Incidence of Radiation Myelitis of the Cervical

Spinal Cord at Doses of 5500 cGy or Greater

Branislav Jeremic, MD, Ljubodrag Djuric, MD,
and Ljiljana Mijatovic, MD

The incidence of permanent damage to the spinal cord as a complication of

radiation therapy generally correlates positively with total radiation dose.
However, several reports have indicated that fraction size is also an important
factor in the development of late damage in normal tissue. To determine the effect
of fraction size on the incidence of radiation-induced spinal cord damage, the
authors reviewed 176 cases of head and neck cancer treated at their department
between 1980 and 1990 with radiation doses of 5500 cGy or greater to a portion of
the cervical spinal cord. Majority of these patients received 6000 cGy or greater
with fraction size ranging from 150 to 200 cGy. Seventy-two of 176 patients have
been observed for 2 years or more. More than one third (26patients) of these
received greater than 6000 cGy with fraction sizes of 157 to 170 cGy. Four of 72
(5,6%) patients had experienced permanent cervical spinal cord damage. The
results of this study suggest that radiation damage to the cervical spinal cord
correlates not only with total radiation dose, but also with fraction size. Low
fraction sizes appear to decrease the incidence of such damage.
Cancer 682138-2141,1991.

P ERMANENT spinal cord damage is one of the gravest

complications in radiation therapy. Its incidence had
increased with the introduction of supervoltage irradia-
tion of the cervical spinal cord, but in majority of cases
with less than conventional dose per fraction.

tion. Radiation-induced transverse myelitis is an irre-

versible process with no effective treatment.’ Symptoms
appear after a latent period of 6 months to 2 years after
completion of radiation therapy and may progress to total
paralysis.2Some patients stabilize after partial neurologic
loss as in Brown-Sequard syndrome, defined as Babinski’s
sign and absence of voluntary movement ipsilaterally to
the spinal cord lesion and loss of pain and temperature
sensation contralaterally. Since Ahlbom’s report on four
patients who had transverse myelitis after receiving ra-
diation for hypopharyngeal ~ a n c e r many
,~ reports have
attempted to define the radiation tolerance of the cervical
spinal In our study we had reviewed a group of
patients with long-term follow-up after radiation therapy,
whose treatment for head and neck cancer involved a
higher total dosage than might be recommended to a por-
From the Department of Oncology, University Hospital of Kragujevac,
Kragujevac, Yugoslavia.
Address for reprints: Branislav Jeremic. MD, KBC Kragujevac. De-
partment of Oncology, Zmaj Jovina b.b., 34000 Kragujevac, Yugoslavia.
Accepted for publication March 19, 1991.
Materials and Methods
Between 1980 and 1990 we had treated 481 patients
with head and neck cancer at the Department of Oncol-
ogy, University Hospital of Kragujevac (Kragujevac, Yu-
goslavia). Careful review of patients’ charts identified 176
patients who received radiation dose of 5500 to 6800 cGy.
One hundred thirty-seven patients were observed for at
least 6 months, 98 for at least 1 year, and 72 patients were
observed for 2 years or more. Head and neck portals en-
compassed the cervical spinal cord in 28 to 40 fractions.
Bilateral parallel opposed fields that included 5 to 15 cm
of cervical spinal cord were used. General policy of the
department was to perform continuous treatment and
only 44 (nine, 13%) patients underwent 2 weeks’ split
given midway through treatments. Fifty-four patients were
treated with 6-meV photons and the rest with 8-meV to
1O-meV photons from linear accelerators. A compensating
filter was used during treatment in 129 (27%) patients to
achieve a homogenous dose within the tissues being ir-
radiated. Irradiation characteristics for 72 patients ob-
served for 2 years or more are given in Table 1.

2138
No. 10 RADIATION AT 5500 CGY OR MORE
MYELITIS - Jeremic et a/. 2139

TABLEI . Total Radiation Dosages, Fraction Sizes, and Length of by a 5 c m long portion of cervical spinal cord was 5670
Spinal Cord Treated for 72 Patients With More Than cGy in 34 fractions, whereas the rest of 8 vm and five
2 Yeam of FoIIow-UP
~~ ~ ~
vertebral bodies received 5000 c G y in 30 fractions. He
Fraction Length of No. of had Brown-Sequard syndrome 6 months after irradiation
Total dose No. of size cord treated Vert No. of but we observed no loss of temperature sensation contra-
(cGy) fractions (cGy) (cm) bodies patients
~~ ~~ ~~

laterally.
5500 28 I96 3-5 2-3 2* Seven patients had Lhermitte’s sign within a few
5670 34 167 3-5 2-3 I* months after completing treatment. In five cases there
6000 30 200 3-5 2-3 5
> 5-10 3-6 14t was complete resolution after a short interval, but in two
32 150-200 3-5 2-3 3 cases major spinal cord damage occurred. First case of
> 5-10 3-6 4 Lhermitte’s sign occurred in a 70-year-old woman with
> 10-15 6-8 2
36 166 3-5 2-3 3 nasopharyngeal cancer. She had received 6000 cGy in 30
> 5-10 3-6 3 fractions of 200 cGy each to a 7-cm portion of cervical
40 150 3-5 2-3 4 spinal cord. She had no medical history of previous or
> 5-10 3-6 3
> 10-15 6-8 2 concomitant diseases and had tolerated irradiation well.
6300 40 157 3-5 2-3 2 Thirteen months after completing irradiation she had left-
> 5-10 3-6 4 sided numbness and right leg weakness. Cranial computed
6500 40 163 3-5 2-3 2
> 5-10 3-6 14 tomography (CT) scan and myelography showed no met-
> 10-15 6-8 3 astatic disease but only slight narrowing and atrophy of
6800 40 I70 3-5 2-3 I the cervical spinal cord. Results of cerebrospinal fluid ex-
Total 72 amination showed no abnormality. Two years and 6
Vert: vertebral.
months years after the onset of symptoms her neurologic
* Cases of radiation myelitis. loss stabilized with a right foot drop only.
Second case of Lhermitte’s sign occurred in a 47-year-
old man who had been treated for an oropharyngeal can-
Results cer and had received 6000 cGy in 30 fractions of 200 cGy
each to an 8-cm segment of cervical spinal cord. A com-
Of 72 patients who were observed for 2 years or more pensating filter was used through the whole irradiation
only 4 (5.6%) patients experienced permanent radiation course. His medical history revealed no previous or con-
damage. comitant diseases. After 6000 cGy his portals were reduced
A 57-year-old man was treated postoperatively with to exclude a portion of cervical spinal cord previously
5500 cGy in 28 fractions, 196 cGy each fraction to a portal irradiated. Subsequent analysis revealed no additional ir-
that encompassed the spinal cord. No history of previous radiation to the cervical spinal cord during Phase I1 of
or concomitant diseases were noticed. His laboratory the irradiation course. Fourteen months after completing
findings remained perfectly normal during the irradiation radiation therapy he had left leg weakness and right-sided
course. The portals included 1 1 cm of cervical spinal cord numbness. Cranial CT scan and myelography showed no
and seven vertebral bodies. A compensating filter was used abnormalities as well as his cerebrospinal fluid remained
after the second week of irradiation. He had Brown-Se- normal. During 3 years of close follow-up he improved
quard syndrome 5 months after irradiation. The patient’s his condition and now has no evidence of disease (NED)
neurologic symptoms progressed no further and he is alive with left leg weakness only.
with no evidence of the disease (NED) 2.5 years after
completion of radiation therapy, given for an oropharyn- Discussion
geal cancer.
A 50-year-old man was treated with curative intent for Since Ahlbom published historic report in 194 1 ,3 there
a hypopharyngeal cancer. Total dose was 5000 cGy in 30 have been several attempts to define the radiation toler-
fractions, 167 c G y each in Phase I to a portal that encom- ance of the cervical spinal cord.
passed cervical spinal cord for 13 cm and eight vertebral In 1948, Boden’ concluded that the tolerance was 3500
bodies. Phase I1 was planned with 2000 cGy in 12 fractions to 4500 cCy in I7 days, depending on the field size. More
with shrinking field technique to include 5 cm of previ- than 10 years later, Pallis ef al.’ lowered the tolerance
ously irradiated spinal cord. But after several treatments level to 3500 cGy in 28 days to 4300 cGy in 42 days with
in Phase I1 he permanently interrupted his irradiation, smaller fields. Several reports had followed evaluating the
with radiation dose given in Phase I1 of 670 cGy in four importance of fraction size and overall time. Philips and
fractions over 5 cm of a cervical spinal cord and three Buschke,8 reporting on the incidence of myelitis after
vertebral bodies. Therefore, total radiation dose received treatment to portions of cervical or thoracic spinal cord,
2 140 November 15 1991
CANCER
concluded that the number of fractions was the most im-
portant factor in the development of late radiation damage
of spinal cord. Correlation of dosage with number of frac-
tions have led them to a line connecting 6000 cGy in 35
fractions in 7 weeks ( 17 1.5 cGy per fraction) with 5000
cGy in 25 fractions in 5 weeks (200 cGy per fraction)
which excluded all of their myelitis cases. Report of van
der Kogel' from 1977 confirmed that the spinal cord tol-
erance depended more on the number of fractions and
less on overall treatment time. Absence of neurologic
damage was also noted by Baekmark' in 1975, who re-
ported on 23 patients who were treated to cervical spinal
cord and in some cases, also to a segment of thoracic
spinal cord with high doses using no more than 200 cGy
per fraction, treatments given daily, five times weekly.
Patients were observed for more than I year and received
more than 5000 cGy to the cord, most receiving 5700
cGy in 6 weeks.
Relation of total radiation dose and volume and the
development of myelitis in patients who underwent ra-
diation therapy to the cervical spinal cord was studied by
Abbatucci ef d4in 1978. They found that 5000 cGy in
25 fractions in 35 days could be tolerated safely if only
three to five vertebrae were irradiated homogeneously. In
that study 1 1 of 12 patients who were irradiated with 300
cGy per fraction had myelitis but there were no data for
patients irradiated with 200 ctiy per fraction.
Kim and Fayos6 reported on a group of patients who
underwent irradiation with doses to the cervical spinal
cord higher than the previously recorded tolerant doses,
In the three cases of permanent myelopathy that occurred,
multiple fields were used, probably overlapping in the re-
gion of spinal cord. They concluded that a short length
of spinal cord could tolerate 6000 cGy in fraction sizes
of 180 cGy to 200 cGy. They suggested that fraction size,
too, was one of the most important factors in the devel-
opment of permanent radiation myelopathy.
In our series, four of 72 (5.670) patients who were ob-
served for 2 years or more had radiation myelopathy after
receiving fairly acceptable dose and fraction schedule to
the spinal cord. Since there were no predisposing medical
conditions that would rendered them more at risk for
complications, except perhaps that a large segments of
cervical spinal cord were treated and maybe higher photon
energy used than usually for head and neck cancer.
Any definite conclusion could not be drawn from this
study alone, but the fact that only four of 72 (5.6%) pa-
tients who received high doses to the cervical spinal cord
had permanent radiation damage is probably related to
the relatively low dose per fraction that was used during
treatment. This is in agreement with reports of Kim and
Fayos,6 Baekmark,' and McCunniff and Liang." who all
concluded that tolerance of the cervical spinal cord was
higher with daily fractions of 180 to 200 cGy.
Vol. 68
The permanent cervical spinal cord damage that occurs
after excessive doses of radiation does not manifest itself
until late, 6 to 24 months after treatment, because the
tissues that are damaged are slowly proliferating.".'* There
are clinical evidence that changes in dose per fraction have
a greater effect on the incidence of late injury than acute
injury in normal tissues.
Montague'' observed that when breast cancer patients
were treated with twice-weekly fractionation, using large
doses per fraction, the late complications were significantly
higher in patients treated to the same total dose admin-
istered in five fractions per week. Also, Withers ef u I . , ' ~
while studying the relationship of early and late responses
of pig skin and comparing treatment regimens of five ver-
sus two times per week, observed that even though there
was a decrease in the acute responses with the larger dose
fractions given twice a week, there was a significant in-
crease in the late responses. In a later study of dose survival
curves for target cells for acute and late responses Thamed
and associates" concluded that the sparing that can be
achieved by the use of doses smaller than the conventional
200 cGy per fraction is greater for late than for early effects.
Our own results, in which 4 of 72 (5.6%) patients re-
ceiving high doses to the cervical spinal cord had per-
manent spinal cord damage, almost three fourths ofthem
being treated with doses lower than 200 cGy per fraction,
confirms sparing of normal tissue from late damage with
lower doses per fraction.
In conclusion, some patients may have permanent cer-
vical spinal cord damage even when acceptable dose levels
are delivered. However, the apparent tolerance of the cer-
vical spinal cord to higher doses of irradiation when using
less than conventional fractions as also seen in our study
lends further support to the theory that late effects in nor-
mal tissue are primarily related to the fraction size. This
may also be of particular interest as we study more about
the use of hyperfractionated irradiation with low dose per
fraction in clinical trials, which already have promising
support in date from animal ~ t u d i e s . ' ~
REFERENCES
I . Pallis CA. Louis S. Morgan RL. Radiation myelopathy. Brain 196 I ;
84:460-47Y.
2. Parsons J T . The effect of radiation on normal tissues of the head
and neck. In: Million RR. Cassisi NJ, eds. Management of Head and
Neck Cancer: A Multidisciplinary Approach. Philadelphia: JB Lippincott,
1980; 173-207.
3. Ahlbom HE. The results of radiotherapy of hypopharyngeal cancer
at the Radiumhemmet. Stockholm, 1930-1939. Aria Radiol [Onco[j
I94 I ; 22: 155-1 7 1 .
4. Abbatucci JS, Delozier T, Quint R, Rousell A, Brune D. Radiation
myelopathy of the cervical spinal cord: Time, dose and volume factors.
In! J Radiut Oncol B i d Phys 1978: 4x239-248.
5. Roden G. Radiation myelitis of the cervical spinal cord. Br J Radiol
1948; 2 1 :464-469.
No. 10 RADIATION AT 5500 ~ G OR
MYELITIS Y MORE
6. Kim YH, Fayos JV. Radiation tolerance ofthe cervical spinal cord.
Radio1og.r’ 198 I ; I39:473-478.
7. van der Kogel AJ. Radiation tolerance of the spinal cord: Depen-
dence on fractionation and extended overall times. In: Radiobiological
Approach and Radiotherapy. vol. I . Vienna: International Atomic Energy
Agency. 1977: 83-90.
8. Philips TL, Buschke F. Radiation tolerance of the thoracic spinal
cord. Am J Rocnrgenol 1969; 105:659-664.
9. Baekmark UB. Neurologic complications after irradiation of the
cervical spinal cord for malignant tumor of the head and neck. Acia
Radial Thrr Phys Biol 1975: I4:33-4 I .
10. McCunniff AJ. Liang MJ. Radiation tolerance ofthe cervical spinal
cord. In! J Radial Oncol Riol Phvs 1989; 16:675-678.
I I . Thamed HD. Withers HR, Peters LJ. Fletcher GH. Changes in
- Jeremic el al. 2141
early and late radiation responses with altered dose fractionation: Im-
plications for dose-survival relationships. In1 J Rodiuf Oncol Biol Phys
1982; 8:219-226.
12. Withers HR, Peters U,Kogelnik HD. The pathobiology of late
effects of irradiation. In: Meyn RE. Withers HR. eds. Radiation Biology
in Cancer Research. New York: Raven Press, 1980; 439-448.
13. Montague ED. Experience with altered fractionation in radiation
therapy of breast cancer. Radiology 1968; 90:962-966.
14. Withers HR. Thames HD, Flow BL. Mason KA. Hussey DH.
The relationship of acute to late skin injury in 2 and 5 fractions/week
y ray therapy. In1 J Radial Oncol Biol Phy.y 1978; 4:591-601.
15. van der Schueren E, Landuyt W. Ang KK, van der Kogel AJ.
From 2 Gy to 1 Gy pre fraction: Sparing effect in rat spinal cord? In1 J
Radial Oncol Biol Phys 1988; 14:297-300.