Case Report: AntiNMDA Receptor Encephalitis

Clinical Presentation

Case 1. Ms. A Presents With Concern for Acute Drug Intoxication

Ms. A, age 22 years, was previously admitted to an outside hospital for approximately 1 month because of slowly escalating behavior, described as bizarre. Subsequently, she presented to the emergency department (ED) after continued bizarre behavior and inappropriate sexual behavior. Ms. A’s mother also reported that Ms. A used the synthetic cannabinoid flakka, making acute drug intoxication a concern. In the ED, Ms. A was noted to be uncooperative, violent, and hypersexual with incoherent speech and nonpurposeful movement. Her physical examination was significant for hypothermia (94.3°F).

Case 2: Ms. B Presents With Witnessed Seizures and Concern for Drug Intoxication

Ms. B, age 26 years, had a medical history of bipolar disorder, depression, and attempted suicide. She presented multiple times over the last 9 days to an outside hospital with chief complaints of headache, a racing heart, chest pain, and an episode of auditory hallucinations, during which she believed she heard a baby crying. Ms. B was repeatedly discharged without a specific diagnosis until she experienced a seizure that was witnessed by her spouse. The first witnessed seizure Ms. B had was characterized by rigid arms and open eyes and mouth. She had a second seizure in the ED of the outside hospital, consisting of 20 to 30 seconds of bilateral arm movement, head nodding, and repetitive blinking, during which time Ms. B was unresponsive to verbal commands. She was intubated, stabilized, sedated, and transferred to our hospital for further care. Initial reports from the outside hospital suggested that Ms. B may have ingested the synthetic cannabinoid, K2, making acute drug intoxication a primary concern. Considering the presence of headache and seizures, meningitis and encephalitis were also initial concerns.

Diagnostic Studies

Case 1: Ms. A’s Laboratory Study Findings

Ms. A had urinary and blood toxicology tests that were negative for synthetic cannabinoids, ruling out acute flakka intoxication, although her urine drug screen was positive for cannabinoids. Blood and urine drug tests were also negative for salicylates, ethanol, acetaminophen, amphetamines, barbiturates, benzodiazepines, opiates, cocaine metabolites, phencyclidine, and methadone. An expanded toxicology screen showed only olanzapine, which had been administered in the ED as a sedative. Ms. A had a lumbar puncture, and cerebrospinal fluid (CSF) analysis revealed an elevated white blood cell (WBC) count (17/mcL) with lymphocytic predominance (91%) and protein and glucose within normal limits. CSF analysis was negative for gram stain, viral culture, and antibodies for herpes simplex virus (HSV)-1, HSV-2, and Lyme disease.

Case 1: Ms. B’s Laboratory Study Findings

A urinary drug screen was only positive for benzodiazepines (antiepileptic/sedative administered at outside hospital) and negative for amphetamines, barbiturates, opiates, cocaine metabolites, cannabinoids, phencyclidine, and methadone. Sedation was weaned overnight, and the following day Ms. B was active and attempted to get out of bed. She was, however, still intubated and not following commands. That afternoon, she again seized, initiating a period of continuous slow, nonpurposeful movement of all extremities, head nodding, blinking, grimacing, and chewing, diagnosed as status epilepticus. She remained intubated and without quantifiable return of consciousness throughout her hospital stay.

Subsequent Diagnostic Testing

Case 1. CSF From Day 8 of Admission Positive for Antibodies to N-methyl D Aspartate Receptor

Ms. A had a relapsing and remitting course characterized by periods in which she was both cooperative and coherent followed by periods of repetitive movements, incoherent speech, and unresponsiveness. She experienced auditory hallucinations (“phone ringing”) and intermittent autonomic instability characterized by elevated blood pressure (diastolic blood pressure >110 mmHg).

On day 8 of this admission, Ms. A had a repeat lumbar puncture and CSF analysis that showed oligoclonal banding, elevated WBCs (22/mcL) with a lymphocytic predominance (79%), normal protein, and normal glucose. Antibody testing of this CSF sample also revealed an N-methyl D aspartate receptor (NMDAR) antibody titer of 1:20 (ARUP Laboratoratories; Salt Lake City, UT; indirect fluorescent antibody test) and was negative for Epstein-Barr virus (EBV), West Nile virus, and syphilis.

Seizure-like activity developed on day 11 of this admission, with sudden unresponsiveness, foaming at the mouth, twitching, and agonal breathing, followed by a period of somnolence. By day 15, Ms. A also developed electrocardiogram (ECG) changes suggestive of a right bundle branch block (RBBB).

Ms. A had an abdominal ultrasound, which identified a right adnexal mass, consistent with an ovarian teratoma rather than a corpus luteum cyst. On day 25 of admission, she had a right oophorectomy, and the mass was diagnosed as a mature solid teratoma. Mature skin and adnexa, adipose tissue, gut epithelium, and neural tissue were identified without the presence of immature neural tissue (Figure 1). Of note, a lymphocytic infiltrate was identified adjacent to the mature neural component.

Case 2. CSF From Day 3 of Admission Positive for NMDAR Antibodies

Ms. B underwent a lumbar puncture and CSF analysis on day 3 of her admission that demonstrated an elevated WBC count (107/mcL) with a lymphocytic predominance (76%), normal glucose and protein, and oligoclonal banding by electrophoresis. The CSF analysis was negative for California encephalitis antibody, Eastern equine encephalitis antibody, St. Louis encephalitis antibody, Western equine encephalitis antibody, West Nile virus antibody, and cryptococcus antigen. Syphilis testing was negative, as was testing for enterovirus, HSV-1, and HSV-2. Bacterial and fungal culture of the CSF was also negative. On day 5, a reserve CSF tube was sent for NMDAR antibody testing (Mayo Clinic Laboratories; Rochester, MN; indirect immunofluorescence assay), which found a titer of 1:1024.

Ms. B had a CT scan that identified a left adnexal mass, consistent with a teratoma. On day 17 of admission, she underwent salpingoophorectomy, and the mass was diagnosed as a grade I immature teratoma with a lymphocytic infiltrate (Figure 2).

Treatment and Outcomes

Case 1: Therapeutic Plasma Exchange and Return to Baseline

Beginning on day 12 of this admission, Ms. A was treated with intravenous immunoglobulin (IVIG), rituximab, and therapeutic plasma exchange (TPE). The TPE regimen was 1.0 plasma volume per procedure with 5% albumin replacement every other day for 10 days. This was followed by weekly TPE at the same volume and albumin replacement for 4 weeks. Ms. A experienced significant improvement after the first TPE and had returned to her baseline by the final procedure, complaining only of a mild hand tremor.

Case 2: TPE and Death

Beginning on day 17, Ms. B was treated with IVIG, rituximab, methylprednisolone, and also had TPE (1.0 plasma volume per procedure with 5% albumin replacement) every other day for 10 days. Ms. B tolerated TPE well but had no significant clinical improvement. She remained persistently unresponsive with nonpurposeful movement. Ms. B’s clinical course was eventually complicated by deep venous thromboses in all extremities, bacteremia, and a pneumothorax. Her family elected for comfort care only and TPE was discontinued. On day 39, Ms. B died, 47 days after her initial presentation to the outside hospital.

Discussion

We describe 2 women who had classic clinical and laboratory findings seen in antiNMDAR encephalitis. In Ms. A’s case the timing from symptom onset to documented seizure activity was 36 days, whereas Ms. B developed seizure activity only 9 days after symptom onset.

Ms. A had a prolonged period of purely psychiatric symptoms before developing autonomic instability and arrhythmia, whereas Ms. B complained of heart palpitations and chest pain at her initial presentation. Both patients had teratoma excision, IVIG, rituximab, and TPE, with Ms. A experiencing complete recovery and Ms. B having no response to treatment. These differences in initial disease course and treatment response suggest a marked difference in disease severity that may also be reflected by antiNMDAR titers (1:20 titer for Ms. A and 1:1024 titer for Ms. B).

Ovarian Teratoma Classification

Previous literature has drawn a correlation between NMDAR antibody titers and disease severity in antiNMDAR encephalitis,^1-5 as described here. The etiology for varying antibody titers, however, has not been reported. Ms. A and Ms. B not only had a large difference in antibody titers, they also had different pathology of their ovarian teratomas (Figures 1 and 2) with Ms. A having a mature teratoma and Ms. B having a teratoma with immature features. Ovarian teratomas are classified as benign mature teratomas or malignant immature teratomas. The presence of immature neural tissue defines malignant immature teratomas, which are graded by the amount of this tissue that is present (Table).^6-8 Grade 1 immature teratomas are clinically low grade, with behavior similar to a benign mature teratoma. Grade 2 and 3 immature teratomas are associated with a more aggressive course of disease.^6-8

Ms. B had a grade 1 immature teratoma but also had a significantly higher antiNMDAR titer compared with Ms. A (1:1024 versus 1:20), who had a mature teratoma. Both women had teratomas with a lymphocytic infiltrate. Immune reaction to a teratoma is a proposed mechanism for antibody induction in antiNMDAR encephalitis,³ and the histologic findings of the 2 described cases support this hypothesis. We postulate that neural tissue in a teratoma may provide an immunogenic nidus leading to the production of antibodies that cross-react with native NMDARs. Furthermore, immature neural tissue may have enhanced immunogenicity relative to a mature neural component, and this could contribute to an increased NMDAR antibody titer in someone with an immature vs mature teratoma.

Response to Treatment

Both Ms. A and Ms. B had many classic symptoms associated with the syndrome of antiNMDAR encephalitis, including behavior changes, hallucinations, seizures, autonomic dysregulation, and abnormal movements. Several of these characteristic symptoms (eg, psychosis and autonomic dysregulation) overlap with the clinical picture of acute drug intoxication, making consideration of this in the differential diagnosis reasonable. Despite this, CSF antibody testing for antiNMDAR encephalitis was ordered on day 8 and day 3 of admission for Ms. A and Ms. B, respectively—relatively early in their respective hospital courses—with directed treatment initiated on day 25 and day 17 of admission, respectively. This suggests the timing of diagnosis and initiation of therapy were not critical to the observed differences in treatment response and outcome. After her first TPE procedure, Ms. A began to show improvement noticeable to her family and clinicians, and her condition returned to baseline after 9 TPEs over 5 to 6 weeks. In contrast, Ms. B had no improvement after 5 TPE procedures over 10 days. Despite a later intervention in relation to symptom onset for Ms. A (53 days) compared with Ms. B, (26 days), Ms. A had a markedly better response to a nearly identical therapy. This likely reflects an enhanced response to treatment for patients with lower antibody titers compared to those with higher titers.

Summary

AntiNMDAR encephalitis can present with a diverse constellation of symptoms including new-onset psychosis, cardiac arrhythmias, and seizures. Despite the prevalence of drug intoxication antiNMDAR encephalitis is an important syndrome to consider, especially when women in their 3rd to 4th decade of life present with psychiatric symptoms and seizures because it is straightforward to diagnose and has effective treatment.

Prompt diagnosis and treatment of antiNMDAR encephalitis by CSF analysis, abdominal ultrasound, oophorectomy, immunotherapy, and TPE may be beneficial, but the anti-NMDAR titer likely represents the most critical factor predicting therapeutic response. The mechanism of antiNMDAR titer variation is still unknown, but these 2 cases suggest the possibility that immature neural tissue present in associated teratomas may have increased immunogenicity leading to increased antibody titers and a more severe disease course.

Although we compare only 2 cases in this report, the similarities in initial presentation and eventual diagnosis and treatment but not teratoma pathology, antibody titer, or response to treatment suggest the hypothesis that antibody titer and immaturity vs maturity of neural tissue within a teratoma may correlate with disease severity. Prognosis, which has been correlated with the magnitude of the NMDAR antibody titer,⁵ may also correlate with the maturity of the associated ovarian teratoma, if a teratoma is identified.