The Bárány Society and the International Headache Society jointly proposed diagnostic consensus criteria for vestibular migraine (VM) included in the International Classification of Headache Disorders (ICHD-3) (Table 1) (1). VM is the most common cause of episodic vertigo in both adults and children, with a lifetime prevalence of 1% and a 1-year prevalence of 0.9% in the general population with a female preponderance (2). The underlying pathophysiology of VM is still unclear; however, possible mechanisms involve hyperactivity within the trigeminovascular system (TVS) and nociceptive brain stem centers, as well as an abnormal sensory modulation or integration in both pain and vestibular processing areas (3). Due to the absence of large, systematic, controlled clinical studies that specifically address VM, most therapeutic recommendations for VM are extrapolated from clinical trials on migraine with and without aura.

Botulinum toxin type A (BTX-A) has been used to treat different types of headache by blocking peripheral signals to the central nervous system, which inhibits central sensitization. The clinical efficacy of BTX-A in treating chronic migraine was previously demonstrated in the PREEMPT (Phase III Research Evaluating Migraine Prophylaxis Therapy) randomized clinical trial (4). BTX-A has also been used for other types of headaches, such as episodic migraine (<15 headache days per month) (5), chronic daily headache (>15 headache days per month) (6), and chronic tension-type headache, in a systematic series of exploratory controlled trials (7). In patients with frequent migraine headaches or vertigo attacks (15 days per month for more than 3 months), a trial of BTX-A might be reasonable; however, to date, there are no data on the management of chronic migraine headache or vertigo attacks in patients with VM.

Resting-state functional connectivity analysis in magnetic resonance imaging (rs-FC MRI) can visualize and measure the brain's intrinsic functional architecture, which aids in understanding the underlying mechanisms of various neurological disorders (8). The amplitude of low-frequency fluctuation (ALFF) changes in rs-FC MRI data is thought to be associated with local neuronal activity, and an ALFF analysis is effective for detecting fluctuations in spontaneous low-frequency oscillations.

The aim of this study was to evaluate the efficacy of BTX-A injection for the prophylactic treatment of VM. We hypothesized that a symptomatic improvement after BTX-A treatment would be reflected in changes in brain connectivity in areas that process and store vestibular and pain information. Therefore, rs-FC MRI data acquired pre- and post-BTX-A treatment were analyzed by comparing ALFF and seed-based rs-FC to detect changes in spontaneous brain activity in patients with VM.

In this study, 155U BTX-A injections significantly reduced both vertigo attacks and migraine headache compared with baseline in our 20 patients with VM. The patients also reported significant improvements in their functionality and the tolerability of migraine headache and vestibular symptoms after BTX-A treatment.

Although various mechanisms have been put forward to explain VM symptoms, interactions between nociceptive and vestibular pathways are broadly considered to represent the underlying pathophysiology of VM. The reciprocal connections between the vestibular nuclei and other brainstem structures, such as the raphe nuclei, the parabrachial nucleus, and the locus coeruleus, could modulate the sensitivity of the trigeminovascular system (TVS) and pain pathways (18), which could explain vestibular symptoms in VM. Altered neural activity within the TVS is considered to be the primary mechanism for migraine headache (19). Thus, TVS neuropeptides such as substance P and calcitonin gene-related peptide (CGRP), which cause vasodilation and neurogenic inflammation, might lead to throbbing pain and central sensitization (20). The TVS also innervates the inner ear, and some neurotransmitters, such as CGRP and serotonin, are expressed in the vestibular and cochlear systems and might also be involved in VM pathophysiology (21). Animal studies have shown increased CGRP expression in a rotational model of motion sickness (22), and CGRP null mice have a reduced vestibulo-ocular reflex (23), which gives more evidence that CGRP is expressed throughout various vestibular nuclei (24). These findings highlight the potential role of elevated CGRP signaling as an underlying mechanism in the association between migraine and vertigo (25). Parallel activation of the vestibular and nociceptive pathways, such as the periaqueductal gray and hypothalamic areas, has been proposed, and reciprocal connections between these systems could modulate neural activity within the TVS and vestibular system (26, 27). Furthermore, the response of both headache and vertigo symptoms to migraine prophylactic medications suggests that migraine headache and vertigo could share a common pathophysiology in the VM population (26).

Although the exact mechanism of action of BTX-A in VM is still a matter of speculation, animal studies suggest that BTX-A can block the release of nociceptive mediators such as substance P, glutamate, and CGRP, which reduces nociceptive input into the central nervous system from the periphery (28). Correspondingly, it has been shown that CGRP levels in chronic migraine patients strongly correlate with their response to BTX-A therapy (29), and CGRP levels after BTX-A treatment were significantly lower than before BTX-A treatment (median, 51.89 vs. 74.09 pg/ml, p<0.001) in patients with chronic migraine (30). Therefore, the activation of the TVS and vestibular system and the release of pain-producing molecules, especially CGRP, could be a potential therapeutic intervention during chronic migraine and VM attacks. Injections of BTX-A at the designated pericranial sites inhibit the release of CGRP, which would otherwise facilitate nociceptive transmission, contribute to the development of peripheral and central sensitization and possibly also contribute to pain processing and vestibular system modulation (20). Therefore, the primary mechanism of action for BTX-A in VM might be reversing peripheral and central sensitization by inhibiting the release of CGRP and (to a lesser degree) other pain-producing molecules from TVS neurons (31). In addition, it was suggested that BTX-A-mediated blockage of excitatory synapses might indirectly enhance inhibitory neurotransmission through an as yet unknown mechanism. Recently, it was reported that such a mechanism might be connected with the opioid and GABA-ergic systems in the central nervous system (32). Therefore, we speculate that the primary mechanism of BTX-A in patients with VM might involve the inhibition of TVS neuropeptide release such as CGRP and the enhancement of endogenous opioidergic and GABA-ergic transmissions, which can modulate neural activity within both the TVS and the vestibular system. Supportively, BTX-A has been shown to be less effective in chronic tension-type headache, a condition in which CGRP levels have been shown to be normal (33).

Cortical areas involved in processing vestibular information include the insular and retroinsular cortex, the parietal operculum, the STG, the temporo-parietal junction and posterior parietal cortex, the postcentral gyrus, the cingulum, the hippocampus, and the frontal cortices (34). These regions are multisensory areas in which the inputs of several sensory modalities, i.e., auditory, somatosensory, and visual, converge with vestibular information. Similarly, pain is a complex, multifactorial subjective experience, and a large distributed brain network is involved in nociceptive processing (35). Nociceptive-specific cortical and subcortical brain regions that are commonly activated by pain stimulation include the anterior cingulate cortex, insula cortex, frontal and pre-frontal cortices, primary and secondary somatosensory cortices, thalamus, basal ganglia, cerebellum, amygdala, hippocampus, and regions within the parietal and temporal cortices (36). The somatosensory and nociceptive cortical projections strongly overlap with the vestibular cortical projections, giving rise to vestibular-somatosensory and vestibular-nociceptive interactions, whose occurrence has been demonstrated by caloric and galvanic vestibular stimulation in humans (37). Therefore, it has been suggested that vestibular input could influence the processing of other tactile and nociception inputs.

In this study, alterations in spontaneous brain activity in VM patients were evaluated before and after BTX-A treatment using two different sequential data-driven analyses, namely, ALFF and a seed-based functional connectivity (FC) analysis. Increased spontaneous activity was indicated by significantly higher ALFF values in the right temporal lobe (STG) after BTX-A treatment compared with the baseline. FC between the right STG and left PSG, right SMG, and right MTG was also increased after treatment. The enhanced regional functional activity after BTX-A treatment, mainly in the right temporal lobe, which contains the vestibular cortical core regions and participating pain processing regions, suggests that BTX-A injection can modulate the functional connectivity of the vestibular and pain cortices. The STG is an important area within the vestibular network and is closely connected to other multisensory brain regions. The posterior STG is consistently described as a core region for central vestibular processing, along with the posterior insula and parietal operculum, and as a secondary processing region for coordination of the eyes and head and spatial orientation. Structural and functional abnormalities of the STG have previously been demonstrated in both migraine patients (38) and patients with vestibular disorders (39). In chronic migraine, hypometabolism was reported in the bilateral temporal regions compared with healthy controls (38). On the contrary, hypermetabolism in the temporal lobe regions, including the temporal pole and STG, has been found in response to painful stimuli (40). In VM patients, an increased metabolism in the temporal areas was reported during a migraine attack (41). During the attack-free interval of VM patients, voxel-based morphometry in MRI showed a decrease in GM volume bilaterally in the inferior and middle temporal gyri, the cingulate cortex, and the posterior insula, as well as in the STG, supramarginal gyrus, dorsolateral prefrontal cortex, and inferior occipital gyrus (39). In this study, alterations in the ALFF values were observed only at the right side STG after treatment, which reflects the right-hemispheric dominance of the vestibular cortical network in right-handers in our VM patients (42).

The FC between the right STG and the right SMG increased after BTX-A treatment, and interestingly, the correlation analysis showed a negative correlation with the headache-related disability in the MIDAS score. The SMG and angular gyrus within the inferior parietal lobule (IPL) have been shown to be activated by vestibular stimulation (43). The IPL is also part of somatosensory processing, functions in the pain response and the sensing of temperature and pressure. In patients with migraine with or without aura, structural abnormalities of the SMG have been demonstrated (44). VM patients also have decreased GM volume in the SMG, the prefrontal cortex, and the posterior insula–operculum regions compared with healthy controls. Increased FC between the right STG and SMG and the negative correlation with the MIDAS score indicate that migrainous headaches improved after BTX-A treatment over time in parallel with the increased FC between the STG and SMG. We also found increased FC between the right STG and the PCG and MTG. These brain regions belong to the multisensory vestibular network, and the areas are activated by exposure to magnetic vestibular stimulation (45). The MTG in particular is considered to be functionally relevant to vestibular compensatory mechanisms and is strongly interconnected with other multisensory cortical areas that form a multisensory-integrative network (39). The increase in FC between the STG and regions of the MTG and PCG thus allows for two interpretations: first, it contributes to vestibular compensation during the recovery phases after BTX-A treatment, and second, the inclusion of somatosensory processing might reflect the integration between the vestibular and nociceptive systems.

This study has several limitations. First, the number of VM patients is relatively small in this longitudinal approach. Second, we choose a controlled before-after study which is a non-randomized design in that outcomes are measured before and after BTX-A treatment, both in a group that receives the treatment and in another comparison group. Typically this design could have a high risk of bias because of other differences between the groups that are being compared (confounders), and the placebo effect cannot be excluded. Third, it is difficult to assert that the change in the resting-state brain connectivity is a VM-specific finding because the comparative analysis was not conducted with a group of patients with migraine other than VM. Fourth, even though we did not change the use of preventive medications, we cannot exclude the possibility that the use of preventive therapies might have influenced the resting state of the brain and led to functional changes. Lastly, the follow-up of the patients for 2 months after the treatment might be too short; longer longitudinal studies are needed to evaluate the long-term effects of BTX-A in patients with VM.

In conclusion, our novel findings that BTX-A is effective in improving vertigo and migraine attacks in patients with VM and improved functional brain activity and connectivity in the vestibular and somatosensory (pain) processing regions after BTX-A treatment are intriguing and promising to expand our understanding of the pathophysiology of VM. With respect to the therapeutic mechanisms, botulinum toxin might enhance intrinsic vestibular, nociceptive, and cross-modal interactions within the multisensory vestibular cortical network in patients with VM. The findings suggest a disease-related dysfunction of multisensory (nociceptive, somatosensory, vestibular) cortical networks in VM patients.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

This study was approved by the Ethics Committee of Jeonbuk National University Hospital (IRB File No. 2020-10-033-001). The patients/participants provided their written informed consent to participate in this study.

S-YO was involved in the study concept and design and drafting a significant portion of the manuscript. J-JK was involved in the acquisition and analysis of data and drafting a significant portion of the manuscript and figures. SK and J-ML were involved in the acquisition and analysis of data. J-SK and MD were involved in the analysis of data and drafting a significant portion of the manuscript.

This study was supported by the Neurological Disorder Research Program of the National Research Foundation of Korea (NRF) grant funded by the Korean government (Ministry of Science and ICT) (Nos. 2022R1A2B5B01001933 and 2020M3E5D9080788) and by the Fund of the Biomedical Research Institute, Jeonbuk National University Hospital.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.