Psychiatric Symptoms of Patients With Anti-NMDA Receptor Encephalitis

Objective: We conducted this study to analyze the clinical characteristics of the psychiatric symptoms of patients with anti-NMDAR encephalitis. Methods: A retrospective study of anti-NMDAR encephalitis in China was performed. The clinical characteristics of the psychiatric symptoms, the relationship between the antibodies titers and clinical characteristics of patients with anti-NMDAR encephalitis were determined. Results: A total of 108 patients with a definitive diagnosis of anti-NMDAR encephalitis were included in this study. 103 patients (95%) developed one or several psychiatric symptoms. The comparison of the high titer group and the low titer group showed that more patients presented psychiatric symptoms as the initial symptom in the high titer group (P = 0.020), the prevalence of the symptoms such as depressive, catatonic, and central hypoventilation were also higher in the high titer group than the low titer group (P = 0.033, 0.031 and 0.006, respectively). Meanwhile, more patients received a combination treatment of IVIg and corticosteroids in the high titer group than the low titer group and patients in high titer group were prescript with anti-psychiatric drugs more often than the patients in low titer group (P = 0.026 and 0.003, respectively). Conclusions: Psychiatric symptoms are the most common clinical characteristics of patients with anti-NMDAR encephalitis. Patients with higher antibodies titers more often presented with psychiatric symptoms as the initial symptom, and showed a more severe clinical feature. Screening for the anti-NMDAR antibodies is essentially important in patients who present psychiatric symptoms with or without other neurological symptoms.

serum titers, whereas those who did not improve maintained high titers in serum and cerebrospinal fluid (CSF) (2). Although immunotherapies have been suggested to be the first-line treatment strategy of patients with anti-NMDAR encephalitis, and 79% patients can reach a good outcome (3), anti-psychotic therapies may still be implanted due to the prominent psychiatric symptoms. Many of the studies to date that have looked at treatment algorithms for anti-NMDR encephalitis have focused on immunotherapy, with only a few reports looking at treatment of the psychiatric manifestations of the disease. Therefore, we conducted this study to further analysis the clinical characteristics of the psychiatric symptoms of patients with anti-NMDAR encephalitis, as well as the relationship between the antibodies titers and clinical characteristics, especially psychiatric symptoms, and the treatment strategy of patients with anti-NMDAR encephalitis.

Patients
Patients who presented with psychiatric symptoms, seizures or focal neurological signs were tested for the presence of NMDAR antibodies in serum or CSF sample at the West China Hospital of Sichuan University between June 2011 and April 2016. All patient who tested positive for anti-NMDAR antibodies in CSF or/and serum were included in this study and patients lacking key clinical data or suspecting of virus encephalitis or other infectious encephalitis were excluded. The diagnosis of anti-NMDAR encephalitis is consistent with previous studies (5,6). Clinical information was obtained by the authors or referring physicians at symptom onset and at regular intervals during the course of the disease. Fifty-one patients have been partly reported (5). We systemically reviewed the psychiatric symptoms of each patient in the following category: aggression, depressive, catatonic, disorganized, anxious, psychotic (hallucination or delusions), manic, suicidal, and insomnia. The diagnosis of each category meet the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) (7,8).
The outcome was assessed using the modified Rankin Scale (mRS) at the last visit (9,10). A diagnosis of anti-NMDAR encephalitis was considered to be definitive when (1) encephalitic signs such as psychiatric symptoms, seizures, or focal neurological signs were shown; (2) anti-NMDAR antibodies were detected in CSF or/and serum. Central hypoventilation was considered if the patient needs respiratory support such as mechanical ventilation. Relapse was defined as the onset or worsening of symptoms at least 2 months after improvement.

Subgroup Analysis
Based on the titer of the antibodies in the serum and CSF, we divided all the patients into two groups: low titer group and high titer group. The psychiatric symptoms were compared between these two groups. The antibody titers ≧1:32 in CSF / ≧1:100 in serum were considered high titer, and the antibody titer <1:32 in CSF/ <1:100 in serum were considered low titer.

Antibodies Study
This study was approved by the Research Ethics Committee of Sichuan University. Written informed consent was obtained from each subject. The serum and /or CSF samples of all patients were sent to two institutions (Oumeng Biotechnology Corporation, or Peking Union Medical College Hospital, Beijing, China) to detect antibodies against NMDAR, contactinassociated protein 2 (CASPR2), GABAR B1/B2, AMPA receptors 1/2 and leucine-rich glioma-inactivated protein 1 (LGI1). The data that support the findings of this study are available from the corresponding author upon reasonable request. Samples were classified as positive or negative by indirect immunofluorescence using EU 90 cells according to previous study (5).

Statistical Analyses
Statistical analyses were performed using SPSS version 20.0. We performed a univariate analysis in continuous variables such as age. Gender and different clinical symptoms were analyzed as categorical variables. The independent t-test or one-way analysis of variance (ANOVA) was used for continuous variables, and the Chi-square test or Fisher's exact test was used for categorical variables. When counts of zero cells were recorded, odds ratios (ORs) were calculated using Haldane's modification, which adds 0.5 to all counts to accommodate possible zero counts (11). P-values <0.05 (two-sided) were considered to be significant.

RESULTS
A total of 108 patients with a definitive diagnosis of anti-NMDAR encephalitis were included in this study. Seventy-eight patients had positive antibody results in both CSF and serum, and 30 patients only had CSF positive antibody results. The test for other autoimmune antibodies, such as antibodies against AMPA receptors 1/2, CASPR2, LGI1, and GABAR B1/B2 were negative. The demographic and general clinical characteristics are summarized in Table 1. Sixty-three patients were female (58%), and the mean age was 27.1 years (range: 9-71).

Clinical Symptoms
Sixty-two patients (57%) presented with psychiatric symptoms as the initial symptom, including mood alteration (anxious or depressive), aggression, manic, delusion, and visual or auditory hallucinations. Forty-two patients (39%) presented with neurological symptoms, such as seizure, movement disorder, and speech disturbances, as the initial symptom.

Subgroup Analysis-Low Titer Group Verses High Titer Group
The results of the comparison between the low titer group and high titer group shown in Table 2 and Figure 1. In total, 58 patients were included in the low titer group and 50 patients were included in the high titer group. Twenty-seven patients presented psychiatric symptoms as the initial symptom in the low titer group while 35 patients in the high titer group presented psychiatric symptoms as the initial symptom (P = 0.020, We also compared the treatment strategies between the high titer group and low titer group. And we found that patients in high titer group were prescript with anti-psychiatric drugs more often than the patients in low titer group (P = 0.003, OR = 0.168, 95% CI: 0.046-0.616), but the frequencies of different types of anti-psychiatric drugs were not significantly different. Meanwhile, more patients received a combination treatment of IVIg and corticosteroids in the high titer group than the low titer group (P = 0.026, OR = 0.415, 95% CI: 0.190-0.906). In addition, the relapse rate, the mean mRS and the frequency of the residual symptoms at the last visit were not shown significant difference between the two groups.

Treatment Strategies
As for the treatment strategies, 105 out of 108 patients (97%) received immunotherapy, including 49 patients treated with intravenous immunoglobulin (IVIg, 0.4 g/kg per day for 5 days) once or several times, seven patients were treated with intravenous methylprednisolone (1 g/day for 5 days) alone, 46 patients received a combination treatment of IVIg and intravenous methylprednisolone, one patient received a combination treatment of IVIg, intravenous methylprednisolone and plasma exchange, and two patients received IVIg, intravenous methylprednisolone and a second-line therapy (one with cyclophosphamide, one with rituximab). Eighty-nine patients (82%) were prescript drugs to control the psychiatric symptoms, including 37 patients with anti-psychotic drugs alone, four patients with anti-anxiety drugs alone, 38 patients with a combination of anti-psychotic drugs and anti-anxiety drugs, one patient with a combination of anti-psychotic and anti-depressive drugs, 9 patients with a combination of anti-psychotic, anti-anxiety and anti-depressive drugs. The antipsychotic drugs include typical (haloperidol and tiapride) and atypical (risperidone, olanzapine, quetiapine, and clozapine). Three patients had electroconvulsive therapy (ECT) isolated or combined with other anti-psychotic drugs. Twelve patients underwent tumor resection.   As for the side effect of the antipsychotics, three patients had extrapyramidal symptoms, one patient had tardive dyskinesia, two patients had decreased consciousness (somnolence) and one patient had liver function damage after using antipsychotics. However, as all patients were prescript with other drugs at the same time, the side effects may also due to the use of other drugs and may also reflect the natural progression of anti-NMDA receptor encephalitis.

DISCUSSION
The present study not only further proved that Chinese patients with anti-NMDAR encephalitis might have a lower incidence of tumor and a relatively better outcome compared with studies from other countries and regions, which is consistent with our previous study (5), but also reveal several features mainly focus on the characteristic and treatment of the psychiatric symptoms of patients with anti-NMDAR encephalitis.
Psychiatric symptoms are the most common clinical characteristic of patients with anti-NMDAR encephalitis; the incidence is around 65-80% (2,9,12). Antibodies against NMDAR may be associated with psychiatric symptoms for several reasons. Studies have shown that dysfunction of NMDAR and the glutamatergic system may be associated with the pathogenesis of schizophrenia, as antagonists of NMDAR, including phencyclidine and ketamine, have been shown to induce psychotic symptoms (positive and negative) and behavioral and cognitive impairments similar to those observed in patients with schizophrenia, a finding which suggests that NMDAR hypofunction may lead to secondary dopaminergic dysregulation (13)(14)(15). Isolated psychiatric episodes can also be found in patients with positive anti-NMDAR antibodies, in a prevalence of ∼4%, either at disease onset or relapse (16,17). And patients with isolated psychiatric symptoms or milder presentations do not necessarily progress to more severe multi-symptom stage, despite prolonged periods without treatment (17).
Recently, several studies explored the relationship between anti-NMDAR antibodies and well-defined psychiatric disorders, such as schizophrenia, major depressive disorder, and borderline personality disorder (18)(19)(20)(21). Although it is reported that several types of serum NMDAR antibodies were present in 9.9% of acutely ill patients who were initially diagnosed with schizophrenia (21), the antibody subtype profile of these patients differed from those of non-schizophrenic anti-NMDAR encephalitis patients, and the frequency of antibody positivity in these patients was similar to that observed in the controls (22)(23)(24). Kawai et al. reviewed a group of patients with mood disorder and found that four out of 13 patients whose CSF was obtained had positive anti-NMDAR antibodies. And all of the four patients developed some neurological symptoms during the clinical course (25). This study presented the relationship between psychiatric disorders (mood disorder) and anti-NMDA receptor antibodies from a different perspective. Positive anti-NMDA receptor antibodies in patients with psychiatric symptoms usually indicate neurological disease rather than purely psychiatric disorders, and the treatment strategy also need to be changed according to the underline etiology.
In the present study, we divided the patients into two groups according to the antibodies titers and found the patients with higher antibodies titers more often presented psychiatric symptoms as the initial symptom and showed a more severe clinical feature. More patients in the high titer group presented depressive and catatonic symptoms, and more patients had central hypoventilation which may need mechanical ventilation or tracheotomy. Previous studies have shown that depressive and catatonia are the most common psychiatric symptoms in patients with anti-NMDAR encephalitis (26). As anti-NMDAR encephalitis is an IgG antibody mediated autoimmune disorder, the severity of the disease is closely related to the titers of the antibodies. Patients who improved had a parallel decrease of serum titers, whereas those who did not improve maintained high titers in CSF and serum (2).
From clinical practice we noticed that patients with a higher titer might have a more severe clinical impression, especially in patients who presented catatonic/stupor and patients who need mechanical ventilation. Although catatonia has commonly been ascribed to schizophrenia, it is more commonly seen in affective disorders and medical and neurological disorders, and anti-NMDAR encephalitis is considered to be an autoimmune type of catatonia. (27). The prevalence of catatonia is estimated to be 10-25% in the mixed inpatient populations of psychiatric institutions, however, there is a lack of systematic investigations on the presentation of catatonia in different patient groups (28,29). It has been proposed that catatonia results from dysregulation in the glutamate, GABA-A, and dopamine pathways (30), and the NMDA receptor antagonist ketamine elicited catatonia-like signs when administered in healthy subjects (31,32). And hypoventilation can be a results of the disruption of NR1 subunit of the NMDA receptor (33).
Previous studies showed that anti-NMDAR encephalitis often has a good outcome if appropriate treatment such as immunotherapy or tumor removal are carried out in time (3), however, as high as 45% of our patients had residual symptoms such as memory deficits, speech disturbances and especially psychiatric symptoms, which may still affect the quality of life of these patients. It is suggested that around 25% of the patients with anti-NMDAR encephalitis are left with memory, cognitive, and motor deficits (34). The mild deficits symptoms such as poor attention, behavioral disinhibition, and sleep dysfunction during the recovery period may be due to the frontallobe dysfunction (2). Notably, although the frequency of the depressive and catatonic symptoms were significantly different due to the different antibodies titers in the early stage of the disease course, the frequency of the residual symptoms was no significant difference between the low titer group and high titer group, which suggests a relative good treatment results despite the former severe symptoms presented in the high titer group.
Although immunotherapies have been suggested to be the first-line treatment strategy of patients with anti-NMDAR encephalitis, and 79% patients can reach a good outcome (3), anti-psychotic therapies may still be implanted due to the prominent psychiatric symptoms, and a small portion of the patients only had psychiatric symptoms (17). Many of the studies to date that have looked at treatment algorithms for anti-NMDR encephalitis have focused on immunotherapy, with only a few reports looking at treatment of the psychiatric manifestations of the disease (27,35). No specific guidelines exist for treatment of psychiatric symptoms in this setting. Our results showed that eighty-nine patients (82%) were prescript drugs in order to control the prominent psychiatric symptoms. Previous reports suggested that ECT may help stabilizing the psychiatric symptoms of patients with anti-NMDAR encephalitis (27). But some suggested that use of highly sedating medications such as quetiapine, chlorpromazine, valproic acid, and benzodiazepines; high-potency antipsychotics like haloperidol may exacerbate motor symptoms in patients with anti-NMDAR encephalitis (36). As all of our patients were prescript with anti-psychiatric drugs as an additional therapy to the immunotherapy, and the categories and the dosages of the drugs are too various, it is difficult to further confirm whether these anti-psychiatric drugs is truly effective or not. Further studies are needed to analyze the underlying mechanisms and effectiveness of the antipsychiatric drugs in the setting of patients with anti-NMDAR encephalitis, and if effective, a standard treatment algorithm is also required.
There are some limitations in the present study. Given the retrospective nature of this study, there may have possible selection bias. All patients were recruited from the department of neurology, therefore, some of the patients who only presented with psychiatric symptoms may be underestimated and the evaluation of some psychiatric symptoms may be insufficient. And the prevalence of memory deficits may also be underestimated because patients with severe psychosis or cognitive disorders at the early stage of the disease cannot be evaluated. Meanwhile, the psychiatric symptoms were evaluated at the first admission and during the clinical course (when patients can cooperate) in the present study, so the evaluation time point of each patient may be different.
In conclusion, psychiatric symptoms are the most common clinical characteristics of patients with anti-NMDAR encephalitis. Patients with higher antibodies titers more often presented with psychiatric symptoms as the initial symptom, and showed a more severe clinical feature. Screening for the anti-NMDAR antibodies is essentially important in patients who present psychiatric symptoms with or without other neurological symptoms.

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the article/supplementary material.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by This study was approved by the Research Ethics Committee of Sichuan University. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

AUTHOR CONTRIBUTIONS
WW wrote the main manuscript, analyzed the data, made the Tables 1, 2, and Figure 1. WW, LZ, X-SC, and LH collected the data. DZ and J-ML came up with the main ideas, lead the research group, and arranged the work of all authors. All authors reviewed the manuscript.

FUNDING
This study was supported by the National Natural Science Foundation of China (Grant Nos. 81571272 and 81801285). This two funds supported studies related to encephalitis and epilepsy. The funds (Grant No. 81571272) will pay for the open access publication fee.