirect effects upon nutrient arterioles which feed adjacent normal spinal cord. The pathogenesis of chronic radiation myelo- pathy is also felt to be both a direct neurotoxic effect and an indirect effect mediated by the obliterative irradiation- induced changes in the vasculoconnective stroma.2.3.s

Table 2. Post-irradiation changes in neurological status

Type of deficit*

Permanent return to

completely normal

Permanent Permanent improvement no change Deterioration’!

Motor 12119 2119 O/l9 5119 Sensory IO/IS O/l5 l/l5 4115 Urinary S/IO O/IO 2110 3110

Pain 61 7 01 7 01 ’ l/ 7 Total 33/5 1 (65%) 2/5l (4%) 3/5l (6%) l3/5l (25%)

*Motor (weakness), sensory (paresthesia and/or sensory level), urinary (retention), and pain. tSee Table 3.

928 Radiation Oncology 0 Biology 0 Physics May 1982, Volume 8, Number 5

Table 3. Causes of post-irradiation neurological deterioration in 5 patients (all biopsy-proven)

Table 4. Neurological status of 14 patients who did not deteriorate neurologically post-irradiation

No. patients

Local in-field tumor recurrence Intracranial subarachnoid tumor

dissemination Pantopaque arachnoiditis

There is justifiable theoretical concern, therefore, that in a spinal cord previously damaged by tumor and surgery, the radiation tolerance dose might be less than that stated above because both the prior damage and the possible future irradiation damage are mediated through the same pathophysiological mechanisms described above. In addi- tion, spinal cord segments longer than 10 cm are usually irradiated in spinal cord glioma patients to insure ade- quate rostra] and caudal intramedullary tumor exten- sion6,’ (Table I), and such longer segments have a decreased irradiation tolerance.‘.”

Most recent Post-irradiation neurological follow-up

No. patients status (years)

Living 2 Gardening 2,9 I Tennis 4 I Mountain Climbing I6 I Gymnastics 3 3 Normal 16.6, 2

Dead 6 Normal, DID* 2, 4, 5, 6, 9, 15

*Dead of intercurrent death. Each had normal neurological examination prior to death with 3/3 having normal spinal cords at autopsy.

damaged by tumor/surgery, these doses can result in long-term neurological recovery without necessarily caus- ing radiation myelopathy.

No patient in the present series, however, has demon- strated thus far any evidence of chronic radiation my- elopathy. Each patient has survived at least beyond the time period in which the onset of this complication is at its peak.2,3 ‘.’ Follow-up periods have been, in fact, quite long

(Table 4). In this series all patients did well even when their spinal cord doses equalled, approached or occasion- ally even exceeded estimated tolerance levels. Regardless of which time-dose-fractionation regimen is used (Fig 1 ), and regardless of which tolerance level is used.

Patients with intraspinal ependymomas and astrocyto- mas have high survival rates.‘.’ High post-irradiation local tumor control rates have been achieved after approximately 4500-5000 rad in 180-200 rad fractions for astrocytoma patients, and after 4000-4500 rad in 180-200 rad fractions for ependymoma patients.’ The latter group also showed evidence of a dose-response relationship for local tumor control.’ Patients with spinal cord glioblastomas, however, continue to do very poorly with present therapy.’

The present study does nor define the 5% tolerance level for the previously-damaged spinal cord because none of these patients developed this complication. It shows, instead, that doses approaching, equalling, or occasion- ally even exceeding various definitions of tolerance may be given to a spinal cord damaged by tumor with no increased risk of radiation myelopathy. For the radiation therapist who wishes to give higher doses to the spinal cord in an attempt to cure intraspinal glioblastomas, however, this paper does hot determine the spinal cord tolerance limit.

CONCLUSION

Therefore a favorable therapeutic ratio (tumor con- trol/complications) may exist for patients with spinal cord astrocytomas and ependymomas. When these patients are irradiated to the above doses, local tumor control can be expected for the vast majority of patients. Even in these spinal cords, which have been previously

Based upon 19 patients irradiated for intraspinal glio- mas, tumoricidal doses’ of approximately 4500-5000 rad in 180-200 rad fractions can be given to astrocytoma patients and 4000-4500 rad in 180-200 rad fractions to ependymoma patients with the expectation of no increased risk of chronic radiation myelopathy, even though long spinal segments (often involving thoracic/ lumbar segments) are irradiated. Although an occasional patient whose dose exceeded tolerance has done well neurologically, higher doses at present have not increased the cure rate of these neoplasms. The true 5% tolerance level of the previously-damaged spinal cord remains to be defined.

REFERENCES Abbatucci, J.S., Delozier, T., Quint, R., er al.; Radiation myelopathy of the cervical spinal cord: Time, dose, and volume factors. Int. J. Radiat. Oncol. Biol. Phys. 4: 239- 248, 1978. Atkins, H.L.. Tretter, P.: Time-dose considerations in radiation myelopathy. Acta Radiat. Ther. 5: 79-94, 1966. Boden, G.: Radiation myelitis of the cervical spinal cord. Br. J. Radiol. 21: 464-469, 1948. Douglas, M.A., Parks, C.C., Bebin, J: Sudden myelopathy secondary to therapeutic total-body hyperthermia after

5.

6.

7.

spinal cord irradiation. N. Engl. J. Med. 304: 583-585, 1981. Eyster, E.F., Wilson, C.B.: Radiation myelopathy. J. Neu- rosurg. 32: 4 14-420, 1970. Kopelson, G.. Linggood, R.M.: lntramedullary spinal cord astrocytoma vs. glioblastoma: The prognostic importance of histologic grade. Cancer (In press). Kopelson, G., Linggood, R.M., Kleinman, G.M., ef al.. Management of intramedullary spinal cord tumors. Radiol- ogy 135: 473-479, 1980.

8

9

IO

II

Radiation tolerance of the damaged spinal cord 0 G. KOPELSON 929

Lambert, P.M.: Radiation myelopathy of the thoracic spi- nal cord in long term survivors treated with radical radio- therapy using conventional fractionation. Cancer 41: 175 I - 1760. 1978.

Maier. J.G.: Radiation myelitis of the spinal cord. Front. Rad. Ther. Oncol. 6: 404-41 I, 1972. Pallis, C.A., Louis, S., Morgan, R.L.: Radiation myelopa- thy. Brain 84: 460-479, I96 I.

Penner, R.D.. Archambeau, J.O.: Brain tolerance unit: A method to estimate risk of radiation brain injury for various

12.

13.

14.

dose schedules. Int. J. Radial. Oncol. Biol. Phys. 7: 397- 402, 1981. Phillips, T.L., Buschke, F.: Radiation tolerance of the thoracic spinal cord. Am. J. Roentgenol. 105: 6599665, 1969. Sheline, G.E., Wara, W.M., Smith, V.: Therapeutic irra- diation and brain injury. Int. J. Radiat. Oncol. Biol. Phys. 6: 1215-1228, 1980. Wara, W.M., Phillips, T.L., Sheline, G.E., Schwade, J.G.: Radiation tolerance of the spinal cord. Cancer 35: l558- 1562. 1975.