Dysphagia in cerebral hypoxia

1. Introduction

The process of swallowing involves the complex sequence of neuromuscular events of transporting a bolus from mouth to stomach while ensuring protection of the airway. Neural control of the swallowing act is commonly divided into three elements – an afferent system (involving cranial nerves V, IX and X) providing swallow sensory feedback; the brainstem swallowing centers coordinating output of motor nuclei of cranial nerves V, IX, X and XII; and various cortical regions, which initiate and modulate volitional swallowing [1–5].

Physiologically, there are three distinct phases of swallowing: oral, pharyngeal, and esophageal; these phases have been proven to be independent of each other [6 6]. In healthy individuals, after initiation of swallowing, the three phases occur sequentially and always in the same sequence [6]. The oral phase of swallowing is initiated voluntarily, the pharyngeal and esophageal phases occur secondary to stimulation of the pharynx and esophagus after a certain threshold volume or mass of material accumulates [6].

Neurophysiological assessment, positron emission tomography and magnetic resonance imaging studies in young and older adults indicate that the initiation of swallowing as a voluntary action depends on activation of multiple cortical regions, most prominently the lateral pericentral, perisylvian, anterior cingulate cortex and the adjacent supplementary motor area and the left cerebellum [1, 5, 7–10]. These regions have been implicated in voluntary swallowing of a water bolus, voluntary swallowing of saliva and “autonomic” swallowing of saliva [3]. Unilateral cerebral lesions indicate that swallowing centers have an asymmetrical distribution and that activation of the postcentral gyrus and anterior parietal cortex appears to be lateralized to the left hemisphere, independent of handedness [1–3, 7].

The reflexive component of swallowing is controlled by central pattern-generating circuitry of the brainstem [6]. The central pattern generators for the oral phase are located in the reticular formation and trigeminal nucleus, whereas the central pattern generators for the pharyngeal and esophageal phases are located in the nucleus ambiguus and nucleus tractus solitarius (mainly the dorsal motor and ventromedial nucleus) [6].

Dysphagia is a frequent problem in various neurological disorders and has been investigated most extensively in patients with stroke and multiple sclerosis, where swallowing impairment is present in up to 78% and 40% of patients, respectively [2, 11–14].

Clinical experience suggests that nearly all patients with cerebral hypoxia develop severe dysphagia, irrespective of age. Hypoxia is known to affect mainly cerebral structures especially the basal ganglia (pallidum, subthalamic nucleus), hippocampus, parieto-occipital cortex, and cerebellar cortex [15]. Brainstem and spinal cord are much more resistant to hypoxia and therefore less likely to be affected. Hence, it may be expected that cerebral hypoxia would be associated with dysphagia, especially affecting the voluntary oral phase of the swallowing act.

In 2008, Seidl et al. published data of 101 patients with neurological disorders presenting with dysphagia, 17 of whom had cerebral hypoxia, yet data of this specific subgroup is not presented individually [16]. To our knowledge, swallowing function has not yet been studied specifically in young patients with cerebral hypoxia. The objective of this study is to report the development of swallowing function in a series of adolescent and young-adult patients with cerebral hypoxia.

2. Methods

Eight consecutive patients of similar age, 1 male, 7 females, who were admitted to our institution after the acute phase following cerebral hypoxia were included in this study. Mean age±standard deviation (range) was 19.9±3.6 (16–25) years. Detailed patient demographics and clinical features are summarized in Table 1.

Abbreviations: F, female; M, male; DIC, disseminated intravascular coagulation; GCS, Glasgow Coma Scale; min, minutes; MRI, magnetic resonance imaging; yrs, years.

Each patient underwent clinical neurological evaluation, MRI, a standardized neurophysiological assessment, and repeated clinical and fiber-endoscopic evaluation of swallowing. Remission stages of apallic syndrome were classified according to Gerstenbrand Remission Scale [17].

MRI studies were performed using 1.5 T scanner with diffusion weighted images, intracranial time-of-flight angiography, T1 (with and without fat saturation) and T2 sequences. In certain conditions MR angiography and proton MR spectroscopy was performed.

Standardized neurophysiological assessment included evaluation of somatosensory, motor, and brainstem auditory evoked potentials according to clinical neurophysiological standards as well as auditory startle responses [18, 19].

Clinical evaluation of swallowing was performed by experienced speech therapists and included recording of feeding modalities (oral intake, nasogastric tube, percutaneous gastrostomy), respiration, cooperation level, visual and palpatory evaluation of laryngeal elevation, direct observation of morphology and motility of the lips and tongue, analysis of orofacial movement and sensibility, and observation of trial feeding with different boluses. Standardized and consistent amounts of trial boluses were used, specifically 2 ml for pureed bolus and 5 ml for liquid bolus.

Fiber-optic endoscopic evaluation of swallowing was performed by an experienced rater according to the procedure by Langmore et al. which has been considered the “gold standard” for detection of swallowing impairment [20–22]. Food and liquid boluses were given to the patients to determine the integrity of the pharyngeal swallow based on the patient’s ability to protect their airway, sustain airway protection for a period of several seconds, and initiate a prompt swallow without spillage of material into the vallecula and pyriform sinuses. Timing and direction of movement of the bolus through the hypopharynx and the patient’s ability to clear the bolus during the swallow was recorded. Special attention was paid to presence of pooling and residue of material in the laryngopharynx. Sensitivity of the pharyngeal and laryngeal structures was carefully tested by touching laryngeal structures, i.e. epiglottis, with the tip of the endoscope physiologically resulting in muscular contractions (glottic occlusion) and coughing. If present, the patient’s response to residue, penetration, and aspiration of the bolus was noted. All observations were made over several swallows in order to assess a physiological eating environment.

Data were analyzed using the statistical packages SPSS version 18.0 and STATA version 10.1. Metric variables were presented as means±standard deviation and categorical variables as numbers and percentages.

Content and structure of this study were aligned according to STROBE recommendations to improve the quality of reporting of observational studies [23].

Our case series consists of eight consecutive patients with severe cerebral hypoxia. Etiology of cerebral hypoxia was similar in five patients who were victims of a mass panic during a sports competition with cardiac and/or respiratory arrest and additional severe polytrauma, but without significant direct trauma to swallowing-related cranial nerves or muscles. Two other patients suffered fulminant pulmonary embolism, and one patient was hypoxic as a result of attempted suicide. Initial Glasgow Coma Score was 3 in all patients. Duration of resuscitation was unknown in five patients and ranged from 10 to 45 minutes in the others (see Table 1).

All patients were initially treated on an intensive care unit and were subsequently transferred to our institution in the post-acute phase for further neurological rehabilitation. Time between the cerebral hypoxia and admission to our institution was 48±21 (21–79) days, and duration of rehabilitation at our institution was 168±81 (68–292) days.

On admission, all patients presented with the clinical features of various remission stages from a hypoxic apallic syndrome according to Gerstenbrand (see Table 1) [17].

Standardized neurophysiological assessment was performed in order to establish location and amount of functional impairment of underlying pathways. Testing revealed normal auditory evoked potentials in seven patients and an increased interpeak latency in one patient, consistent with a mesencephalic lesion. Normal or exaggerated (disinhibited) auditory startle reactions were recorded in all patients. In contrast, only one patient had normal somatosensory and motor evoked potentials, while four patients hadnormal motor evoked potentials. Detailed results of the various neurophysiological investigations are presented in Table 2.

MRI revealed typical hypoxic lesions in thalamus and basal ganglia in all patients. Cortical lesions were present in five patients, a callosal lesion in one patient. Notably, only one patient had a hypoxic lesion in the pontomesencephalic area on the left side, consistent with the pathological findings in the neurophysiological assessment.

All patients underwent intensive daily neurorehabilitative therapy including physical and occupational therapy and intensive logopedic swallowing therapy with extra- and intraoral stimulation. Logopedic therapy was performed daily for a mean of 45±15 (30–60) minutes and consisted of Facial-Oral Tract Therapy (FOTT), orofacial regulation therapy (ORT), and functional dysphagia therapy [24–26]. Initially, six patients presented with intense automative oral patterns (chewing, biting, smacking, etc.), mainly due to lack of cortical inhibition.

The first clinical and fiber-optic endoscopic evaluation of deglutition was performed at 1.5 –5.5 months and repeat studies were completed 3 to 12 months post-hypoxia. Initially, all eight patients displayed severe global swallowing dysfunction, mainly consistent of a massively delayed oral phase and delayed intra-oral bolus transportation. Non-volitional swallowing reflex was intact in the seven patients without brainstem lesions. Notably, the only patient with a brainstem lesion initially suffered from absence of an intact swallowing reflex including silent aspiration in addition to the delayed intra-oral bolus transportation. However, follow-up examinations performed 4 to 10 months after cerebral hypoxia revealed intact swallowing reflexes in all eight patients. In addition, post-deglutitive aspiration was noted in three patients, and two patients showed notably reduced tongue motility. The relationship between lesion site and dysphagia is depicted in Table 3.

Six patients required a tracheal canula after the acute phase due to persistent dysphagia. Three of these patients – including the patient with brainstem lesions – achieved weaning from the canula 4 to 10 months after endotracheal placement, followed by surgical revision of the tracheostoma.

Initially, seven patients required percutaneous endoscopic gastrostomy tubes for sufficient nutrition, two of these achieved adequate oral nutrition so that the gastrostomy tube could be removed.

During the course of the hospital stay, seven patients gradually improved clinically and at discharge, presented with remission stages 7 (n = 3), 4 (n = 1), 3 (n = 2) and 1 (n = 1) of apallic syndrome according to Gerstenbrand. One patient showed no signs of remission at all.

Seven patients were eventually discharged to parental care at home, while one was transferred back to the local hospital near her home.

Various neurological disorders associated with dysphagia have been well documented in the literature, in particular cerebrovascular and neuroimmunological diseases [1, 2, 11–14]. Previous studies suggest that the voluntary action of swallowing represented in the oral phase requires the concerted action of various supratentorial motor cortical areas, most prominently the lateral pericentral, perisylvian, and anterior cingulate cortex, the adjacent supplementary motor area together with the cerebellum. The non-volitional reflexive component of swallowing is controlled by central pattern-generating circuitry of the brainstem, located in the reticular formation, trigeminal nucleus, nucleus ambiguous, and nucleus tractus solitarius [6]. Cerebral hypoxia leads to severe dysfunction of cerebral cortex, corticobulbar tracts, and cerebellar motor modulation.

Our study suggests that dysphagia is common in patients with cerebral hypoxia. All patients included in our study showed a delayed oral phase, consistent with dysfunction of the volitional execution of swallowing. Notably, patients without clinical, imaging, or neurophysiological evidence of brainstem lesions presented with an intact non-volitional swallowing reflex, whereas the only patient with a brainstem lesion initially presented with an absent swallowing reflex. A possible explanation could be that hypoxic lesions in the brainstem affect the integrity of the central pattern-generating circuitry resulting in dysfunction of the reflexive component of swallowing. However, four months after hypoxia an intact swallowing reflex was noted in this patient as well, suggesting that neuronal reorganization processes in the brainstem could recondition this circuitry necessary fordeglutition.

Strengths of our study are similar patho-etiologies in a similar group of young adults allowing comparison of the clinical features. The different investigations were carried out by the same raters throughout the observation period, i.e. the same clinicians cared for all patients in the post-acute phase, the same neuroradiologist assessed the MRI studies, the same neurophysiologist carried out the neurophysiological evaluation, and the same speech therapist treated the patients. Main limitations of this study are the small sample size and different durations of observation periods and number and timing of follow-up examinations.

Despite initial similarities noted in all patients, their development differed markedly, corroborating the difficulty in the clinical setting to predict outcome of hypoxic patients early in the course of treatment. Normal brainstem auditory evoked potentials and intact startle reactions (even if disinhibited) concur with intact brainstem function also with respect to deglutition and are thus good predictors for swallowing capabilities later on. Supraspinal lesions, documented by clinical symptoms, imaging, or neurophysiology, concur with delayed and impaired oral transport and hence impaired oral phase ofswallowing.

In conclusion, dysphagia as seen in patients with cerebral hypoxia may share similar pathophysiological mechanisms as seen and reported in patients with cerebrovascular and neuroimmunological disorders. Although brainstem lesions are rare in patients with cerebral hypoxia, dysphagia appears to be a common complication particularly in the early stages of remission which needs to be closely considered in the management of these patients. Long-term prognosis with respect to swallowing is occasionally surprisingly good in patients with acquired brain injury, highlighting the importance of a multidisciplinary approach to facilitate an optimal outcome and prevent secondary complications.

Conflict of interest

None to report.