Severe tooth loss secondary to orofacial dyskinesias in anti-NMDA receptor encephalitis

Abstract

A 24-year-old woman with no significant medical or psychiatric history was brought to the emergency department due to altered mental status and bizarre behaviour. Physical examination was remarkable for decreased speech output and orofacial dyskinesia. Upon further evaluation, electroencephalogram showed extreme delta brush waves and cerebrospinal fluid was positive for anti-NMDA receptor antibodies. Despite aggressive treatment with steroids and immunosuppressive therapy, her dyskinesia was severe enough to cause tooth loss, tongue and lip laceration.

Background

Anti-N-methyl-D-aspartate (NMDA) receptor encephalitis is a rare autoimmune disorder increasingly recognised as an important differential diagnosis to be considered in patients presenting with acute to subacute encephalopathy. Here, we discuss the clinical features of a patient with this condition and report an unusual association with tooth loss secondary to orofacial dyskinesia. We believe it is important for all practitioners to recognise the presentation of this condition and the appropriate diagnostic workup, allowing for more rapid treatment initiation and hopefully better outcomes.

Case presentation

A 24-year-old woman was brought to the emergency department by friends due to altered mental status. In the months prior to presentation, she had developed an eating disorder. Three days prior she began exhibiting bizarre behaviours such as perseverative house cleaning and social isolation. Gradually, she developed decreased speech output, which progressed to mutism.

A complete history was elicited. She had no previous medical or psychiatric diagnosis, had no prior surgeries and took no illicit drugs, supplements or medications. Prior to presentation, our patient was functional and independent for all activities of daily livings. She had one child and was not sexually active. There was no known significant family history or recent sick contacts.

At the time of admission, the patient was encephalopathic, unable to follow commands, and with no comprehensible speech output. She was afebrile and her neck was supple. She exhibited orofacial dyskinesias (video 1) and choreoathetotic movements in her right arm. Otherwise, there was no focality to her neurological exam.

Video 1.

Download video file ^{(3.1MB, mp4)}

DOI: 10.1136/bcr-2018-228380.video01

Exhibiting orofacial dyskinesia which became so severe the patient suffered tongue lacerations and shattered teeth.

Investigations

Serum chemistry and blood counts were within normal limits and a urine drug screen performed in the emergency department was unrevealing. A CT and MRI brain with and without contrast did not demonstrate a clear aetiology for her symptoms. Due to a high suspicion for seizures (possibly secondary to an encephalitis), an electroencephalogram (EEG) was performed and showed extreme delta-brush pattern (figure 1). A lumbar puncture demonstrated cerebrospinal fluid (CSF) pleocytosis (71 cells, 100% lymphocytes) with normal glucose and was positive for anti-NMDA receptor antibodies. Other CSF and serum laboratory data, including an extensive autoimmune and infectious panel was negative. Transvaginal ultrasound did not visualise ovarian pathology. Contrast-enhanced CT scans of her chest, abdomen and pelvis did not detect any evidence of malignancy.

Figure 1.

Representative portable 30-min electroencephalogram epoch of the patient performed on day 3 of hospitalisation showing extreme delta-brush pattern. Bipolar montage remarkable for delta waves with superimposed beta frequency activity (arrows) and well-organised background.

Differential diagnosis

At the time of admission to the emergency department, and before neurological evaluation, her symptoms were thought to be secondary to a psychiatric condition or to drug use. The lack of a prior psychiatric illness made this hypothesis less likely; and a negative urine drug screen lessened the possibility of drug intoxication. Additionally, the course of her symptoms was subacute. Her language impairment had a broad differential diagnosis including ischaemia, seizure, infection and inflammatory processes involving language centres. The presence of psychiatric symptoms and associated movement disorders (orofacial dyskinesia) raised concern for an autoimmune encephalopathy, specifically anti-NMDA receptor encephalitis. Other aetiological possibilities included viral or bacterial meningoencephalitis such as herpes infection however, CSF analysis and cultures were not suggestive. Other autoimmune or paraneoplastic causes for limbic encephalitis including anti-LGI1, anti-AMPA, anti-GAD65 and anti-Hu related syndromes were considered.

Treatment

Regarding immunologic therapies, on day 2 of admission, treatment was initiated with intravenous methylprednisolone 1-gram daily for 5 days. Prednisone 60 mg daily was continued for 1 month, then slowly tapered off throughout the hospitalised. Starting on day 7 of admission, further treatment with five sessions of plasma exchange over 10 days was initiated. On day 21 of hospitalisation, as there was no clear clinical response to treatment, 2 g/kg of intravenous immunoglobulin (IVIG) was given over 5 days. Finally, two 1-gram infusions of rituximab were given, separated by 2 weeks, on days 60 and 74 of admission.

Outcome and follow-up

Despite aggressive treatment early in the hospitalisation, the patient’s condition deteriorated, with persistent encephalopathy, severe orofacial dyskinesia and dystonic posturing of upper and lower extremities. She remained intubated and sedated in the neuroscience intensive care unit (ICU). Despite sedation, the recalcitrant dyskinetic facial movements caused forceful teeth grinding and jaw clenching resulting in tongue and lip laceration and tooth loss (figure 2). As she had not yet responded to immunosuppressive therapy 6 weeks into the admission, a bilateral salpingo-oophorectomy was performed empirically, considering microscopic signs of malignancy would be missed by transvaginal ultrasound or contrast-enhanced CT. Subsequent pathological examination was performed and did not reveal any evidence of tumour or malignant cells.

Figure 2.

Tooth loss, tongue and lip laceration secondary to orofacial dyskinesia.

Botulinum toxin injections were employed to help control the orofacial dyskinesia, which were not responding to high doses of haloperidol or lorazepam. Each masseter received 100 units and each temporalis muscle 20 units of onabotulinumtoxin A. In the weeks following the rituximab infusions, our patient experience marked improvement of both encephalopathy and dyskinesia. On day 87 of hospitalisation, she began following simple commands. Some verbal output occurred on day 91. At the time of this writing (24 months after initial presentation), the patient has recovered to her baseline mental status, remains on levetiracetam, and is amnestic to the whole account. She received a third infusion of rituximab 8 months after the second infusion, and her CD19 T cell counts was 0 prior to the infusion.

Discussion

Anti-NMDA receptor encephalitis is a rare autoimmune encephalitis syndrome that classically presents with psychiatric manifestations such as bizarre behaviour, psychosis, hallucinations and decreased level of consciousness.¹ In many patients, this syndrome is also associated with movement disorders, which include orofacial dyskinesia, dystonia, rigidity and opisthotonos-like posturing.² Seizures, autonomic instability and central hypoventilation can also complicate a patient’s course.³ In an observational study of 577 patients with anti-NMDA receptor encephalitis, 95% were younger than 45 years old, and 37% were younger than 18 years old. A female to male predominance of 4:1 was reported.⁴ In contrast, in a cohort of patients over 45 years in age, 45% were male.⁵

In our patient, the association of subacute psychiatric symptoms with orofacial dyskinesia raised concern for this disorder. An EEG showed the highly suggestive extreme delta-brush waves⁶ and the diagnosis was confirmed with CSF anti-NMDA receptor antibodies.⁴ This EEG finding is better characterised as rhythmic delta activity at 1–3 Hz with superimposed bursts of rhythmic 20–30 Hz beta frequency activity ‘riding’ on each delta wave. Although highly suggestive of NMDA receptor encephalitis, other conditions have also been associated with this pattern, including other encephalitides, mesial temporal lobe epilepsy, and even ICU patients suggesting a high mortality rate.⁷ MRI brain may show non-specific changes such as transient fluid attenuation inversion recovery (FLAIR) or contrast enhancing abnormalities in cortical or subcortical regions however it was unremarkable in this patient.³

This autoimmune diagnosis should always be suspected in patients with the above symptoms and other associated findings including CSF with lymphocytic pleocytosis and/or oligoclonal bands, abnormal EEG (most specifically delta-brush waves, but including slow and epileptic activity) and abnormal MRI. Of 100 cases, 95 had abnormal CSF, 92 had an abnormal EEG and 55 an abnormal MRI.³ Although the final diagnosis is made by the presence of the anti-NMDA receptor antibodies in serum or CSF, false-positive cases have been described when analysing serum only.⁸

The detection of an ovarian teratomas in association with this syndrome, is also well described, especially in young women.⁹ It is thought that up to 50 percent of all female patients older than 18 years have ovarian teratomas, while in patients younger than 14 years, the likelihood of finding a malignancy is much lower up to nine percent. Imaging studies such as transvaginal ultrasound, CT and MRI can be used as a diagnostic modality.¹⁰ In our patient, even after a thorough evaluation that included a salpingo-oopherectomy and pathologic analysis, there was no demonstration of malignancy.

There are no published guidelines regarding the immunotherapy for NMDA encephalitis. Though as described in prior publications, the most accepted treatment includes an initial trial of intravenous solumedrol followed by IVIG or plasma exchange. In selected refractory cases, rituximab and/or cyclophosphamide are often employed. In the presence of a tumour, careful examination is required and tumour removal is often necessary.^{3 11} In our case, it is unclear if the surgical procedure changed this patient’s prognosis. Retrospectively, it is possible that immunosuppressive therapy only and no surgery would have been the most appropriate course of action in our patient, although the current evidence is unclear.

In a case series of 100 patients with anti-NMDA receptor encephalitis, 92 were treated with immunotherapy and 47 of 53 comorbid ovarian tumours were removed.³ The median time to symptom improvement was reported as 10 weeks for patients without tumours, 11 weeks for patients with tumours that were treated in a delayed fashion or not at all, and 8 weeks for those with tumours treated early; with a range of 2–50 weeks, 4–40 weeks and 2–24 weeks in the respective cohorts.³ The median duration of hospitalisation was 2.5 months, ranging from 1 to 14 months.³ Even with early immunotherapy initiation, as seen in our patient, the natural history of this condition involves a delayed response to treatment. This must be taken into consideration when making invasive treatment decisions.

The abnormal movements are usually controlled with sedative medications and treatment of the encephalitis. Here, we describe severe and nearly intractable orofacial dyskinesia causing the loss of teeth. Dental and perioral complications of orofacial dyskinesias and bruxism in NMDA encephalitis are well described, including severe teeth grinding, loss of teeth, and mutilation of lips and tongue. Most likely mechanisms are secondary to the abnormal movements, which may also result in grimacing, masticatory-like movements, and forceful jaw opening and closing with repeated shear stress and bruxism.^{3 11} Although common in NMDA receptor encephalitis and other movement disorders, the dental complications of orofacial dyskinesias are rarely discussed in the literature. To date, there is no consensus regarding prophylaxis or symptomatic treatment. Some reports have encouraged the use of rubber dam and bite blocks for dental protection in patients with severe orofacial dyskinesias, although no specific treatment have been previously described.¹²

We provide video and pictorial demonstration of the involuntary movements and dental sequelae as well as the strongly suggestive EEG findings of extreme delta brush, which we hope will aid other practitioners in the early recognition and treatment of this condition. We also demonstrate that local botulinum toxin injections may partially control dyskinesia in anti-NMDA encephalitis.

Learning points.

• Anti-NMDA receptor encephalitis should be suspected in young patients with acute psychiatric symptoms and associated involuntary movements.

• Extreme delta-brushes waves on electroencephalogram are highly suggestive of anti-NMDA receptor encephalitis.

• Although rare, tooth loss may be a complication in anti-NMDA encephalitis.

• Local botulinum injections may be used as a complimentary approach to control orofacial dyskinesia in anti-NMDA receptor encephalitis.