Functional magnetic resonance and swallowing: critical literature review

Introduction Aspects of the neuroanatomical representation of swallowing have been investigated in humans through brain mapping techniques, such as functional magnetic resonance imaging (fMRI). Objective This critical qualitative review of the literature analyzed international scientific publications in the PubMed database that investigated the activation of the central nervous system in humans during the act of swallowing. Methods This investigation was limited to articles that investigated adults older than 18 years, published in English or Portuguese, between January 2002 and December 2013. Publications that did not have access to the full text, that were repeated by overlapping keywords, case studies, literature reviews, letters to the editor, and those not directly related to the topic of the investigation were excluded. Results A total of 649 articles were identified, of which 21 matched the inclusion criteria. Conclusion The main purpose of the manuscripts that investigate the swallowing process through fMRI were: to characterize swallowing in different pathologies; to compare swallowing in different age groups; to describe brain activation in different stimulation conditions. These studies indicate multiple cortical regions involved in swallowing control. Overall, the studies indicate that fMRI is a non-invasive and quantitative method that allows the investigation of characteristics that are quite often not clinically visible.


Introduction
Swallowing is a complex sensory-motor process that involves several physiological stages. 1 The hypothesis has been proven that multiple brain areas are activated during swallowing, both in children and adults, reflecting regions that are responsible for different aspects of the swallowing process. 2 It is known that the cerebral cortex plays an important functional role in the regulation of swallowing. 3 The reflexive components of swallowing depend on the brainstem swallowing centers; the onset of swallowing is a voluntary action, which depends on the integrity of the motor areas of the cortex. 4 Some aspects of the neuroanatomical representation of cortical function that controls swallowing have been investigated and identified in humans, using brain imaging techniques such as the functional magnetic resonance imaging (fMRI), considered one of the most recent and advanced methods of functional neuroimaging without the use of ionizing radiation. 5 The advent of functional magnetic resonance imaging has facilitated or a better detection and quantification of organizational changes in cortical activation, with improved spatial and temporal resolution. 5 fMRI is a safe and non-invasive method to investigate the human brain, and has been indicated in the investigation of dysphagia after brain damage. 6 The aim of this study was to perform a critical analytic summary of relevant articles on the fMRI findings during swallowing in different groups studied in the international literature.

Methods
Since this was a non-experimental study, there was no need for consent forms or institutional ethics committee approval. The Cochrane Handbook precepts were followed to establish the research method. 7 The articles used in this study were selected through the PubMed database using the keywords: ''deglutition'', ''deglutition disorders'', ''magnetic resonance'', and ''magnetic resonance spectroscopy'', limited to studies performed in adult individuals, published in English and Portuguese, carried out from January 2002 to December 2013.
The search for publications in the database was independently conducted by the researchers to minimize possible citation losses. Each article recovered from the database was independently analyzed by the researchers to determine its relevance for selection and inclusion in the study. Articles in languages other than Portuguese and English were excluded, as well as publications that did not allow access to the full text (obtained from the CAPES Journal Portal) and those repeated due to overlapping keywords. Of the full texts obtained, those related to case studies, literature reviews, studies with animals, letters to the editor, and those that were not directly related to the topic of investigation were excluded. Texts that were effectively related to the research topic were analyzed. All stages of the study were independently conducted by the researchers. In case of disagreement among the researchers, only texts with a consensual final opinion were included. Due to its nature, this was not a single-blind study. After crossing ''deglutition × magnetic resonance'', ''deglutition × magnetic resonance spectroscopy'', ''deglutition disorders × magnetic resonance'', and ''deglutition disorders × magnetic resonance spectroscopy'', a total of 649 articles were found, of which 151 had unavailable summaries. Of the remaining 498, 189 were repeated. Thus, 309 articles were evaluated and of these, only 21 were included in this study.
The reason for the exclusion of 288 articles from the study is shown in Table 1. The 21 selected articles were critically evaluated regarding the following: objectives; number and gender of participants; age range; criteria and evaluation methods; results; and conclusions.

Results
The study results are briefly described in Table 2.

Discussion
Thirteen different groups of authors were identified. There was no prevalence or differentiation of results between genders. All items had quantitative analysis, supported by statistical results. Of the 21 selected articles, only one 23 included the participation of raters, to give reliability to the videofluoroscopic evaluation.
Of the assessed articles, nine 2,10,15,17---24 used control groups to make comparisons. In 20 articles, 2,8---27 the study design was cross-sectional, with evaluation/performance of a single examination, and in only one 21 the design was longitudinal, including the evaluation/performance of examination at two different moments, i.e., pre-and postoperatively. According to the methodological design of this study, all articles used the fMRI exam, but other exams were also used in association, such as electromyography (EMG) and videofluoroscopic swallowing study (VFSS).
The parameters considered for evaluation during the fMRI examination were not consistent. Some studies evaluated the role of swallowing saliva, water, and barium 15,16,23,24 ; others, the swallowing of saliva and/or water 2,9,10,13,17---21 ; whereas another assessed the swallowing of solids, liquids, and semisolids to differentiate consistencies. 12 Regarding the instructions to perform the tasks, verbal commands were used, such as ''swallow'' and ''do not swallow'', 8,11 as well as the visual and/or auditory and/or gustatory stimulation. 14,22,25---27 In general, to analyze the results of the fMRI examination, the blood oxygenation level dependent (BOLD), regions of interest (ROI), cortical lateralization, and diffusion tensor imaging (DTI) criteria were used, as well as time latency.
After a critical review of the articles, it was observed that studies that presented the findings of fMRI during swallowing tasks can be divided into three groups: studies describing brain activation in healthy individuals in different conditions and with different stimuli 8,11---14,16,19,22,26,27 ; studies of patients with diseases/comorbidities, who were compared to healthy individuals, 10,17,18,21,23---25 with the following comorbidities: Alzheimer's disease (AD), 23,24 obesity, 25 tongue cancer, 10,21 amyotrophic lateral sclerosis (ALS), 17 and ischemic cerebrovascular accident (CVA) 18 ; and studies comparing brain activation at different ages. 2,15,20,24 For better presentation and discussion of the study findings, thematic agglutination was performed according to the abovementioned groups. In cases of AD, in brief, significantly lower response (BOLD) was observed in several cortical areas that are traditionally involved in normal swallowing. One of the studies 23 concluded that, while the swallowing disorder is generally observed in the final stages of AD, changes in cortical control of swallowing can start long before the dysphagia becomes apparent.
Studies of patients with tongue cancer 10,21 sought to determine the adaptation mechanisms of the nervous system, after the glossectomy. The fMRI findings of glossectomized patients showed greater activation in the parietal cortex and adaptive responses (in cortex areas associated with tongue movement planning during swallowing) of the CNS after glossectomy.
The study that explored the structural and functional changes in patients with amyotrophic lateral sclerosis, with or without dysphagia 17 found that, during voluntary saliva swallowing, all individuals had the same activated cortical area, but the dysphagic patients showed a decrease in activation. It was also demonstrated that, even in cases in which the disease cannot be detected by conventional MRI, fMRI may disclose the alterations in brain function during swallowing, showing possible swallowing disorders that may develop in the future.
The findings of articles 8,9,11---14,16,19,22,26,27 that only studied brain activation in healthy patients in the presence of different stimuli ---consistency, flavor, different instructions Participants had to swallow, in the supine position, 5 mL of water at 15 • C, a capsule, and agar (solid) without aid of water and without chewing.
The size of the activated area during the swallowing of water was higher than for agar. The task of swallowing a capsule may require more from oral swallowing and motor coordination than that for solid or liquid food. Cortical representations of swallowing vary by type of food and can explain the responses of variables in the performance of swallowing in patients with oropharyngeal dysphagia. Participants had to swallow the saliva accumulated in the mouth or 3 mL of water injected into the oral cavity, in accordance with the visual command.
There was a four-fold increase in the volume of cerebral activation when swallowing water compared to saliva swallowing, particularly the right premotor cortex and the prefrontal cortex. This specific activation pattern may represent a compensatory response to the demands of swallowing water considering the age-related decrease in oral sensorimotor function. Lowell  Participants had to swallow saliva, water, and barium, alternating with pauses in swallowing, according to the commands.
The findings suggest that the elderly individuals showed a higher cortical involvement to complete the same swallowing tasks when compared to the younger ones. The elderly showed a higher latency to the onset of pharyngeal swallowing and increased residue of the ingested material in the pharynx. These findings suggest that the elderly had a higher cortical involvement to complete the same swallowing tasks than younger individuals. Kaway et al., 2009 16 To investigate brain activation after simultaneous display of stimuli associated with swallowing movements.
12 healthy and right-handed adults, with vision and hearing within the normal range (20---28 years).
Auditory and visual stimuli of the act of swallowing and neck anatomy were presented in association during swallowing. The cortical areas activated with different types of stimuli were analyzed.
According to this study, the audiovisual stimuli associated with the swallowing movement can be applied in the treatment of patients with dysphagia. The activation of areas associated with the swallowing movement (planning and performance) occurred during all stimuli. Water was injected into the oral cavity of the participants, who should follow random visual commands (''swallow'', ''get ready to swallow'', ''touch your tongue'', ''clear your throat'').
Areas activated during each of the task components showed partial differentiation of neural location for the several components of swallowing. The study was able to identify the brain areas involved in the swallowing task, in agreement with previous findings. Barium and water were injected into the oral cavity of the patients and subjects should swallow when they feel the entire content in their mouth; as well as saliva when they saw the visual command.
The group with AD had significantly lower response in many cortical areas that are traditionally involved in normal swallowing. They did not recruit new regions, or compensate in regions that are normally activated during swallowing. Although swallowing disorders are usually observed in the final stages of AD, changes in the cortical control of swallowing can start long before dysphagia become apparent.
Humbert et al., 2011 24 To assess brain responses in swallowing tasks, comparing individuals by age and the presence of Alzheimer's disease (AD), through fMRI. Swallowing of saliva, water, and barium, intercalated with moments of pause in the swallowing task after predetermined commands.
Absence of significant differences in areas of brain activation when comparing elderly individuals, young individuals, and elderly individuals with AD. Individuals with AD required greater effort to perform the pause function during swallowing when the command for this action was requested. Water was injected into the oral cavity of the participants, which should follow the commands that appeared on a television at random (''swallow'', ''get ready to swallow'', ''touch with your tongue'', ''clear your throat'').
There was decreased activation in the elderly compared to young individuals during swallowing and the analyzed tasks. These reductions were significant in several primary somatosensory areas, indicating a decline in neural processing of sensory signals to coordinate the swallowing response. to perform the task of swallowing, different ages, and different central nervous system foci ---showed the following: (1) higher brain activation: during liquid swallowing compared to solids; during water swallowing compared to saliva swallowing; when the food offered had flavor, compared to saliva and water; when there was a verbal command to perform swallowing; when using olfactory, gustatory, and visual stimuli associated with swallowing; and when using sensory stimulation with air injection in the oral cavity during the swallowing (compared to the times in which there was no injection of air); (2) increased activation in areas of interest (programming and performance) when visual and auditory stimuli associated with swallowing were used in combination to carry out the function; (3) that the oral sensorimotor cortex in the left and right hemispheres are not functionally equivalent; and (4) that spontaneous swallowing involves the activation of the cerebellum and basal ganglia, as well as cortical structures. By analyzing the normal standards for brain activation during swallowing tasks, it was observed that although the contrast analysis failed to identify the specific activation foci for swallowing, overlapping activation maps suggest that the most lateral extent of the precentral and anterior parietal cortex, rostral anterior cingulate cortex, precuneus and left parietal operculum are preferentially activated in swallowing. 9 Suzuki et al. (2003), 8 however, found that regions activated during swallowing were observed in the sensorimotor cortex, insula, cerebellum, putamen, globus pallidus, thalamus, anterior cingulate gyrus, supplementary motor area, superior temporal gyrus, and in the substantia nigra; the cerebellum was bilaterally activated, especially on the left side; activation of the globus pallidus and putamen was bilateral.
In the past few years, a growing number of neuroimaging studies in healthy subjects have shown multiple cortical regions involved in the control of swallowing. Functional brain imaging may help to elucidate the relevant neural mechanisms, identifying the neural patterns that control this complex sensory-motor action, providing evidence of functional changes in the cerebral cortex after a comorbidity, such as those cited in this review, helping to identify and the correct rehabilitation of dysphagia.
This critical review showed that the examination is of great importance in the early identification of brain alterations, facilitating the choice of a better rehabilitation approach for dysphagic patients, or those at risk for dysphagia, allowing improvement of the clinical picture or risk prevention.

Conclusion
This study demonstrated that the fMRI is a non-invasive, quantitative method that provides specific answers, which are sometimes not clinically identified.
It can be used associated with and/or in addition to other imaging tests to confirm results or as a validation method. As a negative aspect, this study showed that for all examinations, all patients should swallow in the supine position, which may influence the adequate performance of the swallowing function. Another aspect that must be considered is that there was no standardization of the methodologies used in the assessed articles, both in case selection and in the study materials and methods.