Topiramate as Possible Treatment for Catatonia in Anti-NMDA Receptor Encephalitis

Anti-N-methyl-d-aspartate (anti-NMDA) receptor encephalitis is a clinical syndrome first described in 2005 (1). The specific causative antibody was identified 2 years later by Dalmau et al. (2), who also noted that the syndrome is more prevalent in women because of its frequent association with gynecological neoplasms. The clinical presentation consists of a viral-like prodrome, including lethargy, headache, upper-respiratory symptoms, nausea, diarrhea, myalgias, and fever, followed by an acute phase with neuropsychiatric symptoms (3). These include agitation, psychosis, behavioral changes, seizures, unresponsiveness, abnormal movements, autonomic abnormalities, and hypoventilation, which can necessitate admission to an intensive care unit (4).

Catatonia has been described as a complication of anti-NMDA receptor encephalitis, as well as paraneoplastic limbic encephalitis (5, 6). Currently, there are no published cases of catatonia in this population that were treated by nonbenzodiazepine pharmacotherapies. The current therapeutic recommendations for catatonia include first-line treatment with benzodiazepines and electroconvulsive therapy (ECT) for some refractory cases. There is some evidence for the effectiveness of second-generation antipsychotics and glutamatergic drugs (7). Topiramate, which modulates both glutamatergic and GABAergic activity, has been successful anecdotally in treating catatonic states in patients with schizophrenia and bipolar disorder. To our knowledge, this is the first report of anti-NMDA receptor encephalitis–related catatonia successfully treated with topiramate.

Case

A 20-year-old woman with no past medical or psychiatric history presented to a local hospital with acute onset agitation, psychosis, and paranoia. She reported delusional thoughts and concerns that someone was trying to kill her. A primary psychiatric illness was suspected, but multiple episodes of seizure-like activity complicated her hospital stay. These events were characterized by screaming and thrashing without clear evidence of lateralizing epileptic signs. She did not report an aura. There was no electroencephalogram (EEG) correlation with these events. Furthermore, no delta brush (a unique EEG pattern in adults with anti-NMDA receptor encephalitis associated with a more prolonged illness) was seen (8). Complete blood count, comprehensive metabolic panel, urine drug screen, and urinalysis were unremarkable. Vital signs were stable. Due to suspicion for autoimmune/paraneoplastic encephalitis, lumbar puncture was performed on day 8 of hospitalization. The results revealed glucose of 47 mg/dl (normal range 45–80), protein of 54.3 mg/dl (20–40), white blood cell count of 311/mm³ (<10), red blood cell count of 1,036/mm³ (<10), lymphocytes 88%, monocytes 12%, and anti-NMDA receptor antibody (anti-NMDAr-Ab) present to 1:160 titration. Serum NMDAr-Ab was not sent, because studies have shown cerebrospinal fluid NMDAr-Ab to be more sensitive (9).

The presence of antibodies in CSF was considered diagnostic of anti-NMDA receptor encephalitis. The patient was started on high-dose steroids. Because this condition is frequently a paraneoplastic phenomenon associated with ovarian tumors, specifically teratoma (2), an MRI was performed and revealed a right ovarian cystadenoma. The neoplasm was surgically resected. On day 15, she was started on therapeutic plasma exchange (TPE) and continued for five therapeutic sessions, once every other day for 10 total days. The patient's condition continued to decline. On day 19 she had a 15-minute episode of left gaze deviation with lip-smacking, concerning for a seizure. Intravenous lorazepam administration resolved this episode.

On day 22, because of continued mental fluctuation, the patient was transferred from the outside hospital to this tertiary care facility. Upon arrival she was mute and rigid, with fixed gaze. She was unable to follow commands, did not move spontaneously, and exhibited waxy flexibility. Stuporous catatonia (10) was considered. To confirm the diagnosis, lorazepam 1 mg IV was given. Within minutes the patient started to move and follow commands.

Lorazepam was initially scheduled for 2 mg every 6 hours IV, but it was then discontinued because of the concern strongly voiced by the patient's mother about the addictive potential of benzodiazepines, given a family history of substance abuse. Even though the evidence for using higher doses of benzodiazepines was clearly communicated, and the patient herself had no known history of substance use disorders, her mother expressed her wishes in seeking alternative therapies. Faced with this ethical and scientific dilemma, and in the absence of family members except for the mother, our rationale was to continue with TPE and steroids to address the etiology behind her illness. In parallel, we decided to respect the mother's wishes and look for alternative symptomatic treatment for catatonia, with the understanding that we would restart benzodiazepines if there was no response to the alternative. ECT was not available in the hospital.

Given the patient's mental fluctuation, pseudo-seizures, and need for TPE, it was not possible to transfer her to a psychiatric hospital for ECT. Other pharmacotherapies, such as zolpidem (11) (addictive potential) and valproate (weight gain and teratogenic potential) were excluded.

Even though NMDA receptor antagonists, such as amantadine and memantine (7), have been indicated as potential treatments of catatonia, we hypothesized that these might not be the best option for our patient. It has been shown that the glycine-binding subunit of the NMDA receptor is causative in anti-NMDA receptor encephalitis and that receptor internalization leads to the symptoms (3). Thus we thought that NMDA receptor antagonists, which result in internalization of the glutamate subunit, might worsen the anti-NMDA receptor encephalitis disease course. We hypothesized that the GABA/kainite-acting medication, topiramate, could be more effective and less harmful. Topiramate is a partial blocker of kainate and a-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) glutamate receptors (12). Whereas postsynaptic NMDA receptors are the target of pathological autoantibodies, AMPA receptors are both pre- and postsynaptic, and kainate receptors are presynaptic only (13). The ability to modulate both GABAergic and glutamatergic transmission through receptors other than NMDA makes topiramate mechanistically a good candidate for the treatment of catatonia in anti-NMDA receptor encephalitis.

After obtaining consent, treatment with topiramate was started and up-titrated to 100 mg a day. The patient's verbal and motor abilities began to improve within 24 hours and continued to rapidly improve over the next 2 weeks. The rate of symptomatic improvement, which is expected in anti-NMDA receptor encephalitis with TPE alone, appeared to increase with the addition of topiramate. Catatonia symptom scales, which might inform this clinical impression, were not used. On the 39th day after initial admission, the patient was alert, oriented, and able to walk 50 feet with assistance. She was discharged to inpatient rehabilitation and lost to follow-up.

Discussion

Catatonia is a well-described clinical syndrome with unclear pathophysiology (2). It is typically associated with psychiatric disorders, such as schizophrenia and bipolar disorder (10). The current pathophysiological hypothesis implicates GABA-A, dopamine D2 receptors, and NMDA receptors. It is also known to be associated with disorders with well-characterized organic causes, including anti-NMDA receptor encephalitis (5). Studies have emphasized the importance of recognizing the possibility of underlying organic causes of apparent psychiatric syndromes, including psychosis and catatonia in young females (3). Anti-NMDA receptor encephalitis is of particular importance because of its relatively high prevalence among autoimmune conditions and high morbidity and mortality if left untreated (14). Without treatment, patients with anti-NMDA receptor encephalitis can experience progressive neurologic deterioration and death. With treatment, most patients will make a full recovery within 2 years of disease onset. Predictors of positive outcomes include the presence and subsequent removal of a tumor, quick diagnosis, and aggressive treatment. Although catatonia has not been addressed specifically, the largest published study on treatment and outcomes in anti-NMDA receptor encephalitis (577 patients from 200 centers and 35 countries) indicated that half the patients improved within the first 4 weeks of first-line therapy. By 24 months, 80% were either fully recovered or, even if there was some residual disability, were able to look after their own affairs without assistance (15).

NMDA receptors are ligand-gated cation channels with critical roles in synaptic transmission and plasticity. They are composed of two heteromers, referred to as NR1 and NR2 subunits, which bind glycine and glutamate, respectively (16). A previous study has established a connection between antibodies to the NR1 subunit and anti-NMDA receptor encephalitis (13). Further evidence for the involvement of NMDA receptor blockade comes from the fact that NMDA antagonists, such as phencyclidine, are capable of producing symptoms similar to those seen in patients with limbic encephalitis, including catatonia (17). However, it appears that the decreased NMDA activity in anti-NMDA receptor encephalitis might be mediated by receptor internalization, rather than by direct antagonism (3). Though the management of catatonia is typically focused on using benzodiazepines, there is evidence pointing to glutamatergic involvement. This is especially true when catatonia is associated with anti-NMDA receptor encephalitis. Multiple authors have published work on the utility of glutamate antagonists in the treatment of catatonia unrelated to autoimmune conditions (18). There are currently no clear clinical guidelines beyond the use of benzodiazepines and ECT. In the absence of prospective and randomized data, the team took into account the patient's age, symptom severity, and mother's preference.

Conclusions

Management of anti-NMDA receptor encephalitis focuses on immunotherapy and tumor resection, which leads to the resolution of neuropsychiatric symptoms. We point out the importance of recognizing catatonic symptoms in patients with autoimmune encephalitis. In this case, catatonia remained even after therapy for encephalitis was started and after seizure frequency improved. Catatonia was confirmed via benzodiazepine challenge. The patient's mother strongly opposed benzodiazepines and a number of other treatments. As no other family could be found, the treatment team decided to respect the mother's wish and seek a treatment she would accept on the patient's behalf.

Based on the involvement of the glutamatergic system in the pathophysiology of catatonia associated with autoimmune encephalitis, we started topiramate. We observed that catatonia lagged behind the cognitive state improvement, improving only after the addition of topiramate. Future work should iteratively deploy appropriate symptom scales throughout the clinical course to more concretely assess change over time in cognitive state and catatonia separately.