Our results showed a number of cases with NMDAR antibody positivity in a broader range of psychiatric disorders, such as sleep disorders and schizophrenia (group A: 3 out of 5 cases, B: 3 out of 5 cases & C: 4 out of 51 cases). Although the causative relationship between NMDAR antibody positivity and psychiatric symptoms in these patients are unknown, they exhibit unique demographic and clinical characteristics. Eight (out of 10) are female, the majority of cases are 20–30
year olds, and ovarian tumors are found in 2 patients. Most of their symptoms are resistant to pharmacological treatments but respond relatively well to mECT, the clinical characteristics often seen in psychotic symptoms associated with NMDAR encephalitis [22
NMDAR and psychiatric symptoms
Schizophrenia is a common, heterogeneous, and complex disorder with unknown aetiology [24
]. There is established evidence of NMDAR hypofunction [25
] as a central component of the functional dysconnectivity; this is one of the most accepted models for schizophrenia [26
]. Moreover, autoimmune mechanisms have been proposed to be involved, at least in subgroups of schizophrenia patients [27
]. In the last few years, a number of antibodies to neuronal cell surface antigens have been identified in cases of autoimmune encephalitis that respond to immunotherapy [29
]. Over two-thirds of patients with NMDAR antibody encephalitis have prominent psychiatric symptoms or may present to psychiatric services in the first instance [23
]. The psychiatric symptoms are those seen in schizophrenia including delusions, hallucinations, and catatonic movement disorder.
This characteristic clinical presentation resembles acute psychosis followed by a rapid decline in the level of consciousness, central hypoventilation, seizures, involuntary movements, and autonomic instability. Although anti-NMDAR encephalitis is a potentially fatal condition, if the diagnosis is made rapidly, effective treatments are available [29
In fact, the most favorable outcome occurs with tumor removal (e.g., teratoma), usually in combination with immunotherapy (IV steroids, IV immunoglobulin, or plasma exchange) [9
]. Also, a good clinical outcome was reported in patients treated with immunotherapy without tumor removal [36
]. Besides tumor removal and immunotherapy, symptomatic treatment with antiepileptic drugs and benzodiazepines may partially relieve symptoms [5
Our first three cases in group A had typical clinical pictures of anti-NMDAR encephalitis, beginning with psychiatric symptoms, followed by seizures and disturbances of consciousness (Table , group A). No tumors were found in these cases, but two of them responded well to steroid pulse treatments.
Hypersomnia and encephalitis
As far as we know, there has been no report of narcolepsy with NMDAR antibody positivity. However, two studies reported that some patients with contemporary Encephalitis Lethergica (EL) were positive for NMDAR antibodies [23
]. Ten out of twenty patients were positive and these patients predominantly fit into the dyskinetic form of EL [38
]. The five patients with the somnolent-Parkinsonian form of EL, which is considered to be the classic form of EL, were negative for NMDAR antibodies [38
These results together with the fact that 3 out of 5 narcoleptic subjects who were positive for anti-NMDAR antibodies exhibited severe psychiatric symptoms and that 8 out of 10 conventional narcoleptic subjects studied were negative for NMDAR antibodies (p
0.025, Chi-square test), show that NMDAR antibody positivity may be more specifically related with occurrences of psychiatric symptoms.
Nevertheless, it is possible that the immune mediated mechanisms are more frequently involved in narcolepsy and that these may also be responsible for their associated symptoms, and further studies are warranted.
The prevalence of NMDAR positivity in group C
The occurrence rate of NMDAR positive cases in our group C (4 out of 51 cases) is similar to that of Zandi’s report (3 out of 46 cases)[33
Since there had been no reported cases of NMDAR antibodies identified in patients with purely psychiatric disorders, Zandi et al. [33
] hypothesized that this antibody would be present in a proportion of patients with early schizophrenia, in the absence of overt seizures, movement disorders, or other neurological signs. They found 3 cases in 46 examined cases that fulfilled DSM IV criteria for schizophrenia, and the patients were tested early in the course of their illness. They also described the first case of a patient with NMDAR antibodies and a purely psychiatric presentation, that responded to immunotherapy (plasmapheresis, oral prednisolone).
The term, “atypical psychosis” has been used, especially by Japanese psychiatrists [39
] as a possible clinical entity for acute and transient psychotic disorders which cannot be easily classified as either schizophrenia or a mood disorder with psychotic features. Some of the important clinical characteristics of atypical psychosis include acute onset, emotional disturbances, psychomotor disturbances, alternations of consciousness, high prevalence in women, and oriented premorbid personality, characteristics that mirror those of our psychotic cases. These authors had suspected involvements of brain organic changes in atypical psychosis.
However, atypical psychosis, by its meaning, comprises a widely varied and poorly understood collection of disorders, and atypical psychosis was listed in DSM-III-R under the heading Psychosis Not Otherwise Specified (NOS); this does not define the nosological entity and is rather used as a residual category. Consequently, in DSM-IV, the term atypical psychosis is no longer mentioned as a synonym for this category.
While schizophrenia and affective disorders have dominated the psychiatric literature and research efforts in psychotic disorders, several other atypical psychotic conditions are emerging as significant. Included among these are psychotic disorders secondary to medical conditions [40
]. If NMDAR antibody positivity is functionally involved in pathophysiology of the disease in the patients listed in group C, the disease will fit in the category of psychotic disorders secondary to medical condition.
Together with our present study, further determination of whether the anti-NMDAR antibody plays functional roles in these patients with schizophrenia, schizo-affective disorders and atypical psychosis, is critical, since treatment choices, including immunotherapy, are different from those for classical psychosis and since the immune mediated mechanisms may also be involved more frequently in these psychotic patients.
We present cases to illustrate that anti-NMDAR encephalitis should be considered when diagnosing patients with acute psychosis, and that mECT can possibly be considered as an effective treatment option for these cases.
In our 5 out of 10 anti-NMDAR antibody positive cases, mECT was effective. In one previous case series, one patient was found to respond to ECT [41
]. Two patients have been reported with paraneoplastic catatonia and ovarian teratoma that partially improved with ECT, but full recovery was only obtained after tumor removal [9
]. Still, the remission may have been spontaneous and the temporal association between ECT and recovery may have been coincidental. Recently, Braakman [23
] reported that one patient (47 old male) deteriorated clinically in a period of 10
weeks, and recovered in a period of 3
weeks after ECT was started.
The mechanism of action of ECT remains largely unclear. Still, in animal models ECT has been shown to up-regulate NMDA receptors [42
]. This may, in part, explain the efficacy of ECT in our patients, since NMDAR was down-regulated during periods of anti-NMDAR encephalitis [34
We report remarkable recoveries of our patients following mECT. Psychosis, delusions, stupor, and catatonia rapidly disappeared. Further clinical observation or studies in patients with anti-NMDAR encephalitis are needed to determine the significance of our observations [38
Results of our study together with those by others redefine this new class of psychotic disorders positive for anti- NMDAR antibody. These include narcolepsy with psychosis, EL [38
], schizophrenia accompanied by convulsion [13
], atypical symptoms of psychosis and psychotic patients who are drug resistant but respond relatively well to mECT.
We are also aware of the limitations of the study. (1) We did not include normal controls or typical schizophrenia subjects. (2) The retrospective recollection of the clinical data may favor the description of partial or isolated symptoms of the disease because we could miss subtle signs or symptoms associated with the predominant manifestation. (3) We did not measure the CSF antibody in the majority of our cases and future prospective studies should include paired serum-CSF antibody measures.