Complete cognitive recovery in a severe case of anti-N-methyl-d-aspartate receptor encephalitis treated with electroconvulsive therapy

Abstract

Anti-N-methyl-d-aspartate (NMDA) receptor encephalitis usually presents with prominent neuropsychiatric symptoms and many patients experience cognitive sequelae. Electroconvulsive therapy (ECT) has been suggested as a part of the treatment, particularly for catatonia, but concerns that ECT may worsen the cognitive function and long-term outcome may limit its use. We present a case of anti-NMDA receptor encephalitis with severe neuropsychiatric manifestations including refractory catatonia and behavioural change. A pre-ECT neuropsychological assessment revealed dysfunction in multiple cognitive domains in spite of intensive pharmacological treatment. Twenty days after the ninth and last ECT treatment, the patient underwent the same neuropsychological tests, which showed normalised test results within all cognitive domains and no need of rehabilitation. The case demonstrates that the use of ECT in anti-NMDA receptor encephalitis with severe pretreatment cognitive dysfunction can be associated with a highly favourable cognitive outcome.

Background

In 2007, Dalmau et al identified antibodies in cerebrospinal fluid (CSF), reacting against N-methyl-d-aspartate (NMDA) receptor predominantly in the hippocampus and forebrain, from 12 women who presented with a combination of neurological and psychiatric symptoms, and in which teratomas was found after further examination. This led to the discovery of anti-NMDA receptor encephalitis.¹ The disease has subsequently been described in patients of all ages, both genders and in patients without concurrent neoplastic disease.² Anti-NMDA receptor encephalitis usually presents with behavioural changes, psychosis, catatonia, cognitive dysfunction, seizures, dyskinesis and autonomic dysfunction preceded by non-specific prodromal symptoms.^{2 3}

The treatment of anti-NMDA receptor encephalitis consists of immunotherapy and, if relevant, tumour removal.² Despite therapy, more than 75% of patients treated for anti-NMDA encephalitis experience significant cognitive sequelae of varying severity during recovery, predominantly of the domains of memory, attention and executive functions, and the severity of cognitive sequelae depends on the duration of the acute disease course.³

Electroconvulsive therapy (ECT) has been suggested as a part of the treatment in anti-NMDA receptor encephalitis, particularly for catatonia and in cases where neuroleptic and benzodiazepine treatment fail.⁴ However, fear of worsening the cognitive dysfunction may limit its use. Here, we present a case of severe anti-NMDA receptor encephalitis with prominent global cognitive dysfunction, which resolved completely after ECT.

Case presentation

A highly educated woman, in her late 30s, with no previous neurological or psychiatric history, presented at the psychiatric hospital with altered behaviour, stupor, perseveration, echolalia and delusions. Before hospital admission, the patient had 2–3 weeks with progressive symptoms of disturbances of concentration, memory and sleep, anorexia and loss of energy preceded by a short unspecific viral-like infection in the upper respiratory tract. The patient had no specific neurological symptoms, such as focal or generalised seizures, or autonomic instability at any point during the disease course, neither before nor after hospital admission. The patient was transferred to the department of neurology at the same day of hospital admission to exclude encephalitis or intracerebral tumour.

Analyses of the CSF showed lymphocytic pleocytosis leading to treatment with intravenous aciclovir for presumed viral encephalitis until microbiology studies identified no viral agents in CSF, at which point aciclovir was withdrawn. Brain MRI was unremarkable, and positron emission tomography with CT (PET-CT) showed no underlying malignity. Electroencephalography (EEG) showed an abnormal encephalopathic pattern with 2–3 Hz activity bilateral frontotemporal predominantly in the right hemisphere. Due to the clinical suspicion of autoimmune encephalitis, intravenous high-dose corticosteroids (1 g/day for 2 days followed by tapering) and immunoglobulins (0.4 g/kg/day in 5 days) were administered on hospital day 2.

On hospital day 7, the patient underwent the first neuropsychological assessment revealing cognitive and behavioural dysfunction with a dominating executive dysfunction indicating bilateral frontal affection, as well as affected working memory and memory (table 1). After days with clinical improvements, the patient developed subacute onset increasing and fluctuating agitation, aggression, anxiety, insomnia, visual and auditive hallucinations and delusions on hospital day 13. Plasmapheresis (one cycle every other day for a total of seven cycles) and intravenous high-dose corticosteroids (500 mg/day for 3 days followed by tapering) were initiated, and psychiatric symptoms were attempted treated with relevant doses of atypical (olanzapine, quetiapine) and typical (haloperidol) neuroleptics as well as benzodiazepines (lorazepam, diazepam) alone or in combination with little or no response.

Table 1.

Neuropsychological tests in a patient with anti- receptor encephalitis before and after nine bilateral ECT treatments

	Before ECT		After ECT
Cognitive test	Test scores	Comments	Test scores	Comments
Attention and working memory
Digit span forward	Max and stable five digits	Suspected impairment	Max 6 and stable five digits	Normalised score
Digit span backward	Max and stable two digits	Impaired	Max 4 and stable three digits	Normalised score
Trailmaking A	37 s, 1 error	Normal score	31 s, 0 errors	Normal score
Learning and memory
Rey’s complex figure	7/36	Impaired	24/36	Normalised score
RBANS wordlist, learning	27/40 (9 false positives)	Impaired	NA	NA
RBANS wordlist, recall	2/10	Impaired	NA	NA
RBANS wordlist, recognition	18/20	Impaired	NA	NA
Rey Auditory Verbal Learning Test, learning	NA	NA	56/75	Normalised score
Rey Auditory Verbal Learning Test, recall	NA	NA	12/15	Normalised score
Rey Auditory Verbal Learning Test, recognition	NA	NA	15/15	Normalised score
Processing speed
SDMT (Symbol Digit Modalities Test)	NA	NA	53 to 0 errors	Normalised score
Executive functions
Design fluency	23 figures, 45 repetitions	Impaired	35 figures, 1 repetition, 0 errors	Normalised score
Verbal fluency, categorical	18, 12 repetitions, 0 errors	Impaired	28, 0 repetitions, 0 errors	Normalised score
Verbal fluency, fonological	11, 1 repetition, 2 errors	Impaired	14, 1 repetition, 0 errors	Normalised score
Rey’s complex figure, copy:	23/36 in 105 s	Impaired	36/36 in 120 s	Normalised score
Trailmaking B	Not completed. 4 errors	Impaired	51 s, 0 errors	Normalised score

All scores are severely affected before ECT and normalised after ECT. Note that not exactly the same tests were used at the pre-ECT versus post-ECT session due to differences in clinical cognitive status between the two sessions, that is, the patient was due to severe cognitive deficits not able to perform more complex cognitive tests before ECT. Thus, a more complex verbal memory-test and testing of processing speed were selected after ECT in order to evaluate cognitive status in more detail.

ECT, electroconvulsive therapy; NMDA, N-methyl-d-aspartate.

On hospital day 20, the diagnosis of anti-NMDA receptor encephalitis was confirmed in further analyses of CSF. After terminated treatment with plasmapheresis and still no clinical improvement, second-line treatment with rituximab (1 g/day on day 1, 14 and scheduled 180) was initiated on hospital day 33. Olanzapine up to 40 mg/day and diazepam up to 75 mg/day was administered for psychotic symptoms and presumed catatonia with still very limited response.

The patient was transferred to the psychiatric department on hospital day 34, and on day 36 treatment with ECT began. Following the standard procedure of the department, the patient was anaesthetised with thiopental, and succinylcholine was administered for muscle relaxation. Bitemporal ECT was given with a 0.5–1.0 pulse width, a stimulus duration of 1–8 s, and a potential energy range of 50–1000 mC. The energy dose (in percent of 500 mC) at the first treatment was set to 40%. During the treatment series, no further dose adjustments were deemed necessary based on seizure quality, clinical response and side effects. A total of nine treatments, the first three en bloc (ie, for 3 consecutive days), over the following 20 days were given with excellent clinical response, which began after the first three treatments. After terminated ECT treatment, the patient started olanzapine and benzodiazepine tapering.

Outcome and follow-up

On hospital day 76—20 days after the last ECT treatment—the patient underwent the same neuropsychological tests, performed by the same clinician as the first time, which revealed normalised test results within all cognitive domains and no need of rehabilitation. The neuropsychological assessment implied remission of impaired working memory, memory and executive functions (table 1). On day 78, the patient was discharged from the psychiatric hospital. At a clinical visit 2-week post-hospital discharge, CSF was normalised, and EEG showed diffuse beta activity, but nothing focal or paroxysmal. The patient described no psychiatric symptoms and no disturbances in sleep or cognition, although she was amnestic for the period of severe catatonia and ECT. Now, 5 months after discharge, the patient continues to be well, has taken up full-time employment and experiences only light and occasional memory problems.

Discussion

Our case presents a previously healthy 38-year-old woman with isolated psychiatric and neurological symptoms of confirmed anti-NMDA receptor encephalitis. Although prompt initiation of first-line and subsequently second-line therapy in combination with antipsychotic drugs and benzodiazepines, the patient continued suffering from refractory catatonia, delirium, global cognitive dysfunction and psychotic symptoms. Once ECT began, we observed a robust clinical response and the patient was discharged after nine ECT treatments with no detectable neuropsychiatric deficits.

To our knowledge, this is the first report of an adult with a detailed description of the cognitive functions during and after anti-NMDA receptor encephalitis treated with ECT. Cognitive domains usually affected in anti-NMDA receptor encephalitis are processing speed, memory and executive functions. Cognitive deficits are prevalent at all points of recovery from anti-NMDA receptor encephalitis, but improvement in cognitive outcomes can be expected as patients recover.³

Our case present with cognitive dysfunction as typically described in anti-NMDA receptor encephalitis: affected memory and executive function; processing speed was not assessed at the first neuropsychological assessment but was found normal at the follow-up. After treatment cognitive functions were normalised within all cognitive domains.

The mechanism(s) by which ECT works continues to be unresolved. ECT has powerful effects in phenomenologically different conditions such as psychotic depression, catatonia and delirium. The effects of ECT observed in anti-NMDA receptor encephalitis may relate to the classic theories about ECT mechanism, such as neurotrophic changes, increases of seizures threshold or normalisation of neuroendocrine signalling,⁵ or—perhaps more likely—to the reversal of disturbances in functional brain connectivity, which has recently been demonstrated in anti-NMDA receptor encephalitis.^{6 7}

There are no consensus guidelines on the management of anti-NMDA receptor encephalitis in general or its psychiatric symptoms specifically. Treatment strategies are mostly based on retrospective observational studies and expert opinions.⁸ In case reviews, the use of ECT in anti-NMDA receptor encephalitis has been associated with a rapid improvement on symptoms, particularly for catatonia,⁴ indicating that ECT could help with faster recovery and shorter time spent in the hospital, and may even be lifesaving.⁹ However, ECT has known cognitive side-effects, generally limited to the first few days after the treatment series and usually seen in the domains of anterograde memory, working memory, processing speed and executive functioning,¹⁰ which may render clinicians reluctant to use this treatment modality in a disorder that includes severe cognitive dysfunction in its pretreatment symptomatology. Our case underlines the importance of considering ECT in the treatment of anti-NMDA receptor encephalitis and also clearly illustrates that the cognitive side-effects of ECT can be completely outweighed by the beneficial cognitive effects of the disruption of treatment-resistant catatonia, psychosis and delirium that may occur in severe cases of anti-NMDA receptor encephalitis. The case raises the question whether ECT should not only be a ‘last resort’ after antipsychotic and benzodiazepine treatments fail, but may have a more causal beneficial influence on disease progression and thereby should have a place earlier in the treatment course of severe cases of anti-NMDA receptor encephalitis. This hypothesis would obviously require further study.

Patient’s perspective.

I have no recollection of my hospital admission. I only remember the period up until my admission and the last weeks when I was admitted to the psychiatric ward. The first 2 months after I was discharged I felt heavy and tired in my head after too much activity or noise. Also had a hard time being with my kids for a full day without breaks. The tiredness disappeared after a break with calm/silence. After 2 months at home, I started working part-time and have escalated to full-time over the past 2 months.

Today I find that I forget things faster, and I have to note down more details during the day than before. However, it does not affect me significantly and I’m properly also more aware of it than before. I can still experience tiredness / headaches by very loud noises or many people for a long time. Otherwise, I have no disability of my illness.

Learning points.

• Anti-N-methyl-d-aspartate (NMDA) receptor encephalitis is often associated with neuropsychiatric symptoms including cognitive dysfunction that may be refractory to treatment.

• The use of electroconvulsive therapy (ECT) in anti-NMDA receptor encephalitis with severe pretreatment cognitive dysfunction can be associated with a highly favourable cognitive outcome.

• ECT should possibly have a place earlier in the disease course in complicated cases of anti-NMDA receptor encephalitis with severe neuropsychiatric manifestations.