The present resting-state study was designed to assess, for the first time, the association between the functional integrity of the auditory network in interictal MwoA and neuropsychiatric comorbidities. Importantly, our findings demonstrated increased spontaneous activation within the left auditory cortex, which was independent of structural and microstructural abnormalities and clinical characteristics.
Recent studies have also, however, reported microstructural changes and cortical abnormalities in the temporal lobe in patients with MwoA [20–22]. Specifically,the surface-based morphometry revealed decreased mean cortical thickness of the insula and illustrated a significant correlation with disease duration [21, 23]. However, a voxel-based morphometry study demonstrated opposite volumetric changes in the insula in patients with MwoA during the ictal period [24]. Moreover, morphometric studies have also revealed an increased gyrification index [22] and cortical thickness [25] in the left PoCG using a general linear model approach. Although the structural changes indicated associations with sensory discrimination of pain, information processing and multisensory integration, there has been no consensus on a coupling between structural alterations and functional regulation in the pathological mechanism of migraine. Nonetheless, auditory network functional disruptions were detected in the absence of remarkable GM changes (regional or global) in the present study, possibly indicating that functional changes may precede or induce structural or microstructural abnormalities. An alternative possibility was that the morphological post-processing methodology might have insufficient sensitivity in detecting microstructural changes.
The auditory cortex is a network particularly essential for individual stressful experiences, cognitive processes and adaptive behavior, that allow a particular individual to respond to the environment in a predictive manner [4, 26, 27], which is related to the STG region. The left STG region showed significantly increased connectivity in our study. Previous neuroimaging studies have repeatedly provided evidence of disrupted activation in the STG in migraine patients by using different methodological approaches [28, 29]. However, Schwedt et al [30] illustrated the opposite pattern of neural activation in the STG in episodic migraineurs compared with HC. Thus, the variance in results may, however, be due to the lack of a standard methodological approach, discordant sample size, and the clinical heterogeneity of participants. The increased activation in the STG may reflect damage to modulatory systems related to inhibition of nociceptive drive from trigeminovascular afferents [31]. The sense of pain is linked to the lower threshold for tolerating unpleasant signals. The STG may play a supplementary role in enhancing the sensitivity to pain response or enhancing the perception of pain information from recurrent painful stimuli. In addition, activation in the STG region showed a significant correlation with anxiety. Importantly, similar to our results, previous studies have demonstrated abnormal functional connectivity between limbic areas and the STG in migraineurs [32, 33], indicating an imbalance in the limbic-auditory pathway. Hence, our findings may suggest a possible substitute physiological correlation of resting-state connectivity changes. Another probable explanation could be that the maladaptive brain response due to repeated stress exposure could underlie or be related to the observed auditory dysfunction.
The insula cortex, as a core region in the salience network, is reciprocally connected to multiple brain networks and a wide variety of functions from sensory and affective processing to high-level cognition [34, 35]. Previous fMRI neuroimaging studies, during and between migraine attacks, have provided evidence of neural mechanisms in the insula cortex and considered to be a region involved in non-specific pain modulation [36]. In migraineurs, changes in the intrinsic connectivity between the insula cortex and other networks may set a platform for abnormally intensified responses to sensory stimuli, which are, however, well-tolerated by healthy subjects. In trigeminovascular models, the insula cortex directly or indirectly receives affective and nociceptive inputs, known as pain trigger, from the brainstem, and delivers them to and from the sensory cortex reciprocally [31]. Moreover, calcitonin gene-related peptide (CGRP) is a trigger to migraine in migraineurs, but not in normal persons [37], and CGRP antagonists have emerged as new effective drugs currently in development for migraine [38]. Interestingly, some structures in the central nervous system express CGRP receptors that have been proposed as a “visceral network” in the brain [39, 40], including the insula cortex and limbic system. In addition, nociceptive hypersensitivity might be maintained and induce by the insula-limbic pathway, even in the absence of controlled peripheral noxious stimulation [41]. Also, concerning the physiology of migraine, the subcortical network in question is seldom explored. It may perhaps, however, provide an anatomical framework that underlies hypersensitive responses to spontaneous headache without exogenous triggers. Taken together, it proves that the insula and other brain networks share (i) essential reciprocal topographic connectivity; (ii) form anatomical information pathways, and finally (iii) receive and project into multiple modulation networks.
In the present study, we showed that, between migraine attacks, migraineurs without aura have greater activation in the left PoCG than HC. The PoCG (primary sensory cortex) is a crucial region in the somatosensory network (SMN) and is activated by sensory stimuli, including pain. Painful stimuli activate “pain matrix” regions within the somatosensory cortex, limbic system and subcortical areas [42, 43]. These regions are responsible for descending and ascending modulations, cognitive and motivational aspects, and pain processing [30, 44]. In the absence of external stimuli, resting-state and position emission tomography (PET) studies have demonstrated abnormal connectivity or metabolism in many brain networks in migraineurs compared with controls [5, 45]. These regions of the brain are similar to the hyperactive regions observed during pain-evoked stimulation in migraineurs. Therefore, the disorientations in the left PoCG may reflect overactivity and hypersensitivity during an evoked or spontaneous stimulation in the pain processing of migraineurs. Moreover, the strengthened activation in the left PoCG was linked to anxiety and depression. Additionally, cortico-limbic interactions mediate maladaptive and adaptive responses, which are essential for emotion, motivation, and memory via cognitive control mechanisms [46]. The limbic regions may channel inputs into or receive outputs from the primary sensory cortex, regulating switching between the limbic system and sensory cortex to optimize responses to pain. Hence, the impaired limbic-PoCG pathway in migraine could be a physiological process to impair emotion modulation and enhance pain perception. Notwithstanding, more studies are needed to unravel the direct association between the primary sensory cortex and the limbic regions in migraine patients.
Previous studies have suggested that global functional deficits in multisensory information processing is a characteristic feature of migraineurs during the interictal period [47]. In our study, we applied a component analysis to examine the activation and modulation of the auditory cortex in information transfer in MwoA. Hyperactivation within the core regions of the auditory network and salience network as a marker of attentional shift towards pain, provides direct evidence of disrupted functional segregation and integration in brain networks of migraine patients without aura. Current findings suggest that changes in the limbic and auditory regions of migraineurs are pertinent, as apparent changes occur in these areas [26].
We acknowledge the fact that our study had some limitations. Firstly, the ICA methodological approach used enabled us to evaluate activations within the auditory cortex functionally, but it did not provide substantial information concerning casualties and interactions of the entire brain. Secondly, we only focused on the functional auditory network changes in migraineurs without aura, but did not examine the structural connectivity abnormalities of the brain. Thirdly, the relationship between auditory network functional changes and psychiatric symptoms could be considered a working hypothesis that emerged from our work, and future studies are needed to further elucidate these potential correlations. Finally, our study was restricted to migraineurs without aura during the interictal period.