Naloxone in experimental spinal cord ischemia: Dose-response studies

European Journal of Pharmacology, 103 (1984) 115-120

115

Elsevier

N A L O X O N E IN EXPERIMENTAL SPINAL CORD ISCHEMIA: D O S E - R E S P O N S E STUDIF~S A L A N I. F A D E N *, T l i O M A S P. JACOBS, M I C H A E L T. SMITH * and JUSTIN A. ZIVIN **

Neurobiology Research &'nit and * Department of Pathology, &~iformed Services Unit:ersity of the Health Sciences, Bethesda, Maryland 20814, and ** Department of Neurology, University of Massachusetts, Worcester, Massachusetts 01605, U.S.A. Received 13 March 1984, revised MS received 25 April 1984, accepted 4 May 1984

A.I. FADEN, T.P. JACOBS, M.T. SMITH and J.A. ZIVIN, Naloxone in experimental spinal cord ischemia: dose-response studies, European J. Pharmacol. 103 (1984) 115-120.

Temporary aortic occlusion produces a consistent degree of spinal cord injury in the unanesthetized rabbit. This 'spinal stroke' model was utilized to examine the potential therapeutic effects of the opiate antagonist naloxone in central nervous system ischemia. Naloxone treatment resulted in dose-related enhancement of motor recovery; greatest functional recovery was observed in rabbits treated with a dose of 2 mg/kg per h. This dose compares well with the high doses of naloxone shown to have a beneficial effect in other experimental models of stroke and spinal injury. In contrast, clinical stroke studies, which have been largely unsuccessful, have utilized naloxone doses which are several orders of magnitude lower than those successfully employed in experimental models. Spinal cord ischemia

Spinal stroke

Naloxone

1. Introduction

During the past few years there has been a rapid expansion of experimental literature reporting therapeutic utility for opiate antagonists in the treatment of traumatic and ischemic disease of the central nervous system (Baskin et al., 1982; Faden et al., 1982a,b; 1981a,b; 1983a; Flamm et al., 1982; Hosobuchi et al., 1982; Levy et al., 1982a; Young et al., 1981; Zabramski et al., 1983). The rationale for the use of such compounds has been based on the hypothesis that endogenous opioids are released following injury and contribute to the pathophysiological process by reducing spinal or cerebral blood flow. Several years ago we demonstrated that /3-endorphin-like immunoreactivity was increased more than tenfold following traumatic cervical spinal cord injury in the cat, and that this increase was associated with a reduction in spinal cord blood flow (Faden et al., 1981b). * To whom all correspondence should be addressed: Neurobiology Research Unit, Uniformed Services University of the He',dth Sciences, 4301 Jones Bridge Road, Bethesda, Maryland 20814, U.S.A. 0014-2999/84/$03.00 © 1984 Elsevier Science Publishers B.V.

Endorphins

Moreover, we showed that the opiate receptor antagonist naloxone significantly improved spinal cord blood flow (Faden et al., 1981b) as well as long-term motor recovery (Faden et al., 1982b; 1981a,b). The beneficial effects of naloxone on blood flow and functional neurologic recovery after spinal cord injury were subsequently confirmed by other investigators (Flamm et al., 1982; Young et al., 1981). There are a number of similarities between ischemic and traumatic injury to the central nervous system; the neurologic dysfunction in both conditions appears to result, at least in part, from a critical reduction of blood flow, and the pathological process appears to be reversible for a period of time following the insult (Faden et al., 1982a). These similarities prompted our evaluation of naloxone in a canine model of cerebral ischemia caused by air embolization. In that model, naloxone at high doses (2.0 m g / k g bolus, 2.0 m g / k g per h for 4 h) significantly improved the somatosensory-evoked response and blocked the development of multifocal cerebral infarction which occurred in control animals and which is typically found in this model (Faden et al., 1982a). A num-

116

ber of other investigators have reported beneficial effects of high dose naloxone treatment in experimental stroke (Baskin et al., 1982; Faden et al., 1983a; Hosobuchi et al., 1982; Levy et al., 1982a; Zabramski et al., 1983). However, some controversy has resulted from the inability of other laboratories (Holaday and D'Amato, 1982; Kastin et al., 1982; Levy et al., 1982b; Nels et al., 1983) to demonstrate a therapeutic action for this opiate antagonist. In the various naloxone-stroke studies, some of the reported interlaboratory differences probably have resulted in part from differences in experimental design including treatment dose, time of treatment and the severity of the injury produced in a given model. We have recently addressed some of these methodological variables utilizing a spinal ischemia model in the awake rabbit which yields a highly predictable degree of post-ischemic paralysis (Fadcn et al., 1983a). We found that high doses of naloxone (2.0 m g / k g ) significantly improved functional neurological recovery following 25 min of aortic occlusion, a severe degree of ischemia producing complete hindlimb paralysis in nearly all control animals. However, when the ischemic period was further prolonged to 30 min, naloxone was ineffective in improving outcome at doses up to 10 m g / k g per h. These findings further support a therapeutic role for naloxone in ischemic disease of the central nervous system, but underscore the fact that the severity of the ischemia is an important variable determining whether the therapeutic effects of a pharmacological agent will be observed in a given model. Another critical variable which has as yet received little attention in either experimental spinal injury or experimental cerebral ischemia is naloxone dose. In the present studies we employed a less severe spinal ischemia model in the rabbit (20 rain occlusion) in order to provide information regarding optimal treatment dose and the lowest effective dose for treatment of ischemic central nervous system disease. 2. Materials and methods New Zealand albino rabbits (2.0 + 0.25 kg) were anesthetized with ketamine hydrochloride (50

m g / k g i.m.) and sodium pentobarbital (40 m g / k g i.v.). Under aseptic techniques, a transperitoneal approach was made to expose the abdominal aorta and polyvinyl-chloride tubing (0.75 mm O.D.) was placed around the aorta distal to the renal arteries. This tubing was threaded through plastic buttons (6.0 ram) which were placed dorsal and ventral to the aorta in order to produce a snare ligature. To prevent movement through the incision site, the ligature was passed through a vinyl guide tube (6.25 mm O.D.) which was secured to the abdominal muscles. A canvas jacket was placed around the animal to protect the incision site and the externally accessible ligature. Approximately 18 h post-surgery, when the animals were awake, the aorta was occluded for 20 rain by pulling on the snare ligature. The ligature was subsequently released and removed with the guide tube through the surgical site. Fifteen animals each were randomly assigned to 1 of 4 treatment groups: (1) high-dose naloxone (2.0 m g / k g per h); (2) moderate-dose naloxone (0.2 m g / k g per h); (3) towdose naloxone (0.02 m g / k g per h); or (4) physiological saline. The number of animals per group was based on expected differences as shown in earlier pharmacological studies with this model. Drugs or vehicle were administered in equal volume through a catheter in the lateral ear vein and given as 4 hourly injections at the above dose beginning 15 min pre-occlusion. Thus, naloxone-treated animals received a total dose of 8.0 m g / k g , 0.8 m g / k g or 0.08 m g / k g . The use of naloxone pre-treatment was done to insure that the drug reached the injury site since the present model relies on total ischemia of the spinal cord; thus, this model contrasts with most cerebral ischemia models in which collateral circulation is available to permit access of treatment drug. H~ndlimb function was graded by an investigator unaware of treatment group at 2, 24 and 48 h post-occlusion utilizing an ordinal grading scale as follows: Score 0 2 2 3

Definitions Complete paralysis Some functional movement,markedly impaired Hopping, slightly impaired Normat

At 48 h post-occlusion, animals were killed with

117 s o d i u m p e n t o b a r b i t a l a n d the spinal cords rapidly extracted for histological e x a m i n a t i o n . After removal, spinal cords were fixed in 10% formalin with 10% glycerine by volume for at least 72 h. Tissue from 29 animals was examined without knowledge of prior treatment. Spinal cord segm e n t s from mid-thoracic region to the conus medullaris were cross-sectioned at 2 m m intervals allowing careful m e a s u r e m e n t of craniocaudal extent of lesions present. The cross-sectional slices were dehydrated in graded alcohols, e m b e d d e d in paraffin a n d cut in 10 /~m sections for light microscopy. Sections were stained with hematoxylin a n d eosin a n d luxol-fast blue. The spinal cord lesions were then graded according to the following definitions: Score Definitions Liquefaction necrosis of gray matter with craniocaudal extent of > 4 mm. Liquefaction necrosis of gray matter with craniocaudal extent of 2-4 mm. Liquefaction necrosis of gray matter with craniocaudal extent of < 2 mm. No liquefaction necrosis; some gliosis and neuronal reaction may be evident. Neurological a n d histopathological scores in the various groups were c o m p a r e d utilizing KruskalWallis A N O V A followed by i n d i v i d u a l M a n n W h i t n e y U rank sum tests. S p e a r m a n ' s rank correlation test was used to measure the association

between neurological a n d histopathological scores. A P value less than 0.05 was considered statistically significant.

3. Results

3.1. Clinical

At 24 h post-occlusion, 9 of 15 saline-treated control a n i m a l s showed severe or complete hindlimb paralysis (score = 0 or 1), a n d at 48 h post-inj u r y 10 of 15 control animals showed the same severe degree of injury (table 1). In contrast, naloxone-treated a n i m a l s showed dose-related red u c t i o n in m o t o r dysfunction. At 24 h, severe paralysis was found in 6 of 15 low-dose n a l o x o n e animals, 5 of 15 moderate-dose n a l o x o n e animals, a n d only 4 of 15 high-dose n a l o x o n e a n i m a l s (table 1A). Similar dose-response effects were also evident at 48 h after injury (table 1B). Neurological differences between high-dose n a l o x o n e a n d control rabbits were significant at both 1 a n d 2 days (each P < 0.05). Differences between moderate-dose n a l o x o n e rabbits a n d controls were significantly different at 2 days (P < 0.05), but not at 1 day, after occlusion. Differences between lowdose naloxone-treated rabbits a n d control rabbits failed to reach significance at eigher 1 or 2 days following occlusion (P > 0.05).

TABLE 1 Neurological scores after spinal cord ischemia in the rabbit. Neurological * score

Naloxone ~ 2.0 mg/kg

Naloxone 0.2 mg/kg

Naloxone 0.02 mg/kg

Saline

3

XXXXXXX

XXXXXXX

XXXXXXX

XXX

2

XXXX

XXX

XX

XXX

I

XXX

XXX

XXX

XXXXX

0

X

XX

XXX

XXXX

3 2 1

xxxxxx xxxx xxxx

xxxxxxx xx xxxx

xxxxxx xxx xxx

xx xxx xxxxxx

0

X

XX

XXX

XXXX

(A) 24 h

(B) 48 h

* Indicates clinical score based on ordinal scale (see Materials and methods), x Indicates scores for individual rabbits, t Indicates statistical significance (P < 0.05), Mann-Whitney rank sum test.

118

4. Discussion O

-6 1 ;"

o

~i~: ~-.-',

~;.,-3~, 2

1

i] o

o

0 1 2 3 Histopathologicai Score Fig. 1. Matrix indicating the degree of association between neurological and histopathological scores. The number of animals falling into each box are given. Dark-shaded squares (48.3% of total) indicate perfect correlation; light-shaded squares (41.4%) indicate close correlation; and the non-shaded squares (10.3%) indicate no correlation. Spearman's rank correlation coefficient: r~ = 0.698, n = 29, P < 0.01.

3.2. Histopathology Grossly, the spinal cords displayed myelomalacia involving varying extent of the lower spinal cord from T12 to the conus medullaris. Microscopic evaluation similarly showed changes ranging from normality to cavitary necrosis. Pathological changes, when present, involved gray matter but not white matter. Anterior horn was predominantly affected, with relative sparing of the intermediate gray matter and sparing of the posterior horn. Correlation between neurologic and histopathologic scores were significant (r S= 0.698, P < 0.01) (fig. 1). Histopathologic scores were significantly higher in naloxone-treated animals (Kruskal-Wallis statistic= 9.035, P = 0.029) and appeared to be dose-related (table 2).

In this 'spinal stroke' model in the rabbit, naloxone treatment significantly reduced motor dysfunction and was associated with less spinal pathology. Moreover, these beneficial effects of naloxone appeared to be dose-related, with significant therapeutic effects observed for both highdose (2.0 m g / k g per h) and moderate-dose (0.2 m g / k g per h) treatment. In contrast, low-dose naloxone treatment (0.02 m g / k g per h) failed to show significant improvement when compared with saline controls. These findings underscore the fact that dose considerations may bc critical in evaluating the therapeutic utility of naloxone in experimental ischemia, and indicate that relatively high doses of this opiate antagonist may be required for therapeutic efficacy. It should be noted that best effects observed were produced by a very high dose of naloxone (8.0 m g / k g total dose). In each of the studies in the cat (Lcvy et al., 1982a), dog (Faden et al., 1982a) and baboon (Baskin e~ al., 1982; Zabramski et al., 1983), in which naloxone showed a beneficial effect, treatment doses for naloxone were in the range of 2-10 mg/kg. These doses should be contrasted with the clinical reports, predominantly negative, in which naloxone has been utilized in the range of 0.006 to 0.1 m g / k g (Baskin and Hosobuchi, 1981; Bredesen et al., 1982; Fallis et al., 1983; Jabaily et al., 1982). Of course, these clinical studies also had other potentially confounding problems including variability of stroke etiology, severity of i n j u ~ and time of treatment. Although the major rationale for the use of naloxone in central nervous system ischemia was based on the hypothesis that opiate antagonists

TABLE 2 Histopathology ,scores after spinal cord ischemia in the rabbit. Pathology * Score

Naloxone 2.0 m g / k g

Naloxone 0.2 m g / k g

Naloxone 0.02 m g / k g

Saline -

3

xxxxx

xxx

-

2

X

XXX

X

XX

1

×

XX

X

XXX

0

XX

--

XXXX

X

• Spinal cord tissue was graded by the craniocaudal extent of the lesion (see Materials and methods), x Indicates scores for individual rabbits.

119

would reverse the pathophysiological effects of endogenous opioids, the high doses of naloxone required to demonstrate a therapeutic effect suggest the possibility that its therapeutic actions may be due to a non-opiate receptor-mediated action. At high doses, for example, naloxone appears to have potent antioxidant activity (Koreh et al., 1981). Moreover, at 2 mg/kg, naloxone rapidly reverses post-traumatic reduction in extracellular calcium after experimental spinal injury (Stokes et al., 1984). However, the fact that the therapeutic action of naloxone in spinal shock is stcreospecific (Faden et al., 1980; Holaday and Faden, 1980) is more suggestive of a receptor-mediated mechanism of action. From this perspective, the high doses of naloxone required in the present and previous studies of CNS ischemia are also consistent with the hypothesis that naloxone acts at non-~ (morphine) opiate receptors, such as 8, x or c receptors (Schulz et al., 1981; Zukin and Zukin, 1981). Although naloxone possesses its highest selectivity for the ~ receptor, it is a relatively non-selective opiate receptor antagonist which, at high doses, will antagonize the effects at other classes of opiate receptors (Yaksh and Howe, 1982). The 8 receptor does not appear to be involved since we have found that the highly 8 selective antagonist M154,129 (ICI) fails to improve functional recovery in the present model (Faden et al., 1983b). In contrast, the somewhat x selective opiate antagonist WIN44,441-3 improves neurological outcome after both ischemic and traumatic spinal cord injury (Faden et at., 1984). The hypothesis that the x receptor may play a role in the pathophysiology of spinal injury is consistent with findings that the x receptor is the predominant opiate receptor in the spinal cord (Gouarderes et al., 1981; Traynor et al., 1982). In summary, naloxone treatment of rabbits subjected to spinal 'stroke' (caused by temporary aortic occlusion) resulted in significantly reduced motor dysfunction and diminished spinal pathology. The beneficial actions of this opiate antagonist were dose-related, with the greatest effects observed at the highest dose utilized (2 m g / k g per h). This dose is comparable to that utilized in other experimental models of stroke or spinal injury in

which naloxone has proved beneficial and exceeds, by several orders of magnitude, the naloxone doses utilized to treat cerebral ischemia in humans. Thus, in addition to questions regarding time of treatment and severity of injury, naloxone dose should be a critical variable in the design of future clinical studies.

Acknowledgments This work was supported by the Office of Naval Research (Contrast No. N0001482WR20257). The opinions or assertions contained herein are the private ones of the authors and are n o t be construed as official or reflecting the views of the Department of Defense or the Uniformed Services University of the Health Sciences. The experiments reported herein were conducted according to the principles set forth in the 'Guide for the Care and Use of Laboratory Animals', Institute of Laboratory Animals Resources, National Research Council D H E W Pub. No. (NIH) 78-23. We wish to thank Ms. Cynthia Conoley for her technical assistance, and Mrs. Jacqueline C. Mosely and Miss Eleanor M. Bell for preparing this manuscript.

References Baskin, D.S. and Y. Hosobuchi, 1981, Naloxone reversal of ischemic neurological deficits in man, Lancet ii, 272. Baskin, D.S., C.F. Kieck and Y. Hosobuchi, 1982, Naloxone revers',d of ischemic neurologic deficits in baboons is not mediated by systemic effects, Life. Sci. 31, 2201. Bredesen, D.E., J.R. Cutler and R.P. Simon, 1982, Failure of naloxone to reverse vascular neurologic deficit, Neurology 32, A79. Faden, A.I., J.M. Hallenbeck and C.Q. Brown, 1982a, Treatment of experiment',d stroke: comparison of naloxone and thyrotropin-releasing hormone, Neurology 32, 1083. Faden, A.I., T.P. Jacobs and J.W. Holaday, 1982b, Comparison of early and late naloxone treatment in experimental spinal injury, Neurology 32, 677. Faden, A.I., T.P. Jacobs and J.W. l-loladay, 1980, Endorphinparasympathetic interaction in spinal shock, J. Autonom. Nerv. Sys. 2, 295. Faden, A.I., T.P. Jacobs and J.W. Holaday, 1981a, Opiate antagonist improves neurologic recovery after spinal injury, Science 211,493. Faden, A.I., T.P. Jacobs, E. Mougey and J.W. lloladay, 1981b, Endorphins in experimental spinal injury: therapeutic effect of naloxone, Ann. Neurol. 10, 326. Faden, A.I., T.P. Jacobs, G.P. Smith, B.A. Green and J.A. Zivin, 1983a, Neuropeptides in spinal cord injuty: comparative experimental models, Peptides 4, 631. Faden, A.I., T.P. Jacobs and J.A. Zivin, 1983b, Naloxone but

120 not a delta-selective antagonist improves neurological recovery after spinal 'stroke' in the rabbit, Life Sci. 33(Suppl. I), 707. Faden, A.I., S. Knoblach, C. Mays and T.P. Jacobs, 1984, Motor dysfunction after spinal cord injury is mediated by opiate receptors, Peptides (in press). Fallis, R., M. Fisher and R. Lobo, 1983. A double-blind trial of naloxone in acute stroke, Stroke 14, 124. Flamm, E.S., W. Young, H.B. Demopoulos, V. DeCrescito and J.J. Tomasula, 1982, Experimental spinal cord injury: treatment with naloxone, Neurosurgery 10, 227. Gouarderes, C., B. Attali, Y. Audigier and J. Cros, 1981, Opiate binding sites in the lumbo-sacral cord from a variety of species, in: Advances in Endogenous Opiates, Kodansha Ltd., Tokyo, p. t8. Holaday, J.W. and R.J. D'Amato, 1982, Naloxone or TRII fails to improve neurological deficits in gerbil models of "stroke', Life Sci. 31,385. Holaday, J.W. and A.I. Faden, 1980, Naloxone acts at central opiate receptors to reverse hypotension, hypothermia and hypoventilation in spinal shock, Brain Res. 189, 295. Hosobuchi, Y., D.S. Baskin and S.K. Woo, 1982, Reversal of induced ischemic neurologic deficit in gerbils by the opiate antagonist naloxone, Science 215, 69. Jabaily, J. and J.N. Davis, 1982, Naloxone partially reverses neurologic deficits in some but not all stroke patients, Neurology 32, 1A97. Kastin, A.J., C. Nissen and R.D. Olson, 1982, Failure of MSF-1 or naloxone to reverse ischemic-induced neurologic deficits in gerbils, Pharmacol. Biochem. Behav. 17, 1083. Koreh, K., M.L. Seligman, E.S. Flamm and H.B. Demopoulos, 1981, Lipid antioxidant properties of naloxone in vitro, Biochcm. Biophys. Res. Comm. 102, 1317.

Levy, R., P. Feustel, J. Severinghaus and Y. Hosobuchi, 1982a, Effect of naloxone on neurologic deficit and blood flow during focal cerebral ischemia in cats, Life Sci. 31, 2205. Levy, I).E.. C.L. Pike and D.G. Rawlinson, 1982b, Failure of naloxone to limit clinical or morphological brain damage in gerbils with unilateral carotid artery occlusion, Soc. Neurosci. Abstr. 8, 248. Nels, D.G., C. Gaines and R.M. Crowell, 1983, Effects of naloxone in the treatment of experimental stroke in awake monkeys, Stroke 14, 123. Schulz, R., M. Wuster and A. Hevz, 1981, Pharmacological characterization of the c-opiate receptor. J. Pharmacol. Exp. Ther. 216, 604. Stokes, B.T., P. Fox and G. Hollinden, 1984, Extracellular metabolites: their measurement and role in the acute phase of spinal cord injury, in: Proceedings of the Fifth Neurotrauma Conference, ed. J. Jane (in press). Traynor, J.R., P.D. Kelley and M.J. Rance, 1982, Multiple opiate binding sites in rat spinal cord, Life Sci. 31, 1377. Yaksh. T.L. and J.R. llowe. 1982, Opiate receptors and their definition by antagonists. Anesthesiology 56, 246. Young, W., E.S. Flamm, H.B. l)emopoulos, J.J. "l'omasula and V. DeCrescito, 1981, Naloxone ameliorates post-traumatic ischemia in experimental spinal contusion, J. Neurosurg. 55, 209. Zabramski, J.M., R.F. Spetzler, W.R. Selcman, L.A. Hershey and U. Roessman, 1983, Naloxone improves neurologic function during and outcome after temporary focal cerebral ischemia, Stroke 14, 123. Zukin, R.S. and S.R. Zukin, 1981, Multiple opiate receptors: emerging concepts, Life Sci. 29, 2681.