Hyperthermic injury versus crush injury in the rat sciatic nerve: a comparative functional, histopathological and morphometrical study
J .F. H o o g e v e e n 1, D. T r o o s t 2, j . W o n d e r g e m 3, A . H . W . van de r K r a c h t i a n d J° H a v e m a n i
I Department of Radiotherapy, Unirersity ofAmsterdam, Academisch Medisch Centram, Amsterdam (The Netherlands), 2 Department of Pathology, Subdirision Neuropathology, Aeademisch Medisch Centrum, Amsterdam (The Netherlands) mzd 3 Department of Clinical Oncology,
Academisch Ziekenhuis Leiden, Leiden (The Netherlands)
(Received 31 May, 1991) (Revised, received 17 September, 1991)
(Accepted 24 September, 1991)
Key words: Local hyperthermia; Crush; Sciatic nerve; Neuropathy; Histology; Vasculature
Summary
Functional and morphological changes of the rat sciatic nerve after local hyperthermia (30 min, 45°C) and crush treatment were compared. After hyperthermic injury nerve function loss developed in a time period of about 7 h. Nerve crush led to an immediate loss of nerve function. Nerve function loss was assessed by a motor and a sensory function test. Recovery from function loss took place in both treatment groups and was complete in 4-5 weeks. Early (within 8 h post-treatment) histopathological changes in the nerve after heating included edema, possible blood stasis and changes in the blood vessel wall, like swelling of the media. During this period some axonal changes were observed. Immediate after crushing axons were severly damaged, while many blood vessels remained normal. Within one week after both treatments, degeneration of axons and myelin was observed at the site and distal from the site of the lesion (Wallerian degeneration). Three weeks after treatment a major part of the axons had regenerated and remyelinated. Vascular changes at the site of lesion could still be observed in the heat-treated nerves. Twelve weeks after both treatments, blood vessels appeared to be normal again. Morphometrical analysis of the treated nerves confirmed the histological observations. Three and 12 weeks after treatment average axon diameters were significant smaller and average myelin sheaths were significant thinner compared to untreated nerves. These parameters did not differ significantly when the two treatment groups were compared.
Introduction
Hyperthermia has been used in conjunction with radiotherapy or chemotherapy for the treatment of malignant diseases for several years (Gonzalez Gonza- lez et al. 1988; Overgaard 1989; Herman et al. 1989). To prevent unwanted side effects of this treatment, it is necessary to investigate the effects of heat treatment on normal tissues. Knowledge about the tolerance of peripheral nerves to hyperthermia is important as pe- ripheral nerves are distributed to almost every part of the body. Peripheral neuropathy in patients during or after whole body hyperthermia were reported by Adam
Correspondence to: J.F. Hoogevcen, Dept. of Radiotherapy, Uni- versity of Amsterdam, Academisch Medisch Centrum, Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands. Tel: (33) (020) 5669111; Fax: (33) (020) 5664440.
et al. (1987), Bull et al. (1979), Gerad et al. (1984) and Selker et al. (1983), after (deep) regional hyperthermia by Perez et al. (1984) and during and after superficial hyperthermia by Scott et al. (1985).
Few experimental studies have been published con- cerning the effects of heat on peripheral nerves. A nerve conduction block at temperatures ranging from 41 to 45°C was reported by Brodkey et al. (1964) and Eliasson et al. (1986a,b). Wondergem et al. (1988) investigated the effects of local hyperthermia on the motor function of the rat sciatic nerve. A 50% loss of motor function in 50% of the animals was observed 24 h after 58, 32 and 12 min of heating at 43.0, 44.0 and 45.0°C, respectively. Thermal resistance after fraction- ated hyperthermia was shown by De Vrind et al. (1991).
Histopathologicai studies on the effects of hyper- thermia on peripheral nerves have not been reported yet. We compared the histopathological and functional changes of the rat sciatic nerve after hyperthermic
56
injury and after a crush injury in order to get a better understanding of the changes typical for hyperthermic damage in the nerve. Morphometric methods were used to quantify nerve regeneration.
Materials and methods
Animals Female rats of an inbred Wistar strain, obtained
from the GDIA (Gemeenschappelijk Dieren lnstituut Amsterdam), were used in all experiments. The rats (170-200 g) were housed 8 to a cage and were kept under conventional laboratory conditions. Before all treatments, rats were anaesthetized with a mixture of 98% oxygen and 2% halothane (Fluothane) via a semi- open system. The sciatic nerve of the left hind leg was exposed by cleaving the overlying muscles. A nerve segment from 6 mm distal to the sciatic notch was either crushed or heated. After treatment the wound was closed with surgical staples. Within 30 rain after operation animals were able to react normally. Experi- mental groups consisted minimally of 6 animals.
Heating technique Details of the heating method have been described
earlier (Wondergem et al. 1988). Briefly, heat was administered to the nerve with a thermostatically con- trolled thermode. The exteriorized sciatic nerve was placed on the thermode and immersed in paraffin oil in order to prevent drying of the nerve. Temperature of the thermode was measured with a copper-constan- tan thermocouple. Control studies showed that the temperature in the nerve is equal to the temperature in the thermode +0.20C. In this study the nerve was heated over a length of 5 mm at 44°C or 45°C during 30 rain. These temperatures were chosen on basis of the results presented by Wondergem et al. (1988). A linear temperature/time relationship (heating time has to be changed with a factor 2 when temperature has changed by 1 ° C) was shown when motor function loss was investigated after heating the rat sciatic nerve at differ- ent temperatures during different heating times. Con- trol nerves were sham-treated at 38°C during 30 rain.
Nerve crush The exteriorized sciatic nerve of anaesthetized rats
was crushed using a haemostat (jaw width 3.5 ram) by keeping the nerve between the jaws for 30 s.
The 'toe-spreading' test (motor function) Loss of motor function was assessed using a modi-
fied version of the toe-spreading test (De Medinaceli et al. 1982; Terzis and Smith 1987). We measured the toe-spreading (distance from first to fifth digit) from both hind paws from walking tracks (Wondergem et al.
1988) every half hour after treatment up to 8 h after treatment, and every two days after treatment to inves- tigate recovery from motor function loss. The relative toe spreading of the left foot (heat-treated nerve or crushed nerve) was calculated with the untreated right foot as control. A motor function loss of 100% corre- lates with 30% relative toe-spreading; 50% motor func- tion loss (MFL50) correlates with a relative toe-spread- ing of 65%.
The 'foot sole stimulation test" (sensory function) Sensory function after treatment was investigated by
the 'foot sole stimulation test' which was described by De Koning et al. (1986). The foot sole of a rat was stimulated by means of a small electric current. The strengths of current for testing ranged from 0.1 to 0.6 mA. A rat without loss of sensory function will with- draw its foot instantaneously as soon as an electric stimulus is given. No withdrawal of the foot after stimulation at 0.6 mA indicates a loss of 100% of the sensory function. Withdrawal at intermediate strengths of current is scored as incomplete function loss. A sensory function loss of 50% (SFL50) correlates with no withdrawal at 0.3 mA.
week, 3 weeks and 12 weeks after treatment. The site of heat treatment could be identified easily (slightly swollen, a different colour and connective tissue forma- tion at the site of heating). After crush injury ner~.es were examined 4 h, 1 week, 3 weeks and 12 week:" aft.~r treatment. Before dissection of the treated nerv,; part, the animals were anaesthetized with pentobarbizone sodium (Nembutal, 50 mg/kg i.p.). The treated part of the nerves and the peroneal nerves 10 mm distal from the treated area were removed and stretched on a piece of filter paper and put in Karnovsky fixative containing 0.1% picrine acid (pH 7.3) for at least 24 h. Nerves were post-fixed 2 h in osmium tetroxide 1%, stained with uranyl acetate, dehydrated in acidified 2,2-dimethoxypropane (DMP) and embedded in Epon. After hardening, semi-thin cross-sections through the middle of the dissected nerves were cut (1 /~m) and stained with toluidine blue 0.2% in a sodium carbonate solution 2.5%.
Morphometric analysis The cross-sections used for histological investigation
were also used for morphometric analysis. Photographs (with a final magnification of x 2000) were taken at random from the tibial branch of the nerve 3 and 12 weeks after treatment. The edge of the tibial nerve and large blood vessels were excluded from photographing. The photographs were put on a Genius Tablet (type GT-1212), connected with a personal computer. Not all
fibers on a photograph were measured, only those which fitted in the rectangle (7 × 11 cm) on a sheath we put on each photograph. Fibres touching bottom line and the left side line of the rectangle and cross- sections through the nodes of Ranvier and Schmidt- Lanterman incisures were excluded from counting. With the use of a 'mouse' (connected with the tablet), the outlines of axons and the outlines of axons plus their myelin sheaths were followed. These measured myelinated fibres became visible on the computer screen and axon area and axon plus myelin sheath area were calculated automatically. From these values a special computer program calculated the axon diame- ter and myelin sheath thickness. From these measure- ments we derived the average value per rat and then per treatment group.
The probability values for significance between the individual groups were obtained using the HP-21S cal- culator.
All measured myelinated axons in a group were plotted in scatter diagrams to illustrate the myelinated axon composition of nerves 3 and 12 weeks after treat- ment.
Results
Nerce function after treatment After hyperthermic injury (30 min, 4 5 ° 0 both motor
function and sensory function slowly decreased (Fig. 1). It took approximately 7 h to disappear completely. Heating for 30 rain at 44°C resulted in an average motor function loss of ±20% 24 h after treatment. After nerve crush these functions disappeared immedi- ately. The time course of recovery from motor function loss was the same either after heat t reatment (30 min, 45°C) or nerve crush (Fig. 2), Motor function recovered completely within 9 days when nerves were heated for 30 min at 44°(]. We could not detect motor or sensory, function loss in the sham-treated group (30 rain, 38°C).
Histopathology
Hyperthermia Histological changes in the nerve were absent in the
sham-treated group (Fig. 3). The first histopathological changes at the site of a heat injury of 30 rain at 45°C were merely changes in the blood vessel structure, such as loosening and swelling of the advcntitia and swelling of the media (Fig. 4). Most blood vessels were packed with erythrocytes (blood stasis?). Furthermore the space around some blood vessels was enlarged and axons were lying further apart from each other (edema in the nerve). Eight hours after heating most axons were normal, although some axonal changes occurred. These changes include disruption of myelin sheaths, flattened
.J LL
1 o[ 57
r . , . I . I • , • J . i . , , i , l , , . i •
0 1 2 3 4 5 6 7 8 9 10 11 12
TIME AFTER TREATMENT (hrs)
Fig. !. Motor function loss (MFL) and sensory function loss (SFL) plotted against time after treatment. +, motor function loss after hyperthermia; v, motor function loss a[ter crush; e, sensory function loss after hyperthermia; o, sensory function loss after crush. Each
point represents the mean of 8 rats _+ S.D.
myelin profiles, a few 'darker ' axons and some swollen axons. Twenty-four hours after heating, blood vessel damage was still present and axonal changes (disrup- tion of myelin sheaths, axonal swelling and fragmenta- tion, and the presence of many 'dark' axons) were widely spread in the nerve (Fig. 5) The peroneal nerve 10 mm distal from the heated part of the nerve showed no histopathological changes 24 h after heating.
Within 1 week, complete degeneration of axons and myelin sheaths were seen, both distal and at the site of the lesion (Wallerian degeneration). At this stage there was an abundant cellular reaction existing of macrophages, lymphocytes and (proliferating) Schwann cells in the treated nerve. Three weeks after hyperther- mia many small remyelinated axons were observed (Figs. 6 and 7). Sometimes clusters of small myelinated axons were seen, indicating that sprouts of several proximal nerve fibers were passing the site of the lesion within the original endoneurial tube. At this time changes in the vessel wall structure, such as
125 - -
i 100"
75.
~ 5 0
25' LU o
0 7 14 21 28 35 42 49 56 6'3 70
TIME A~=TER TREATMENT (days)
Fig. 2. Relative toe-spreading after hyperthermic treatment (+) or after crush (o) plotted against time after treatment. Each point
represents the mean of at least 6 rats_+ S.E.M.
58
Fig. 3. Sham-treated nerve. Normal nerve structure, x 400.
massive thickening of the media, are still present at the site of the heat lesion. Twelve weeks after hyperther- mia there was a continuous remyelination of axons and the axon size was larger. BIoodvessels seemed to be essentially normal at this time. Twenty-four hours after heating for 30 rain at 44°C, edema was observed at the site of treatment. Most blood vessels and axons ap- peared to be normal.
Crush Axons of crushed nerves were severely damaged
within 4 h at the site of injury (Fig. 8). Many blood vessels were normal. Some vessels had a thickened media and some vessels were packed with erythrocytes (blood stasis?). Occasionally we observed a hemor- rhagic lesion. Within one week post-treatment the sci- atic nerve at the site of, and distal from the lesion, had
Fig. 4. Four h after hyperthermia. Swelling of the media, loosening of the adventitia, perivascular edema, some disrupted myelin sheaths (arrows). x400.
59
Fig. 5. Twenty-four hours after hyperthermia. Disintegration of axons and myelin sheaths (arrows). Swelling of the media. ×400.
completely degenerated (WaUerian degeneration) as shown in Fig. 9. Three weeks after crushing (Fig. 10) no qualitative differences in regeneration and remyeli- nation of axons could be observed when compared to myelinated axons of heated nerves. Changes in blood vessel structure could not be detected.
Morphometric analysis Three weeks after crushing or heating the sciatic
nerve, average axon diameter and myelin sheath thick- ness were significantly smaller when compared to sham-treated nerves (Table 1). A significant increase in average axon diameter and myelin sheath thickness was found twelve weeks after treatment compared to the 3-week values. At this time average axon diameter is smaller and average myelin sheath thickness is thin- ner compared to sham-treated axons. There is no sig- nificant difference between average axon diameter and
Fig. 6. Three weeks after hyperthermia. Regeneration of small myelinated nerve fibres (arrows), massive vessel wall thickening, possible recanalisation of a thrombus, x 400.
60
Fig. 7. Three weeks after hyperthermia. Extreme thickening of media, debris of unknown origin in vessel wall (arrow). x400.
myelin sheath thickness of crushed and heated nerves. Regeneration rate is apparently the same in both treat- ment groups.
The scatter diagrams composed of all measured myelinated axons (Fig. 11) illustrate the average values listed in Table 1. Three weeks after crush or hyperther- mia only small axons with thin myelin sheaths are present at the site of lesion. Twelve weeks after treat- ment more larger axons and thicker myelin sheaths
were measured in both treatment groups, but myeli- hated axon composition at this time still differs from sham-treated nerves.
Discussion
The effects of crushing or heating the rat sciatic nerve appeared to be completely different the first
Fig. 8. Four h after crush. Severely damaged axons, myelin disruption and normal blood vessel, x400.
61
Fig. 9. One week after crush. Complete degeneration of the nerve, normal blood vessels, many macrophages (arrows) present. ×400.
hours after treatment. A direct loss of motor and sensory function was observed after nerve crush, whereas after hyperthermia a gradual function loss took place.
The severe axonal damage after nerve crush is prob- ably responsible for nerve function loss within 30 min. The first effects of h~erthermic injury do not concern axons and their myelin sheaths, but this treatment affects the vasculature. Degenerative axonal and myelin
changes appear later. From our study we do not know the effects of heating on unmyelinated fibres.
The (damaging) effects of hyperthermia on vascula- ture and vascular permeability in tumors and normal tissues, such as muscle and skin, have been reported by many authors (Emami et al. 1981; Badylak et al. 1985; Reinhold and Endrich 1986; Jansen and Haveman 1990). Damage to endothelial cells by heat might lead to changes in vascular permeability. Powell and Myers
Fig. 10. Three weeks after crush. Regeneration of nerve fibers and remyelination, normal blood vessel structure, x 400.
62
(1989) described how in several experimental models of demyelinating and axonal disease (e.g. experimental allergic neuritis) an altered vascular permeability pre- cedes the development of edema and nerve fibre in- jury, Nerve edema can alter the endoneurial microenvi- foment by increasing pressure, reducing blood flow, or altering electrolyte concentrations in the endoncurial fluid.
Milder heat treatments, for example 30 rain at 44°C, caused a lesser degree of motor function loss than heat treatment for 30 min at 45°C. Histological examination 24 h after heating 30 rain at 44°C showed edema in the
nerve, but no blood stasis (as found after heating for 30 min at 45°C). Ischemia, as a result of blood stasis, might exacerbate neuronal damage after hypcrthermia. Blood flow can bc decreased or can even be arrested by hyperthcrmic treatment (Song et al. 1984). If the restoration of an adequate circulation is delayed, is- chemia reaches a point where it leads to breakdown of the axon and the sequence of changes that constitute Wallerian-like degeneration. Recovery from function loss will be complete if the local circulation is restored before more advanced structural changes in the nerve develop (Sunderland 1978).
S.O0
£ I l 2.SO' ==
shamltreated
0.00 0 10
+ +
+ + ++÷ + +
4. 4- +÷ + + +
++ + ++-~ ~+,+, * + **tf~:~*+ + +
)+* .
;, ; ; o Axon diometer (pro)
8"001 3 wks after 8.00
I hyperthermla
2"601
3 w k s a f t ~ Cru~r~
_=, , i=
2.80' + ++
"~ '++ + I l | "+ + *÷+ ÷
, T e l i ~ . " ~ ÷ .:- l
I o . o o o ~ 0.% _ ;
Axon dimmoter (Jura) Axon-dimn~tor (/Jrn) .°01 '0'[ 12 wks after c rush
i i
I " ° 1 ÷÷"" i = " ~+
| + . ~ + + + + +
0 " 0 0 / - " - " 0 . 0 0 1 I - • - - l l
o = , ; ; ,o o + + . ; ',o
Axon dilln~t4r tam) Axon-dlll i l4,tlr (Min)
Fig. I1, This set of scatter diagrams relates myelin sheath thickness t~ axon diameter. Because no significant differences in average axon diameter and myelin sheath thickness between rats in a group were calculated, all measured axons were put together in these plots. Three weeks after hyperthermia (n = 9) or crush (n = 6) only small axons with thin myelin sheaths were present in the nerves at the side of injury. After twelve weeks (crush: n = 5; hyperthermia: tr = 6) more larger axons and thicker myelin sheaths were measured but myelinated axon variance within the
nerve still differs from sham-treated nerves (n = 4).
TABLE 1
AVERAGE AXON DIAMETER AND MYELIN SHEATH THICKNESS FROM CRUSHED NERVES, HEATED NERVES (BOTH 3 AND 12 WEEKS AFTER TREATMENT) AND SHAM- TREATED NERVES
Significance between different groups were calculated using the t-test. All calculated values are significantly different from the con- trol values (P < 0.005). The 3- and 12-week treatment groups are also significantly different from each other ( P < 0.05).
Other experimental studies concern the effects of hyperthermia on peripheral nerve conduction. Com- paring these studies did not yield an agreement about heat tolerance and sensitivity of different nerve types to hyperthermia (Brodkey et al. 1964; Eliasson et al. 1986a,b). Klumpp and Zimmermann (1980) found an irreversible conduction block in myelinated fibres 110 min after heating at 46.5°C, or 10 rain at 51°C. They consider that decomposition of nerve fibre con- stituents, which are essential for impulse conduction, is the dominant cause of conduction block of myelinated fibres. Local ischemia plays a subordinate role in their opinion. In one experiment using a temperature of 45.8°(2 during 20 min a complete arrest of blood circu- lation was observed in the heated region of the nerve after 20 rain of heating, which, however, did not affect the myelinated and unmyelinated fibre action poten- tials during a subsequent observation of 4 h. Our observations show normal motor and sensory function up to approximately 4 h after heating while blood vessel damage is prominent. If Klumpp and Zimmer- mann (1980) had measured action potentials longer after heating, they might have detected decreased ac- tion potentials after heating.
Similarities in recovery after hyperthermia and nerve crush were demonstrated by morphometric analysis of the nerves and the 'toe-spreading test'. Apparently regeneration capacity is not influenced by crushing or heating a nerve. Neither one of both damaging events causes severe scar formation which would hinder pene- tration of regenerating axons in the endoneurial tubes, and therefore inhibit recovery.
Twelve weeks after both treatments nerve function is fully restored. However, morphometric analysis of treated nerves at 12 weeks shows significantly smaller average axon diameters and thinner myelin sheaths when compared to control values.
In conclusion we assume that the first target of hyperthermia in rat sciatic nerve is the vasculature.
63
Edema together with a transient ischemia are probably responsible for the complete loss of nerve function and subsequent degeneration.
Acknowledgements The authors wish to thank the co-workers of the subdivision of Electronmicroscopy, department of Pathology, for technical advice and assistance. We also thank Mr. A. Klop for his valuable help in developing software for morphometry. This work was supported by a grant from the Netherlands Cancer Foundation (Koningin Wilhelmina Fonds).
References
Adam, A.M., Hughes, R.A.C., Payan, J. and McColl, l. (1987) Peripheral neuropathy and hyperthermia. Lancet, 30: 1270-1271.
Badylak, S.F., Babbs, C.F., Skojac, T.M., Voorhees, W.D. and Richardson, R.C. (1985) Hyperthermia-induced vascular injury in normal and neoplastic tissue. Cancer, 56: 991-1000.
Brodkey, J.S, Miyazaki, Y., Ervin, F.R. and Mark, V.H. 0964) Reversible heat lesions with radiofrequency current. J. Neuro- surg., 21: 49-53.
Bull, J.M., Lees, D., Schuette, W., Whang-Peng, J., Smith, R., Bynum, G., Atkinson, R., Gottdiener, J.S., Gralnickm H.R., Shawker, T.H. and DeVita, V.T. 0979) Whole body hyperther- mia: a phase ! trial of a potential adjuvant to chemotherapy. Ann. Int. Med., 90: 317-323.
De Koning, P., Brakkee, J.H. and Gispen W.H. (1986). Methods for producing a reproducible crush in the sciatic and tibial nerve of the rat and rapid and precise testing of return of sensory func- tion: beneficial effects of melanocortins. J. Neurol. Sci., 74: 237-246.
De Medinaceli, L., Freed, W.J. and Wyatt, R.J. (1982) An index of the functional condition of rat sciatic nerve based on measure- ments tnade from walking tracks. Exp. Neurol., 94: 788-790.
De Vrind, H.H., Haveman, J. and Wondergem, J. (1991) Induction of thermal tolerance of rat sciatic nerve by mild hyperthermia. Int. J. Rad. Biol., 59: 251-259.
Eliasson, S.G., Monafo, W.W., Nussbaum, G. ann OIsen, L. (1986a) Differential effects of in vitro heating on rat sciatic nerve branches and spinal nerve roots. Exp. Neurol., 93: 57-66.
Eliasson, S.G., Monafo, W.W. and Meyr, D. (1986b) Potassium ion channel blockade restores conduction in heat-injured nerve and spinal nerve roots. Exp. Neurol., 93: 128-137.
Emami, B., Nussbaum, G.H., Hahn, N., Piro, A.J., Dritschilo, A. and Quimby, F. (1981) Histopathologic study on the effects of hyper- thermia on microvasculature, int. J. Radiat. Oncol. Biol. Phys., 7" 343-348.
Gerard, H, Van Echo, D.A., Whitacre, M., Ashman, M., Helrich, M., Foy, J., Ostrow, S., Wiernik, P.H. and Aisner, J. (1984) Doxoru- bin, cyclophosphamide and whole body hyperthermia for treat- ment of advanced soft tissue sarcoma. Cancer, 53: 2585-2591.
Gonzalez Gonzalez, D., Van Dijk, J.D.P. and Blank, L.E.C.M. (1988) ChestwaU recurrences of breast cancer: results of combined treat- ment with radiation and hyperthermia. Radiother. Oncol., 12: 95-103.
Herman, T.S., Jochelson, M.S., Teicher, B.A., Scott, P.J., Hansen, J., Clark, J.R., Pfeffer, M.R., Gelwan, L.E., Molnar-Griffin, B.J., Fraser, S.M., Svennson, G., Borustein, B.A., Ryan, L. and Cole- man, C.N. (1989) A phase l-ll trial of cisplatin, hyperthermia and radiation in patients with locally and advanced malignancies. Int. J. Rad. Oncol. Biol. Phys., 17: 1272-1279.
Jansen, W. and Haveman, J. (1990) Histopathological changes in the skin and subcutaneous tissues of mouse legs after treatment with hyperthennia. Pathol. Res. Pract., 186: 247-253.
64
Klumpp, D. and Zimmermann, M. (1980) Irreversible differential block of A- and C-fibers following local nerve heating in the cat. J- Physiol., 298: 471-482.
Overgaard, J. (1989) The current and potential role of hyperthermia in radiotherapy. Int. J. Radiat. Oncol. Biol. Phys., 16: 535-549.
Perez, C.A., Emami, B. and Nussbaum, G.H. (1984) Regional (deep) heating, clinical studies in progress. Front. Radiat. Ther. One., 18:108-125 (Karger, Basel).
PoweU, H.C. and Myers, R.R. (1989) The blood-nerve barrier and the pathological significance of nerve edema. In: E.A. Neuvelt (Ed.), Implications of the blood-brain barrier and its manipula- tion. Vol. 1. Basic Science Aspects. Plenum Medical Book Com- pany, New York, pp. 199-222.
Reinhold, H.S. and Endrich, B. (1986) Invited Review. Tumour microcirculation as a target for hyperthermia. Int. J. Hypertherm., 2: 111-137.
sient microwave-induced neurosensory reactions during superfi- cial hyperthermia treatment. Int. J. Rad. Oncol. Biol. Phys., 11: 561-566.
Selker, R.G., Jacobs, S.A, Stoller, R., Kristofik, M. and Randeli, M. (1983). On the limits of human tolerance of whole body hyper- thermia and simultaneous chemotherapy. Rad. Res, 94: 550.
Song, C.W. (1984) Effect of local hyperthermia on blood flow and microenviroment: a review. Cancer Res., (Suppl.)44, 4721s-4730s.
Sunderland, S. (1978) lschemia of nerves. In Nerve and Nerve Injuries. 2nd edn, Churchill Livingstone, Edinburgh, pp. 164-173.
Terzis, J.K. and Smith, KJ. (1987) Repair of severed peripheral nerves: comparison of the 'de Medinaceli' and standard mierosu- ture methods. Exp. Neurol., 96, 672-680.
Wondergem, J., Haveman, J., Rusman, V., Sminia, P. and Van Dijk J.D.P. (1988) Effects of local hyperthermia on the motor function of the rat sciatic nerve. Int, J. Rad. Biol., 53: 429-438.
