Lower Cranial Dystonia with Inflated Cheeks: A Case of Dystonic Respiratory Failure

Abstract

We herein report a 76-year-old woman who developed lower cranial dystonia with a peculiar appearance of cheek inflation. The patient showed strong contraction of the orbicularis oris muscles. Consequently, her cheeks were passively inflated by expiration without exit. When the dystonic attack persisted, she developed cyanosis but recovered immediately after passive mouth opening. An autopsy revealed progressive supranuclear palsy. We tentatively named this characteristic dystonia “lower cranial dystonia with inflated cheeks" because of its peculiar appearance of inflated cheeks. This dystonia can cause respiratory failure. Therefore, neurologists should recognize such dystonia as a movement disorder emergency.

Introduction

Lower cranial dystonia (LCrD) is focal dystonia in which involuntary and sustained contraction of the muscles of the lower face, jaw, tongue, and pharynx causes persistent abnormal posture or movement. Examples of LCrD include jaw opening or closing, jaw protrusion or lateral deviation, tongue protrusion, or a combination of these abnormal postures (1,2). LCrD rarely but occasionally causes respiratory distress.

We herein report a case of LCrD with a peculiar presentation of cheek inflation and denting due to tight lip closure, which led to hypoxemia and cyanosis and placed the patient in a critical situation.

Case Report

A 76-year-old woman presented to the neurology department with an 8-year history of parkinsonism. Her parents had no consanguineous marriages, but her younger brother had been clinically diagnosed with Parkinson's disease. Her medical history was unremarkable, and her regular medication included levodopa/benserazide 200/50 mg daily, without antipsychotics. On an examination, she had mandibular tremor, retrocollis, neck and trunk muscle rigidity, bradykinesia, and gait disturbance.

At 78 years old, she developed dementia and dysphagia and underwent gastrostomy due to aspiration pneumonia. Brain magnetic resonance imaging (MRI) revealed atrophy of the midbrain tegmentum. Occasionally thereafter, the patient developed breathing with bilateral orbicularis oris muscles strongly contracted, the mouth closed, and both cheeks inflated and dented.

This breathing pattern persisted for several minutes to half an hour. Eventually, oxygen saturation gradually decreased to a minimum of 40%, and cyanosis occurred. At that time, she had neither stridor nor abnormal muscle contraction in the chest or abdomen. Dystonia, a sustained contraction of the orbicularis oris muscle, aborted when the mouth was passively opened. The passive inflating movement of her cheeks with breathing also disappeared, and her oxygen saturation quickly recovered. However, the dystonic attacks sometimes repeated several times successively.

We attempted sensory tricks, such as making her suck on a straw, but these were ineffective. Continuous monitoring of oxygen saturation was necessary because the dystonic attacks occurred almost daily. However, the attacks occurred only while awake, and the patient did not have nocturnal hypoxemia in the all-night oxygen saturation monitoring test. This dystonic attack was always detected by a hypoxemia alarm, except once, when we happened to visit her room. When we detected the dystonic attack, we were unable to perform video recording or electromyography (EMG) at the time because saving the patient's life was our top priority. In addition, botulinum toxin treatment for LCrD was not approved in Japan at the time; therefore, it was not performed. Because of the life-threatening situation, we considered performing tracheotomy. However, the family members did not agree.

The following year, bilateral vertical ophthalmoplegia developed. At 82 years old, although we continued levodopa/benserazide at the same dose of 200/50 mg daily, cheek inflation became asymmetrical and predominant on her right cheek. In addition, the contraction of the orbicularis oris muscle during dystonic attacks became weaker than before, and we were able instead to observe mild protrusion of the tongue (Supplementary material). When we held a hand in front of her face, we were able to confirm the outflow of expiration, and the patient no longer became hypoxemic or cyanotic. The patient died of aspiration pneumonia at 89 years old. Her family agreed to an autopsy.

A brain autopsy revealed neuropathological features of progressive supranuclear palsy (PSP) (Fig. 1A). Microscopically, the subthalamic nucleus, midbrain tegmentum, and substantia nigra showed degeneration. The brain stem, cerebellar nucleus, basal ganglia, and cerebral cortex widely demonstrated neurofibrillary tangles (Fig. 1B) and tufted astrocytes (Fig. 1C). The distribution of neuronal degeneration and tau pathology was not significantly different from that in usual PSP cases. There were no characteristic findings, such as Lewy bodies, glial cytoplasmic inclusions, or astrocytic plaques, suggesting Parkinson's disease, multiple system atrophy, corticobasal degeneration, or other diseases that present with parkinsonism.

Figure 1.

(A) Transverse section of the midbrain and upper pons show atrophy of the tegmentum and depigmentation of the substantia nigra and locus ceruleus. A 5-mm-unit ruler is attached. (B) A microphotograph of superior colliculus. Neurofibrillary tangles are scattered. Gallyas stain. Scale bar: 20 μm. (C) Microphotograph of globus pallidus. Tufted astrocytes are seen. Gallyas stain. Scale bar: 20 μm.

Discussion

The following three characteristic findings were observed in this patient. 1) The patient contracted the bilateral orbicularis oris muscles strongly and closed her mouth, and at the same time, she repeatedly inflated and dented her cheeks with respiration. 2) As this condition persisted, oxygen saturation gradually decreased, and cyanosis occurred. 3) Opening of her mouth alone recovered her hypoxemia. The most noticeable external finding was inflated and dented cheeks. However, as shown in the video, no contraction of the cheek muscles was observed. This cheek movement itself was not dystonia but rather a passive movement due to expiration and inspiration without exit.

If only the orbicularis oris muscle contracts continuously, it is the same as the usual oromandibular dystonia, and hypoxemia does not occur because the patient can breathe through the velopharyngeal port (Fig. 2A). Anatomically, there are three possible mechanisms underlying hypoxemia due to dystonia: 1) obstruction of both the mouth and velopharyngeal port (Fig. 2B), 2) obstruction of the middle or lower pharynx or larynx (3-6), and 3) dysfunction of the respiratory muscles (diaphragm or accessory respiratory muscles) (7). Of these three mechanisms, obstruction of the middle pharynx, lower pharynx, or larynx and dysfunction of the respiratory muscles were ruled out, as passive mouth opening alone recovered the hypoxemia in our case. Therefore, the simultaneous obstruction of the mouth and velopharyngeal port was strongly suggested. In this situation, the air that had lost its outlet moved back and forth between the oral cavity and lungs, causing the cheeks to alternately inflate and dent. Furthermore, hypoxemia was able to be recovered simply by opening the mouth, as the respiratory movement itself is normal. Thus, when the mouth and velopharyngeal port contracted simultaneously, all three characteristic findings in our case could be reasonably explained.

Figure 2.

A schematic illustration of a sagittal section of the mouth and pharynx. ×: no flow. (A) Usual oromandibular dystonia without respiratory failure. Because the velopharyngeal port is open, patients do not develop hypoxemia or cyanosis. (B) Simultaneous obstruction of the mouth and velopharyngeal port caused hypoxemia and cyanosis in our patient. In this situation, the patient’s cheeks were passively inflated by expiration without exit. Hypoxemia was able to be resolved by simple passive opening of the mouth because the respiratory movement itself was normal.

Six similar cases have been reported (Table), including four with Parkinson's disease, one with PSP, and one with multiple system atrophy (8-10). None of these attacks occurred during sleep, and one patient had obvious sensory trick. These findings were consistent with the features of dystonia. A sensory trick is a characteristic finding of dystonia but it is not present in all cases. Only 51.4% (198/385) of oromandibular dystonia (11) and 38.5% (5/13) of dystonia with respiratory impairment (7) had sensory tricks.

Table.

Previously Reported Cases Similar to the Present Study.

References	(8)	(9)	(10)	Present study
Age at onset of the dystonic attacks	66 years	n.d.	79 years	78 years
Sex	Male	n.d.	Female	Female
Underlying disease	PSP 1 case	MSA 1 case PD 3 cases	PD 1 case	PSP 1 case
Time interval between onset of underlying disease and onset of the dystonic attacks	4 years	Advanced stage	4 years	10 years
Sensory trick	- Mouthpiece	n.d.	+ Contact with mandible	- Straw
Occurrence during sleep	-	n.d.	n.d.	-
Hypoxemia	-	+	+	+
Confirmation of velopharyngeal port obstruction	+	n.d.	n.d.	Not done
Medications at onset of the dystonic attack (mg/day)	n.d.	n.d.	Levodopa/benserazid 450/112.5 mg Pramipexole 2.5 mg Cabergoline 0.5 mg Selegiline 2.5 mg	Levodopa/benserazide 200/50 mg

MSA: multiple system atrophy, n.d.: not documented, PD: Parkinson’s disease, PSP: progressive supranuclear palsy

We tentatively named this characteristic dystonia “lower cranial dystonia with inflated cheeks" because of its peculiar appearance of inflated cheeks. The term “lower cranial dystonia" was chosen instead of “oromandibular dystonia" because tongue protrusion dystonia was also present in our case, and involvement of the pharyngeal muscles was strongly suggested. Three studies are needed to establish this unique form of dystonia as a distinctive sub-phenotype of LCrD.

The first is the nasopharyngeal fiberscope performed during dystonic attacks. Only Hayashi et al. confirmed the obstruction of the velopharyngeal port directly using a nasopharyngeal fiberscope (8). However, it is not easy to perform nasopharyngeal fiberscope promptly because this dystonia is not always present, occurring intermittently. In addition, when dystonia causes hypoxemia, it is ethically unacceptable to perform the examination, as the first priority is patient life support.

The second is an EMG study of the pharyngeal muscles, such as the levator veli palatini (LVP) muscle. In healthy individuals, this muscle contracts during swallowing to close the velopharyngeal port and prevent the backflow of food into the upper pharynx (12,13). Because the LVP muscle is not located on the body surface, an EMG study requires a specially designed recording electrode (14).

Third, further pharmacological studies are required. Kato et al. reported that discontinuation of selegiline exacerbated these dystonic attacks, but resumption resolved them (10). In contrast, there are reports of PSP cases in which levodopa causes oromandibular dystonia (15) or facial dystonia (16). In our case, increasing or decreasing the dosage of antiparkinsonian drugs or adding antipsychotic drugs carries a risk of inducing death by asphyxia. Therefore, we continued levodopa/benserazide at the same dosage of 200/50 mg daily and did not add any other drugs. However, over the next four years, mild morphological changes in the dystonia occurred, and hypoxemia and cyanosis no longer occurred. Whether these dystonic attacks were caused by the disease itself or induced by antiparkinsonian drugs should be investigated by accumulating a large number of cases in the future.

In conclusion, we experienced a case of LCrD with a peculiar form of inflated cheeks. LCrD with inflated cheeks can lead to respiratory failure. Therefore, neurologists should be aware of this dystonia as a movement disorder emergency.

The authors state that they have no Conflict of Interest (COI).

Supplementary Material

Supplementary Video

Video recordings obtained at 83 years old. The patient inflated her cheeks during expiration, and dents appeared in the cheeks during inspiration. When we recorded this video, the contraction of the orbicularis oris muscle was weaker than it had been before, and we were able to observe mild tongue protrusion. Cheek inflation was asymmetrical, and there was a small amount of expiratory outflow. By that time, the patient had no longer become hypoxemic or cyanotic; therefore, we were able to safely record the video.