Leksell Gamma Knife lesioning of the rat hippocampus: the relationship between radiation dose and functional and structural damage

Abstract

The goals of the study were to determine at what dosage and after what interval impairment of hippocampal function occurs after Leksell gamma knife radiosurgery (GKS) of the rat hippocampus and to assess the associated structural changes.
Long-Evans rats were irradiated with maximum doses of 25, 50, 75, 100, and 150 Gy, and four 4-mm isocenters were used to cover the hippocampus bilaterally. The impairment of hippocampal function, which is associated with a loss of memory, was measured by testing the impairment of the rats' orientation in a Morris water maze. Changes in the irradiated tissue were measured using magnetic resonance imaging (Bruker 4.7/20 experimental spectrometer). The data were compared with histologically demonstrated changes. Significantly higher incidences of edema, necrosis, and behavioral changes were observed following administration of doses higher than 50 Gy. No edema, necrosis, or behavioral changes were observed when doses were 25 Gy.
It would seem that rats can be used for experiments involving the induction of complex brain lesions by using four 4-mm isocenters. Testing retention memory for behavioral changes after bilateral GKS of the whole hippocampus proved insensitive; acquisition memory should be tested to assess functional changes of hippocampus. Significantly higher incidences of edema, necrosis, and behavioral changes were observed for doses higher than 50 Gy. There seems to be a therapeutic window during which doses may affect epilepsy without impairing the memory of the rat.

Full text

ESPITE limited experience with this procedure, ra-
diosurgery for epilepsy represents a promising
treatment.1,3,8,20,21 An experimental basis for a ra-
tional application of this method, however, must be estab-
lished in detail. Previous experimental radiosurgical stud-
ies in animals have been focused on histopathological
changes of normal or neoplastic tissue.2,10–12,14 Experimen-
tal studies on the anticonvulsive effect of radiosurgery
applied in a rat epilepsy model, however, have been pub-
lished during the last 2 years.6,13,15 Radiosurgery performed
using subnecrotic doses has been shown to control epilep-
sy in a rat model without causing subsequent behavioral
impairment.13 Only unilateral lesioning of the dorsal hip-
pocampus has been studied. The aim of the present study
was to clarify at what dosage level and after what latency
interval impairment of hippocampal function occurs fol-
lowing radiosurgery of the rat hippocampus. The dose was
delivered using the Leksell gamma knife, and associated
structural changes were also studied.
Materials and Methods
The experiment was performed with the approval of the ethical
committee of the Na Homolce Hospital and the Animal Care Com-
mittee of the 2nd Medical Faculty of Charles University, Prague.
Radiosurgical Lesioning of the Hippocampus
Bilateral irradiation of the hippocampus was performed in 64
Long–Evans rats. Their age at the time of irradiation was 3 months,
and their weight ranged from 320 to 380 g. Intraperitoneal
thiopenthal (40 mg/kg) was used for anesthesia. Rats were fixed to
a specially designed MR imaging–compatible stereotactic apparatus
(Fig. 1). Stereotactic MR images were then obtained using a
Siemens Expert 1-tesla imager in which a T1-weighted three-dimen-
sional mode was used. The reconstructed slice thickness was 0.8
mm and the matrix was 154 256. The images were exported to
the GammaPlan treatment planning workstation, and radiosurgical
lesions were made.
A pilot group of six rats underwent bilateral irradiation. The dose
was delivered to the region of the dorsal hippocampus alone. Two
4-mm collimator isocenters were used. The six rats were divided
into two groups of three, which received a maximum dose of 100
and 150 Gy, respectively. In all the remaining rats the entire hip-
pocampus was covered bilaterally with the 70% isodose by using
four 4-mm isocenters.
J Neurosurg (Suppl 5) 97:666–673, 2002
666
Leksell gamma knife lesioning of the rat hippocampus: the
relationship between radiation dose and functional and
structural damage
ROMAN LIS˘C
˘ÁK, M.D., PH.D., VILIBALD VLADYKA, M.D., PH.D., JOSEF NOVOTNY
´ JR., M.SC.,
GUSTAV BROZ˘EK, M.D., D.SC., KATER
˘INA NÁME˘STKOVÁ, M.D., VLADISLAV MARES˘, M.D., D.SC.,
VÍT HERYNEK, M.SC., PH.D., DANIEL JIRÁK, M.SC., MILAN HÁJEK, M.SC., D.SC.,
AND EVA SYKOVÁ, M.D., D.SC.
Department of Stereotactic and Radiation Neurosurgery, Na Homolce Hospital; Department of
Physiology, Charles University; Institute of Physiology; Academy of Sciences; MR Laboratory; and
Institute of Clinical and Experimental Medicine, Prague, Czech Republic
Object. The goals of the study were to determine at what dosage and after what interval impairment of hippocampal
function occurs after Leksell gamma knife radiosurgery (GKS) of the rat hippocampus and to assess the associated
structural changes.
Methods. Long–Evans rats were irradiated with maximum doses of 25, 50, 75, 100, and 150 Gy, and four 4-mm
isocenters were used to cover the hippocampus bilaterally. The impairment of hippocampal function, which is associ-
ated with a loss of memory, was measured by testing the impairment of the rats’ orientation in a Morris water maze.
Changes in the irradiated tissue were measured using magnetic resonance imaging (Bruker 4.7/20 experimental spec-
trometer). The data were compared with histologically demonstrated changes.
Significantly higher incidences of edema, necrosis, and behavioral changes were observed following administration of
doses higher than 50 Gy. No edema, necrosis, or behavioral changes were observed when doses were 25 Gy.
Conclusions. It would seem that rats can be used for experiments involving the induction of complex brain lesions
by using four 4-mm isocenters. Testing retention memory for behavioral changes after bilateral GKS of the whole hip-
pocampus proved insensitive; acquisition memory should be tested to assess functional changes of hippocampus.
Significantly higher incidences of edema, necrosis, and behavioral changes were observed for doses higher than 50 Gy.
There seems to be a therapeutic window during which doses may affect epilepsy without impairing the memory of
the rat.
KEY WORDS •radiosurgery • gamma knife • rat • memory • radiation necrosis •
edema
D
J. Neurosurg. (Suppl 5) / Volume 97 / December, 2002
Abbreviations used in this paper: GKS = gamma knife radio-
surgery; MR = magnetic resonance.

A second pilot group of six animals (maximum of 100 Gy in
three and 150 Gy in three) underwent irradiation, with the 70% iso-
dose corresponding to the medial border of the hippocampus (the
coordinates of the isocenters covering the dorsal part of the hip-
pocampus were 1 mm above the center of hippocampus). The
experimental groups of rats were irradiated in such a way that the
coordinates of the isocenters covering the dorsal part of the hip-
pocampus were centered right in the middle of hippocampus (Fig.
2). The maximum dose was 150 Gy in three rats, 100 Gy in nine, 75
Gy in 12, 50 Gy in 16, and 25 Gy in 12. The rats were divided,
together with nonirradiated control rats, into groups for testing at 1,
3, 6, and 12 months after radiosurgery. A group of four rats irradi-
ated with 50 Gy, together with two control rats, were tested repeat-
edly for hippocampal function. Magnetic resonance imaging was
also performed for a 14-month period after irradiation to track the
lesion evolution over time.
Doses reported in this paper are those prescribed in the Gam-
maPlan treatment planning system. Additional study, which fo-
cused on the evaluation of inaccuracies in the stereotactic irradiation
of the rat brain, demonstrated that it is necessary to apply a correc-
tion factor for the absolute dose. This means the exact values of
applied doses may be obtained by multiplying the calculated doses
by a factor of 1.078. Detailed findings of the study, which deter-
mined this factor, will be published elsewhere.
Testing of Hippocampal Function
Behavioral testing of learning was performed in the Morris water
maze.16 The Morris water maze is a circular pool, 196 cm in diam-
eter, filled with water at a temperature between 20 and 23˚C. Escape
is possible on a hidden platform with a diameter of 10 cm, which is
submerged 2 cm below the water surface so the rat cannot see it.
The platform remained at the same position during the 5 consecu-
tive days of testing. Multiple extramaze landmarks on the walls and
ceiling of the room provided cues for the spatial relationship of the
hidden platform. The testing was performed eight times each day.
During the first two trials, the rat swam around the edges of the
pool. This is natural behavior because the greatest chance of escape
from water in nature is usually to be found at the shore. The rat then
started to cruise the pool in different directions and would finally
find the platform by chance. If this did not happen within 60 sec-
onds the rat was led to the platform by hand. The rat then was
allowed to rest on the platform for 15 seconds while trying to deter-
mine its spatial coordinates and to remember them. On the 5th day
a normal unimpaired rat would find the island in about 7 seconds if
released from any location. A rat with impaired hippocampal func-
tion would search for the island for significantly longer or even fail
to find it.9,17,18 From the eight measurements performed during a sin-
gle day, the result of the first trial was omitted and the sum of the
seven remaining latencies was calculated. The theoretical value of
the maximum sum of latencies was taken to be 420 seconds for each
rat. When the group of rats treated with the same dose was evaluat-
ed, the group mean of these sums of latencies was calculated.
Magnetic Resonance Imaging
After the functional behavioral testing was completed, all the
experimental rats together with 10 controls underwent imaging in a
Bruker Biospec 47/20 spectrometer (4.7-tesla/200 MHz) with a 200
mT/m gradient system and equipped with a surface coil made in-
house. During the examination, rats received inhalational isofluran
anesthesia (Forane; Abbott, Czech Republic). The anesthesia was
induced by a 3% concentration of the anesthetic in air and main-
tained at a concentration of 1.5% during the whole examination. A
set of T2-weighted images was acquired using a standard turbo–spin
echo sequence (TR = 2000 msec, TE = 35.1 msec, FOV = 4 cm,
matrix 256 256, slice thickness/distance = 1:1 mm). The volume
of any edema was evaluated by manual segmentation using an
MaZda program. The T2relaxation times were evaluated from a set
of T2-weighted MR images obtained using a CPMG sequence with
30 echoes (TE = 8.63 msec, TR = 2500 msec).
Histological and Cytochemical Examinations
After functional testing and MR imaging, the rats were killed by
exsanguination in deep whole-body anesthesia (intraperitoneal
injection of 5% narkamon, 0.2 ml/100 g body weight). The brains
were transcardially perfused with 0.9% NaCl and flushed with 4%
buffered formaldehyde. For basic histological examination either
the whole brains or only the right halves were postfixed in 4%
buffered formaldehyde or Carnoy solution (6:3:1) and embedded in
Paraplast-Plus (Polysciences Inc., Warrington, PA). The 7-m-
thick coronal sections were stained with toluidine blue. Cyto-
chemical analysis was performed on the coronal cryostat sections,
prepared from the remaining unfixed or prefixed left hemispheres.
It included immunocytochemical detection of glial fibrillary acidic
protein (Mabs GF-01, 1:25; Exbio, Prague, Czech Republic), synap-
tophysin, and syntaxin, (Mabs Clone SPV-38 and HPC-1; Sigma,
St. Louis, MO) revealed by biotinylated goat anti–mouse im-
munoglobulin G (Fab fragment 1:200) and extravidin–peroxidase
conjugate (1:100, both Sigma) and was terminated by a conven-
tional diaminobenzidine reaction. Morphometric evaluations,
including determination of the width of the neocortex and the hip-
pocampus, the population density of pyramidal hippocampal neu-
J. Neurosurg. (Suppl 5) / Volume 97 / December, 2002
Radiosurgical lesioning of the rat hippocampus
667
FIG. 1. Photograph showing the stereotactic MR imaging–com-
patible apparatus for rat fixation.
FIG. 2. The whole hippocampus was covered bilaterally by the
70% isodose (70, 60, 50, and 40% isodoses are shown).

rons, and the concentration of the immunocytochemical staining
product were performed using an eye-piece micrometer and an ocu-
lar grid (20 20) or the Advanced Image Data Analyzer Program
(AIDA 2.11; Raytest Isotopengeraete GmbH, Germany) applied to
digitalized microscopic pictures (Olympus Provis).
Statistical Analysis
To evaluate the changes in the rat brain over time postirradiation
three events were studied as a function of the maximum dose deliv-
ered: MR imaging–documented edema, necrotic changes verified
by the histological examination, and behavioral changes in the
Morris water maze. The log-rank test was used to determine the
level of the peripheral dose causing a significantly higher incidence
of the three aforementioned clinical events. Results were considered
to be significant if log-rank probability values were less than 0.05.
All statistical analyses in this study were performed with the SPSS
statistical software (version 10.0).
Results
Behavioral Testing of Memory Retention
Initially, the two pilot groups of rats were tested for re-
tention of memory. The rats in these groups were trained
to find the hidden platform in the Morris water maze and
the GKS was performed thereafter. The rats were tested
twice a week postirradiation until memory impairment
was detected. The first pilot groups consisted of six rats in
which only the dorsal part of the hippocampus was irradi-
ated using two 4-mm isocenters. The maximum dose was
100 Gy in three rats and 150 Gy in another three rats. Rats
irradiated with 100 Gy showed no sign of memory impair-
ment until the 78th day after GKS, and the latency of
seven daily measurements did not exceed 70 seconds.
Two rats irradiated with a maximum 150-Gy dose exhib-
ited memory impairment on the 35th day postirradiation
and one after the 50th day, when latency exceeded 70
seconds.
The second pilot group consisted of six rats and the
whole hippocampus was irradiated using four 4-mm iso-
centers. The maximum dose was again 100 Gy in three
rats and 150 Gy in the other three. The rats irradiated with
100 Gy had no sign of memory impairment until the 50th
day postradiosurgery and the latency of seven daily meas-
urements did not exceed 70 seconds. The rats irradiated
with a maximum dose of 150 Gy showed memory impair-
ment on the 32nd day postirradiation (one rat) and no
impairment was shown even 36 days postirradiation in the
other two rats.
Behavioral Testing of Memory Acquisition
Tests for memory retention induced by preirradiation
training proved to be insensitive for the purpose of this ex-
periment. Consequently, the remaining rats were tested for
acquisition of spatial memory. Irradiation of the hip-
R. Lis˘c˘ák, et al.
668 J. Neurosurg. (Suppl 5). / Volume 97 / December, 2002
FIG. 3. Graphs showing acquisition curves in Morris water maze–tested rats. Upper Left: One month after radio-
surgery. Upper Right: Three months after radiosurgery. Lower Left: Six months after radiosurgery. Lower Right:
Twelve months after radiosurgery.

pocampus was performed in untrained rats and the whole
hippocampus was irradiated using four isocenters (4-mm
collimator). The sensitivity of this functional testing was
verified in the group of three rats irradiated with the max-
imum dose of 150 Gy. This group was tested 1 month
postirradiation, and their latencies in finding the hidden
platform were two- to threefold worse compared with the
control group (Fig. 3 upper left).
The first group of 12 untrained rats was irradiated with
maximum doses of 25, 50, 75, and 100 Gy (three animals
for each dose) and tested 1 month after the radiosurgery,
when no differences were observed in comparison with
the control group (Fig. 3 upper left). The second group of
untrained rats was irradiated with maximum doses of 25,
50, 75, and 100 Gy (three animals for each dose), tested 3
months after the radiosurgery, and no differences com-
pared with the control group were observed (Fig. 3 upper
right). The third group of 12 untrained rats was irradiated
with 25, 50, 75, and 100 Gy to the maximum (three ani-
mals for each dose). Rats irradiated with the 100-Gy dose
died between the 4th and 5th month after radiosurgery and
the rest were tested 6 months after radiosurgery. No dif-
ference was observed between rats irradiated with 25 and
50 Gy and those in the control group. Rats irradiated with
75 Gy were significantly worse (Fig. 3 lower left). The
fourth group of nine untrained rats was irradiated with
maximum doses of 25, 50, and 75 Gy (three animals for
each dose). The 100-Gy dose was omitted because rats
receiving this dose had died within the 6 previous months.
Rats in this group were tested 12 months after radiosur-
gery. No differences compared with the control group
were observed in rats irradiated with the 25- and 50-Gy
doses. Rats irradiated with 75 Gy were significantly worse
(Fig. 3 lower right).
Lesion Evolution Observed on Magnetic Resonance
Imaging
A pilot study performed in rats irradiated with doses of
100 and 150 Gy and two isocenters revealed massive
lesioning with edema in the hippocampi and adjacent tis-
sue 3 months after irradiation. No structural changes were
observed on the T2-weighted images of hippocampi dur-
ing the first 6 months in rats irradiated with doses up to 75
Gy. Hippocampal T2-weighted values were the same in the
control rats (mean 65.3 msec). The rats irradiated with 25
Gy had no visible lesions by 12 months postirradiation. In
the case of the rats irradiated with higher doses (50 and 75
Gy) response to the treatment was not uniform.
All four rats receiving 50 Gy suffered edema, which de-
veloped approximately 8 months postirradiation. Its max-
imum volume was observed after 9 months. In two cases
the edema slowly disappeared, with only a small necrotic
scar remaining after 14 months. In the other two edema
persisted. Its volume did not change, although it was filled
with a connective tissue with a similar T2relaxation time
to that found in the surrounding tissue as shown in Fig. 4.
A lesion was observed in one rat 6 months after GKS with
50 Gy and its edema volume was 13.8 mm3. No edema
was observed in the group of rats evaluated 12 months af-
ter irradiation, although a hypointense signal was demon-
strated in the hippocampus of one rat, probably indicating
necrotic tissue after edema absorption.
In two of five rats irradiated with 75 Gy a large asym-
metrical lesion (edema volume 345.9 176.3 mm3) cov-
ering both hippocampi and a considerable part of the cor-
tex was revealed 6 months after irradiation (Fig. 5). The
two rats with these lesions were examined once again 14
months after irradiation, and both were found to have a
massive edema of up to 624.9 mm.3
J. Neurosurg. (Suppl 5) / Volume 97 / December, 2002
Radiosurgical lesioning of the rat hippocampus
669
FIG. 4. Evolution of the lesion in the hippocampus of a rat irradiated with a dose of 50 Gy at 6 months (A), 8 months
(B), 9 months (C), 10 months (D), 12 months (E), and 14 months (F).

Another group of three rats evaluated 12 months af-
ter irradiation had small lesions of 16.3 mm3or less. Ede-
ma was always accompanied by a considerable change of
T2relaxation time. The T2values in the edema increased
up to 148 msec, whereas those in the necrotic tissue re-
vealed after the disappearance of the edema were low at
29.8 msec.
Results of the Morris water maze test, assessed by la-
tency value, were correlated with the volume of edema at
6 and 12 months after irradiation. A significant functional
impairment was found in the rats suffering from edema.
Figure 6 shows the dependence of functional impairment
on the edema volume.
Morphology and Cytochemistry
Severe necrotic lesions were found 3 months after GKS
in the hippocampus as well as in the adjacent dorsooccip-
ital cortex and corpus callosum in all rats irradiated with
100 and 150 Gy applied with two isocenters. At that time
point, irradiation with a dose of 100 Gy, applied with four
isocenters, induced only mild reduction in the thickness of
the neocortex (30%) above the hippocampus, a slight
reduction in the number of pyramidal cells (16.3%), di-
lation of capillaries of the lateral ventricles, and a diffuse
astrogliosis. After the 50-Gy dose, the choroid plexus of
the lateral ventricles appeared slightly hypertrophic in one
of two rats.
After a 6-month postirradiation interval, the dose of 50
Gy applied with four isocenters produced small focal
necrotic lesions in one of three rats in the medial parts of
the dorsooccipital cortex and the corpus callosum above
the hippocampus. The lateral ventricles were also dilated
in some segments. After administration of a 75-Gy dose,
postnecrotic lesions and/or cavities were more extensive
and most evident in the dorsooccipital cortex. Moreover,
partial ablation occurred in some segments of the hip-
pocampus. The layer of the pyramidal cells of the hip-
pocampus was slightly disfigured, and there was a mild
astroglial reaction, which was independent of the dose.
Apart from a mild astrocytic reaction, no changes were
found after 25 Gy. The same result remained true of the
25-Gy group 1 year after radiosurgery.
At 12 to 14 months after 50-Gy irradiation, small, soli-
tary foci of necrosis occurred in one of four rats. After 75-
Gy irradiation, we found necroses and/or postnecrotic
cavities. In all the rats, the caudal parts of the hippocam-
pus were more affected than the rostral segments. The size
of the hippocampus was reduced in both the 50- and 75-
Gy groups (18 to 30%) and the pyramidal neurons
were hyperchromic. In some rats, there was dilation of
capillaries, a pericapillary edema, and distension and de-
formation of the lateral ventricles. An astrocytic reaction,
although not entirely proportional to the radiation dose,
was apparent in all these rats. No significant changes in
the expression of synapsin and syntaxin, markers of the
pre- and postsynaptic vesicles, were found except for
a less diffuse distribution of the immunohistochemical-
staining product. Representative histological findings are
shown in Fig. 7.
Significantly higher incidences of edema (p 0.001),
necrosis (p 0.001), and behavioral changes (p = 0.031)
were observed when maximum doses were in excess of 50
Gy. There was no edema, necrosis, or behavioral change
observed when the dose was 25 Gy.
Discussion
The current surgical treatment of epilepsy consists of
the resection of the epileptogenic focus. Radiosurgery has
been studied as a noninvasive alternative for the micro-
surgical removal of an epileptogenic focus. The first clin-
ical studies were performed in patients with temporal lobe
epilepsy and mesiotemporal sclerosis,20 and experimental
studies were performed in animals with focal epilepsy.6,7,15
The results of these animal experiments indicated that a
decreased frequency of epileptic seizures can be achieved
with subnecrotic doses. In a rat model of epilepsy, the
delivery of 20 Gy to the ventral part of the hippocampus
led to a decrease in the frequency of epileptic seizures
within 2 months after GKS, and this effect was even more
R. Lis˘c˘ák, et al.
670 J. Neurosurg. (Suppl 5). / Volume 97 / December, 2002
FIG. 5. Edema and necrotic lesion in the hippocampus of a rat
irradiated with a dose of 75 Gy 6 months after radiosurgery.
FIG. 6. The Morris water maze test results (latency) correlated
to the edema size. These data were acquired 6 and 12 months after
irradiation.

pronounced 4 to 6 months after radiosurgery. The dose
of 40 Gy led to a decrease in epileptic seizures within
1 month, producing a maximum effect during the 2nd
month after irradiation. The dose of 20 to 40 Gy applied
to the ventral hippocampus using two 4-mm isocenters did
not cause necrosis 10 months after the radiosurgery.6In
another study the positive effect of a 20-Gy GKS dose in
a rat with focal epilepsy was observed 2 weeks after radio-
surgery.15 Although the behavioral testing in the Morris
water maze was also performed,13 only one half of the hip-
pocampus unilaterally was irradiated using one 4-mm iso-
center. These experiments6,13,15 would appear to show that
a positive antiepileptic effect can be achieved using a sub-
necrotic, low dose of 20 Gy. This is in agreement with
the results of Barcia-Salorio and Barcia3who observed
the antiepileptic effect of a 20-Gy dose in patients with
temporal lobe epilepsy. Experience with the radiosurgical
treatment of brain tumors and arteriovenous malforma-
tions with secondary epilepsy shows that the effect on sec-
ondary epilepsy can also be achieved with relatively low
doses of radiation.
The hippocampus is a crucial structure responsible for
the generation of epileptic seizures in mesiotemporal scle-
rosis, and radiosurgery of the amygdalohippocampal com-
plex should bring relief to the patient without anatomical
ablation of this structure. Despite this promising use of
radiosurgery, the effect of radiation on the normal func-
tioning of the hippocampus, particularly in terms of mem-
ory, remains unknown. The range of doses that can sup-
press epileptic activity and not impair the memory
function of hippocampus can affect the indication criteria
for radiosurgery. Currently, patients with bilateral epilep-
tic foci are disqualified from either epilepsy surgery or
radiosurgery. The same applies to patients whose amobar-
bital and neuropsychological test results indicate that the
affected hippocampus may not be sacrificed. Knowledge
of the effects of radiosurgery on the normal hippocampus
is important in these cases. Investigation of the conse-
quences of a radiosurgical lesion induction on the normal
hippocampus is a logical and hitherto absent component
of the theoretical basis of GKS for epilepsy.
There has been no previous publication concerning the
creation of complex lesions in rat brains, which could be
used for the purposes of this experiment. Usually a single
4-mm isocenter has been used to investigate changes in rat
brains.10,12 Because the effect of radiosurgery depends on
the irradiated volume, the tolerance of the rat brain to the
use of more isocenters (and consequently a higher volume
of irradiated brain tissue) has not been investigated. One
group of investigators used two isocenters.6We have used
four 4-mm isocenters for the irradiation of the whole hip-
pocampus bilaterally and found that the rat could be used
as an animal model in a wide range of doses creating com-
plex lesions in the brain. This animal could also be tested
over at least 1 year.
Damage to the hippocampus from different causes leads
to memory impairment. Therefore, testing the memory
impairment in the Morris water maze was chosen as a sen-
sitive indicator of hippocampal function. A platform is
hidden so that the rat cannot perceive it with any of its
senses. The position of platform can be remembered only
by correlating it with its surroundings. It is generally as-
sumed that the rat (in common with humans) has a cogni-
tive map of the surrounding world in its memory.19 The rat
can perform operations by using this map, enabling it to
find the hidden platform. Navigation in the water maze is
strongly impaired after bilateral damage of the hippocam-
pus and interruption of its input. Bilateral and partial uni-
lateral loss of hippocampal function by tetradotoxin sig-
nificantly impairs the navigation of rats in a Morris water
maze.7Ibotenic acid lesions of the hippocampus selective-
ly damage the cognitive part of an animal’s orientation in
a water maze.17 Damage of the hippocampus after pro-
tracted pilocarpine seizures is proportional to the impair-
ment of navigation in a Morris water maze.4,5
J. Neurosurg. (Suppl 5) / Volume 97 / December, 2002
Radiosurgical lesioning of the rat hippocampus
671
FIG. 7. Morphological features of the dorsal and ventral hip-
pocampi in irradiated rats at 6 and 12 months. a: Dorsal hippo-
campus control specimen at 6 months. b: Dorsal hippocampus
25-Gy group specimen at 6 months. c: Dorsal hippocampus 50-
Gy group specimen at 6 months. d: Dorsal hippocampus 75-
Gy group specimen at 6 months. e: Dorsal hippocampus 25-Gy
group specimen at 12 months. f: Dorsal hippocampus 50-Gy
group specimen at 12 months. g: Ventral hippocampus 50-Gy
group specimen at 12 months. h: Dorsal hippocampus 75-Gy
group specimen at 12 months. i: Ventral hippocampus 75-
Gy group specimen at 12 months. Toluidine blue, original magni-
fication 25.

Based on previously published results,9,18 we had as-
sumed that irradiation of the dorsal part of the hippocam-
pus would be sufficient. Therefore, in the first pilot group
of rats, radiosurgical lesioning was performed using two
4-mm isocenters only in this dorsal part of the hippocam-
pus. According to published radiobiological studies, with
a single 4-mm isocenter 150 Gy produces a necrotic lesion
within 1 month and 100 Gy does the same in about half
the rats within 3 months after the radiosurgery.10,12 Rats
irradiated with a maximum of 150 Gy to the dorsal part
of the hippocampus alone showed no sign of memory im-
pairment within 1 month when tested twice a week. Two
animals, however, showed signs of impairment after 35
and 50 days, respectively. Therefore, in the second pilot
group the whole hippocampus was irradiated bilaterally
using four 4-mm isocenters, with 150- and 100-Gy doses.
The 100-Gy dose was used because we did not know if
the irradiated volume would cause diffuse postirradiation
edema in the rat brain. Although radionecrosis should
have developed within 30 days after the 150-Gy dose,
these rats showed no impairment in the 36 days following
radiosurgery. In these pilot groups, memory retention was
tested after the rats had been trained to find the hidden
platform and radiosurgery was performed thereafter. No
memory impairment was observed in rats irradiated with
150 Gy before the 35th day after radiosurgery and none in
rats irradiated with 100 Gy before 70 to 80 days after
radiosurgery, when extensive necrotic lesions involving
the neocortex were created. We found an explanation for
the late impairment in the fact that repeated testing of the
retention of memory track in the Morris water maze
served as a stimulation of the plasticity processes, and the
memory track was probably restored in residual preserved
regions in the case where the lesion developed gradual-
ly. Therefore, this test could not be viewed as sensitive
enough, and the acquisition of the memory track, rather
than its retention, was tested in untrained rats for the re-
mainder of the experiments, and irradiation was delivered
to the whole hippocampus by using four 4-mm isocenters.
When untrained rats were irradiated with a maximum
dose of 150 Gy and tested 1 month thereafter, outcome
was two- to threefold worse compared with that in the
control group, and histology revealed necrotic changes.
The untrained rats irradiated with 25 to 100 Gy had no im-
pairment of memory acquisition 1 month after radiosur-
gery. Likewise no MR imaging–detected changes were
observed and the histology showed no necrosis.
Three months after the irradiation of untrained rats with
doses of 25 to 100 Gy, there were no changes in the behav-
ioral acquisition test. Magnetic resonance imaging re-
vealed no changes and histology showed no necrosis, al-
though an astrocytic reaction and a deficit of PCNA-
positive cells were present in rats irradiated with 25 Gy
and with increasing intensity in higher doses.
None of the rats irradiated with the 100-Gy dose sur-
vived for 6 months after radiosurgery to be tested. With
the exception of a single rat receiving 50 Gy, no untrained
rats irradiated with 25- and 50-Gy doses showed any
changes in the behavioral acquisition test after 6 months.
Magnetic resonance imaging revealed no changes and his-
tology showed no necrosis, except in the case of the one
impaired rat in the 50-Gy group. A different situation was
observed in the group of four rats irradiated with a maxi-
mum of 50 Gy, which underwent MR imaging repeatedly,
whereas the rest of the tested animals were tested only
once. In this group of four rats the MR imaging changes
were more pronounced. These rats differed from the oth-
ers in that repeated use of anesthesia was performed dur-
ing the MR imaging examination, which lasted for several
hours. We believe that the effect of protracted anesthesia
with eventual hypoxia can increase the potential for the
development of structural and functional changes in the
irradiated tissue, and this hypothesis should be verified in
further studies. In all rats irradiated with 75 Gy significant
impairment was shown in the behavioral acquisition test,
MR imaging revealed brain edema, and the histologi-
cal examination showed necrotic cavities. Similar results
were observed in the groups of animals tested 12 months
after irradiation.
Conclusions
Rats can be used for animal experiments in which com-
plex brain lesions are created using four 4-mm isocenters.
With maximum doses less than 75 Gy, the rats survive and
can be tested for at least 12 months after radiosurgery. Re-
tention memory for the testing of behavioral changes af-
ter bilateral radiosurgery of the whole hippocampus was
insensitive, and acquisition memory should be tested
to track the functional changes of the hippocampus. Sig-
nificantly higher incidences of edema, necrosis, and be-
havioral changes were observed in rats receiving doses
above 50 Gy, but none was demonstrated when using
doses of 25 Gy.
There appears to be a therapeutic window in which the
dose affecting epilepsy does not impair memory of the
rat. This therapeutic window should be further tested by
studying the epileptic rats to verify the sensitivity to mem-
ory impairment of a hippocampus harboring an epileptic
focus.
Acknowledgments
The authors are grateful for the contributions of Josef Vymazal,
M.D., Ph.D., Josef Novotn˘y, M.Sc., Ph.D., Gabriela ˘Simonová,
M.D., Ph.D, Daniela Tlachaová, and Martin Burian, M.Sc.
References
1. Alexander E III, Lindquist C: Special indications: radiosurgery
for functional neurosurgery and epilepsy, in Alexander E III,
Loeffler JS, Lunsford LD (eds): Stereotactic Radiosurgery.
New York: McGraw-Hill, 1993, pp 221–225
2. Altschuler E, Lunsford LD, Kondziolka D, et al: Radiobiolog-
ic models for radiosurgery. Neurosurg Clin N Am 3:61–77,
1992
3. Barcia-Salorio JL, Barcia JA: The role of radiosurgery in the
treatment of epilepsy, in Gildenberg PL, Tasker RR (ed): Text-
book of Stereotactic and Functional Neurosurgery. New
York: McGraw-Hill, 1998, pp 1925–1932
4. Broz˘ek G, Langmeier M, Hort J, et al: Correlation between
navigation deficit in the Morris water maze and the extent of
hippocampal damage after long lasting status epilepticus. Eur
J Neurosci 9 (Suppl):172, 1996
5. Cavalheiro EA, Leite JP, Bortolotto ZA, et al: Long-term ef-
fects of pilocarpine in rats: structural damage of the brain trig-
gers kindling and spontaneously recurrent seizures. Epilepsia
32:778–782, 1991
R. Lis˘c˘ák, et al.
672 J. Neurosurg. (Suppl 5). / Volume 97 / December, 2002

6. Chen Z, Kamiryo T, Henson SL, et al: Anticonvulsant effects of
gamma surgery in a model of chronic spontaneous limbic
epilepsy in rats. J Neurosurg 94:270–280, 2001
7. Fenton A, Bure J: Place navigation in rats with unilateral
tetradotoxin inactivation of the dorsal hippocampus: place but
not procedural learning can be lateralized to one hippocampus.
Behav Neurosci 107:552–564, 1993
8. Heikkinen ER, Heikkinen MI, Sotaniemi K: Stereotactic radio-
therapy instead of conventional epilepsy surgery. Acta Neu-
rochir 119:159–160, 1992
9. Hock BJ, Bunsey MD: Differential effects of dorsal and ventral
hippocampal lesions. J Neurosci 18:7027–7032, 1998
10. Kamiryo T, Kassell NF, Thai QA, et al: Histological changes in
the normal rat brain after gamma irradiation. Acta Neurochir
138:451–459, 1996
11. Kamiryo T, Lopes MBS, Berr SS, et al: Occlusion of the ante-
rior cerebral artery after gamma knife irradiation in a rat. Acta
Neurochir 138:983–991, 1996
12. Kondziolka D, Lunsford LD, Claassen D, et al: Radiobiology
of radiosurgery: Part I. The normal rat brain model. Neurosur-
gery 31:271–279, 1992
13. Maesawa S, Kondziolka D, Dixon CE, et al: Subnecrotic stereo-
tactic radiosurgery controlling epilepsy produced by kainic acid
injection in rats. J Neurosurg 93:1033–1040, 2000
14. Marks LB, Spencer DP, Acker JC, et al: Radiosurgery in rat
brain, in Kondziolka D (ed): Radiosurgery 1995. Basel: Kar-
ger, 1996, Vol 1, pp 308–315
15. Mori Y, Kondziolka D, Balzer J, et al: Effects of stereotac-
tic radiosurgery on an animal model of hippocampal epilepsy.
Neurosurgery 46:157–168, 2000
16. Morris RGM: Spatial localization does not require the presence
of local cues. Learn Motiv 12:239–261, 1981
17. Morris RGM, Schenk F, Tweedie F, et al: Ibotenate lesions
of hippocampus and/or subiculum: Dissociating components
of allocentric spatial learning. Eur J Neurosci 2:1016–1028,
1990
18. Moser M, Moser EI, Forrest E, et al: Spatial learning with a
minislab in the dorsal hippocampus. Proc Natl Acad Sci USA
92:9697–9701, 1995
19. O’Keefe J, Nadel L: The Hippocampus as a Cognitive Map.
Oxford: Oxford University Press, 1978
20. Régis J, Peragut JC, Rey M, et al: First selective amygdalohip-
pocampal radiosurgery for mesial temporal lobe epilepsy. Ste-
reotac Funct Neurosurg 64 (Suppl 1):193–201, 1995
21. Whang CJ, Kwon Y: Long-term follow-up of stereotactic gam-
ma knife radiosurgery in epilepsy. Stereotact Funct Neuro-
surg 66 (Suppl 1):349–356, 1996
This experiment was supported by IGA Grant No. MZ CR NF
5161-3/1999.
Address reprint requests to: Roman Lis˘c˘ák, M.D., Hospital Na
Homolce, Roentgenova 2, Prague 150 30, Czech Republic. email:
Roman.Liscak@homolka.cz.
J. Neurosurg. (Suppl 5) / Volume 97 / December, 2002
Radiosurgical lesioning of the rat hippocampus
673