The actions of fentanyl to inhibit drug-induced emesis

Neuropharmocology

Vol. 30, No. 10,pp. 1073-1083,1991

002%3908/91 $3.00 + 0.00 Copyright 0 1991 Pergamon Press pit

Printed in Great Britain. All rights reserved

THE ACTIONS OF FENTANYL TO INHIBIT DRUG-INDUCED EMESIS N. M. BARNES,’ K. T. BUNCE,’R. J. NAYLOR’ and J. A. RUDD’* ‘Postgraduate Studies in Pharmacology, The School of Pharmacy, University of Bradford, Bradford BD7 IDP and 2Department of Gastrointestinal Pharmacology, Glaxo Group Research Ltd, Ware, Hertfordshire SG12 ODP, U.K. (Accepted 22 April 1991) Summary-The ability of fentanyl to inhibit drug-induced emesis was investigated in the ferret. Initial studies established that morphine, in small doses (0.025-0.5 mg/kg s.c.), induced emesis in the ferret that decreased at the larger doses of 1 and 2 mg/kg (s.c.). Fentanyl (IO-80 pg/kg s.c.) failed to induce emesis

but in this dose range prevented the emesis induced by morphine (0.5 mg/kg s.c.), apomorphine (0.25 mg/kg s.c.), copper sulphate (100 mg/kg intragastric) and cisplatin (10 mg/kg iv.). The antiemetic effects could be obtained in the absence of sedation or motor impairment. The antagonism by fentanyl of apomorphine-, copper sulphate- and cisplatin-induced emesis was inhibited by naloxone (0.1 or 0.5 mg/kg s.c.). It is concluded that fentanyl exerts a broad spectrum of actions to inhibit drug-induced emesis. An autoradiographic study of the binding of [3H]DAG0 to the brainstem of the ferret indicated high densities of n recognition sites in the area postrema, nucleus tractus solitarius, dorsal motor nucleus of the vagus, reticular medulla and other sites. The results are discussed in terms of balanced facilitatory and inhibitory opioid systems, regulating emesis and that the antiemetic actions of fentanyl reflect an important, although not necessarily an exclusive, action at p opioid receptors. Key wordr-emesis,

ferret, fentanyl, apomorphine, copper sulphate, cisplatin.

Morphine has a complex action to influence emesis. In cats and dogs the acute administration of morphine can induce emesis (Wang and Glaviano, 1954; Costello and Borison, 1977; Bhargava, Dixit and Gupta, 1981) yet when administered as a pretreatment it is an effective antiemetic against subsequent challenge with morphine (Costello and Borison, 1977), nicotine (Beleslin, Krstic, Stefanovic-Denic, Strbac and Micic, 198 l), apomorphine (Costello and Borison, 1977; Blancquaert, Lefebrvre and Willems, 1986), copper sulphate (Blanquaert et nl., 1986; Wang and Glaviano, 1954), staphylococcus enterotoxin (Bayliss, 1940) and veratrum, acetyl strophanthidine and dibutyl CAMP (Costello and Borison, 1977). The emetic and antiemetic effects of morphine are antagonised by naloxone (Costello and Borison, 1977; Blancquaert et al., 1986), indicating that both effects are mediated by opioid receptors. However, whereas the emesis is probably mediated by the area postrema (Wang and Glaviano, 1954), the antiemetic effects may occur in the “vomiting centre” (Borison and Wang, 1953; Costello and Borison, 1977). Certainly, the breadth of the antiemetic spectrum of action of morphine would indicate an action close to or at the final output from the “vomiting centre”. The ~1 opioid receptor may be involved in the antiemetic effects of morphine (Harris, 1982), although this is difficult to evaluate using morphine *To whom correspondence and reprint requests should be addressed.

with its non-selective actions on opioid receptors. However, fentanyl is a highly potent p receptor agonist, lacking the ability to induce emesis in its own right but antagonising apomorphine-induced emesis in the dog (Niemegeers, Schellekens, Van Bever and Janssen, 1976; Blancquaert et al., 1986). In the present study the spectrum of the antiemetic action of fentanyl was investigated in the ferret and the location of p receptors within the brain stem was determined. METHODS Animals

Albino or fitch ferrets of either sex (0.8-1.4 kg), bred at the University of Bradford, were housed individually at 22 +_ 1°C under artificial lighting, with lights on between 07.00 and 21.00 hr. Chronic cannulation of the jugular vein

The animals were anaesthetised with halothane (N,O/O, carrier) and the fur on the ventral and dorsal area of the neck was shaved and swabbed with 5% chlorhexidine gluconate solution in 10% alcohol. An incision, approximately 2-3 cm long, was made in the skin above the left jugular vein, which was dissected free of fatty tissue and ligated at the cephalic end. A cannula was then inserted and externalised at the dorsal region of the neck by the use of a trochar. The cannula was filled with 250 IU/ml sodium heparin/ saline (0.9% w/v) solution and was secured with a cotton suture at the region of the ventral incision. The

1073

N. M. BARNESet al.

1074

outer wound was then closed with a polyamide suture. A purse string suture (polyamide) was used to secure the cannula at the dorsal exit point. Benzylpenicillin (50 mg/kg i.m.) was administered

as an antibiotic cover on the day of the operation. All animals were allowed a minimum recovery period of 48 hr before any further experimental procedure. Patency of the cannulae was maintained by daily flushing with 0.2 ml of 250 IU/ml sodium heparin/saline (0.9% w/v). Induction and measurement of emesis

All experiments were performed in the afternoon (13.00-18.00 hr), with the exception of those involving morphine, which were performed in the morning between 08.30-13.00 hr. Apomorphine (0.25 mg/kg), morphine (0.5 mg/kg) or the respective vehicle, was administered subcutaneously; copper sulphate (100 mg/kg) or vehicle was administered intragastritally, under light halothane anaesthesia (N,O/Or carrier), as previously described (Costall, Domeney, Naylor, Owera-Atepo, Rudd and Tattersall, 1990). Cisplatin (10 mg/kg) or vehicle was administered through the chronically indwelling cannulae in the jugular vein and flushed in with 0.5 ml saline. After treatment with drug or vehicle, the animals were placed in individual observation cages for the assessment of emesis. For experiments involving the

administration of apomorphine and copper sulphate, naloxone or vehicle was administered subcutaneously as a pretreatment at 20 min and fentanyl or saline was administered as a pretreatment at 15 min. Where animals were challenged with morphine, both naloxone and fentanyl were administered as pretreatments at 15 min. In experiments using cisplatin, naloxone, fentanyl, or vehicle was given subcutaneously 45 min after injection of cisplatin. Emesis was characterised by rhythmic abdominal contractions, which were either associated with the oral expulsion of solid or liquid material from the gastrointestinal tract (i.e. vomiting) or not associated with the passage of material (i.e. retching movements). The time of onset of emesis after challenge with drug was recorded for each animal, as were the numbers of retches and vomits and their frequency during the subsequent period of 30-240 min, following the onset of emesis. The significance of differences between treatments was assessed using the Mann-Whitney U-test. IdentiJication of p receptor recognition sites by autoradiography

Animals were anaesthetised with halothane (N,O/O, carrier) and exsanguinated. The brains were rapidly removed and frozen by being submerged in hexane (-5OC) for approximately 10 sec. The frozen tissue was set in embedding medium (OCT

301 T

Fig. 1. The emesis induced by morphine in the ferret. Animals were observed for a period of 30 min after the subcutaneous injection of morphine and values are the means f SEMs of 4 determinations. If an animal failed to either retch or vomit, then the time to onset was taken as 30min.

Fentanyl inhibits drug-induced emesis compound, Miles Scientific) before 20 ,um sections of the hindbrain were cut in the transverse plane using a cryotome (AS600, Anglia Scientific, specimen temperature minus 15-20°C chamber temperature minus 20-25°C) and thaw-mounted onto gelatincoated glass microscope slides and stored (dessicated) overnight at -20°C. For binding studies, sections of hindbrain of the ferret were pre-incubated in Tris buffer (170.0 mM, pH 7.4) containing sodium chloride (100 mM) at 22°C for 20 min. To remove the sodium ions the sections were washed twice (5 min each) in incubation buffer (50.0 mM Tris buffer; 5.0 mM magnesium chloride; 2.0 gl-’ bovine serum albumin; 20.0 mgl-r bacitracin) before adjacent sections were incubated at 22°C for 60 min in incubation buffer with the addition of 4.0 nM [3H]Tyr-D-AlaGly-NMePhe-Gly-ol ([3H]DAGO) in the absence (total binding) and presence (non-specific binding) of 1.0 PM naloxone. The sections were then washed twice (7fmin each) in incubation buffer (minus bacitracin) at 4°C before being rinsed in ice-cold distilled water for 1 sec. The sections were rapidly dried in a stream of cold dry air and then exposed to Hyperfilm-[3H) (Amersham) in X-ray cassettes together with tritium standards (Amersham) for 6 weeks (stored at minus 20°C). Films were developed using Kodak LX24 developer (5 min) and Kodak FX40 fixer (5 min) and the autoradiographs generated were analysed and quantified (with reference to Amersham tritium standards; equivalent values of tissue for intact grey matter), using a Bio-Quant System IV image analysis system (R and M Biometrics). To aid the identification of brain-stem nuclei, the slide-mounted sections were histologically stained with Luxol Fast Blue G and Cresyl Fast Violet, as previously reported (Barnes, Barnes, Costall, Naylor, Naylor and Rudd, 1990). The data represents measurements of density from a minimum of 3 bilateral measurements of total and nonspecific binding, which were made for each of three animals. Drugs

Cisplatin (Lederle) was prepared in normal saline at 70-75°C followed by gradual cooling to 40-50°C and administered immediately. Apomorphine.HCl (Sigma) was prepared in 0.01% sodium metabisulphite. Morphine sulphate (British Drug Houses), and fentanyl.HCl (Janssen Pharmaceutics) were prepared in saline (0.9% w/v). Copper sulphate.5H,O (British Drug Houses) was prepared in distilled water and administered in a volume of 1 ml/kg. All subcutaneous injections were given as OSml/kg. RESULTS

Drug-induced emesis Morphine. The subcutaneous injection of 0.025 mg/kg morphine failed to cause any reliable

retching or vomiting movements,

only l/4 animals

1075

showing a barely discernible emetic response. Increasing the dose to 0.125 mg/kg (s.c.) caused emesis in all animals tested with an onset of action of 3.9-5.6 min, causing 2-13 episodes, 1l-49 retches and l-4 vomits. There was a consistent trend using 0.25 mg/kg (s.c.), for a slight reduction in the onset of emesis (1.5-3.7 min) and an increase in the number of episodes (6-15) retches (19-49) and vomits (3-5). A dose of 0.5 mg/kg (s.c.) caused a similar and maximum response, which was not increased by the use of larger doses at 1.O and 2.0 mg/kg (s.c.). Indeed, using the latter doses, there was a reduction in the number of retches and vomits (Fig. 1). A pretreatment for 15 min with naloxone (0.1 and 1.0 mg/kg s.c.) prevented the emesis induced by morphine (0.5 mg/kg s.c.) (Fig. 2). Fentanyi. Fentanyl (10, 20, 40, and 80pg/kg s.c., n = 4) failed to induce retching and vomiting during a period of 240 min. At doses of 10 and 20 pgg/kg (s.c.) the behaviour of the ferrets was indistinguishable from saline treated controls. However, at 40pg/kg

*

30

.*

12 1

1I 86

0

_C 3 M

45

7

I

0.1 1.0 Naloxone

Naloxone

40 35 $30 5 g25 s 20 $15 I10 5 aI--

M

0.1 1.0 Naloxone

Fig. 2. The effect of naloxone to prevent morphine-induced emesis in the ferret. Naloxone (0. I and 1.0 mg/kg s.c.) was administered 15 min before the injection of morphine (0.5 mg/kg s.c.). Values are the means k SEMs of 4 determinations. If an animal failed to either retch or vomit, then the time to onset of emesis was taken as 30min. Significant differences, relative to morphine-treated control (M) animals, are indicated *P < 0.05, l*P < 0.01 (Mann-Whitney U-test).

1076

N. M. BARNESet al.

(s.c.), the animals appeared quiet and, at 80 pg/kg (s.c.) the animals were prostrate with respiratory depression, which lasted for approximately 120 min. Apomorphine. Based on previous work (Costall, Naylor, Owera-Atepo, Rudd and Tattersall, 1989) and preliminary experiments, a dose of 0.25 mg/kg (s.c.) apomorphine was selected as the dose to produce a reliable emetic response in the ferret. The onset of the emetic action was within 1.2-2.8 min, producing 4-10 episodes of 11-47 retches and 3-7 vomits, with a duration of action of approximately 22 min. Cisplatin. An intravenous dose of 10 mg/kg cisplatin was selected as producing a reliable emetic response in the feret (Costall, Domeney, Naylor and Tattersall, 1986). In the present study, using chronically cannulated animals, the intravenous injection of cisplatin or vehicle did not appear to cause overt discomfort. Cisplatin evoked emesis with a latency of action of 46-96 min and comprised 7-20 episodes of 20-124 retches and 1-13 vomits. Copper sulphate. An intragastric (i.g.) dose of copper sulphate (100 mg/kg) was selected on the basis of previous studies (Costa11 et al., 1990). This dose produced a reliable response in all ferrets with an onset of action of 3-10 min, 7-13 episodes of 3496

retches and 5-13 vomits, with a duration of action of approximately 22 min. The profile of the emetic responses of apomorphine, morphine, cisplatin and copper sulphate is shown in Fig. 3. Fentanyl-induced antagonism of drug induced emesis. There was a clear trend for fentanyl lOpg/kg

(s.c.), to delay the onset of apomorphine (0.25 mg/kg s.c.)-induced emesis and to reduce the numbers of episodes; the numbers of retches and vomits were significantly reduced. Fentanyl 20 pg/kg (s.c.) abolished apomorphine-induced emesis (Fig. 4) and the motor behaviour of the ferrets was not impaired by these small doses of fentanyl. It was an unusual occurrence that during the 15 min pretreatment period, prior to the administration of apomorphine, fentanyl 10 pgg/kg (s.c.) induced in 2/4 ferrets 1 or 2 brief episodes of retching (1-4 retches) occurring 10-l 1 min after treatment with fentanyl. Fentanyl 20pg/kg (s.c.) did not induce emesis during the pretreatment period (Fig. 4). Fentanyl 10 pg/kg (s.c.) completely protected 4/6 ferrets from the emetic actions of morphine (0.5 mg/kg s.c.). In the remaining 2 animals retching was reduced and vomiting abolished. A larger dose

of fentanyl

20 gg/kg

(s.c.)

caused

a similar

MORPHINE

APOMORPHINE

CISPLATIN

COPPER SULPHATE

O-

, 0

.

I 40

-

I. 50

8. 120

1. 150

*. 200

TIME (min) Fig. 3. The emetic profile of apomorphine and other emetogens in the ferret. Apomorphine (0.25 mg/kg) and morphine (0.5 mg/kg) were injected subcutaneously, copper sulphate (100 mg/kg) was administered directly into the stomach and cisplatin (lOmg/kg) was injected intravenously. Values are the means + SEMs of 6-7 determinations.

' 240

Fentanyl inhibits drug-induced emesis MORPHINE

APOMORPHINE

COPPER SULPHATE

CISPLATIN

161

61 T

6

60

51

1

‘1 T

‘21

4

10

7

6

6

4

6 2 1

+rc*+ c

10

20 40

3

*

4 *

IT

5

*

2

2

1

3/

0!I Cl020

-

*

C

10

20 40

60

ii C

10

&A 20 40

60

Fentanyl ~g/ks(s.c.) Fig. 4. The effect of fentanyl to prevent drug-induced emesis in the ferret. Fentanyl (l&80 pg/kg s.c.) was administered 15min before the injection of morphine (0.5 mg/kg s.c.), apomorphine (0.25 mg/kg s.c.), copper sulphate (lOOmg/kg intragastric) and 45min after cisplatin (lOmg/kg i.v.). Values are the mean f SEMs of 46 determinations. If an animal failed to retch or vomit, then the time to onset of emesis was taken as 30min (morphine, apomorphine, copper sulphate) or 240min (cisplatin). Significant differences, relative to emetogen-treated control animals (C), are indicated as *P < 0.05, **P < 0.01 (Mann-Whitney U-test).

antagonism and a dose of 40~g/kg (s.c.) prevented morphine-induced emesis in all animals (Fig. 4). Fentanyl at doses of 10 and 20 pg/kg (s.c.) caused a similar response to modify the emesis induced by cisplatin (lOmg/kg i.v.), with a trend to reduce the number of episodes and retches and a significant reduction in vomiting. The effect of treatment with 40 and 80 pg/kg (CC.) was the same (Fig. 4), completely NP M,k?-0

preventing the emesis in 2/4 and 3/4 animals respectively, and markedly reducing all the parameters in the remaining ferret. Fentanyl, at a dose of lOpg/kg (s.c.) reduced the number of emetic episodes and significantly reduced the numbers of retches and vomits to intragastric administration of copper sulphate. A dose of 20 pg/kg (s.c.) had a similar effect and, at 40 ,ug/kg

N. M. BARNESer al.

1078

(s.c.) fentanyl protected 3/4 animals from retching and vomiting; the animal that was not protected exhibited 39 retches and 4 vomits in 10 episodes, 9min after administration of copper sulphate. At 8O,ng/kg (s.c.), fentanyl completely protected 314 ferrets from retching and vomiting, the ferret that did respond had 1 retch 27 min after administration of copper sulphate (Fig. 4). I~~i~iiio~ by nufoxo~e of the anti-emetic effect of fenranyl. Doses of apomorphine (0.25 mg/kg s.c.), copper sulphate (lOOmg/kg i.g.) and cisplatin

APOMORPHINE

(10 mg/kg i.v.) were selected as causing marked emesis, and fentanyl at a relatively large dose of 80 &kg (s.c.) as antagonising the emesis induced by copper sulphate and cisplatin; a dose of 20pg/kg (s.c.) of fentanyl was selected as antagonising apomorphine-induced emesis. Morphine was not included as an cmetogen in this study since naloxone antagoniscd the morphine-induced cmesis, in its own right (Fig. 2). The effect of a~~stration of naloxone (0.1-1.0 mg/kg s.c.) alone was first determined in non-treated ferrets; naloxone failed to induce emesis

COPPER SULPHATE

CISPLATIN

Fig. 5. The effect of naloxone to antagonise the anti-emetic effects of fentanyl in the ferret. Naloxone (0.1 mg/kg SC.) was given as a 20min pretreatment and fentanyl (20-80pg/kg se.) as a 15min pretreatment, prior to the injection of apomorphine (0.25mg/kg s.c.) or copper sulphate (100 mg/kg intra~stric). Fentanyl and naloxone were administered su~utaneous~y 45 min after cispiatin (10 m&kg i.v.) injection. Values are the means f SEMs of 4-8 determinations. If an animal failed to retch or vomit, then the time to onset of emesis was taken as 30 min (apomorphine and copper sulphate) or 240 min (cisplatin). Significant difference, relative to emetogen-treated controls (APO, CIJ, CP) are indicated as fP < 0.05, **P ~0.01 (Map-Whi~ey U-test); si~ifi~nt antagoni~ by naIoxone of the effects of fentanyl is indicated tP < 0.01 (Mann-Whitney U-test).

Fentanyl

inhibits

(n = 4). The effect of administration of naloxone to modify drug-induced emesis was then determined and there was a consistent trend for naloxone, 0.1 mg/kg (s.c.) to increase the number of retches and significantly increase the numbers of emetic episodes and vomits in animals challenged with apomorphine, and to significantly increase the numbers of episodes, retches and vomits in animals treated with copper sulphate. The numbers of episodes, retches and vomits to cisplatin were also increased by some lO-30% in the naloxone treated animals, although this failed to achieve statistical significance (Fig. 5). Naloxone significantly inhibited the actions of fentanyl to reduce the numbers of episodes, retches and vomits induced by apomorphine, the values returning to control levels. Naloxone was also 40.0-79.9 603

T

T

v-

u

s

2

-r

min

60.0-l 19.9

min

100

:

lz z 2 8

E z

zj

60 1 60 40 20

O 120-240

min

drug-induced

emesis

1079

Table 1. Differential distribution of specific binding of [‘HJDAGO (4.0 nM) (defined by 1.0 pM naloxone) in the brain stem of the ferret. Each value represents the mean f SEM from at least 3 separate animals Specific binding of [‘HJDAGO (fmol/mg tissue)

Area Hypoglossal nucleus Nucleus tractus solitarius Area postrema Inferior olivary nucleus Commisural nucleus tractus solitarius Spinal trigeminal nerve complex Dorsal motor nucleus of the vagus nerve Reticular medulla Cerebellum

30.7 * 5.3 27.2 k 5.2 23.6 f 1.5 20.8 f 1.7 19.5 f 0.6 18.9+_ 1.5 15.3 * 1.9 12.8 f 1.7 6.2 f 0.8

effective in inhibiting the actions of fentanyl to antagonise copper sulphate-induced emesis, although the numbers of vomits remained below the value for control animals. However, whilst there was a trend for naloxone to antagonise the fentanyl (80pg/kg)induced inhibition of cisplatin-induced emesis, the number of emetic episodes and retches and vomits remained well below the control values for cisplatin (Fig. 5). This prompted the use of smaller doses of fentanyl (40pgg/kg) that still retained an effective antiemetic action and a larger dose of naloxone (0.5 mg/kg). Naloxone (0.1 and 0.5 mg/kg) successfully antagonised the inhibitory effects of fentanyl (40 pg/kg) during the first 40 min period of observation, the values being similar to those obtained in animals treated with cisplatin or naloxone alone. However, during the subsequent period of 40 min, the effectiveness of the smaller dose of naloxone had declined (probably reflecting the short half-life of naloxone), whilst the antagonist effects of the larger dose of 0.5 mg/kg naloxone were still evident. Similar comments apply to observations made during the last 2 hr period of recording (Fig. 6). Autoradiographic analysis of the distribution of recognition sites for [3H]DAG0 in the medulla of the ferret

Fig. 6. The ability of naloxone to antagonise the effects of fentanyl to inhibit cisplatin-induced emesis in the ferret. Fentanyl (F) and naloxone (Nal) were administered subcutaneously 45 min after cisplatin (SP, 10 mg/kg i.v.). Values are the means &-SEMs of 4-8 determinations and are expressed as the total number of retches plus vomits occur-

ring during 40-79.9, 80-199.9 and 120-240 min after treatment with cisplatin. Significant differences between controls treated with cisplatin (CP) and each treatment with drug, during each period of time is indicated as *P < 0.05, **P c 0.01; significant antagonism of the effects of fentanyl (40 pg/kg s.c.) by naloxone (0.1 and 0.5 mg/kg s.c.), during each period of time is indicated as tP < 0.05; significant antagonism of the effects of fentanyl (80pg/kg s.c.) by naloxone (0.1 mg/kg s.c.), during each period of time is indicated as AP < 0.05 (Mann-Whitney U-test).

The [‘HIDAGO (4.0nM) was used to identify p opioid recognition sites in the brainstem of the ferret as defined by 1.0 PM naloxone. The hypoglossal nucleus, nucleus tractus solitarius and area postrema contained the highest density of specific binding sites (> 20 fmol/mg tissue), with the inferior olivary nucleus, commisural nucleus tractus solitarius, spinal trigeminal nerve complex, dorsal motor nucleus and reticular medulla, containing at least twice the density of specific binding sites detected in the cerebellum (6.2 + 8 fmol/mg tissue) (Table 1, Fig. 7). DISCUSSION

In the present study, small doses of morphine induced a reliable emetic response in the ferret, the emetic potential decreasing at larger doses. The ability of lesions of the area postrema to protect against morphine-induced emesis in the cat and dog may indicate that the emetic site of action is at the

1080

N. M. BARNESet al.

chemoreceptor trigger zone (Wang and Glaviano, 1954; Costello and Borison, 1977). The latter region is highly vascularised and, being outside the bloodbrain barrier, is readily accessible to many agents including morphine, which has a modest lipophilicity. The emetic action of small doses of morphine in the cat and dog (Costello and Borison, 1977; Blancquaert et al., 1986) and in the ferret (present studies) is mediated through a naloxone-sensitive mechanism, indicating an involvement of opioid receptors. The studies of Blancquaert et al. (1986), using selective opioid receptor agonists, indicate that the opioid ‘emetic’ receptor is likely to be of the 6- and/or K receptor subtype. However, as the dose of morphine is increased, a penetration into the hindbrain structures, mediating the emetic reflex, would permit an action of morphine on the p receptor, which is considered to mediate the antiemetic effect of opioid agonists (Harris, 1982; Blancquaert et al., 1986). Hence, the relative propensity of opioid drugs to affect S, K and p receptors, and their lipophilicity, will determine an emetic-antiemetic profile. Fentanyl is lipophilic and a potent p receptor agonist, with an antiemetic profile in the dog (Blancquaert et al., 1986). This was confirmed in the ferret, fentanyl being effective in doses as small as 10 fig per kg to antagonise morphine-induced emesis. However, the interest of the present study was the additional actions of fentanyl, to prevent emesis induced by apomorphine, copper sulphate and cisplatin and the ability of naloxone to antagonise the effects of fentanyl indicates an involvement of opioid receptors. The three emetogens induce their effects through different mechanisms: apomorphine induces emesis through dopamine receptors in the area postrema, which has traditionally been considered to relay information to the “vomiting centre” (Borison, Borison and McCarthy, 1981; 1984); cisplatin-induced emesis may reflect a 5-hydroxytryptaminergic activation, mediated by 5-HT, receptor mechanisms in the area postrema (Higgins, Kilpatrick Bunce, Jones and Tyers, 1989; Smith, Callahan and Alphin, 1988) although stimulation of the enteric nervous system by cisplatin through vagal mechanisms may cause information to be relayed to the area postrema and the nucleus tractus solitarius (Andrews, Hawthorn and Sanger, 1987; Gunning, Hagen and Tyers, 1987; Andrews, Rapeport and Sanger, 1988); copper sulphate-induced emesis is believed to reflect mainly gastric irritation, information being relayed by the vagus and splanchnic nerves to the central emetic circuits and, importantly, directly to the “vomiting centre” (Wang and Borison, 195la and b; 1952). The ability of fentanyl to inhibit emesis induced by these three agents and morphine, indicates an action beyond the area postrema, at the “vomiting centre” or its efferent systems. The “vomiting centre” was originally identified by stimulation experiments, as an area located within the reticular formation (Borison and Wang, 1949).

However, in later experiments it has proved difficult to clearly identify any one area as the “vomiting centre” (Miller and Wilson, 1983), which is perhaps better understood as a function of a number of separate effector nuclei. This would involve the area postrema, dorsal vagal motor nucleus, nucleus tractus solitarius and reticular formation, where vomiting would occur as a summation of many different effector stimuli, finally triggering the event (Davis, Harding, Leslie and Andrews, 1986). Therefore, fentanyl could be envisaged to influence the emetic circuit at a number of sites and the present autoradiographic data, using [3H]DAG0 supports previous studies, showing the presence of p opiate receptors in the nucleus tractus solitarii (Dashwood, Muddle and Spyer, 1988) but also in the area postrema, hypoglossal nucleus, dorsal motor nucleus of the vagus nerve and reticular formation (see also Palkovitz, 1985; Leslie, 1985). Therefore, the antiemetic actions of fentanyl may reflect an agonist interaction at the p receptor in such systems. Such receptors are located on the vagal afferents in the nucleus tractus solitarii, at least in the cat, (Dashwood et al., 1988) and, whilst the detailed mechanisms for the initiation and control of vomiting are not known, the inputs to the nucleus tractus solitarii may well be concerned with the reflex control of cardiovascular, respiratory and gastrointestinal functions, including emesis. The p receptor, mediating antagonism of emesis, may be subject to an endogenous agonist tone since, in the ferret, there was a clear trend for naloxone to enhance emesis induced by apomorphine (see also Costello and Borison, 1977), copper sulphate and cisplatin. Also, naloxone has been reported to induce emesis after intracerebroventricular injection (McCarthy and Borison, 1974) and naloxone can increase nausea and vomiting in patients receiving chemotherapy (Kobrinsky, Pruden, Cheang, Levitt, Bishop and Tenenbein, 1988). Recently, using in vivo binding techniques in mice, the existence of an in vivo tonic control of p opioid receptors by endogenous opioid peptides has been reported by Meucci, Delay-Goyet, Roques and Zajac (1989). The profile of action of fentanyl as a ‘broad spectum’ antiemetic agent is distinguished from other agents, which have more restricted actions, e.g. the neuroleptics inhibit emesis induced by dopamine agonists and the 5-HT, receptor antagonists inhibit chemotherapy- and radiation-induced emesis (see Costa11 et al., 1990). However, fentanyl is also distinguished from the dopamine and 5-HT, receptor antagonists by the closeness of the doses antagonising emesis and causing sedation. Also, to establish the full extent of its antiemetic activity, its profile should be assessed using other types of emetic stimuli, e.g. motion sickness, anticipatory emesis. In addition, although its effects have been interpreted in terms of an agonism of p receptors a more detailed characterisation, using 6 and K agonists and antagonists, is

Fig. 7. Autoradiographs of the distribution of binding sites for [3HjDAG0 in transverse sections of the brainstem of the ferret. (A) and (I3) Binding of rH]DAGO (4.0 nM, total binding); (C) and @) binding of rEIjDAG0 (4.OnM) in the presence of nafoxone (l.OpM, non-specific binding); (E) and (F) histochemical staining of the section used to generate the total autoradiogram; X, dorsal motor nucleus of the vagus nerve; XII, hypoglossal nucleus; Cb, cerebellum; cNTS, commusural nucleus tractus solitarius; ION, interior olivary nucleus; RM, reticular medulla; STNC, spinal trigeminal nerve complex. Scale bar represents 2 mm.

1081

Fentanyl inhibits drug-induced emesis

required to fully identify the involvement of opioid receptors. In summary, fentanyl is revealed as an antiemetic agent with a broad spectrum to inhibit drug-induced emesis and a useful tool to elucidate the involvement of opioid receptors in emesis. The possibility of its clinical use as an antiemetic agent is made problematical by the p agonist-related respiratory, cardiovascular and other behavioural effects. The latter may include nausea and/or vomiting in some patients, at least when used as an adjunct to anaesthesia (Sinclair and Cooper, 1983). With this in mind, it remains a major challenge to dissociate the potentially useful antiemetic properties of opioid agonists from the serious and unwanted side effects. Acknowledgemenrs-The authors gratefully acknowledge gifts of fentanyl from Janssen Pharmaceutics and cisplatin from Lederle. J. A. Rudd is supported by a Glaxo studentship. REFERENCES

Andrews P. L. R., Hawthorn J. and Sanger G.J. (1987). The effects of abdominal visceral nerve lesions and a novel S-HT-M receptor antagonist on cytotoxic and radiation induced emesis in the ferret. I. Physiol. 382: 47P.

Andrews P. L. R., Rapeport W. G. and Sanger G. J. (1988) Neuropharmacology of emesis induced by anti-cancer therapy. Trends Pharmac. Sci. 9: 334-341 Barnes J. M., Barnes N. M., Costall B., Naylor I. L., Naylor R. J. and Rudd J. A. (1990) Topographical distribution of 5-HT, receptor recognition sites in the ferret brain stem. Naunyn -Schmiedebergs Arch. Pharmac. 342: 17-2 1. Bayliss (1940). Studies on the mechanisms of vomiting produced by staphylococcus enterotoxin. I. exp. Med. 7: 669684.

Beleslin D. B., Krstic S. K., Stefanovic-Denic K., Strbac M. and Micic D. (1981). Inhibition of morphine and morphine-like drugs of nicotine-induced emesis in cats. Brain Res. Bull. 6: 451-453.

Bhargava K. P., Dixit K. S. and Gupta Y. K. (1981). Encephahn receptors in the emetic chemoreceptor trigger zone of the dog. Br. J. Pharmac. 72: 471475. Blancquaert J.-P., Lefebvre R. A. and Willems J. L. (1986) Emetic and antiemetic effects of opioids in the dog. Eur. J. Pharmac. 128: 143-150.

Borison H. L., Borison R. and McCarthy L. (1981) Phylogenie and neurologic aspects of the vomiting process. J. c/in. Phannac. 21: 23S-29s.

Borison H. L., Borison R. and McCarthy L. E. (1984) Role of the area postrema in vomiting and related functions. Fed. proc. 43: 2955-2958.

Borison, H. L. and Wang S. C. (1949) Functional localisation of the central coordinating mechanism for emesis in cat. J. Neurophysiol. 12: 304-313. Borison, H. L. and Wang S. C. (1953) Physiology and pharmacology of vomiting. Pharmac. Rev. 5: 193-230. Costall B.. Domenev A. M.. Navlor R. J., Owera-Ateno J. B., ‘Rudd J.- A. and Tattersall F. D. (1990) Fluphenazine, ICS205-930 and dl-fenfluramine differentially antagonise drug-induced emesis in the ferret. Neurohharmacology 29: 453-462.

Costa11 B.. Domenev A. M., Navlor R. J. and Tattersall F. D. (1986) 5-Hydroxytryptamine M. receptor antagonism to prevent cisplatin-induced emesis. Neuropharmacology 25: 959-961.

1083

Costa11 B., Naylor R. J., Owera-Atepo J. B. and Tattersall F. D. (1989) The responsiveness of the ferret to apomorphine induced emesis. Br. J. Pharmac. 96: 329P.

Costello D. J. and Borison H. L. (1977) Naloxone antagonises narcotic self-blockade of emesis in the cat. J. Pharmac. exp. Ther. 203: 222-230.

Dashwood M. R., Muddle J. R. and Spyer K. M. (1988) Opiate receptor subtypes in the nucleus tractus sohtarii of the cat: the effect of vagal section. Eur. J. Pharmac. 155: 85-92. Davis C. J., Harding R. K., Leslie R. A. and Andrews P. L. R. (1986) The organisation of vomiting as a protective reflex. In: Nausea and Vomiting: Mechanisms and Treatment (Davis C. J., Lake-Bakaar G. V. and Grahame-Smith D. G., Eds) pp. 65-75. Springer, Berlin. Gunning S. J., Hagan R. and Tyers M. B. (1987) Cisplatin induces biochemical and histological changes in the small intestine of the ferret. Br. J. Pharmac. 90: 135P. Harris A. L. (1982) Cytotoxic-therapy-induced vomiting is mediated via enkephalin pathways. The Lancer I, 714-716. Higgins G. A., Kilpatrick G. J., Bunce K. T., Jones B. J. and Tyers M. B. (1989) 5-HT, receptor antagonists injected into the area postrema inhibit cisplatin-induced emesis in the ferret. Br. J. Pharmac. 97: 247-255. Kobrinsky N. L., Pruden P. B., Cheang M. S., Levitt M., Bishop A. J. and Tenet&in M. (1988) Increased nausea and vomiting induced by naloxone in patients receiving cancer chemotherapy. Am. J. Ped. Hem. Oncol. 10: 206208. Leslie R. A. (1985) Neuroactive substances in the dorsal vagal complex of the medulla oblongata: nucleus of the tractus solitarius, area postrema, and dorsal motor nucleus of the vagus. Neurochem. Int. 7: 191-211. McCarthy L. and Borison H. L. (1974) Vomiting evoked by naloxone compared with apomorphine injected into the cerebrospinal fluid in cats. Fed. Proc. 23: 516. Meucci E., Delay-Goyet P., Roques B. P. and Zajac J-M. (1989) Binding in vivo of selective p opioid receptor agonists: Opioid receptor occupancy by endogenous enkephalins. Eur. J. Pharmac. 171: 167-178. Miller A. D. and Wilson V. J. (1983) ‘Vomiting centre’ reanalyzed: an electrical stimulation study. Brain Res. 270: 154-158.

Niemegeers C. J. E., Schellekens K. H. L., Van Bever W. F. M. and Janssen P. A. J. (1976) Sufentanil, a very potent and extremely safe intravenous morphine-like compound in mice, rats and dogs. Arzneim.-Forsch. 26: 1551-1556. Palkovitz M. (1985) Distribution of neuroactive substances in the dorsal vagal complex of the medulla oblongata. Neurochem. Int. 7: 213-219.

Sinclair M. E. and Cooper G. M. (1983) Alfentanil and recovery. Anaesthesia 38: 435437. Smith W. L., Callahan E. M. and Alphin R. S. (1988) The emetic activity of centrally administered cisplatin in cats and its antagonism by zacopride. J. Pharm. Pharmac. 40: 142-143.

Wang S. C. and Borison H. L. (1951a) The vomiting centre: its destruction by radon implantation in dog medulla oblongata. Am. J. Physiol. 166: 712-717. Wang S. C. and Borison H. L. (1951b) Copper sulphate emesis: a study of efferent pathways from the gastrointestinal tract. Am. J. Physiol. 164: 520-526. Wang S. C. and Borison H. L. (1952) A new concept of organisation of the central emetic mechanism: recent studies on the sites of action of apomorphine, copper sulphate and cardiac glycosides. Gastroenrerology 22: I-12.

Wang S. C. and Glaviano V. V. (1954) Locus of emetic action of morphine and hydergine in dogs. J. Pharmac. exp. Ther. 111: 329-334.