Swallowing disorders occur in about 30%–50% of patients who have survived stroke. They are risk factors for aspiration pneumonia, malnutrition, and dehydration [1]. They also increase the length of hospital stay, interfere with rehabilitation, and increase the risk of death in severe cases. The risk of aspiration pneumonia in a stroke patient with dysphagia is three times higher than that in a stroke patient without dysphagia. In addition, the mortality rate is three times higher in stroke patients with pneumonia [2,3]. Therefore, early detection and treatment of swallowing disorders should be a priority in rehabilitation of stroke patients to reduce morbidity, malnutrition, and the risk of death. Before the first oral intake of fluids or food after a stroke, a dysphagia test is recommended regardless of stroke severity. Clinical guidelines based on systematic reviews should be used [1,2].

However, unlike dysarthria or aphasia that can be quickly diagnosed as early symptoms of stroke, symptoms of dysphagia are less prominent. Thus, evaluation and treatment of dysphagia may be delayed [4]. Therefore, magnetic resonance imaging (MRI)-based predictors of swallowing disorders are useful for early management. Several studies have attempted to identify lesion patterns that can predict swallowing disorders. However, these failed to yield meaningful results due to small sample sizes and roughly classified brain regions [5,6]. Recent, larger studies that overcame these limitations examined effects of lesion location on dysphagia incidence, pattern, and complications in acute stroke [7-9]. However, most studies were focused on the incidence of dysphagia. Reports on the association between lesion location and duration of dysphagia recovery are lacking [10]. Because the occurrence of dysphagia is associated with various swallowing mechanisms, reported parameters have limitations to identify primary brain lesion in dysphagia. We hypothesized that primary brain lesion location could be associated with severity of dysphagia or time to recovery. Thus, the aim of this study was to investigate the association between brain lesion location on MRI performed after infarction and the duration of dysphagia recovery in patients with middle cerebral artery (MCA) infarction.

Mean age, mRS score, NIHSS score, and recovery time for all patients were 65.84±10.1 years, 3.1±1.5, 7.7±6.5, and 26.71±16.39 days, respectively. Dysphagia recovery time was longer in patients with lesions of the PLIC or caudate nucleus (Table 2). The average recovery time was 36.65±15.83 days in patients with PLIC lesions and 32.6±17.27 days in patients with caudate nucleus lesions. The mean recovery time in patients without lesions of both PLIC and caudate nucleus, those with a lesion of the PLIC only, those with a lesion of the caudate nucleus only, and those with lesions of both the PLIC and caudate nucleus were 12.9±6.7, 24.7±14.0, 24.6±22.9, and 39.4±20.0 days, respectively (Table 3). One-way ANOVA showed statistically significant differences in mean recovery time between groups (p<0.001). Scheffe multiple comparison test showed significant differences between patients with and without lesions in PLIC and caudate (p<0.001) (Table 3).

Various studies have attempted to identify the location of the swallowing center. Recently, the right cerebral hemisphere has been reported to have a greater effect on swallowing function than the left cerebral hemisphere [7,12]. However, swallowing requires the interaction of both cerebral hemispheres. Left hemisphere lesions are associated with oral phase dysfunction while right hemisphere lesions are associated with pharyngeal phase impairment [13,14]. Studies on voluntary and reflexive swallowing centers have shown various findings. Kern et al. [15] have reported that swallowing centers are in the bilateral primary motor and sensory regions, insular cortex, and prefrontal lobe. Flowers et al. [16] have reported that pontine and medullary lesions are associated with dysphagia. Some studies have suggested that internal capsular lesions are associated with dysphagia [17,18].

These results suggest that the bilateral neural network from the medulla to cerebral hemispheres mediates swallowing. The internal capsule is a pathway for large thalamocortical, corticothalamic, corticopontine, corticobulbar, and corticospinal fibers. The PLIC lies between the lenticular nucleus and thalamus. It is the pathway for corticospinal and corticobulbar tracts originated in the motor cortex, ventral posterolateral nucleus, and ventral posteromedial. The corticobulbar tract travels from the ventral motor cortex to the midbrain pons and medulla oblongata. It forms synapses directly or indirectly with the nuclei of facial, hypoglossal, and trigeminal cranial nerves to affect swallowing [19]. Therefore, activation of these nerves may affect swallowing muscles [19]. A study on the treatment of dysphagia by exciting the corticobulbar tract has reported positive results [20]. Therefore, lesions in the PLIC may cause dysphagia after MCA infarction and affect the response to recovery.

The swallowing mechanism can be affected by many neurons and muscles. Dysphagia can occur even when there is damage to only one of these. In this respect, it would be difficult to identify a single brain center associated with dysphagia. Therefore, among many lesions that can cause dysphagia, we think that the primary lesion most involved in the swallowing mechanism will become clearer after evaluating recovery from dysphagia rather than the incidence. A primary lesion with greater impact on recovery from dysphagia will be associated with longer duration of recovery regardless of the presence of absence of other lesions. However, classification of MCA infarction by lesion site has statistical limitations because independence between groups is not ensured as most lesions show overlap. Therefore, we evaluated lesions with longer mean recovery time without any difference in mean age, mRS score, or NIHSS score for all patients. We found that patients with PLIC and caudate nucleus lesions had longer mean recovery time compared to all patients (Table 2). In addition, the mean NIHSS score was lower for these lesions compared to that for other lesions while there was no significant difference in mean age, mRS score, or NIHSS score for all patients (Table 2). We classified groups according to overlap of these two lesions so that independence between groups was ensured (Table 3). As a result, patients with MCA infarction involving both the PLIC and caudate nucleus had significantly longer recovery time than patients without both lesions. There was no significant difference in recovery time for patients with a lesion in only one of these sites (Table 3).

In diffusion tensor imaging analysis of patients with MCA infarction, apparent diffusion coefficient value of the bilateral PLIC can differ significantly according to the presence or absence of dysphagia [11]. The caudate nucleus is associated with spatial working memory and directed movement [21,22]. Goal-oriented executive functioning is also closely associated with cognitive functions such as memory [23]. In contrast with the PLIC, the relationship between swallowing and the caudate nucleus is unclear. While further research is needed, it should be noted that these two regions receive similar blood supply [24].

We used the third stage of the dysphagia diet because it was the primary nutritional source for patients with cerebral infarction. These patients require 25 kcal/kg to prevent malnutrition [25]. Therefore, to provide adequate calories, a thin rice gruel diet is insufficient. The third stage diet contains 8 g of porridge and 10 g of thickener for a total of 1,765 kcal. This diet provides sufficient nutrition for rehabilitation therapy in patients with cerebral infarction. Therefore, a third stage diet can be the primary nutritional goal. In addition, unlike the thin rice gruel diet, the third stage diet requires sufficient recovery of oral and laryngeal phases. Therefore, the ability to tolerate the third stage diet is a sign that overall swallowing function has recovered.

Establishing a treatment plan for patients with dysphagia is an important part of rehabilitation therapy for stroke. Goal diet, determination of the proper diet for swallowing status, treatment method, duration of treatment, and more information can be determined with VFSS. However, VFSS may be delayed or omitted. Therefore, in the initial evaluation of patients with stroke, obtaining information related to dysphagia would help plan rehabilitation. In this study, the duration of dysphagia recovery increased with infarction of PLIC and caudate nucleus, providing important clues to the duration of recovery and the presence of dysphagia.

In patients with MCA infarction, the duration of dysphagia recovery was longer in patients with lesions of both the PLIC and caudate nucleus than that in patients without these lesions. Therefore, longer dysphagia treatment may be required for a patient with lesions of both the PLIC and caudate nucleus after an MCA infarction.