вход по аккаунту


Cisplatin-induced Neurotoxicity with Seizures in Frogs.

код для вставкиСкачать
Cisplatin-induced Neurotoxicity
with Seizures in Frogs
Karen S. Blisard, PhD, MD," and Deborah A. Harringtont
Cisdiamminedichloroplatinum I1 (cisplatin) given by injection to adult frogs (Rzna pipiens) resulted in tonic-clonic
seizures 3 to 5 weeks later. The seizures could be induced multiple times; the animals appeared entirely normal
between seizures. In the spinal cord there was vacuolation in the anterior grey horns. Ultrastructurally, the vacuoles
consisted of swollen astrocytic processes in the neuropil and around neurons. Generalized edema with swelling of
perivascular astrocyte foot processes was not seen. Systemic administration of cisplatin to frogs results in neurotoxicity
with seizures and astrocytic swelling in the spinal cord.
Blisard KS, Harrington DA. Cisplatin-induced neurotoxicity with seizures in frogs.
Ann Neurol 1989;26:336-341
Cisdiamminedichloroplatinum I1 (cisplatin) is a valuable cancer chemotherapeutic agent, but its usefulness
is often limited by toxic side effects [I, 2). Sites of
serious toxicity include the kidney, the gastrointestinal
tract, and the nervous system. In the peripheral nervous system a peripheral sensory neuropathy often develops {3, 41. A few patients experience signs of central nervous system toxicity, including seizures [ S , 61,
blindness or other ophthalrmc symptoms E6-111, or
other neurological deficits [ 11- 131. Previous attempts
to develop an animal model for the neurotoxicity have
been unsuccessful {14]. This report describes the development of cisplatin-induced seizures and central
nervous system toxicity in frogs.
Materials and Methods
A total of 124 leopard frogs (Rumpipiens) were used in this
study. Ninety-eght frogs were injected with 10, 20, or 40
mgkg cisplatin in the dorsal lymph sac. Cisplatin, 1 mg/ml
mixed with 10 mglml mannitol in normal saline, was supplied
by Bristol-Myers Company (Wallingford,CT). The 10- and
20-mgkg doses were administered in a single injection. The
40-mgikg dose was administered in two injections to prevent
fluid overload; the interval between doses did not affect the
results of this study. Twenty-six control animals received an
injection of normal saline. To exclude the possibility that the
combination of cisplatin and mannitol was responsible for
the findings, 3 frogs were injected with 10 mg/kg cisplatin
(Sigma Chemical Corp, St Louis, MO) in 5% glucose. Identical clinical and histological findings were obtained with both
preparations of cisplatin.
The frogs were maintained at room temperature (2122°C) in stainless steel cages appropriate for frogs. At vari-
From the *Research Service, Veterans Administration Medical Center, and the ?Pathology Department, University of New Mexico
Me&cal School, Albuquerque, NM.
ous times after the injection, the frogs were sacrificed by
perfusion with modified Karnovsky's fixative after pentobarb i d anesthesia. The skull was opened to expose the brain,
and the vertebral column was removed intact. After further
fixation, the brain and spinal cord were dissected out and
sectioned. Tissue from some of the animals was processed in
paraffin, cut in 5-wm slices, and stained with hematoxylineosin and with a trichrome stain. Tissue from other animals
was postfixed in osmium, processed in Spurr's resin, cut in
l-pm sections, and stained with methylene blue-azure IIbasic fuchsin. Thin sections were cut at 100 nm, stained with
lead citrate and uranyl acetate, and examined with a Zeiss
EM10 electron microscope (Oberkochen, West Germany).
Observations in the Animals
The animals tolerated the injections well. Because of
the nephrotoxic effect of cisplatin, most of the frogs
developed renal failure with generalized edema and
increased blood urea nitrogen (Blisard and Harrington,
unpublished data, 1987). The edema was often reversible at lower doses. Histologically, the kidneys showed
tubular necrosis and cystic tubular dilatation, changes
analogous to those seen in mammals [15). Cisplatin
administration was eventually lethal in some animals
depending on the dose: 20%; of frogs died or were
moribund at the time of sacrifice at 10 mg/kg (median
time to death = 30 days), 60% at 20 mgkg (median
time to death = 21 days), and 100% of the frogs died
at 40 mg/kg (median time to death = 14 days).
Three to five weeks after cisplatin injection, many of
the frogs began to have seizures. The seizures were
Received Sep 22,1988, and in revised form Feb 20,1989. Accepted
for publication Feb 21, 1989.
Address correspondence to Dr Blisard, Research Service (151), Vete m s Administration Medical Center, Albuquerque, NM 87 108.
336 Copyright 0 1989 by the American Neurological Association
Fig I. Leopard frog (Rana pipiens) during generalized seizm
induced &y ~ y s t m i cisplatin.
most often initiated by handling, tapping the cage, or
feeding. Acoustic stimulation or photic stimulation at
8, 12, or 15 Hz did not cause seizures. Seizure episodes began with tonic extension of all extremities (Fig
1) and lasted from 10 to 60 seconds, during which time
the eyelids closed and breathing stopped. A clonic
phase followed, lasting for 30 to 120 seconds. After a
period of decreased responsiveness (<5 min) the frogs
appeared completely normal. Although additional seizures could be induced after recovery, the frogs never
developed status epilepticus. Paralysis or weakness
were not observed, and stimulation of their limbs resulted in prompt, appropriate withdrawal.
The percentage of treated frogs that had seizures
varied inversely with the dose of cisplatin. At 10 mgl
kg, 35 of 40 frogs (88%) had seizures; at 20 mgkg 8
of 15 (53%) had seizures; and at 40 mglkg, 2 of 19
(10%) convulsed (24 frogs sacrificed early to evaluate
the development of morphological changes are not included in these numbers). A pilot study performed
to investigate the effects of lower doses of cisplatin
showed that at 5 m g k g , only 2 of 8 frogs had seizures
and at 2 mgkg 0 of 5 frogs had seizures.
The time of onset of the seizures was related to the
dose of cisplatin (Fig 2 ) . The median time elapsed before seizures occurred was 26 days at 10 mglkg, 20
days at 20 mg/kg, and 18 to 19 days at 40 mglkg. Only
2 frogs given a dose of 40 mglkg cisplatin had seizures;
the animals apparently succumbed to renal toxicity at
this dose before seizures developed.
Four frogs given 10 mdkg cisplatin were studied
Fig 2. Relationship between time to first seizure fmrn cisplatin
and dose.
chronically. One of these animals never developed seizures. The other 3 frogs had repeated seizures and
were sacrificed 8 weeks after their first seizure.
Histological Filadings
The brain and spinal cord were examined in 88 treated
frogs and 26 control animals. Few histological changes
were seen in the brain parenchyma of the treated animals. In 3 frogs, however, there was vacuolation of
neurons in the thalamus, and in 1 there were vacuoles
in the preoptic nucleus.
Consistent changes were found in the spinal cords of
treated frogs sacrificed 0 to 10 days after the initiation
of seizures. Normal architecture was always seen in the
anterior horns of control frogs (Fig 3A). The predominant finding in the cisplatin-treated frogs was vacuoBlisard and Harrington: Cisplatin Neurotoxicity 337
Fig 3. Spinal cord pathological findings in frogs treated with
cisplatin. (A)Spinal cord of a controlfrog showing the normal
appearance of the motor neurons. (B) Spinal cord from a frog
with seizures who had been treated with 10 mglkg cisplatin.
Several ofthe motor neurom show vacuoles of various sizes. The
neuropil is also slightly vacuolated. (C) Spinal cord of another
frog treated with 10 mglkg who hadseizures. All of the motor
neurons are affected,and the nearopil is severely vacuolated. (0)
Spinal cord fmm a frog treated with 20 rngikg cisplatin who
had seizures. There is severe vacuolation ofthe motor neurons
and neuropil. (Hematoxylin-eosin;A, x 400; B, C, and D,
x 400 before 4% (B), 2% (C), and 6% (0)reduction.)
lation in the gray matter of the anterior horns (Fig
3B,C,D). The vacuoles measured 4 to 18 pm in diameter. They occurred peripherally around motor neuron
cell bodies and also in the surrounding neuropil.
Vacuolation was dispersed throughout the cord at all
levels and extended into the upper brainstem in a few
The severity of the changes varied between animals
from single scattered vacuoles at a single level to severe involvement of the entire anterior horn throughout the length of the cord (see Fig 3C). There was no
evidence of motor neuron degeneration, neuronophagia, or decrease in the number of motor neuron cells.
No demyelination was observed. No correlation was
apparent between the extent of vacuolation and the
dose of cisplatin, the time elapsed since the injection,
or the duration of the seizures.
Vacuolation of the anterior horns was seen in 33 of
338 Annals of Neurology Vol 26 No 3 September 1989
Fig 4. Ultrastrclcturalfindin@ in cisplatin neurotoxicity. (A)
Motor neuronfrom a frog treated with 40 mg/kg cisplatin who
had seizures. There is peripheral vacuolation.Some ofthe vacuoles contain mitochondria. (B) Segment of motor nerve axon from
a frog with seizlrres afier 40 mglkg cisplatin. There are numerous vacuoles along the length of the axon. (C) Spinal cord
neuropilfrom a fmg treated with 10 mglkg cisplatin who had
seizures. A swollen process with an irregular outline contains a
bundle of intermediate$lamets, most likely an astrocyticprocess.
(bad citrate and uranyl acetate; A, X 3,100 before 20%
reduction; B, x 4,500 before 15% reduction; C, x 5,000.)
the 38 frogs (87%) that were sacrificed from 0 to 10
days after their seizures began. Of 23 frogs that did not
convulse but were sacrificed after similar periods of
time, 11 (48%) had vacuoles. No vacuoles were seen
in control animals.
Ultrastructurally, most of the vacuoles were round,
appeared to be fluid filled, and seemed to be peripheral to the neuronal cytoplasm or located in the neuropil (Fig 4A,B). A few of the vacuoles appeared to
be swollen cell processes with highly irregular outlines containing bundles of intermediate filaments (Fig
4C), consistent with astrocytic processes. No vacuoles
containing microtubules were identified, and welldefined synaptic processes were never seen in associaBlisard and Harrington: Cisplatin Neurotoxiciry 339
The development of the neuropathological changes
was evaluated in 24 frogs that were sacrificed at 3, 5, or
10 days after the injections. No vacuolation was visible
by light or electron microscopy in these animals, and
the meningeal infiltrate was also much less prominent.
Of the 3 frogs that survived for 8 weeks after their
seizures began, only 1 had vacuoles in the spinal cord.
The meningeal infiltrate was also markedly diminished.
Two of these frogs, however, showed noticeable enlargement of d ventricles. There was parenchymal atrophy with loss of neurons and white matter, which
was most marked in the telencephalon (Fig 5).
Fig 5 . Atrophy of the telencephalon of a frog treated with 10
mgikg cisplatin who had seizures for 8 week. (A) Full-thickness
section of the lateral telencephalon ofa normalfrog. The ventricular sudace, lined by neurons, is at the top of the figure, whereas
the su?$ace of the brain with meninges is at the bottom. (B) Same
area fmm the treated frog. There is parenchymrzl loss in the
white mtter as well as loss of neurons. (Hematoxylin-eosin,
x 100 before 4% reduction.)
tion with either swollen processes or vacuoles. Thus,
most of the vacuoles probably arose from swollen astroglial processes. Astrocyte cell bodies occasionally
showed cytoplasmic expansion and increased numbers
of mitochondria. Swelling of perivascular astrocyte
foot processes was not seen.
The nuclei of motor neurons from treated animals
were more irregular in outline that those from control
animals, and often contained deep clefts (see Fig 4A).
Cup-shaped or ring-shaped mitochondria were occa
sionally seen in processes. The significance of these
mitochondrial shapes was not clear 1161. In motor
neurons, mitochondrial pleomorphism also was observed. Dense bodies occurred frequently in motor
neurons of normal frogs and did not seem to be increased in treated animals.
A mononuclear inflammatory infiltrate was seen in
the meninges of frogs treated with cisplatin. This infiltrate occurred in the meninges around both brain
and spinal cord. It was present to a variable extent in
34 of the 38 frogs (89%) that had seizures and in 6 of
the 23 frogs (26%) who did not. Again, there was no
apparent correlation between the severity of the meningeal infiltrate and the dose of cisplatin, the time
elapsed since the injection, or the duration of the seizures. No infiltrates were seen in the control animals.
Small foci of chronic inflammatory cells or reactive glia
were seen in the neural parenchyma of 7 animals. U1trasmctural examination of cells in the meningeal inflammatory infiltrate failed to reveal the presence of
infecaous agents.
Although the frog is a well-described animal model in
neurobiology, neurological disease in frogs has been
described relatively rarely. An amphibian model has
been described using 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), in which animals given the
drug displayed rigidity, akinesia, and tremors resembling symptoms in patients with idiopathic Parkinson’s
disease 117, 181. Generalized seizures have been induced in frogs with electroshock [19}. This report
presents the frog as an animal model for the neurotoxicity caused by cisplatin, an important cancer chemotherapeutic agent.
The pathogenesis of the seizures induced by cisplatin administration cannot be definitively determined
from these studies. Several possibilities can be considered. One cause of the seizures could be uremia, since
cisplatin is a nephrotoxin {151. Although in many of
the frogs the drug caused renal failure, it often resolved before the onset of seizures. In another model
of uremia in frogs, seizures were not apparent 120f. By
analogy, uremia as a source of epileptogenesis in this
model seems unlikely.
A second explanation is that activation of a latent
infection in the frog, resulting from cisplatin or immunosuppression secondary to its administration,
could cause both seizures and meningeal inflammation.
Such a situation has been described in mice, in whom
it has been shown that a latent viral infection that becomes activated with increasing age results in paralysis
221). These animals have spongiform changes in the
brain and spinal cord. Wild frogs are known to be
infested with parasites and viruses. The possibility that
a latent infection caused the seizures cannot be excluded but seems unlikely because no other morphological evidence of an infectious agent was seen.
The third possibility is that cisplatin damage to the
frog nervous system could result in seizures. The fact
that vacuoles occurred in the anterior horns of the
majority of the frogs that had seizures suggests that the
s p i d cord may well have contributed to epdeptogenesis. Generalized seizures originating in the spinal cord have been described in cats with transected
340 Annals of Neurology Vol 26 No 3 September 1989
spinal cords 122). Occasional human patients with
spinal cord transections have sustained grand mal seizures as part of the syndrome of autonomic dysreflexia
{ 2 3 } . Tetraethylammonium chloride injected intrathecally in frogs resulted in generalized seizures that appeared similar to those that we have observed, and the
authors of that study thought the effects were due in
part to stimulation of motor neurons [24}.
The systemic administration of cisplatin in frogs results in a model of epilepsy that appears unique among
models of chemically induced seizures. Many agents
that induce seizures are more effective after intracranial administration; even those that are effective when
given systemically-such as kainic acid, bicuculline,
and pentylenetetrazole-cause seizures within minutes
to hours after administration 1251. The seizures from
cisplatin occurred 3 to 5 weeks after injection, and
repeated episodes could be induced after intervening
normal periods. The ultrastructural findings of swelling
of astrocytic processes in this model are similar to the
observations made following administration of other
epileptogenic agents such as kainic acid 1261, aminopyridine 1271, bicuculline 1281, and pentylenetetrazole
E291. Thus, this model may represent a new opportunity for experimental analysis of the pathophysiological
features of epilepsy.
This work was supported by a Career Development Award from the
Veterans Administration and by a grant from the Pharmaceutical
Research and Development Division of the Bristol-Myers Company
to K.S.B.
The authors would like to express appreciation to Drs Larry E.
Davis, Oscar U. Scrernin, Mario Kornfeld, and Cecilia M. FenoglicPreiser for helpful discussions.
Portions of this work were presented previously in abstract form
(FASEB J 1988;2:a1069).
1. Rosenberg B. Fundamental studies with cisplatin. Cancer
1985;55:2303-23 16
2. Loehrer PJ, Einhorn LH. Cisplatin. Ann Intern Med 1984;lOO:
3. Thompson SW, Davis LE,Kornfeld M, et al. Cisplatin neuropathy. Clinical, electrophysiologic, morphologic, and toxicologic
studies. Cancer 1984;54:1269-1275
4. Roelofs RI, Hmhesky W, Rogin J, Rosenberg L..Peripheral
sensory neuropathy and cisplatin chemotherapy. Neurology
5. Mead GM, Arnold AM, Green JA, et al. Epileptic seizures
associated with cisplatin administration. Cancer Treat Rep
1982;66:1719-1 722
6. Berman IJ, Mann MP. Seizures and transient cortical blindness
associated with cis-platinum (11) diamminedichloride (PDD)
therapy in a thirty-yeardd man. Cancer 1980;45:764-766
7. Ostmw S,Hahn D, Wiernik PH, Richards RD. Ophthalmologic
toxicity after cis-dichlorodiammineplatinum(11) therapy. Cancer
Treat Rep 1978;62:1591-1594
8. Becher R, Schutt P, Osieka R, et al. Peripheral neuropathy and
ophthalmologic toxicity after treatment with cis-dichlorodiammineplatinum (11). J Cancer Rcs Clin Oncol 1980;96:21922 1
9. Pippitt CH, Muss HB, Homesly HD, Jobson VW. Cisplatinassociated cortical blindness. Gynecol Oncol 1981;12:253-255
10. Diamond SB, Rudolph SH, Lubicz SS, et al. Cerebral blindness
in association with cis-platinum chemotherapy for advanced carcinoma of the fallopian tube. Obstet Gynecol 1982;59:84~-86s
11. Cohen RJ, Cuneo RA, Cruciger MP, Jackman AE. Transient
left homonymous hemianopsia and encephalopathy following
treatment of testicular carcinoma with cisplatinum, vinblastine,
and bleomycin. J Clin Oncol 1983;1:392-393
12. Walsh TJ,Clark AW, Parhad IM, Green WR. Neurotoxic effects of cisplatin therapy. Arch Neurol 1982;39:719-720
13. Kapp JP, Sanford RA. Neurological deficit after carotid infusion
of cisplatin and 1,3-bis(2-chloroethyl)-l-nitrosourea (BCNU)
for malignant ghoma: an analysis of risk factors. Neurosurgery
14. Tomiwa K, Nolan C, Cavanagh JB. The effects of cisplatin on
rat spinal ganglia: a study by hght and electron microscopy and
by morphometry. Acta Neuropathol 1986;69295-308
15. Choie DD, Longnecker DS, del Camp0 AA. Acute and chronic
cisplatin nephropathy in rats. Lab Invest 1981;44:397-402
16. Ghadidy FN.Ultrastructural pathology of the cell and matrix.
3rd ed. London: Butterworths, 1988:278-281
17. Barbeau A, Dallaire L, Buu NT, et al. New amphibian models
for the study of l-methyl-4-phenyl-l,2,3,6-tetrahydropyridine
(MPTP). Life Sci 1985;36:1125-1134
18. Barbeau A, Dallaire L, Buu NT, et al. Comparative behavioral,
biochemical and pigmentary effects of MPTP, MPP + and paraquat in Ranapipiens. Life sci 1985;37:1529-1538
19. Johnson SW, Riker WK. Anticonvulsant effects of acute and
chronic phenytoin treatment in Ram pipiens. Proc West Pharmacol SOC1980;23:413-417
20. Fox KE, Russell NJ. Pharmacokinetics of kanamycin in
the bullfrog, Rana catedeiana. Comp Biochem Physiol 1987;
2 1. Gardner MB. Retroviral spongiform polioencephalomyelopathy. Rev Infect Dis 1985;7:99-110
22. Yu J, Chambers WW, Liu CN, et al. Induction of spinal seizures
by natural stimulation in cats. Brain Res 1984;299:323-330
23. Yarkony GM, Katz RT, Wu YC. Seizures secondary to autonomic dysreflexia. Arch Phys Med Rehabil 1986;67:834835
24. Mazella H,Gines F, Reyes A, Planel LR.Effects of injection of
tetraethylammonium into the rachis. Arch Int Pharmacodyn
Ther 1967;166:87-92
25. Ben-Ari Y, Trembley E,Riche D, et al. Electrographic, clinical
and pathological alterations following systemic administration of
kainic acid, bicuculline or pentylenetetrazole: metabolic mapping using the deoxyglucose method with special reference to
the pathology of epilepsy. Neuroscience 1981;6:1361-1391
26. Herndon RM,Coyle JT, Addicks E. Ultrastructural analysis of
kainic acid lesion to cerebellar cortex. Neuroscience 1980;5:
27. Mihaly A, Joo F, Stente M. Neuropathological alterations in the
neocortex of rats subjected to focal aminopyridine seizures.
Acta Neuropathol 1983;61:85-94
28. Atillo A, Soderfeldt B, Kahmo H, et al. Pathogenesis of brain
lesions caused by experimental epilepsy. Acta Neuropathol
29. de Robertis E, Alberici M, de Lores Arnaiz GR. Astroglial
swelling and phosphohydrolases in cerebral cortex of metrazol
convulsant rats. Brain Res 1969;12:461-466
Blisard and Harrington: Cisplatin Neurotoxicity 341
Без категории
Размер файла
759 Кб
neurotoxicity, cisplatin, induced, seizure, frog
Пожаловаться на содержимое документа