The objective of this study was to evaluate the anti-nausea efficacy of ondansetron, a 5HT3 receptor antagonist, in dogs with vestibular disease. In the current open-label prospective multicentre study in 16 dogs with vestibular disease, the degree of nausea and clinical signs suggestive of nausea-like behaviour were significantly reduced after ondansetron administration. Around a third of the patients in this study also showed vomiting in addition to nausea, which also did stop after ondansetron was given. Despite the limitations of an open-label prospective study, this study provides the first evidence of the potential benefit of ondansetron in the treatment of nausea induced by vestibular syndrome and highlights the importance of assessing the degree of nausea in these neurological patients.
In veterinary care, there is a need for a better awareness of veterinary surgeons for the difference between a drug’s antiemetic and/or antinausea effect. Nausea is relatively easy to measure in people, as they can report the severity of nausea they are experiencing. In dogs, however, behaviour and facial expression must be observed precisely in order to adequately evaluate their level of nausea.8,29 Vomiting is often seen and described in vestibular disorders,31 however, the prevalence of preceding nausea has not been described. Radulescu et al. (2020) and Schunk et al. (1988) described a prevalence of vomiting of 25.7% and 40%, respectively.2,32 This is consistent with our results showing an overall prevalence of vomiting in 31.3% of dogs. Nevertheless, all dogs in our study showed clinical signs and behaviour suggestive of nausea. This supports reports that nausea and vomiting can occur separately and that it should not be assumed that a dog that is not vomiting cannot be nauseous. Studies of maropitant, an antiemetic drug developed for the use in veterinary medicine, described the cessation of vomiting but still measurable signs of nausea.11–18 For some owners or veterinarians the absence of vomiting might be an acceptable and satisfying treatment outcome. However, human patients being treated with chemotherapeutics usually consider nausea control more important than emesis control.33,34 Furthermore, one could assume when nausea leads to vomiting, there is temporary relief of the nausea sensation, but blocking vomiting without relieving the nausea may be burdensome. This highlights the importance of nausea research in veterinary patients. Kraus et al. (2019) also determined that many canine owners are aware of the discomfort and negative influence of nausea on animal welfare and would accept higher costs and longer hospital stays in return for an effective relief for their pets.35
Golden Retriever (n = 4) and mixed-breed dogs (n = 2) were most likely to have a diagnosis of ‘vestibular syndrome’ in this case series. Bongartz et al. (2020) also found mixed-breed dogs (15.1%) to be most likely affected.36 Similar breeds were mentioned by some recently published studies consistent with our findings.2,37 In the present study, the median age at onset of neurological deficits was 115 months (9.58 years). Median age reported recently elsewhere ranged from 6.8 years to 12.68 years.2,32,38−40 This apparently wide range in age could represent the different study inclusion criteria. Idiopathic vestibular syndrome tends to appear more frequently in older dogs in contrast to meningoencephalitis of unknown origin, which is a frequent cause of central vestibular syndrome. It mainly affects female brachycephalic dog breeds younger than 5.5 years.40 Neurological examination is a very accurate clinical tool to determine the presumed neuroanatomical localisation of neurological diseases within the vestibular system in more than 90% of cases, according to Bongartz et al. (2020).36 The main clinical findings recorded in the neurological examination of the 16 dogs were head tilt and nystagmus, followed by ataxia. These signs are the most frequently observed clinical findings consistent with vestibular syndrome.2,32,37,41 In summary, although the sample size in this case series was small, the cases included appear to be representative for the commonly affected canine population by vestibular syndrome.
The pathogenesis of nausea is complex.9,10,21,22,27,42 It is not completely understood and includes central and peripheral stimuli. Initiating stimuli originate from the cerebral cortex and limbic system but also from gastric dysrhythmias, the vestibular system, the cerebellum and the area prostema. The stimuli terminate in the nucleus tractus solitarius (NTS) which is located in the brainstem and forms the dorsal vagal complex, together with the area prostema and the dorsal motor nucleus of the vagus. The NTS integrates stimuli of the afferent neurons, which are mediated by serotonin/dopamine, histamine/acetylcholine in combination with serotonine/dopamine and substance P.21,22 Stern et al. (2011) hypothesised that the brainstem is not involved in the genesis of nausea but is required for the prodromal signs (salivation, lip licking, restlessness, lethargy, vocalisation, emesis). Therefore, the pathway of vomiting and nausea can be assumed from this point on separated.42 For the induction of the sensation nausea and the associated increased release of arginine vasopressin (AVP)8,9,42−45, ‘higher’ regions of the brain are required, while for vomitus the functions of the brainstem as the ‘vomiting centre’ are sufficient.42 Projections from the rostral NTS target ‘higher brain areas’ like the hypothalamus via the lateral parabrachial nucleus, in order to initiate the release of AVP from the posterior pituitary. Another hypothalamic effect is the reduction of food intake and the modulation of autonomic outflow via activation of the sympathetic system leading to an increase in epinephrine and serotonin (5-HT) followed by vasoconstriction. Other ‘higher’ brain areas activated by the NTS probably lead to the conscious awareness of the sensation nausea.42
The vestibular system influences not only the cerebellum, spinal cord and extraocular muscles but also the cranial nerve III and the vestibular nuclei as direct pathways, which could induce nausea. By activation of projections to the dorsal vagal complex and ascending projections to higher brain areas like thalamus, lateral postcentral gyrus, insular cortex and temporoparietal cortex the induction of nausea can be modulated. The complexity and multidimensional nature of nausea makes the pharmacological management challenging.34
The 5-HT3 receptors form serotonin-gated ion channels that interact with further 5-HT- receptor subtypes as well as with other neurotransmitters.24 They have a wide distribution and range of function in the central and peripheral nervous system.46 Neurons expressing immunoreactivity for the 5-HT3 receptor subtypes are located in the forebrain, brainstem (especially trigeminal motor and facial nuclei) as well as in the spinal cord.47 In addition to other regions in the central nervous system, the distribution of 5-HT3 receptors in the area prostema (the chemoreceptor trigger zone), limbic and cortical regions,48 and the NTS (especially the rostral part, which interacts with higher brain regions42) is particularly important with regard to pharmacological antinausea intervention.49–53 Balaban et al. (2014) examined neuronal activation after galvanic vestibular stimulation in cats with the help of the detection of c-Fos expression.28 They concluded two networks being positively correlated with the severity of motion sickness signs. The networks include medial, lateral, superior and inferior vestibular nuclei, the lateral nuclei of the NTS, parabrachial nucleus and in general serotonergic and non-serotonergic projections.
Miller et al. (1996) found a region in the human inferior frontal gyrus activated via head movement during yaw-axis rotation in the matter of number of dipoles related to the intensity of nausea.54 The same region was detected in functional MRI after galvanic vestibular stimulation or caloric testing. This specific region was only activated through motion sickness associated nausea and not via speech or finger movement, which are the normal activating stimuli of this region. The same region was also activated after the ingestion of the syrup of ipecac, a well-known herbal substance that induces nausea and vomiting. After ondansetron was applied, the intensity of dipoles in this region was significantly reduced.42,54 Components of the vestibular system, representing one possible origin of nausea stimuli, have a widespread distribution of 5-HT-receptors. These regions of receptor accumulation serve as possible anchor points in the axis of nausea-induction to be targeted pharmacologically. The serotonin receptor antagonist ondansetron induces a greater antinausea effect than maropitant or metoclopramide. Neurokinin-1-receptor and dopamine-2-receptor are represented within the brain, but not being involved in the release of AVP for example.55,56
Nausea is a subjective condition and its presence and magnitude are difficult to assess in both animals and humans, as there still is a lack of a direct measurement method until today. In humans we have to rely on self-reports while in animals we have to rely on a person's interpretation of the observed behaviour. In human medicine, there are several types of validated scales including visual analogue scales (VAS) and numerical rating scales (NRS) in terms of self-reporting.57,58 Although comparing a subjective sensation between several individuals remains challenging, the recognition of nausea in humans itself is much easier than in animals. Therefore, the first step to improve the welfare of canine patients suffering from nausea is to simplify and to improve the recognition of the condition in the first place. Therefore valid methods for measurement are needed. Kenward et al. (2015) refer in their publication to pain as an analogy to nausea and compare the concept of nociception with ‘nausiception’.8
Both, nausea and pain, are subjective sensations which require a conscious perception of the individual and serve as a protective function. These parallels allow access to pain rating scales as possible approach to evaluate nausea, as there are no validated scores for nausea in animals. Pain has been often measured by means of simple descriptive scale (SDS), NRS or VAS with only focus on intensity.59 Holton et al. (1998) and Downie et al. (1978) concluded that the NRS, which is extensively used in human medicine, represents a reliable compromise between a SDS with a lower level of sensitivity as well as the VAS and its more complicated use.60,61
The used NRS in the present study is a modified version of the one established by Rau et al. (2010) and further validated by Kenward et al (2015).17,29 In this case series, we have chosen scale descriptions relying on typical prodromal signs of nausea,10,13,17,27,28 which offers the possibility of making a statement about the dogs’ overall level of nausea. In addition to assessment of nauseaevoked behaviour patterns, the used scale also included evaluation of dog-human-interactions by evaluating lethargy and, therefore, the dogs’ reaction to external stimuli. In general, the scale descriptions facilitated and standardised evaluation although it is less sensitive than the VAS from de la Puente et al. (2007) as the scores are limited to the descriptions on scale and does not allow ratings in between.62 Nevertheless, in order to minimize inconsistencies in interpretation, every dog was assessed by the same trained observer at time points before (T0) and 2 h after (T2) ondansetron administration. Both observers at the RVC and the TiHo found a decrease in nausea-like-behaviour after ondansetron administration. The NRS with the given gradations helped to achieve a higher level of objectivity, comparability and reproducibility in this case series.
Some limitations of the present study should be noted. No blinding of observers was performed and the analyses took place at two different test centres. The observed patients in this cases series were part of a diseased population of elderly dogs (median age of onset 115 months) with a variety of different breeds and potentially long medical history with interfering pre-existing illnesses. These dogs were exposed to a stressful environment (hospital), where they were handled by unfamiliar people. Stress will certainly have been reflected in the dogs’ behaviour and thus may have influenced the behavioural assessment. Efforts were made in the study design to try to reduce the effect of stress by giving dogs a period of 1 h of acclimatisation. In addition to the small sample size, the absence of an untreated control group must be taken into account. As the intensity of the signs of nausea may decrease over time due to the patient's habituation to the state of vestibular disorder, resulting in clinical improvement. As nausea varies greatly between individuals, only a placebo-controlled cross-over study can correctly compare the state of nausea in each individual. Individual adaptation to vestibular disorders is difficult to assess in clinical trials; only observation extending beyond the actual study period and effective plasma concentration can lead to preliminary evidence. However, the paired comparisons conducted between individuals showed an overall significant decrease of nausea scores in the participating dogs with every dog experiencing an improvement in the scores after ondansetron treatment.