Introduction

The pharmacological control of nausea and vomiting produced by chemotherapy treatment, such as cisplatin, is essential to ensure adherence to the therapeutic regime. The most typically employed antiemetic treatments are those that act as antagonists of the serotonin subtype-3 (5-HT3) receptor. Indeed, 5-HT3 antagonists also suppress vomiting in a variety of species, including cats (Rudd et al. 2000), ferrets (Ozaki and Sukamoto 1999) and shrews (Torii et al. 1991; Ito et al. 1995; Andrews et al. 1996, 2000; Matsuki et al. 1997; Darmani 1998). In human clinical trials, although 5-HT3 antagonists are effective in treating the acute phase of chemotherapy-induced emesis, considerable evidence indicates that the 5-HT3 antagonists are relatively ineffective in attenuating the delayed phase of chemotherapy induced emesis (Fabi et al. 2003) as well as anticipatory nausea and vomiting upon re-exposure to the cues associated with treatment (Tyc et al. 1997; Morrow et al. 1998).

The results of clinical trials also demonstrated that cannabinoids, including Δ9-tetrahydrocannabinol (Δ9-THC), can prevent not only acute nausea and vomiting in chemotherapy patients (for recent review, see Tramer et al. 2001), but also the delayed phase (Abrahamov et al. 1995). Experimental evidence with animals confirms the anti-emetic potential of cannabinoids in cats (McCarthy and Borison 1981), ferrets (Simoneau et al. 2001; Van Sickle et al. 2001), pigeons (Feigenbaum et al. 1989), and shrews (Darmani, 1998, 2001a, 2001b, 2001c; Parker et al. 2004). The anti-emetic effect of cannabinoids may be mediated by action at specific cannabinoid (CB1 and CB2) receptors (Darmani 2001b, 2001c; van Sickle et al. 2001). Cannabinoids have also been reported to be more effective than other antiemetics in suppressing anticipatory nausea and vomiting associated with chemotherapy (Weddington et al. 1982; Gurley et al. 1998). In an animal model of anticipatory nausea and vomiting, Parker and Kemp (2001) reported that Δ9-THC suppressed the expression of conditioned retching in the Suncus murinus (house musk shrew) elicited by an environment that had previously been associated with lithium-induced vomiting. This suppressed retching was apparent at doses that did not suppress general activity level.

The cannabinoid and serotonin systems have been shown to interact in the emetic systems of the brain (Himmi et al. 1996) and the gut (Fan 1995). In the brainstem, both CB1 and 5-HT3 receptors have been found in the solitary tract nucleus, an important structure in the emetic pathway (Himmi et al. 1996, 1998). Furthermore, most Δ9-THC-sensitive neurons in this structure showed opposite responses to Δ9-THC and the 5-HT3 receptor agonist, 1-phenylbiguanide, which induces vomiting in ferrets (Higgins et al. 1989). Using the patch-clamp technique, Fan (1995) reported that cannabinoid receptor agonists inhibit the inward current induced by 5-HT3 receptors in the cell bodies of vagal afferent neurons. It has also been shown that the pharmacological activity of the endocannabinoid, anandamide, may be partially mediated through serotonin receptors (Kimura et al. 1998). Finally, Hermann et al. (2002) reported extensive CB1 coexpression with 5-HT3 receptors in several cortical regions.

Since serotonin and cannabinoid systems have been shown to interact in the brain and the gut, it is possible that Δ9-THC and ondansetron (5-HT3 receptor antagonist, OND) may have a synergistic or additive effect on the control of vomiting and retching. Recently, Soderpalm et al. (2001) report that ondansetron may be more effective than smoked marijuana in the control of nausea, but they are both effective in the control of vomiting produced by syrup of ipecac in humans; however, only limited doses were evaluated. There are no experimental studies that directly compare the potential of cannabinoids and 5-HT3 antagonists to suppress vomiting produced by cisplatin. Furthermore, the effect of co-administration of a cannabinoid and a 5-HT3 receptor antagonist on emesis has not been investigated. Therefore, the following experiments were designed to evaluate the relative efficacy of Δ9-THC and OND as well as their combined efficacy in preventing cisplatin-induced vomiting and retching in the Suncus murinus. Additionally, the potential of the CB1 receptor antagonist, SR141716, to reverse OND-induced inhibition of vomiting was evaluated. Although SR has been reported to produce vomiting in shrews on its own at doses of 5 mg/kg or greater (Darmani 2001a), it did not produce vomiting at the 2.5 mg/kg dose used here (Darmani 2001a; Parker et al. 2004). Finally, the potential of cannabidiol (CBD), the principal non-intoxicating cannabinoid found in marijuana, to prevent cisplatin-induced emesis was evaluated. Parker et al. (in press) observed a suppression of lithium chloride-induced vomiting in the Suncus murinus (house musk shrew) not only by Δ9-THC pretreatment, but also by pretreatment with CBD. CBD produced a biphasic effect: Lower doses (5 and 10 mg/kg) suppressed lithium-induced vomiting while higher doses (24 and 40 mg/kg) potentiated lithium-induced vomiting. The CB1 receptor antagonist, SR-141617A (SR), reversed the effect of Δ9-THC, but not the effect of CBD on lithium-induced vomiting, suggesting that the effect of CBD was not CB1 receptor mediated.

Materials and methods

Subjects

The subjects were 63 male (30–52 g) and 64 female (21–31 g) Suncus murinus bred and raised at the Wilfrid Laurier University colony. Each group consisted of approximately half males and half females. Because of the toxicity of cisplatin, the animals were euthanized immediately following each trial. The animals were housed individually in polyethylene cages (25×16×12 cm) with pine wood shavings and shredded paper towels for nesting material. The colony room was maintained at 22±1°C at 14:10 h light-dark schedule (light on at 0700 hours). The shrews had free access to dry cat food plus mink diet and tap water. Normal HCl (1 ml/l) was added to the drinking water to maintain a pH of 5.5 in order to reduce the risk of gastrointestinal disease. All procedures were approved by the Wilfrid Laurier University Animal Care Committee, in accordance with the regulations of the Canadian Council on Animal Care.

Drugs

All pretreatment drugs were prepared in a solution of 1 ml ethanol/1 ml emulsifier (Sigma)/18 ml saline. Δ9-THC was prepared as a 1 mg/ml solution of the vehicle, and was administered in volumes of 1.25, 2.5, 5 and 10 ml/kg. OND was prepared as a 1 mg/ml solution and was administered in volumes of 0.6, 1.5, 3 and 6 ml/kg. For lower doses, OND was prepared as 0.02 mg/ml and 0.2 mg/ml solutions and administered at a volume of 1 ml/kg. SR141716 was prepared as a 1 mg/ml solution and administered in a volume of 2.5 ml/kg. CBD was prepared as a 1 mg/ml solution and was administered at volumes of 5 ml/kg and 40 ml/kg. Vehicle was administered at volumes of 10 ml/kg when compared with THC and OND and at a volume of 40 ml/kg when compared with CBD. The treatment drug was cisplatin prepared as a 1 mg/ml solution of saline and administered in a volume of 20 ml/kg. All drugs were injected intraperitoneally.

Apparatus

The clear Plexiglas observation chambers (22.5×26×20 cm) were illuminated by four 60 W lights suspended from the chamber’s floor. A mirror was mounted at a 45° angle beneath the chamber floor, which allowed for the observation of the ventral surface for the recording of activity and behaviours of the animal.

Procedure

Each animal was offered four mealworms (Tenebrio sp.) in its home cage 15 min prior to pretreatment injections. The shrews received two pretreatment injections 30 min prior to receiving an injection of cisplatin (20 mg/kg) or saline (20 ml/kg), and 15 min later were placed in the observation chamber for 60 min. The test chambers were thoroughly cleaned between trials.

The groups differed on the basis of the pretreatment injections. The Δ9-THC groups received an injection of 1.25 (n=6), 2.5 (n=4), 5.0 (n=4), or 10.0 (n=4) mg/kg Δ9-THC immediately prior to receiving an injection of vehicle (5 ml/kg). The OND groups received an injection of 0.02 (n=6), 0.2 (n=6), 0.6 (n=8), 1.5 (n=8), 3.0 (n=4), or 6.0 (n=4) mg/kg of OND immediately prior to receiving an injection of vehicle (5 ml/kg). The vehicle group (n=8) received two injections of vehicle (5 ml/kg each). The OND+Δ9-THC group (n=7) received an injection of 1.25 mg/kg Δ9-THC immediately prior to receiving an injection of 0.02 mg/kg OND. The SR-OND groups received an injection of 2.5 mg/kg SR141716 immediately prior to an injection of vehicle (5 ml/kg; n=4), 0.2 mg/kg (n=6), 0.6 mg/kg (n=4) or 1.5 mg/kg (n=4) OND.

Additionally, the potential of CBD to suppress cisplatin-induced vomiting was evaluated. The CBD groups were treated as the Δ9-THC and OND groups except that they received a single pretreatment injection 30 min prior to receiving an injection of 20 mg/kg cisplatin or 20 ml/kg saline. The groups included: vehicle (40 ml/kg)-saline (n=7), 40 mg/kg CBD-saline (n=7), vehicle (40 ml/kg)-cisplatin (n=8), 5 mg/kg CBD-cisplatin (n=6), 40 mg/kg CBD-cisplatin (n=6).

During the 60-min observation period, an observer recorded the frequency of retching and vomiting as well as the latency to the first episode of retching or vomiting. Retching was characterized by abdominal contractions and wide opening of the mouth, without the expulsion of gastrointestinal material. Vomiting was defined as abdominal contractions and expulsion of gastrointestinal substance.

Data analysis

The dose-response data for OND and THC alone and were each evaluated by a one-way analysis of variance (ANOVA) and subsequent least significant difference (LSD) pairwise comparison tests. The vehicle group (10 ml/kg) was the same for both analyses. Subthreshold doses of OND and THC were then combined and the data were compared with that for each of the doses separately in a one-way ANOVA and subsequent LSD pairwise comparison tests. The ability of SR 141716 to reverse the effects of OND was evaluated in a 2 (pretreatment-vehicle or SR) by 4 (dose of OND [0.0, 0.2, 0.6, 1.5 mg/kg]) ANOVA. Finally, the ability of CBD to suppress cisplatin induced vomiting and retching was evaluated as a one-way ANOVA with LSD pairwise comparison tests.

Results

Ondansetron

The dose-response data revealed that OND pretreatment suppressed cisplatin-induced vomiting and retching. The upper section of Fig. 1 depicts the mean number (±SEM) of vomiting episodes among groups pretreated with various doses of OND. The ANOVA revealed that the groups significantly differed [F(6,37)=7.5; P<0.01]. Least significant difference (LSD) pairwise comparison tests revealed group vehicle (0.0 mg/kg OND) and 0.02 mg/kg OND vomited more than any other OND pretreated group (P<0.05). The middle section of Fig. 1 presents the mean (±SEM) latency (s) of first vomiting episode the groups pretreated with OND; there was a significant dose effect [F(6,37)=12.7; P<0.01]. As was evident with vomiting frequency, pairwise comparisons tests showed that groups vehicle and 0.02 mg/kg OND displayed shorter vomiting latency then any other OND group (P<0.05). The bottom portion of Fig. 1 shows the mean number of retches displayed by the OND groups. There was a significant dose effect in the number of retches displayed during the observation session [F(6,37)=5.7; P<0.01]. LSD pairwise comparison tests indicated that groups vehicle and 0.02 mg/kg displayed more retching than all groups except group 6.0 mg/kg (P<0.05). Furthermore, group 6.0 mg/kg displayed more retching than groups 0.2 and 1.5 mg/kg (P<0.05). Therefore, as measured by retching, OND produced a bell shaped dose-response curve, with the high dose (6.0 mg/kg) no longer reversing the cisplatin-induced retching.

Fig. 1
figure 1

Effect of OND on cisplatin-induced vomiting and retching. Mean (±SEM) number of vomiting episodes (top section), latency to the first vomiting episode (middle section) and number of retches (bottom section) displayed by shrews pretreated with various doses of OND, 30 min prior to an injection of cisplatin (20 mg/kg)

Full size image

Δ9-THC

Δ9-THC inhibited cisplatin-induced vomiting and retching in a dose-dependent manner. The upper section of Fig. 2 presents the mean (±SEM) number of vomiting episodes displayed by the groups pretreated with Δ9-THC. The single factor ANOVA revealed a significant effect [F(4,21)=3.3; P<0.05]. LSD pairwise comparisons tests revealed that the vehicle group vomited more frequently than any other group (P<0.05), except group 1.25 Δ9-THC. Furthermore, 1.25 mg/kg Δ9-THC group vomited more frequently than group 10 Δ9-THC (P<0.05) and marginally differed from groups 2.5 and 5 Δ9-THC (P=0.061). The middle section of Fig. 2 shows the mean latency (s) to the first vomiting episode for the Δ9-THC groups. There was a significant dose effect for latency to first emetic episode [F(4,21)=9.1; P<0.01]. LSD post hoc comparison tests show that vehicle and 1.25 mg/kg groups showed a shorter latency to the first vomiting episode than all other groups (P<0.05). The bottom section of Fig. 2 displays the mean number of retches during the observation session for the Δ9-THC groups. The groups differed significantly [F(4,21)=5.8; P<0.01]. Post hoc comparison tests showed that both vehicle and 1.25 mg/kg Δ9-THC groups retched more frequently than any other group (P<0.05).

Fig. 2
figure 2

Effect of Δ9-THC on cisplatin-induced vomiting and retching. Mean (±SEM) number of vomiting episodes (top section), latency to the first vomiting episode (middle section) and number of retches (bottom section) displayed by shrews pretreated with various doses of Δ9-THC, 30 min prior to an injection of cisplatin (20 mg/kg)

Full size image

Ondansetron and Δ9-THC

The combined pretreatment of OND+Δ9-THC, at doses ineffective on their own, effectively suppressed cisplatin-induced vomiting and retching. Figure 3 (±SEM) presents the mean number of vomiting episodes, mean latency to first vomiting episode, and mean number of retches elicited by cisplatin in the shrews pretreated with vehicle, 0.02 mg/kg OND, 1.25 mg/kg Δ9-THC and 0.02 mg/kg OND+1.25 mg/kg Δ9-THC. The one-way ANOVA revealed that the number of vomiting episodes differed between groups [F(3,23)=5.7; P<0.01]. LSD post hoc comparison tests show that the combined OND+Δ9-THC group (0.02 mg/kg OND and 1.25 mg/kg Δ9-THC) vomited significantly fewer times than all other groups (P<0.01). The groups also differed on the basis of the latency to the first vomiting episode [F(3,23)=8.0; P<0.01]. LSD post hoc comparison tests revealed that group 0.02 OND+1.25 Δ9-THC displayed a longer latency to the first vomiting episode than all other groups (P<0.001). Finally, the number of retches differed between groups [F(3,23)=7.2; P<0.01]. Post hoc LSD comparison tests show that the combined 0.25 OND+1.25 Δ9-THC group retched fewer times then all other groups (P<0.05).

Fig. 3
figure 3

Effect of a combination of OND+Δ9-THC on cisplatin-induced vomiting and retching. Mean (±SEM) number of vomiting episodes (top section), latency to the first vomiting episode (middle section) and number of retches (bottom section) displayed by shrews pretreated with ineffective doses of OND (0.02 mg/kg) and Δ9-THC (1.25 mg/kg) and an effective combined dose of 0.02 mg/kg OND and 1.25 mg/kg Δ9-THC, 30 min prior to an injection of cisplatin (20 mg/kg)

Full size image

SR141716 and OND

SR did not reverse the effect on OND on cisplatin-induced vomiting. Figure 4 presents the mean frequency of vomiting episodes displayed by the relevant groups pretreated with vehicle prior to OND (solid bars) and the groups pretreated with SR prior to OND (open bars). There was only a significant effect of OND dose [F(3,40)=9.90; P<0.001]; shrews pretreated with vehicle displayed significantly more vomiting than any other group (P<0.01).

Fig. 4
figure 4

Mean (±SEM) number of vomiting episodes displayed by shrews pretreated with vehicle (solid bars) or 2.5 mg/kg SR141716 and one of various doses of OND, 30 min prior to an injection of cisplatin (20 mg/kg)

Full size image

CBD

CBD produced a biphasic effect on cisplatin-induced vomiting; a dose of 5 mg/kg suppressed vomiting, but a dose of 40 mg/kg potentiated vomiting (see Fig. 5). The ANOVA revealed a significant group effect [F(4,29)=25.6; P<0.001]; LSD post-hoc comparison tests revealed that group Veh-Cisplatin displayed significantly more vomiting than groups Veh(40)-Saline (P<0.001), CBD(40)-Saline (P<0.01) or CBD(5)-Cisplatin (P<0.001) and significantly less vomiting than group CBD(40)-Cisplatin (P<0.001). It is also interesting to note that even when administered at a volume of 40 ml/kg, the vehicle did not produce emetic behaviors on its own.

Fig. 5
figure 5

Mean (±SEM) number of vomiting episodes displayed by shrews pretreated with CBD (5 or 40 mg/kg) or VEH 30 min prior to an injection of cisplatin (20 mg/kg) or saline

Full size image

Discussion

Both ondansetron and Δ9-THC dose-dependently suppressed cisplatin-induced vomiting in the Suncus murinus. Furthermore, doses of the two drugs that were ineffective alone (0.02 mg/kg OND and 1.25 mg/kg Δ9-THC) produced complete suppression of vomiting and retching when they were combined. This suggests that a combination treatment of lower doses of these agents may be an effective alternative treatment for vomiting in chemotherapy patients.

Since the cannabinoid and serotonin systems have been shown to interact in the solitary tract nucleus (Himmi et al. 1996, 1998) as well as in the gut (Fan 1995) and that CB1 receptors are coexpressed with 5-HT3 receptors (Herman et al. 2002), we evaluated the potential of the CB1 antagonist SR 141716 to reverse the anti-emetic effects of OND. However, we found that the antiemetic effect of ondansetron was not reversed by a dose of SR141716A that reversed the anti-emetic effect of the agonist WIN 55,212 (Darmani 2001c) and Δ9-THC (Parker et al. 2004), suggesting that the effect of OND is not CB1 receptor mediated. It is not known whether the combined effect of THC and OND is mediated by a synergistic action on a single system or by an additive action of their effects on several systems.

Interestingly, not only did the intoxicating component of marijuana, Δ9- THC, reduce vomiting, but also the non-intoxicating component, CBD. As we have previously demonstrated with lithium-induced vomiting, CBD produced a biphasic effect with a low dose (5 mg/kg) suppressing cisplatin-induced vomiting and a high dose (40 mg/kg) potentiating cisplatin-induced vomiting. Since CBD does not bind with known CB receptors, this anti-emetic effect does not appear to be mediated by CB1 or CB2 activity.

This report is the first direct comparison of the effectiveness of combined doses of cannabinoids and serotonin antagonists in the control of acute vomiting and retching. Our results suggest that a combination of subthreshold doses of these agents can prevent acute vomiting and may be of use in clinical treatment. Additionally, the non-intoxicating component of cannabis, CBD, also prevented acute cisplatin-induced vomiting in the shrew. Our results, however, do not address the effects of these agents (or their combination) on phases of nausea and vomiting that are resistant to treatment with serotonin antagonists: a) anticipatory nausea and vomiting experienced during re-exposure to chemotherapy related cues, and b) the delayed phase of nausea and vomiting produced by chemotherapy treatment (Tyc et al. 1997; Morrow et al. 1998; Fabi et al. 2003). It is conceivable that these phases of nausea and vomiting may be more effectively treated by THC or CBD than OND. Indeed, in an animal model of anticipatory nausea and vomiting, Parker and Kemp (2001) demonstrated that THC suppressed anticipatory retching in the Suncus murinus. It is not known if OND would produce a similar effect. Animal models for delayed nausea and vomiting are difficult to develop because of the individual variability inherent in the onset of bouts of vomiting and retching that follow the acute phase. Future experiments will attempt to shed light on better treatment regimes for these phases of chemotherapy-induced nausea and vomiting.