Temporal rank of clinical characteristics and prognosis of anti‐N‐methyl‐d‐aspartate receptor encephalitis

Abstract Objectives Early recognition and intervention of patients with the anti‐N‐methyl‐d‐aspartate receptor (NMDAR) encephalitis are important to achieve a better prognosis. The study aims to summarize the real‐world perspectives of anti‐NMDAR encephalitis patients in China via electronic medical records (EMRs). Methods Using EMRs of patients from 2013 to 2019 from West China Hospital in China, a retrospective research was conducted to demonstrate the temporary rank of clinical characteristics and disease prognosis of anti‐NMDAR encephalitis. The modified Rankin Scale (mRS) scores were used to divide the anti‐NMDAR‐encephalitis into two groups (poor prognosis vs. good prognosis). Chi‐square test and logistic regression were used to analyze factors associated with prognosis. Results Here, 78 patients were included. The most common clinical characteristics are cognitive dysfunction (86.0%) and thought disorder (86.0%). Cognitive dysfunction, thought disorder, and seizures tended to appear soon after prodrome symptoms. Logistics analysis results showed that cognitive dysfunction (OR = 4.48, 95% CI = 1.09–18.47), the score of (GCS ≤ 8) (OR = 4.52, 95% CI = 1.18–17.32), positive antibodies in serum (OR = 4.89, 95% CI = 1.19–20.13) and delay immunotherapy (OR = 4.76, 95% CI = 1.79–12.60) were risk factors of poor clinical outcomes. Conclusions There are two peaks in the development of autoimmune encephalitis (AE). The first peak is cognitive dysfunction, and the second peak is autonomic dysfunction. Cognitive dysfunction and GCS score ≤8 at admission, antibodies positive in serum, and delay immunotherapy were risk factors for a poor prognosis at discharge.


INTRODUCTION
Encephalitis is an inflammatory disease of the brain caused by an infectious pathogen or by autoimmune processes. Autoimmune encephalitis (AE) can be associated with specific autoantibodies, such as classical onconeuronal antibodies (e.g., anti-Hu,Yo,Ri,Ma2,CV2),which targets intracellular antigens and are often related to underlying cancer. They can be associated with T-cell-mediated cytotoxicity (Bien et al., 2012).
Generally speaking, onconeuronal antibodies were considered to be related with classical limbic encephalitis (LE). However, the antibodies against neuronal cell surface antigens were discovered in the studies of limbic encephalitis, referred to neuronal surface antibody syndromes (NSAS; Zuliani et al., 2019 with LE had voltage-gated potassium channel (VGKC) antibody while their onconeuronal antibody was negative. Subsequent works identified that VGKC-antibody-associated encephalopathy is a common form of autoimmune, non-paraneoplastic (Vincent et al., 2004) and reversible disease (Thieben et al., 2004). AE had gradually entered the public eye since the first case of anti-N-methyl-D-aspartate receptor (NMDAR) encephalitis was reported in 2007 (Dalmau et al., 2007).
Although the AE is rare, with an estimated incidence of 0.8/100,000 per year in the western population (Dubey et al., 2018), the influence of this disease in neurology and psychiatry is considered remarkable (Dalmau & Graus, 2018). Moreover, anti-NMDAR encephalitis is the most common form of AE (Dubey et al., 2018). Given that patients with anti-NMDAR encephalitis present a constellation of symptoms that are usually atypical and varied (Dalmau et al., 2008), this disease is difficult to be diagnosed at an early stage. Therefore, providing timely diagnosis and identified risk factors is very important (Vollmer & Mccarthy, 2016).
The anti-NMDAR encephalitis usually progresses rapidly over days or weeks, usually starting with atypical psychiatric symptoms (e.g., alter mood, memory deficit or sleep disturbance) or prodrome symptoms (e.g., fever or headache). Dalmau's study found that only 23% of patients with anti-NMDAR encephalitis were initially inspected by a neurologist, while 77% were first seen by a psychiatrist (Dalmau et al., 2008). Not handling anti-NMDAR encephalitis timely can worsen psychiatric symptoms. In turn, it can lead to delay in correct diagnosis, which affects the identification by psychiatrists. Although previous researches have demonstrated that 81% of patients with anti-NMDAR encephalitis have a good prognosis (Titulaer et al., 2013), 86% of patients will have long-term neurological deficits (e.g., fatigue and emotional lability; Yeshokumar et al., 2017) and 5−11% of the anti-NMDAR encephalitis will die (Chi et al., 2017). Thus, a comprehensive understanding of what factors may affect the prognosis of anti-NMDAR encephalitis can potentially influence treatment regimens and is essential in offering a beneficial perspective to clinicians, patients, and family members.
Capturing and using clinical information to ensure a safe, high quality, and sustainable healthcare service is vital. Information from electronic medical records (EMRs) has been important to decisionmaking on the disease (Fennelly et al., 2020). EMRs can provide real-world clinical information on disease development, progression, and intervention strategies. Effectively using the EMRs contributes to reduce repetition of tests and work and promotes the safety and quality of healthcare provided (Castillo et al., 2010;O'Donnell et al., 2018).
Thus, this survey aimed to answer the following questions based on EMRs. What are the frequency and temporal rank of anti-NMDAR encephalitis? What factors can be used to predict the prognosis?
Answers from the above questions can be used as real-world evidence to timely identify and effectively manage anti-NMDAR encephalitis.

Study variables
The demographic (gender and age) is directly extracted from EMRs.
Chart review was conducted to gain data regarding incidence trend and clinical characteristics at admission. The incidence trend was calculated by the annual number of diagnosed cases of anti-NMDAR encephalitis and the time for primary diagnosis (no-anti-NMDAR encephalitis) to diagnosis correcting (anti-NMDAR encephalitis). The clinical characteristics include the score of Glasgow coma scale (GCS), altered behavior, cognitive dysfunction, disturbance of perception, thought disorder, seizure, movement disorder, sleep disturbance, F I G U R E 1 The process of data extraction emotional lability, autonomic dysfunction, and viral prodrome. The GCS score was used to assess the level of consciousness of patients at admission. Autonomic dysfunction was detected by sustained atrial tachycardia or bradycardia, orthostatic hypotension (≥20 mmHg fall in systolic pressure or ≥10 mmHg falls in diastolic pressure within 3 min of standing), hyperhidrosis, persistently labile blood pressure, ventricular tachycardia, or cardiac asystole (Dubey et al., 2017). Prodromal symptoms were defined as patients presenting the following symptoms: rhinorrhea, sore throat, diarrhea, and fever (Dubey et al., 2017). The time of "interval to immunotherapy" is defined as the time from admission to start using immunotherapy.
Antibody tests for all patients were accomplished in the same lab-

Statistics
This research was primarily descriptive, and as such, data were summarized and presented as percentages or medians with interquartile ranges as appropriate for the data types (continuous variables and categorical variables). Difference between groups (good prognosis vs. poor prognosis) is identified by the Chi-square test for categorical variables. Estimated coefficients (βs) and odds ratios (ORs) were calculated via logistic regression. p < 0.05 was considered as statistical significance. All statistical analyses were performed using R-studio 3.6. were misdiagnosed as no-AE (e.g., acute and transient psychotic disorder, schizophrenia, manic episode, and major depressive disorder) at admission, accounting for 17.95%.

Clinical characteristics
The temporal sequence where each clinical characteristic was first reported in EMRs was coded numerically as ranks 1, 2, 3, 4, 5, etc.
(1 = first, 2 = second, 3 = third, 4 = fourth, 5 = fifth, etc. Figure 4 illustrates that patients whose disease process was related to clinical characteristics shows prodrome symptoms that dominate the earliest phase of the anti-NMDAR encephalitis phase.

F I G U R E 4
Temporal sequence of the first appearance. The x-axis represents the temporal rank of each clinical characteristic. The y-axis represents the frequency that each clinical characteristic was sorted into different ranks Cognitive dysfunction, thought disorder, and seizures tended to appear soon after the prodrome symptoms. Autonomic dysfunction and movement disorder suggested a secondary and a third peak, respectively.

Logistics analysis of a poor prognosis
The logistics analysis results demonstrated that cognitive dysfunc-  Table 2.

Demographic and incidence trend of anti-NMDAR encephalitis patients
Previous studies found that nearly 80% of AE are seen in women, of which half of the cases occurred in women above the age of 18 (Dalmau et al., 2019). However, in the current research, the sex ratio for anti-NMDAR encephalitis was nearly balanced. Previous researchers have found that the median age of anti-NMDAR encephalitis patients was between 21 and 28 (Chi et al., 2017;Titulaer et al., 2013), while in this research the median age is 29. This may have contributed to our data bias. We collected data only from West China Hospital, and most children and adolescents may choose to go to a specialist hospital (e.g., West China Women's and Children's Hospital). From 2013 to 2019, the number of diagnosed anti-NMDAR encephalitis cases has increased annually, which may be associated with the antibodies' detection methods. The time for primary diagnosis (no-anti-NMDAR encephalitis) to diagnosis correcting (anti-NMDAR encephalitis) shows a downward trend, which may be associated with increased awareness of anti-NMDAR encephalitis among psychiatrists and neurologists.

Frequency and temporal rank of anti-NMDAR encephalitis patients
In 70% of patients, anti-NMDAR encephalitis starts with prodromal symptoms Luca et al., 2011).We found that 54% of patients have represented prodromal symptoms that appeared earlier Abbreviations: CSF = cerebrospinal fluid; EEG = electroencephalography; GCS = Glasgow Coma Scale; ICU = intensive care unit; MRI = magnetic resonance imaging.
relative to other symptoms. The prodromal infection may be regarded as the antigenic trigger for the inflation of anti-NMDAR-specific lymphocytes by molecular mimicry (Peery et al., 2012). The temporal rank of clinical characteristics demonstrated in this research is consistent with Irani's research (Irani et al., 2010), which showed that psychiatric and cognitive disorders appeared at an early stage. While in Gurrera's study (Gurrera, 2019), the present sample size collected from EMRs is much more extensive. Sleep disturbance may play the role as a bridge symptom in the progression of the disease. After the sleep disturbance, the second peak (autonomic dysfunction) and the third peak (movement disorder) appeared. Also, Blattner's study found that untreated sleep complaints may have adversely influenced the automatic and cognitive functions (Blattner & Day, 2020).

Factors for the prognosis of anti-NMDAR encephalitis patients
In this research, we reported 33 patients who represent a good prognosis and evaluated the risk factors of poor prognosis. The good outcomes of anti-NMDAR encephalitis in this study was 42.30%, which was much higher than 41% in the United States (Singh et al., 2015) and 38.3% in China (Mo et al., 2020). The variability may be explained by the differential medical model in different regions. Results showed that patients with cognitive dysfunction (95% CI = 1.09-18.47), seizure (95% CI = 0.98-6.18), movement disorder (95% CI = 0.82-58.34), and GCS ≤ 8 (95% CI = 1.18-17.32) were risk factors for poor clinical outcomes, while seizure and movement disorder have not achieved significant differences in logistics analysis. It is cautious that the CI is relatively wider in this study. There may still be enough precision to make decisions about factors of prognosis, as the width of the CI for an individual study depends to a large extent on the sample size (Julian Higgins, 2020) and our study have a smaller one. Some studies have found that the scores of GCS ≤8 can be regarded as a risk factor for a poor short-term (at discharge) and or long-term (3 months and 6 months after the discharge) clinical outcome (Chi et al., 2017;Mo et al., 2020). The most common reasons for ICU admission were seizure and altered mental state (Hacohen et al., 2013). Moreover, patients who were transferred to ICU often had a poor clinical outcome despite using timely immunotherapy (Mittal et al., 2016). In this study, most

Clinical implications
Anti-NMDAR encephalitis increase the economic burden on both patients and society (Cohen et al., 2019), but the management of anti-NMDAR encephalitis patients (e.g., delays in diagnosis and transportation to specialized centers) in China is still poor. Patients with "red flag symptoms" (e.g., prodromal symptoms, acute or subacute psychiatric symptoms, and seizure; Pollak et al., 2020) should be handled urgently and undergo antibodies examination. Furthermore, the CSF antibody examination should be considered in the initial diagnostic testing. Examination of only the serum is insufficient, although serum testing is more sensitive for several specific antibodies (Leypoldt et al., 2015) and titers in serum are related to a clinical outcome at discharge.
The risk factors of poor clinical outcomes can help clinicians evaluate the possible prognosis at the early stage and provide early monitoring and supportive treatment strategies. Besides, clinicians can interpret the possible clinical outcomes for patients and patients' family members to promote the therapeutic relationship.

Strengths and limitations
The current study has some potential strengths, which mainly includes: (1) the data is directly extract from the EMRs. Real-world evidence has the potential to offer useful information about the use and prognosis of a given clinical treatment within a setting most relevant to routine clinical practice in China; (2) we only included initial anti-NMDAR encephalitis patients so the effect of the medication on anti-NMDAR encephalitis was ruled out. However, our study had some limitations, including (1) the sample size of the current study was insufficient. (2) Patients with anti-NMDAR encephalitis coming from only one specific hospital might limit the generalizability of our findings; (3) we only included "definite" anti-NMDAR encephalitis patients whose anti-NMDAR was detected in the CSF in this study. Probable anti-NMDAR encephalitis was not included. (4) We only evaluated the prognosis at discharge, and cannot rule out the patients who were relapsed or dead.
There is a need for prospective and multicenter trails in the future. (5) In the retrospective study, for the clinical symptoms that if it did not appear in the EMR was scored as "absent" (Gurrera, 2019), ignoring the probably missing data.

CONCLUSION
In summary, we analyzed the temporary rank of clinical characteristics and risk factors for prognosis among anti-NMDAR encephalitis. Although the spectrum of features at admission and course of anti-NMDAR encephalitis is wide, new psychiatric symptoms and accompanied prodromal symptoms can be regarded as "red flag" for anti-NMDAR encephalitis. Cognitive dysfunction, antibodies titers in serum, GCS scores (≤8), and interval to immunotherapy (≥5 days) were risk factors for a poor prognosis.

AUTHOR CONTRIBUTIONS
Wei Zhang conceived the study, contributed to the planning, draft, revision of the manuscript. Fenfen Ge, Runnan Yang, and Yue Wang wrote the first draft of this paper. Fenfen Ge and Jingwen Jiang analyzed and interpreted the data. Fenfen Ge and Mengtong Wan revised the draft critically for important intellectual content. All authors contributed to the revision of the manuscript. All authors gave the final approval of the version to be published.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study can be acquired from corresponding author.