60
Radiotherapy and Ontology, 22 (1991) 60-67 © 1991 Elsevier Science Publishers B.V. All rights reserved 0167-8140/91/$03.50
RADION 00882
Effects of hyperthermia applied to previously irradiated cervical spinal cord in the rat P. Sminia, J. H a v e m a n a n d C. K o e d o o d e r Department of Radiotherapy, Academisch Medisch Centrurn,Amsterdam, The Netherlands (Received 3 November 1989, revision received 18 March 1991, accepted 26 June 1991)
Key words: Retreatment; Hyperthermia; Irradiation; Spinal cord
Summary Rat cervical spinal cord was X-ray irradiated at doses of 15, 18, 20 and 26 Gy. Ninety days later, approximately the same part of the spinal cord was heated at 42.3 + 0.4 °C for 50, 60, 75 or 90 min by means of a 434 MHz microwave applicator. After treatment, animals were observed over a period of 18 months for expression of neurological complications. These complications could either be the result of the heat or of the radiation treatment. The time course showed three distinct peaks in the incidence of neurological symptoms. The first peak was due to the acute response to hyperthermia. The ED5o value for neurological complications one day after treatment at 42.3 + 0.4 °C was 74 + 2 min. Previous X-ray irradiation of the spinal cord with 18, 20 and 26 Gy reduced the EDso to 57 + 7, 65 + 4 and 55 + 5 min (12-26~ of control), respectively. Recovery from heat-induced neurological complications was diminished in previously irradiated animals. The second peak (150-300 days after X-rays) concerned the expression of"early delayed" radiation damage. Hyperthermia given 90 days after irradiation did not influence either the percentage of animals with paralysis or the latent period. Neurological symptoms developing after day 300 were due to the "late delayed" radiation response. No significant difference was observed in the data on paralysis induced by radiation alone or radiation followed by heat. The late radiation-induced minor neurological symptoms were, however, influenced by retreatment with heat.
Introduction Several clinical studies have shown that a significant improvement of local tumour control can be obtained by addition of hyperthermia to conventional cancer treatment [4,5,17,25,28]. In the clinic, hyperthermia is often applied to patients with recurrence of the primary tumour or metastases after initial treatment with radiotherapy or chemotherapy. In these cases, particular caution is required as residual radiation injury might result in a decreased normal tissue tolerance to hyperthermia. Clinical data on retreatment of superficial tumours indicate that the tolerance of previously irradiated skin to subsequent hyperthermia was not significantly affected [ 13,14,19]. There are only limited clinical data available on retreatment tolerance of slowly
proliferating tissues, however. Douglas etal. [2] reported on severe responses of the spinal cord after retreatment by hyperthermia. They observed 3 out of 12 patients who developed acute myelopathy when whole body hyperthermia (WBH) was applied 1-2 months after irradiation combined with chemotherapy. Parks et al. [ 18] observed one of 17 patients with myelopathy as a result of combined treatment of radiation and WBH. Guidelines for retreatment tolerance of normal tissues may be obtained from laboratory investigations using animals. Information on the tolerance of previously irradiated tissues to hyperthermia alone or combined with radiotherapy is available for rapidly proliferating tissues, skin [ 1,9,10,11,29,30] and intestine [7,8]. These studies show that the thermal response
Address for correspondence: P. Sminia, Department of Radiotherapy, Academisch Medisch Centrum, Meibergdreef9, 1105 AZ Amsterdam, The Netherlands.
61 may be increased if heat was applied to previously irradiated tissue. No data are available on retreatment tolerance to hyperthermia of the previously irradiated spinal cord. In the studies of Miller et al. [ 15], Neville et al. [16] and Sminia et al. [24], hyperthermia was applied shortly after local irradiation of the rat spinal cord to investigate the effects of heat on the radiation response. Goffinet et al. [3] studied myelitis of the mouse thoraco-lumbar spinal cord after heat treatment applied shortly before and after irradiation, and 2 weeks prior to irradiation. In the present paper, we report on the effects of irradiation of the rat cervical spinal cord (dose range 15-26 Gy) on the acute response to local hyperthermia (50-90 min that 42.3 °C) applied to approximately the same site 90 days later. The heat treatment was applied at a time where radiation-induced damage was progressing but still not expressed. The influence of hyperthermia on the expression of "early delayed" and "late delayed" radiation damage is investigated. Materials and methods
Animals Female Wistar rats, aged 12-13 weeks (obtained from Harlan CPB, Zeist, The Netherlands) were used in all experiments. Animals were anaesthetized with sodium pentobarbital (Nembutal ®, 50 mg/kg body weight, i.p.) about 15 rain before heat treatment. The analgesic pentazocine (9 mg/kg body weight, i.p.) was administered 5-10 min later. Before irradiation, light anaesthesia of the animals was obtained using about half a dose of the pentobarbital; no pentazocine was given. After both the irradiation and heat treatment, animals recovered from anaesthesia in an infant incubator (Air-Shields Europe, Shannon, Ireland) at a temperature of 32 °C and a relative humidity of about 70 ~o. The number of animals included in this study was 251; 35 were treated with hyperthermia alone, 58 with X-rays alone and 158 with X-rays followed by hyperthermia after 90 days. Treatment groups consisted of 6-13 animals.
Irradiation and dosimetry The cervical spinal cord region of the animals was irradiated from the dorsal side with a Siemens Stabilipan X-ray generator operated at 250 kV and 15 mA, filtered with 1 mm Cu. The anaesthetized animal was placed in supine position on a 10 mm thick lead shield. Local irradiation of the spinal cord, cervical 5-thoracic 2 (C5-T2), was via a rectangular field of
15 mm × 20 mm. Focus to skin distance was 24.2 cm. The actual dose delivered to the spinal cord was checked by measurement both in a tissue equivalent phantom and in a sacrificed rat. First, the dose rate was checked in the rat phantom using a cylindrical ionization chamber (Baldwin and Farmer, type 2570), at a depth representative for the spinal canal. This measurement resulted in a dose rate of 326 cGy/min. Using this value for the dose rate, a sacrificed rat was irradiated with doses of 24, 26 and 28 Gy, respectively, which are doses around the EDs0 value for "early delayed" paralysis [24]. These nominal doses were verified with TLD measurements (TLD-100, Harshaw). By means of a 1.7 mm diameter catheter, three TLD rods were inserted into the spinal canal. The exact position of the rods was checked by radiography. Measurements were performed twice for each dose to eliminate possible variations due to the position of the volume relative to the radiation field. In all cases, the results showed that the difference between the doses estimated by the ionization chamber and the TLD method was less than 1 ~o. Animals were irradiated in air at doses of 15, 18, 20 and 26 Gy. The temperature in the irradiation room was kept at 26 ° C to avoid a drop in body temperature of the anaesthetized animals during the irradiation.
Hyperthermia On day 90 after irradiation, the cervical region in the rat, including the spinal cord (cervical 5 - thoracic 2) was heated using a 434 M H z microwave applicator. Technical aspects of the heating system and details on thermometry were described previously [20,21]. It was shown that the cervical part of the vertebral column could be heated whereas surrounding tissues were only slightly heated. In all experiments, the temperature was regulated using a reference thermocouple probe which was placed against one of the cervical vertebrae 6, 7 or thoracic 1. Temperature measurements inside the vertebral canal were done in animals of about 200 g as well as in animals of about 250 g (90 days older). In the experiments, the cervical vertebral column was heated for 50-90min at a reference temperature of 43.0 + 0.1 ° C (mean + SD) which resulted in an average maximum temperature of 42.3 + 0.4 °C in the vertebral canal at the level of the vertebrae cervical 5 thoracic 2 [20].
Follow-up Up to 18 months after treatment, animals were regularly inspected and neurological complications were scored using an arbitrary scoring system (Table I). The
62 TABLE I
15.
~
2
3
The arbitrary scoring system a. Score
Visible signs
Ability to turn in the Geotaxis test b
~' 10"6
1 I
"~_:~
k~~uL~
5-
0 1
No visible complications Minor neurological symptoms such as unco-ordinated use of the forelegs Reduced ability to walk on the forelegs, paresis, monoplegia Reduced ability to walk on the forelegs, paresis, monoplegia Paralysis of both forelegs Death
3/3
O-
3/3
alone
r
0 l x
-
-
90 1
180 Foil . . . .
270
360
~
450
x
N 90 days
540
p (days)
N
2/3 1/3 0/3
Used for evaluation of neurological complications resulting from treatment of the cervical region in the rat, including the spinal cord between the vertebra cervical 5 and thoracic 2. b See [20]. a
incidence of paralysis, paresis and minor neurological symptoms over this period were used as endpoints, and quantal data were used to establish dose-effect curves which were fitted by logistic regression analysis. From this analysis, the heat or radiation doses required to produce paralysis or severe paresis and minor neurological symptoms in 5 0 ~ of the animals (EDso + SE) were established using BMDP statistical software (Cork, Ireland). The modification in the response to hyperthermia produced by previous irradiation and the modification of the radiation response produced by hyperthermia were compared on the basis of these EDso values. A significant number of animals could not be followed up completely for late effects, because they developed a tumour inside the treated volume [23]. Results
Figure 1 shows the time course of the incidence of paralysis as a result of treatment of the rat cervical spinal cord. In the 90-day period between X-rays and heat, no neurological complications were observed in the irradiated animals. The time course (Fig. 1) showed three more or less distinct peaks in the incidence of neurological symptoms: (1)the acute response to hyperthermia (90-120 days); (2) the "early delayed" response to irradiation (150-300 days) and, (3)the "late delayed" response to irradiation (300-540 days). The distinction between the "early delayed" and "late delayed" radiation response at day 300 was based on the bimodal pattern of the radiation response in the data published by Sminia et al. [24].
Fig. 1. Percentage of paralyzed animals of the initial number of animals plotted as a function of the follow-up period. Animals were irradiated at day 0 (15, 18, 20 and 26 Gy) and heated at day 90 (50, 60, 75 and 90 min at 42.3 °C). (1)Acute response to hyperthermia; (2) "early delayed" response to radiation; (3) "late delayed" response to radiation. The total number of animals included in this histogram is 51.
The effect of previous irradiation on the acute response to hypertherrnia
Figure 2 shows the percentage of animals with minor neurological symptoms (score 1), paralysis (score4) and lethality (score 5) one day after heat treatment at 42.3 °C as a function of X-ray dose. The percentage of animals with minor neurological symptoms was increased in animals that were heated 90 days after irradiation of the spinal cord with 18, 20 or 26 Gy, relative to control animals. This was most pronounced at lower heat doses, 50 and 60 rain at 42.3 °C. At 75 and 90 min at 42.3 °C, the percentage of animals with minor neurological symptoms was already 40-55~o, even in the non-irradiated groups. After 26 Gy followed by 90 min at 42.3 °C 90 days later, the percentage of animals with severe complications, such as paralysis and lethal damage, was greatly increased. For all other treatment groups, the percentage of animals with paralysis and lethality was slightly enhanced by previous irradiation. Figure 3 shows that in previously irradiated animals a shorter heating time is required to obtain neurological complications. The EDso, day 1, for heat alone was 74 + 2rain at 42.3 °C (mean + SD). After 15, 18, 20 and 26 Gy previous irradiation the EDso was 7 8 + 3 , 5 7 + 7 , 6 5 + 4 and 5 5 + 5 m i n , respectively (i.e. reductions of 12-26 %). Data from day 14 (Fig. 4) and 60 (Fig. 5) after hyperthermia show that recovery from heat-induced neurological symptoms took place. All non-irradiated animals, and almost all animals previously irradiated with 15 and 18 Gy, had completely recovered. Animals previously irradiated with 20 or 26 Gy were however, less able to recover, e.g. 50~o (3/6) of the surviving animals preirradiated with 26 Gy, still suffered from minor neurological symptoms. Animals from the 26 Gy
63 Pretrradiated
[~::~0 Gy ~ 1 5
spinal
Gyl~fl8
cord
loo
[]
Gylg;X:~20 Gy~E~I26 GY
8O Symptoms
(score 1 ) Day
1
70 60 5O 40 o
3O
-6
2O
g
10
g
0
z_ i1)
50
60
75
90
50
4O Paralysis o
( s c o r e 4 ) Day 1
30
E
II 50
20
60
Duration of t r e a t m e n t
75
*
EDSO
ml A O []
sham 18Gy 20Gy 26 Gy
9O
at 4 2 3 °C (rain.)
o 10
0
50
60
75
90
50 Lethality
(score
5)
Day 1
40
1° i o L - - - 50
Duration
60 of t h e
Fig. 3. Percentage of animals with neurological complications (score 1-4 pooled and corrected for lethality) one day after hyperthermia as a function of the duration of the heat treatment at 42.3 °C. Animals were sham treated or X-irradiated at a dose of 15, 18, 20 or 26 Gy, 90 days before hyperthermia.
75 heat
treatment
90 at 42.3 °C
Fig. 2. Percentage of animals with minor neurological symptoms, paralysis and lethality one day after hyperthermia applied 90 days after irradiation of the spinal cord. Treatment groups consisted of 7-11 animals.
group that were not heated, did not yet express neurological symptoms at this time. Lethality can occur after hyperthermia alone probably as the result of respiratory paralysis. Figures 2, 4 and 5 show that lethality was increased in 26 Gy preirradiated animals relative to other groups. If animals died as a result of heat injury to the spinal cord, this occurred within 14 days after hyperthermia.
The effect of hyperthermia 90 days after irradiation on the "'early delayed" radiation response of the spinal cord Starting 6 months after X-rays, the effects of radiation treatment became apparent. "Early delayed" paralysis was not observed (with a single exception) in the 15, 18
and 20 Gy groups, either with or without heat. After 26 Gy alone, "early delayed" paralysis occurred in 64?o (7/11) of animals. Recovery from thermal damage was diminished after preirradiation of the spinal cord with 2 6 G y (Fig. 5). In these animals, heat-induced neurological complications were immediately followed by radiation-induced neurological complications. The early response was, however, not increased: 617o (17/28) paralysis. The latent period for early delayed paralysis after 26 Gy alone was 238 + 45 days (n = 7) and after 26 Gy followed by hyperthermia it was 234 + 20 days (n -- 17).
The effect of hyperthermia 90 days after irradiation on the expression of late myelitis No effect of the addition of heat 90 days after irradiation was found on the number of animals with late paralysis (Table II). The latent period before development of late paralysis varied from 322 to 522 days, mean 420 _+ 74 days (n = 15). The number and percentage of animals with late minor neurological symptoms are given in Table III. The data show that heat influences the radiation response, the Dose Modifying Factor (DMF) varied from 1.1 to 1.3. Figure 6 shows a significant shortening in latency for minor neurological symptoms, particularly at higher heat doses, 75 and 90 min at 42.3 °C, and at low X-ray doses (15 and
18 Gy).
64 Preirradioted
[~
0 Gy ~
spinal
Prei rradiated
cord
[
15 Gy ~ ' / ~ I 8 GyE;~;~ 20 Gyl~7/~ 26 Gy
Gy
BO
BO 70
spinal cord
I 0 Oy I ~ I U 15 GyI~F'z~IB G y I ~ F ~ 2 0 O y ~ 2 6
( s c o r e 1) D a y 14
Symptoms
70
60
60
50
50
40
40
30
30
Symptoms ( s c o r e 1) D a y
60
20
10 0
10 50
60
75
0
gO
5O
4O
60
75
g0
75
90
40
Paralysis (score 4 )
Day
Paralysis (score 4 )
14
30
Day
60
30
E E o 20
20
&
o
o
10
0
50
60
75
0
90
50
50 Lethality
(score 5) Day 14
Lethality
40
40
30
30
20
20
10
10
50
60
D u r a t i o n of t h e
75 heat
0
90
t r e a t m e n t at 42.3*C
50
60
15 Gy 480' 440
..~ 4 0 0
TABLE II The number of animals per total in the treatment group expressing "late delayed" paralysis*.
60
75
90
h e a t t r e a t m e n t a t 42.3 °C
Fig. 5. Percentage of animals with minor neurological symptoms, paralysis and lethality 60 days after hyperthermia applied 90 days after irradiation of the spinal cord.
520
Sham
(score 5 ) D a y
D u r a t i o n of t h e
Fig. 4. Percentage of animals with minor neurological symptoms, paralysis and lethality 14 days after hyperthermia applied 90 days after irradiation of the spinal cord.
X-Dose (Gy)
60
50
18 Gy
20 Gy x - r a y s (X)
alone
75' 42.3 "C 90 d after X 90' 42.3"C 90 d Qfter X
"~ 3 6 0 320 _~ 2 8 0 240
Min at 42.3 °C
200
20 26
0/10 2/4
50
60
75
90
1/7 2/4
1/6 2/2
2/6 3/4
0/7 1/1
*After 20 and 26 Gy followed by hyperthermia at 42.3 °C for 50-90 min, 90 days later. In the groups with 15 and 18 Gy, no animals with "late delayed" paralysis were observed.
Fig. 6. The latent period (days) for expression of "late" minor neurological symptoms after 15, 18 and 20 Gy of X-rays followed by 75 and 90 min at 42.3 °C 90 days later. The latent period was not affected after X-rays followed by 50 and 60 rain at 42.3 °C. Probability values for significanee between X-rays alone and X-rays followed by heat were calculated using the Students' t-test. *p < 0.05; ** p < 0.01. Error bars represent standard deviation.
65 TABLE III The number and percentage of animals at risk expressing "late" minor neurological symptoms a. X-Dose (Gy)
Sham
Min at 42.3 °C 50
60
75
90
15 N P 18 N P 20 N P 26 N P
3/9 33 (7-70) 4/10 40 (12-74) 4/11 36 (11-69) 3/3 100 (29-100)
3/8 38 (9-76) 2/6 33 (4-78) 4/7 57 (18-90) 2/2 100 (16-100)
3/7 43 (10-82) 3/6 50 (12-88) 5/6 83 (36-100) -
2/6 33 (4-78) 3/4 75 (19-99) 5/6 83 (36-100) 1/1 100 (0-100)
3/6 50 (12-88) 2/4 50 (7-93) 6/8 75 (35-97) -
EDso b (Gy) DMF b p-value
20.0 + 1.9 -
18.5 + 1.8 1.08 + 0.10 0.567
16.4 _+ 1.7 1.22 + 0.10 0.158
16.3 _+ 1.3 1.23 +_ 0.09 0.108
15.6 _+ 3.3 1.28 +_ 0.16 0.248
After irradiation (15-26 Gy) followed by hyperthermia at 42.3 °C for 50-90 min 90 days later. N = number per total in treatment group; P = per cent; 95 ~o confidence limits between brackets; D M F = Dose Modifying Factor. Probability values for significance were calculated using the normal distribution. b Mean _+ SE.
Discussion
We have studied the effects of hyperthermia applied to the spinal cord at a time where latent radiation damage was present. The results demonstrate that: (1) The sensitivity to hyperthermia of the previously irradiated spinal cord was enhanced (cf. Figs. 2-5) and recovery from heat-induced neurological complications was diminished (cf. Figs. 4 and 5). (2) Neither the percentage of animals expressing radiation induced "early delayed" or "late delayed" foreleg paralysis or paresis, nor the latent period was affected by hyperthermia given after 90 days (cf. Table II). (3) The data on late minor neurological symptoms indicate a heat dose related enhancement of the radiation response, this was, however, not very significant (cf. Table III). The latent period for minor neurological symptoms was significantly shortened (cf. Fig. 6). The effect of previous irradiation on the acute response to hyperthermia
Radiation damage to the spinal cord with doses above 20 Gy is characterized by slowly developing demyelination due to loss of oligodendrocytes [26,27]. Our observations show that, if hyperthermia is applied during development of radiation damage, the number of animals with neurological complications as a result of
heat is significantly enhanced and recovery from neurological symptoms is diminished. The large amount of scatter in the data on neurological symptoms after hyperthermia may be due to the temperature variations of the heated spinal cord. Some animals did not recover after 26 Gy followed 90 days later by 42.3 °C hyperthermia for 60-90 rain. In the literature, the only experimental data concerning thermal sensitivity of previously irradiated normal tissues are for skin and intestine, tissues that proliferate relatively rapidly. These data show a decreased tolerance to hyperthermia due to residual injury. The tolerance of previously irradiated skin to 70-90 rain at 44 °C is reduced by 7-43 ~o at 90 days [30] and by 36~o after 25-40 rain at 43.5 °C given 10 months after irradiation [ 11 ]. Data on mouse intestine show that 20-45 min at 43 °C applied 15 days after irradiation resulted in the same effect as hyperthermia alone at 43.5°C for 20-45 min. This enhanced sensitivity returned to normal within 120 days. A maximum reduction in the crypt cell number of 60~o was found after 60 rain at 42.3 °C, applied 15 days after 8-10 Gy [8]. This effect returned to a subthreshold level within 20 days to 7 months. Comparison of our data on retreatment tolerance of the spinal cord with the data mentioned above for the skin and intestine, shows that the thermal dose where a "memory" of previous treatment was observed differs considerably. Recovery from heat injury in both skin and spinal cord was diminished after previous irradiation.
66
The effect of hyperthermia 90 days after irradiation on the "early delayed" and "'late delayed" radiation response of the spinal cord Hyperthermia given 90 days after irradiation neither influenced the percentage of animals with early delayed radiation paralysis, nor the latent period. This indicates that, once animals had recovered from thermal damage, there was no further influence on the radiation-induced early response. Similarly, no effect ofhyperthermia was found on the percentage of animals with late radiationinduced paralysis (cf. Table III). For radiation-induced minor neurological symptoms the EDs0 was 20.0 _+ 1.9 Gy, which is in accordance with data reported by Hopewell et al. [6], Van der Kogel [26] and another study from our laboratory [24]. Minor neurological symptoms were enhanced in groups that received hyperthermia 90 days later, the maximum D M F was 1.28.
Clinical implications
ratory data show that the maximum tolerated heat dose by mammalian brain and spinal cord is much less: 50-60min at 42.0-42.5°C or at 4 3 ° C for only 10-20 min [ 12,22]. The present results indicate that for previously irradiated nervous tissue, the maximum tolerated heat dose might be about 10-30 ~'o lower than for untreated tissue. Since our data show that the late radiation response may also be enhanced with regard to minor neurological symptoms and shortening of the latent period, caution is required in application of heat to previously irradiated nervous tissue in patients.
Acknowledgements The authors would like to thank Mr. J . J . G . W . Hendriks for his skilful technical assistance and Mrs. A. van der Graaff for typing and editing of the manuscript. We gratefully acknowledge Dr. F. A. Stewart for correcting the English. We would also like to thank Mr. S. Tee for dosimetry and Mr. J. Sybrands for constructing the microwave applicator.
In clinical application of hyperthermia, the aim is generally to heat the turnout at 43 ° C for 30-60 min. LaboReferences 1 Baker, D.G., Sager, H.T., Constable, W. and Goodchild, N. The response of previously irradiated skin to combinations of X-irradiation and ultrasound-induced hyperthermia. Radiat. Res. 96: 367-373. 2 Douglas, M. A., Parks, L. C. and Bebin, J. Sudden myelopathy secundary to therapeutic total-body hyperthermia after spinal cord irradiation. N. Engl. J. Med. 304: 583-585, 1981. 3 Goffinet, D. R., Choi, K.Y. and Brown J.M. The combined effects ofhyperthermia and ionisingradiation on the adult mouse spinal cord. Radiat. Res. 72: 238-245, 1977. 4 Gonzalez Gonzalez, D., Van Dijk, J. D. P., Blank, L. E. C. M. and Rt~mke, P. Combined treatment with radiation and hyperthermia in metastatic malignant melanoma. Radiother. Oncol. 6: 105-113, 1986. 5 Gonzalez Gonzalez, D., Van Dijk, J.D.P. and Blank, L. E. C.M. Chestwall recurrences of breast cancer: results of combined treatment with radiation and hyperthermia. Radiother. Oncol. 12: 95-103, 1988. 6 Hopewell, J.W., Morris, A.D. and Dixon-Brown, A. The influence of field size on the late tolerance of the rat spinal cord to single doses of X-rays. Br. J. Radiol. 60: 1099-1108, 1987. 7 Hume, S.P. and Marigold, J. C.L. Increased hyperthermal response of previously irradiated mouse intestine. Br. J. Radiol. 55: 438-443, 1982. 8 Hume, S. P. and Marigold, J. C.L. Thermal enhancement of radiation damage in previously irradiated mouse intestine. Br. J. Radiol. 59: 53-59, 1986. 9 Law, M. P., Ahier, R. G. and Somaia, S. A long term effect of X-rays on thermal sensitivity of the mouse ear. Br. J. Radiol. 57: 729-731, 1984. 10 Law, M. P., Ahier, R. G. and Somaia, S. Thermal enhancement
of radiation damage in previously irradiated skin. Br. J. Radiol. 58: 161-167, 1985. 11 Law, M. P. and Ahier R.G. A long term effect of prior irradiation on the thermal enhancement of radiation damage in the mouse ear. Int. J. Hyperthermia 3: 167-175, 1987. 12 Lyons, B.E., Britt, R.H. and Strohbehn, J.W. Localized hyperthermia in the treatment of malignant brain tumours using an interstitial microwave antenna array. IEEE Trans. Biomed. Engin. 31: 53-62, 1984. 13 Marmor, J. B. and Hahn, G.M. Ultrasound heating in previously irradiated sites. Int. J. Radiat. Oncol. Biol. Phys. 4: 1029-1032, 1978. 14 Marmor, J.B., Pounds, D., Postic, T.B. and Hahn, G.M. Treatment of superficial human neoplasms by local hyperthermia induced by ultrasound. Cancer 43: 188-197, 1979. 15 Miller, R.C., Leith, J.T., Veomett, R.C. and Gerner, E.W. Potentiation of radiation myelitis in rats by hyperthermia. Br. J. Radiol. 49: 895-896, 1976. 16 Neville, A. J., Robins, H. I., Martin, P., Gilchrist, K. W., Dennis, W. H. and Steeves, R.A. Effect of whole body hyperthermia and BCNU on the development of radiation myelitis in the rat. Int. J. Radiat. Biol. 46: 417-420, 1984. 17 Overgaard, J. The current and potential role ofhyperthermia in radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 16: 535-549, 1989. 18 Parks, L.C., Minaberry, D., Smith, D.P. and Neely, W.A. Treatment of far-advanced bronchogenie carcinoma by extracorporeally induced systemic hyperthermia. J. Thorac. Cardiovasc. Surg. 78: 883-892, 1979. 19 Perez, C. A., Kopeeky, W., Rao, D. V., Baglan, R. and Mann, J. Local microwave hyperthermia and irradiation in cancer
67
20
21
22
23
24
therapy: preliminary observations and directions for future clinical trials. Int. J. Radiat. Oncol. Biol. Phys. 7: 765-772, 1981. Sminia, P., Haveman, J., Wondergem, J., Van Dijk, J. D. P. and Lebesque, J.V. Effects of 434 MHz microwave hyperthermia applied to the rat in the region of the cervical spinal cord. Int. J. Hyperthermia 3: 441-452, 1987. Sminia, P., Van Dijk, J. D. P. and Haveman, J. Developpement d'un radiateur coaxial de forme circulaire a microondes, 434 MHz, permettant une hyperthermie localisee profonde: resultats d'application dans la moelle epiniere y comprise chez les rats. Innovat. Technol. Biol. Med. 9 (no 2 special): 79-89, 1988. Sminia, P., Haveman, J. and Ongerboer de Visser, B.W. What is a safe heat dose which can be applied to normal brain tissue? Int. J. Hyperthermia 5: 115-117, 1989. Sminia, P., Jansen, W., Haveman, J. and Van Dijk, J. D.P. Incidence of tumours in the cervical region of the rat after treatment with radiation and hyperthermia. Int. J. Radiat. Biol. 57: 425-436, 1990. Sminia, P., Haveman, J., Van Dijk, J, D. P. and Hendriks, J, J. G.W. Enhancement by hyperthermia of the "early delayed" and "late delayed" radiation response of the rat cervical spinal cord. Int. J. Radiat. Biol. 59: 259-271, 1991.
25 Valdagni, R. and Amichetti, M. Clinical hyperthermia: five year's experience. Strahlenther. Onkol. 163: 443-445, 1987. 26 Van der Kogel, A.J. Late effects of radiation on the spinal cord. Dose-effect relationships and pathogenesis. Thesis, University of Amsterdam, Edited by Uitgeverij Meinema, Delft, 1979. 27 Van der Kogel, A.J. The cellular basis of radiation-induced damage in the central nervous system. In: Cytotoxic Insult to Tissue. Effects on Cell Lineages, pp. 329-352. Editors: C. S. Potten and J. H. Hendry. Churchill Livingstone, Edinburgh London - Melbourne - New York, 1983. 28 Van der Zee, J., Van Rhoon, G. C., Wike-Hooley, J. L., Faithfull, N. S. and Reinhold, H. S. Whole-body hyperthermia in cancer therapy: a report of a phase I-II study. Eur. J. Cancer Clin. Oncol. 19: 1189-1200. 29 Wondergem, J. and Haveman, J. The response of previously irradiated mouse skin to heat alone or combined with irradiation: influence of thermotolerance. Int. J. Radiat. Biol. 6: 539-552, 1983. 30 Wondergem, J. and Haveman, J. The effect of previous treatment on the response of mouse feet to irradiation and hyperthermia. Radiother. Oncol. 10: 253-261, 1987.
Radiotherapy and Ontology, 22 (1991) 60-67 © 1991 Elsevier Science Publishers B.V. All rights reserved 0167-8140/91/$03.50
RADION 00882
Effects of hyperthermia applied to previously irradiated cervical spinal cord in the rat P. Sminia, J. H a v e m a n a n d C. K o e d o o d e r Department of Radiotherapy, Academisch Medisch Centrurn,Amsterdam, The Netherlands (Received 3 November 1989, revision received 18 March 1991, accepted 26 June 1991)
Key words: Retreatment; Hyperthermia; Irradiation; Spinal cord
Summary Rat cervical spinal cord was X-ray irradiated at doses of 15, 18, 20 and 26 Gy. Ninety days later, approximately the same part of the spinal cord was heated at 42.3 + 0.4 °C for 50, 60, 75 or 90 min by means of a 434 MHz microwave applicator. After treatment, animals were observed over a period of 18 months for expression of neurological complications. These complications could either be the result of the heat or of the radiation treatment. The time course showed three distinct peaks in the incidence of neurological symptoms. The first peak was due to the acute response to hyperthermia. The ED5o value for neurological complications one day after treatment at 42.3 + 0.4 °C was 74 + 2 min. Previous X-ray irradiation of the spinal cord with 18, 20 and 26 Gy reduced the EDso to 57 + 7, 65 + 4 and 55 + 5 min (12-26~ of control), respectively. Recovery from heat-induced neurological complications was diminished in previously irradiated animals. The second peak (150-300 days after X-rays) concerned the expression of"early delayed" radiation damage. Hyperthermia given 90 days after irradiation did not influence either the percentage of animals with paralysis or the latent period. Neurological symptoms developing after day 300 were due to the "late delayed" radiation response. No significant difference was observed in the data on paralysis induced by radiation alone or radiation followed by heat. The late radiation-induced minor neurological symptoms were, however, influenced by retreatment with heat.
Introduction Several clinical studies have shown that a significant improvement of local tumour control can be obtained by addition of hyperthermia to conventional cancer treatment [4,5,17,25,28]. In the clinic, hyperthermia is often applied to patients with recurrence of the primary tumour or metastases after initial treatment with radiotherapy or chemotherapy. In these cases, particular caution is required as residual radiation injury might result in a decreased normal tissue tolerance to hyperthermia. Clinical data on retreatment of superficial tumours indicate that the tolerance of previously irradiated skin to subsequent hyperthermia was not significantly affected [ 13,14,19]. There are only limited clinical data available on retreatment tolerance of slowly
proliferating tissues, however. Douglas etal. [2] reported on severe responses of the spinal cord after retreatment by hyperthermia. They observed 3 out of 12 patients who developed acute myelopathy when whole body hyperthermia (WBH) was applied 1-2 months after irradiation combined with chemotherapy. Parks et al. [ 18] observed one of 17 patients with myelopathy as a result of combined treatment of radiation and WBH. Guidelines for retreatment tolerance of normal tissues may be obtained from laboratory investigations using animals. Information on the tolerance of previously irradiated tissues to hyperthermia alone or combined with radiotherapy is available for rapidly proliferating tissues, skin [ 1,9,10,11,29,30] and intestine [7,8]. These studies show that the thermal response
Address for correspondence: P. Sminia, Department of Radiotherapy, Academisch Medisch Centrum, Meibergdreef9, 1105 AZ Amsterdam, The Netherlands.
61 may be increased if heat was applied to previously irradiated tissue. No data are available on retreatment tolerance to hyperthermia of the previously irradiated spinal cord. In the studies of Miller et al. [ 15], Neville et al. [16] and Sminia et al. [24], hyperthermia was applied shortly after local irradiation of the rat spinal cord to investigate the effects of heat on the radiation response. Goffinet et al. [3] studied myelitis of the mouse thoraco-lumbar spinal cord after heat treatment applied shortly before and after irradiation, and 2 weeks prior to irradiation. In the present paper, we report on the effects of irradiation of the rat cervical spinal cord (dose range 15-26 Gy) on the acute response to local hyperthermia (50-90 min that 42.3 °C) applied to approximately the same site 90 days later. The heat treatment was applied at a time where radiation-induced damage was progressing but still not expressed. The influence of hyperthermia on the expression of "early delayed" and "late delayed" radiation damage is investigated. Materials and methods
Animals Female Wistar rats, aged 12-13 weeks (obtained from Harlan CPB, Zeist, The Netherlands) were used in all experiments. Animals were anaesthetized with sodium pentobarbital (Nembutal ®, 50 mg/kg body weight, i.p.) about 15 rain before heat treatment. The analgesic pentazocine (9 mg/kg body weight, i.p.) was administered 5-10 min later. Before irradiation, light anaesthesia of the animals was obtained using about half a dose of the pentobarbital; no pentazocine was given. After both the irradiation and heat treatment, animals recovered from anaesthesia in an infant incubator (Air-Shields Europe, Shannon, Ireland) at a temperature of 32 °C and a relative humidity of about 70 ~o. The number of animals included in this study was 251; 35 were treated with hyperthermia alone, 58 with X-rays alone and 158 with X-rays followed by hyperthermia after 90 days. Treatment groups consisted of 6-13 animals.
Irradiation and dosimetry The cervical spinal cord region of the animals was irradiated from the dorsal side with a Siemens Stabilipan X-ray generator operated at 250 kV and 15 mA, filtered with 1 mm Cu. The anaesthetized animal was placed in supine position on a 10 mm thick lead shield. Local irradiation of the spinal cord, cervical 5-thoracic 2 (C5-T2), was via a rectangular field of
15 mm × 20 mm. Focus to skin distance was 24.2 cm. The actual dose delivered to the spinal cord was checked by measurement both in a tissue equivalent phantom and in a sacrificed rat. First, the dose rate was checked in the rat phantom using a cylindrical ionization chamber (Baldwin and Farmer, type 2570), at a depth representative for the spinal canal. This measurement resulted in a dose rate of 326 cGy/min. Using this value for the dose rate, a sacrificed rat was irradiated with doses of 24, 26 and 28 Gy, respectively, which are doses around the EDs0 value for "early delayed" paralysis [24]. These nominal doses were verified with TLD measurements (TLD-100, Harshaw). By means of a 1.7 mm diameter catheter, three TLD rods were inserted into the spinal canal. The exact position of the rods was checked by radiography. Measurements were performed twice for each dose to eliminate possible variations due to the position of the volume relative to the radiation field. In all cases, the results showed that the difference between the doses estimated by the ionization chamber and the TLD method was less than 1 ~o. Animals were irradiated in air at doses of 15, 18, 20 and 26 Gy. The temperature in the irradiation room was kept at 26 ° C to avoid a drop in body temperature of the anaesthetized animals during the irradiation.
Hyperthermia On day 90 after irradiation, the cervical region in the rat, including the spinal cord (cervical 5 - thoracic 2) was heated using a 434 M H z microwave applicator. Technical aspects of the heating system and details on thermometry were described previously [20,21]. It was shown that the cervical part of the vertebral column could be heated whereas surrounding tissues were only slightly heated. In all experiments, the temperature was regulated using a reference thermocouple probe which was placed against one of the cervical vertebrae 6, 7 or thoracic 1. Temperature measurements inside the vertebral canal were done in animals of about 200 g as well as in animals of about 250 g (90 days older). In the experiments, the cervical vertebral column was heated for 50-90min at a reference temperature of 43.0 + 0.1 ° C (mean + SD) which resulted in an average maximum temperature of 42.3 + 0.4 °C in the vertebral canal at the level of the vertebrae cervical 5 thoracic 2 [20].
Follow-up Up to 18 months after treatment, animals were regularly inspected and neurological complications were scored using an arbitrary scoring system (Table I). The
62 TABLE I
15.
~
2
3
The arbitrary scoring system a. Score
Visible signs
Ability to turn in the Geotaxis test b
~' 10"6
1 I
"~_:~
k~~uL~
5-
0 1
No visible complications Minor neurological symptoms such as unco-ordinated use of the forelegs Reduced ability to walk on the forelegs, paresis, monoplegia Reduced ability to walk on the forelegs, paresis, monoplegia Paralysis of both forelegs Death
3/3
O-
3/3
alone
r
0 l x
-
-
90 1
180 Foil . . . .
270
360
~
450
x
N 90 days
540
p (days)
N
2/3 1/3 0/3
Used for evaluation of neurological complications resulting from treatment of the cervical region in the rat, including the spinal cord between the vertebra cervical 5 and thoracic 2. b See [20]. a
incidence of paralysis, paresis and minor neurological symptoms over this period were used as endpoints, and quantal data were used to establish dose-effect curves which were fitted by logistic regression analysis. From this analysis, the heat or radiation doses required to produce paralysis or severe paresis and minor neurological symptoms in 5 0 ~ of the animals (EDso + SE) were established using BMDP statistical software (Cork, Ireland). The modification in the response to hyperthermia produced by previous irradiation and the modification of the radiation response produced by hyperthermia were compared on the basis of these EDso values. A significant number of animals could not be followed up completely for late effects, because they developed a tumour inside the treated volume [23]. Results
Figure 1 shows the time course of the incidence of paralysis as a result of treatment of the rat cervical spinal cord. In the 90-day period between X-rays and heat, no neurological complications were observed in the irradiated animals. The time course (Fig. 1) showed three more or less distinct peaks in the incidence of neurological symptoms: (1)the acute response to hyperthermia (90-120 days); (2) the "early delayed" response to irradiation (150-300 days) and, (3)the "late delayed" response to irradiation (300-540 days). The distinction between the "early delayed" and "late delayed" radiation response at day 300 was based on the bimodal pattern of the radiation response in the data published by Sminia et al. [24].
Fig. 1. Percentage of paralyzed animals of the initial number of animals plotted as a function of the follow-up period. Animals were irradiated at day 0 (15, 18, 20 and 26 Gy) and heated at day 90 (50, 60, 75 and 90 min at 42.3 °C). (1)Acute response to hyperthermia; (2) "early delayed" response to radiation; (3) "late delayed" response to radiation. The total number of animals included in this histogram is 51.
The effect of previous irradiation on the acute response to hypertherrnia
Figure 2 shows the percentage of animals with minor neurological symptoms (score 1), paralysis (score4) and lethality (score 5) one day after heat treatment at 42.3 °C as a function of X-ray dose. The percentage of animals with minor neurological symptoms was increased in animals that were heated 90 days after irradiation of the spinal cord with 18, 20 or 26 Gy, relative to control animals. This was most pronounced at lower heat doses, 50 and 60 rain at 42.3 °C. At 75 and 90 min at 42.3 °C, the percentage of animals with minor neurological symptoms was already 40-55~o, even in the non-irradiated groups. After 26 Gy followed by 90 min at 42.3 °C 90 days later, the percentage of animals with severe complications, such as paralysis and lethal damage, was greatly increased. For all other treatment groups, the percentage of animals with paralysis and lethality was slightly enhanced by previous irradiation. Figure 3 shows that in previously irradiated animals a shorter heating time is required to obtain neurological complications. The EDso, day 1, for heat alone was 74 + 2rain at 42.3 °C (mean + SD). After 15, 18, 20 and 26 Gy previous irradiation the EDso was 7 8 + 3 , 5 7 + 7 , 6 5 + 4 and 5 5 + 5 m i n , respectively (i.e. reductions of 12-26 %). Data from day 14 (Fig. 4) and 60 (Fig. 5) after hyperthermia show that recovery from heat-induced neurological symptoms took place. All non-irradiated animals, and almost all animals previously irradiated with 15 and 18 Gy, had completely recovered. Animals previously irradiated with 20 or 26 Gy were however, less able to recover, e.g. 50~o (3/6) of the surviving animals preirradiated with 26 Gy, still suffered from minor neurological symptoms. Animals from the 26 Gy
63 Pretrradiated
[~::~0 Gy ~ 1 5
spinal
Gyl~fl8
cord
loo
[]
Gylg;X:~20 Gy~E~I26 GY
8O Symptoms
(score 1 ) Day
1
70 60 5O 40 o
3O
-6
2O
g
10
g
0
z_ i1)
50
60
75
90
50
4O Paralysis o
( s c o r e 4 ) Day 1
30
E
II 50
20
60
Duration of t r e a t m e n t
75
*
EDSO
ml A O []
sham 18Gy 20Gy 26 Gy
9O
at 4 2 3 °C (rain.)
o 10
0
50
60
75
90
50 Lethality
(score
5)
Day 1
40
1° i o L - - - 50
Duration
60 of t h e
Fig. 3. Percentage of animals with neurological complications (score 1-4 pooled and corrected for lethality) one day after hyperthermia as a function of the duration of the heat treatment at 42.3 °C. Animals were sham treated or X-irradiated at a dose of 15, 18, 20 or 26 Gy, 90 days before hyperthermia.
75 heat
treatment
90 at 42.3 °C
Fig. 2. Percentage of animals with minor neurological symptoms, paralysis and lethality one day after hyperthermia applied 90 days after irradiation of the spinal cord. Treatment groups consisted of 7-11 animals.
group that were not heated, did not yet express neurological symptoms at this time. Lethality can occur after hyperthermia alone probably as the result of respiratory paralysis. Figures 2, 4 and 5 show that lethality was increased in 26 Gy preirradiated animals relative to other groups. If animals died as a result of heat injury to the spinal cord, this occurred within 14 days after hyperthermia.
The effect of hyperthermia 90 days after irradiation on the "'early delayed" radiation response of the spinal cord Starting 6 months after X-rays, the effects of radiation treatment became apparent. "Early delayed" paralysis was not observed (with a single exception) in the 15, 18
and 20 Gy groups, either with or without heat. After 26 Gy alone, "early delayed" paralysis occurred in 64?o (7/11) of animals. Recovery from thermal damage was diminished after preirradiation of the spinal cord with 2 6 G y (Fig. 5). In these animals, heat-induced neurological complications were immediately followed by radiation-induced neurological complications. The early response was, however, not increased: 617o (17/28) paralysis. The latent period for early delayed paralysis after 26 Gy alone was 238 + 45 days (n = 7) and after 26 Gy followed by hyperthermia it was 234 + 20 days (n -- 17).
The effect of hyperthermia 90 days after irradiation on the expression of late myelitis No effect of the addition of heat 90 days after irradiation was found on the number of animals with late paralysis (Table II). The latent period before development of late paralysis varied from 322 to 522 days, mean 420 _+ 74 days (n = 15). The number and percentage of animals with late minor neurological symptoms are given in Table III. The data show that heat influences the radiation response, the Dose Modifying Factor (DMF) varied from 1.1 to 1.3. Figure 6 shows a significant shortening in latency for minor neurological symptoms, particularly at higher heat doses, 75 and 90 min at 42.3 °C, and at low X-ray doses (15 and
18 Gy).
64 Preirradioted
[~
0 Gy ~
spinal
Prei rradiated
cord
[
15 Gy ~ ' / ~ I 8 GyE;~;~ 20 Gyl~7/~ 26 Gy
Gy
BO
BO 70
spinal cord
I 0 Oy I ~ I U 15 GyI~F'z~IB G y I ~ F ~ 2 0 O y ~ 2 6
( s c o r e 1) D a y 14
Symptoms
70
60
60
50
50
40
40
30
30
Symptoms ( s c o r e 1) D a y
60
20
10 0
10 50
60
75
0
gO
5O
4O
60
75
g0
75
90
40
Paralysis (score 4 )
Day
Paralysis (score 4 )
14
30
Day
60
30
E E o 20
20
&
o
o
10
0
50
60
75
0
90
50
50 Lethality
(score 5) Day 14
Lethality
40
40
30
30
20
20
10
10
50
60
D u r a t i o n of t h e
75 heat
0
90
t r e a t m e n t at 42.3*C
50
60
15 Gy 480' 440
..~ 4 0 0
TABLE II The number of animals per total in the treatment group expressing "late delayed" paralysis*.
60
75
90
h e a t t r e a t m e n t a t 42.3 °C
Fig. 5. Percentage of animals with minor neurological symptoms, paralysis and lethality 60 days after hyperthermia applied 90 days after irradiation of the spinal cord.
520
Sham
(score 5 ) D a y
D u r a t i o n of t h e
Fig. 4. Percentage of animals with minor neurological symptoms, paralysis and lethality 14 days after hyperthermia applied 90 days after irradiation of the spinal cord.
X-Dose (Gy)
60
50
18 Gy
20 Gy x - r a y s (X)
alone
75' 42.3 "C 90 d after X 90' 42.3"C 90 d Qfter X
"~ 3 6 0 320 _~ 2 8 0 240
Min at 42.3 °C
200
20 26
0/10 2/4
50
60
75
90
1/7 2/4
1/6 2/2
2/6 3/4
0/7 1/1
*After 20 and 26 Gy followed by hyperthermia at 42.3 °C for 50-90 min, 90 days later. In the groups with 15 and 18 Gy, no animals with "late delayed" paralysis were observed.
Fig. 6. The latent period (days) for expression of "late" minor neurological symptoms after 15, 18 and 20 Gy of X-rays followed by 75 and 90 min at 42.3 °C 90 days later. The latent period was not affected after X-rays followed by 50 and 60 rain at 42.3 °C. Probability values for significanee between X-rays alone and X-rays followed by heat were calculated using the Students' t-test. *p < 0.05; ** p < 0.01. Error bars represent standard deviation.
65 TABLE III The number and percentage of animals at risk expressing "late" minor neurological symptoms a. X-Dose (Gy)
Sham
Min at 42.3 °C 50
60
75
90
15 N P 18 N P 20 N P 26 N P
3/9 33 (7-70) 4/10 40 (12-74) 4/11 36 (11-69) 3/3 100 (29-100)
3/8 38 (9-76) 2/6 33 (4-78) 4/7 57 (18-90) 2/2 100 (16-100)
3/7 43 (10-82) 3/6 50 (12-88) 5/6 83 (36-100) -
2/6 33 (4-78) 3/4 75 (19-99) 5/6 83 (36-100) 1/1 100 (0-100)
3/6 50 (12-88) 2/4 50 (7-93) 6/8 75 (35-97) -
EDso b (Gy) DMF b p-value
20.0 + 1.9 -
18.5 + 1.8 1.08 + 0.10 0.567
16.4 _+ 1.7 1.22 + 0.10 0.158
16.3 _+ 1.3 1.23 +_ 0.09 0.108
15.6 _+ 3.3 1.28 +_ 0.16 0.248
After irradiation (15-26 Gy) followed by hyperthermia at 42.3 °C for 50-90 min 90 days later. N = number per total in treatment group; P = per cent; 95 ~o confidence limits between brackets; D M F = Dose Modifying Factor. Probability values for significance were calculated using the normal distribution. b Mean _+ SE.
Discussion
We have studied the effects of hyperthermia applied to the spinal cord at a time where latent radiation damage was present. The results demonstrate that: (1) The sensitivity to hyperthermia of the previously irradiated spinal cord was enhanced (cf. Figs. 2-5) and recovery from heat-induced neurological complications was diminished (cf. Figs. 4 and 5). (2) Neither the percentage of animals expressing radiation induced "early delayed" or "late delayed" foreleg paralysis or paresis, nor the latent period was affected by hyperthermia given after 90 days (cf. Table II). (3) The data on late minor neurological symptoms indicate a heat dose related enhancement of the radiation response, this was, however, not very significant (cf. Table III). The latent period for minor neurological symptoms was significantly shortened (cf. Fig. 6). The effect of previous irradiation on the acute response to hyperthermia
Radiation damage to the spinal cord with doses above 20 Gy is characterized by slowly developing demyelination due to loss of oligodendrocytes [26,27]. Our observations show that, if hyperthermia is applied during development of radiation damage, the number of animals with neurological complications as a result of
heat is significantly enhanced and recovery from neurological symptoms is diminished. The large amount of scatter in the data on neurological symptoms after hyperthermia may be due to the temperature variations of the heated spinal cord. Some animals did not recover after 26 Gy followed 90 days later by 42.3 °C hyperthermia for 60-90 rain. In the literature, the only experimental data concerning thermal sensitivity of previously irradiated normal tissues are for skin and intestine, tissues that proliferate relatively rapidly. These data show a decreased tolerance to hyperthermia due to residual injury. The tolerance of previously irradiated skin to 70-90 rain at 44 °C is reduced by 7-43 ~o at 90 days [30] and by 36~o after 25-40 rain at 43.5 °C given 10 months after irradiation [ 11 ]. Data on mouse intestine show that 20-45 min at 43 °C applied 15 days after irradiation resulted in the same effect as hyperthermia alone at 43.5°C for 20-45 min. This enhanced sensitivity returned to normal within 120 days. A maximum reduction in the crypt cell number of 60~o was found after 60 rain at 42.3 °C, applied 15 days after 8-10 Gy [8]. This effect returned to a subthreshold level within 20 days to 7 months. Comparison of our data on retreatment tolerance of the spinal cord with the data mentioned above for the skin and intestine, shows that the thermal dose where a "memory" of previous treatment was observed differs considerably. Recovery from heat injury in both skin and spinal cord was diminished after previous irradiation.
66
The effect of hyperthermia 90 days after irradiation on the "early delayed" and "'late delayed" radiation response of the spinal cord Hyperthermia given 90 days after irradiation neither influenced the percentage of animals with early delayed radiation paralysis, nor the latent period. This indicates that, once animals had recovered from thermal damage, there was no further influence on the radiation-induced early response. Similarly, no effect ofhyperthermia was found on the percentage of animals with late radiationinduced paralysis (cf. Table III). For radiation-induced minor neurological symptoms the EDs0 was 20.0 _+ 1.9 Gy, which is in accordance with data reported by Hopewell et al. [6], Van der Kogel [26] and another study from our laboratory [24]. Minor neurological symptoms were enhanced in groups that received hyperthermia 90 days later, the maximum D M F was 1.28.
Clinical implications
ratory data show that the maximum tolerated heat dose by mammalian brain and spinal cord is much less: 50-60min at 42.0-42.5°C or at 4 3 ° C for only 10-20 min [ 12,22]. The present results indicate that for previously irradiated nervous tissue, the maximum tolerated heat dose might be about 10-30 ~'o lower than for untreated tissue. Since our data show that the late radiation response may also be enhanced with regard to minor neurological symptoms and shortening of the latent period, caution is required in application of heat to previously irradiated nervous tissue in patients.
Acknowledgements The authors would like to thank Mr. J . J . G . W . Hendriks for his skilful technical assistance and Mrs. A. van der Graaff for typing and editing of the manuscript. We gratefully acknowledge Dr. F. A. Stewart for correcting the English. We would also like to thank Mr. S. Tee for dosimetry and Mr. J. Sybrands for constructing the microwave applicator.
In clinical application of hyperthermia, the aim is generally to heat the turnout at 43 ° C for 30-60 min. LaboReferences 1 Baker, D.G., Sager, H.T., Constable, W. and Goodchild, N. The response of previously irradiated skin to combinations of X-irradiation and ultrasound-induced hyperthermia. Radiat. Res. 96: 367-373. 2 Douglas, M. A., Parks, L. C. and Bebin, J. Sudden myelopathy secundary to therapeutic total-body hyperthermia after spinal cord irradiation. N. Engl. J. Med. 304: 583-585, 1981. 3 Goffinet, D. R., Choi, K.Y. and Brown J.M. The combined effects ofhyperthermia and ionisingradiation on the adult mouse spinal cord. Radiat. Res. 72: 238-245, 1977. 4 Gonzalez Gonzalez, D., Van Dijk, J. D. P., Blank, L. E. C. M. and Rt~mke, P. Combined treatment with radiation and hyperthermia in metastatic malignant melanoma. Radiother. Oncol. 6: 105-113, 1986. 5 Gonzalez Gonzalez, D., Van Dijk, J.D.P. and Blank, L. E. C.M. Chestwall recurrences of breast cancer: results of combined treatment with radiation and hyperthermia. Radiother. Oncol. 12: 95-103, 1988. 6 Hopewell, J.W., Morris, A.D. and Dixon-Brown, A. The influence of field size on the late tolerance of the rat spinal cord to single doses of X-rays. Br. J. Radiol. 60: 1099-1108, 1987. 7 Hume, S.P. and Marigold, J. C.L. Increased hyperthermal response of previously irradiated mouse intestine. Br. J. Radiol. 55: 438-443, 1982. 8 Hume, S. P. and Marigold, J. C.L. Thermal enhancement of radiation damage in previously irradiated mouse intestine. Br. J. Radiol. 59: 53-59, 1986. 9 Law, M. P., Ahier, R. G. and Somaia, S. A long term effect of X-rays on thermal sensitivity of the mouse ear. Br. J. Radiol. 57: 729-731, 1984. 10 Law, M. P., Ahier, R. G. and Somaia, S. Thermal enhancement
of radiation damage in previously irradiated skin. Br. J. Radiol. 58: 161-167, 1985. 11 Law, M. P. and Ahier R.G. A long term effect of prior irradiation on the thermal enhancement of radiation damage in the mouse ear. Int. J. Hyperthermia 3: 167-175, 1987. 12 Lyons, B.E., Britt, R.H. and Strohbehn, J.W. Localized hyperthermia in the treatment of malignant brain tumours using an interstitial microwave antenna array. IEEE Trans. Biomed. Engin. 31: 53-62, 1984. 13 Marmor, J. B. and Hahn, G.M. Ultrasound heating in previously irradiated sites. Int. J. Radiat. Oncol. Biol. Phys. 4: 1029-1032, 1978. 14 Marmor, J.B., Pounds, D., Postic, T.B. and Hahn, G.M. Treatment of superficial human neoplasms by local hyperthermia induced by ultrasound. Cancer 43: 188-197, 1979. 15 Miller, R.C., Leith, J.T., Veomett, R.C. and Gerner, E.W. Potentiation of radiation myelitis in rats by hyperthermia. Br. J. Radiol. 49: 895-896, 1976. 16 Neville, A. J., Robins, H. I., Martin, P., Gilchrist, K. W., Dennis, W. H. and Steeves, R.A. Effect of whole body hyperthermia and BCNU on the development of radiation myelitis in the rat. Int. J. Radiat. Biol. 46: 417-420, 1984. 17 Overgaard, J. The current and potential role ofhyperthermia in radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 16: 535-549, 1989. 18 Parks, L.C., Minaberry, D., Smith, D.P. and Neely, W.A. Treatment of far-advanced bronchogenie carcinoma by extracorporeally induced systemic hyperthermia. J. Thorac. Cardiovasc. Surg. 78: 883-892, 1979. 19 Perez, C. A., Kopeeky, W., Rao, D. V., Baglan, R. and Mann, J. Local microwave hyperthermia and irradiation in cancer
67
20
21
22
23
24
therapy: preliminary observations and directions for future clinical trials. Int. J. Radiat. Oncol. Biol. Phys. 7: 765-772, 1981. Sminia, P., Haveman, J., Wondergem, J., Van Dijk, J. D. P. and Lebesque, J.V. Effects of 434 MHz microwave hyperthermia applied to the rat in the region of the cervical spinal cord. Int. J. Hyperthermia 3: 441-452, 1987. Sminia, P., Van Dijk, J. D. P. and Haveman, J. Developpement d'un radiateur coaxial de forme circulaire a microondes, 434 MHz, permettant une hyperthermie localisee profonde: resultats d'application dans la moelle epiniere y comprise chez les rats. Innovat. Technol. Biol. Med. 9 (no 2 special): 79-89, 1988. Sminia, P., Haveman, J. and Ongerboer de Visser, B.W. What is a safe heat dose which can be applied to normal brain tissue? Int. J. Hyperthermia 5: 115-117, 1989. Sminia, P., Jansen, W., Haveman, J. and Van Dijk, J. D.P. Incidence of tumours in the cervical region of the rat after treatment with radiation and hyperthermia. Int. J. Radiat. Biol. 57: 425-436, 1990. Sminia, P., Haveman, J., Van Dijk, J, D. P. and Hendriks, J, J. G.W. Enhancement by hyperthermia of the "early delayed" and "late delayed" radiation response of the rat cervical spinal cord. Int. J. Radiat. Biol. 59: 259-271, 1991.
25 Valdagni, R. and Amichetti, M. Clinical hyperthermia: five year's experience. Strahlenther. Onkol. 163: 443-445, 1987. 26 Van der Kogel, A.J. Late effects of radiation on the spinal cord. Dose-effect relationships and pathogenesis. Thesis, University of Amsterdam, Edited by Uitgeverij Meinema, Delft, 1979. 27 Van der Kogel, A.J. The cellular basis of radiation-induced damage in the central nervous system. In: Cytotoxic Insult to Tissue. Effects on Cell Lineages, pp. 329-352. Editors: C. S. Potten and J. H. Hendry. Churchill Livingstone, Edinburgh London - Melbourne - New York, 1983. 28 Van der Zee, J., Van Rhoon, G. C., Wike-Hooley, J. L., Faithfull, N. S. and Reinhold, H. S. Whole-body hyperthermia in cancer therapy: a report of a phase I-II study. Eur. J. Cancer Clin. Oncol. 19: 1189-1200. 29 Wondergem, J. and Haveman, J. The response of previously irradiated mouse skin to heat alone or combined with irradiation: influence of thermotolerance. Int. J. Radiat. Biol. 6: 539-552, 1983. 30 Wondergem, J. and Haveman, J. The effect of previous treatment on the response of mouse feet to irradiation and hyperthermia. Radiother. Oncol. 10: 253-261, 1987.
