key: cord-1012554-lxsdqwkv
authors: Parwani, Jashank; Ortiz, Juan Fernando; Alli, Ammar; Lalwani, Ayushi; Ruxmohan, Samir; Tamton, Hyder; Cuenca, Victor D; Gonzalez, Dina; Anwer, Fatima; Eissa-Garcés, Ahmed; Alzamora, Ivan Mateo; Paez, Maria
title: Understanding Seizures and Prognosis of the Extreme Delta Brush Pattern in Anti-N-Methyl-D-Aspartate (NMDA) Receptor Encephalitis: A Systematic Review
date: 2021-09-21
journal: Cureus
DOI: 10.7759/cureus.18154
sha: e02b5a145a00a7922ca17bbba19049b8d65510a7
doc_id: 1012554
cord_uid: lxsdqwkv

Anti-N-methyl-d-aspartate (NMDA) receptor encephalitis (ANMDARE) is an autoimmune disorder with neurological and psychiatric features. The disease presents with a viral prodrome, followed by psychiatric manifestations. In the next phase, movement disorders or/and seizures occur. Finally, in the last phase, there is a decrease in the level of consciousness. Central hypoventilation and autonomic dysfunction can occur. Recently a unique EEG (electroencephalogram) pattern has been associated with anti-NMDA receptor encephalitis, the extreme delta brush (EDB). Although the association of the EDB with ANMDARE is known by the medical community, its significance is mainly unknown. A systematic review on NMDARE is also scarce. We decided to conduct a systematic review on this topic to consolidate the knowledge and establish the importance of the EDB as a prognostic factor. To conduct this systematic review, we used only studies conducted in humans, written in English, and published in the last 20 years. We used PubMed as a database and searched the following search terms: ("NMDA encephalitis"[Title/Abstract] AND "Epilepsy"[Title/Abstract]) OR (NMDA encephalitis"[Title/Abstract] AND "seizures" [Title/Abstract]) OR ("NMDA encephalitis"[Title/Abstract] AND "extreme delta brush"[Title/Abstract]). The protocol used for this systematic review was the Meta-analyses Of Observational Studies in Epidemiology (MOOSE) protocol, and to analyze the bias of the studies, we used the ROBINS-1 tool. Eight studies were collected from our search strategy. Our data pulling showed that seizures were present in 178/249 (71.48%) patients. Status Epilepticus was reported in 29/96 (30.20%), and the EBD was seen in 30.89% (55/178) patients with seizures. The range of EDB was 5.9%-33% among the studies. Because the sample size was small, the statistical power was decreased. We had a low overall risk of bias. The wide range in the results could be related to the timing of the EEG recording. EDB was associated overall with increased length of hospital stay, increased ICU admission, and incidence of status epilepticus. The etiology of the EDB remains mainly unknown. However, it has been postulated that in NMDAR encephalitis, there is a disruption of the rhythmic neuronal activity. When antibodies block/target the NMDAR, the rhythmic neuronal activity is disrupted, leading to the unique EDB pattern. Another theory suggests that delta activity is caused because of focal abnormalities in the brain, and the superimposition of the beta waves is related to the alterations of the NMDA receptors.

Anti-N-methyl-d-aspartate (Anti-NMDA) encephalitis is an autoimmune disorder with neurological and psychiatric features [1] . The incidence of the disease is 0.15 per 100,000 persons per year [1] . The condition affects both sexes, with a 4:1 ratio in favor of females, and it has a bimodal age distribution, with children (as young as eight months) and adults (20 to 40 years) being the most affected. [1] . Although patients tend to have a longer hospital stay, ICU admission, and hospital costs than in other encephalitides, they are more likely to have a full recovery or recovery with minor sequelae (75%) than severe CNS deficits or death (25%) [2, 3] . Mortality from ANMDARE ranges from 2.7% to 11.45% [4] .

The NMDA receptor is a glutamate-gated ion channel that structurally contains NR1 and NR2 subunits and plays a critical role in synaptic plasticity and memory formation, primarily in frontotemporal, limbic and hypothalamic areas [5, 6] . Antibodies to this receptor have been associated with tumors, mainly with ovarian teratoma (50% of female patients older than 18 years) and testicular germ cell tumors (male) [2] . About 20-27% of patients with herpes simplex encephalitis (HSE) develop anti-NMDA antibodies after 2-16 weeks, a phenomenon known as "post-HSE neurological relapsing symptoms" and is related to CNS infections [7] . Cases of anti-NMDA encephalitis have now been reported in patients with recent or current COVID-19 infection [8] . Possible triggers (e.g., tumors or infections) may explain the pathogenesis of the disease [2] .

There are several phases in ANMDARE. It begins with a prodromal phase characterized by headache, fever, or a week-long viral-like process in 70% of patients [1, 5] . The illness phase is characterized by psychiatric symptoms such as psychosis, agitation, hallucinations, mania, and seizures (the latter can appear at any time during the disease) [1, 5] . If the receptor antagonism persists, a third phase develops (weeks-months) which consists of orofacial dyskinesias, choreoathetosis, dystonia, rigidity, catatonia, and mutism [1, 5] . Finally (months-years) it may present with executive function deficits, memory deficits, and decreased level of consciousness. Although psychiatric symptoms predominate in adults and older adolescents, in children, movement disorders and neurological symptoms are more common. [1, 5] .

The EDB is a pattern usually seen in premature infants. Besides being found in infants, it seems to be unique and specific in anti-NMDAR encephalitis [9] . The pattern has synchronous and symmetric 1-3 Hz waveforms of 1-3 Hz with a superimposed burst of rhythmic 12-30-Hz activity, which is present continuously on the EEG recording [9] . The pattern is usually seen in frontotemporal regions. This pattern usually disappears after treatment with anti-epileptics, but in some cases may persist for several months [10] .

While the association of the EDB with ANMDARE is known by the medical community, its significance is mainly unknown. A systematic review on NMDAR is also scarce. We decided to conduct a systematic review on this topic to consolidate the knowledge and establish the importance of the EDB as a prognostic factor.

We carried out a systematic review using Meta-analyses Of Observational Studies in Epidemiology (MOOSE) protocol [11] .

For the Systematic review, only observational studies were included. The studies had to be conducted on humans, written in English, published in the last 20 years. After screening the studies, we only included papers with one of the following characteristics:

(1) Population: Patients diagnosed with NMDAR encephalitis (2) Intervention: none.

(3) Comparator: Presence or not of extreme delta brush.

(4) Outcomes: Functional outcome, mortality.

We used PubMed as a database for the systematic review. The search was conducted between July 1 2021 and September 1 2021. We used an advanced search strategy with the following terms: ("nmda encephalitis" [Title/Abstract] AND "Epilepsy"[Title/Abstract]) OR ("nmda encephalitis"[Title/Abstract] AND "seizures" [Title/Abstract]) OR ("nmda encephalitis"[Title/Abstract] AND "extreme delta brush"[Title/Abstract]) Data Extraction and Analysis:

We collected the following information from each paper: 1) Autor(s); 2) year of publication ; 3) Patients characteristics); 4) Use of antiepileptics; 5) EEG changes, including the EDB; 6) Prognosis and scoring.

Bias Assessment:

We used the ROBINS-1 tool for bias analysis of this systematic review [12] . Figure 1 shows the results of the study using a Preferred Reporting Items for Systematic Reviews and Meta- 

We found six observational studies discussing the extreme delta brush and seizures on NMDA encephalitis. Table 1 discusses the main findings in these studies [4, 10, 13, 14, 15, 16] . 2021 

The most common EEG pattern that was noted was diffuse slowing (n= 20)

followed by generalized rhythmic delta activity (n= 9), focal spikes, and slowing (n= 8 each). EDB was seen in only 3 EEGs. MRI abnormalities were more common in patients without EDB than those with EDB. No differences were observed in evolution after 6 months. No patients had reported non provoked seizures in the follow-up, but 1 patient (6%) had a persistent abnormal EEG with focal epileptiform activity.

Espinola-Nadurille et al. [13] (2019)

Status epilepticus was more frequent in patients without catatonia. Mortality was present in 10% of the total sample. It was associated with status epilepticus and was less frequent in the catatonia group.

Schmitt et al. [14] (2012)

None of the patients with EDB responded clinically or electrographically, despite the use of benzodiazepines and other IV antiepileptic drugs. Patients with EDB were associated with a trend toward worse outcomes at discharge, with a mean modified Rankin Scale score of 4.0 +/-0.8 compared with 3.1+/-1. 

EEG patterns were not related to worse prognosis, but the study failed to differentiate if there between patients with EDB and without EDB.

Wang et al. The EEG patterns occurred in a definite chronological organization, with EBA appearing first, followed by EDB and then GRDA. None of these patterns was associated with seizures, but GRDA was associated with abnormal movements. In the study, EDB was not associated with epileptic seizures or status epilepticus, as reported by Veciana et al. (2015) , and was not correlated with bad outcomes.

Viswanathan (2020) [18] The mRS scores were recorded at first admission, discharge, and at last follow-up. On admission, 33 patients had mRS scores of 4-5. At last follow-up (24 months), 37 patients had mRS scores < 2. Both patients who had EDB in our study presented with alteration in consciousness and thereafter improved with treatment. After six months of treatment, both were asymptomatic. The prognostic value of EDB is not clear. 

In Figure 2 we see the prevalence of seizure and delta brush in each of the studies. Additionally, the authors mentioned other limitations: According to Aungsumart et al., their study was limited mainly due to a retrospective study design [14] .

Limitations in Wang et al.'s study were due to sample size and study design that can lead to selection bias. Additionally, this study did not focus on mild cases of anti-NMDAR encephalitis and had difficulty following up with patients [15] .

Zhang et al.'s study was limited due to selecting the sample only at Children's Hospital of Fudan University in Shanghai, including diagnosis and treatment. Besides, it has the characteristics of a retrospective study [16] .

According to Viswanathan et al., this study was limited mainly because the study design is retrospective and possibly due to no use of cEEG; they did not report enough data about the abnormalities [17] .

Limitations in Jeannin-Mayer et al.'s study were a small median of cerebral MRI performed in every patient and the lack of monitoring each patient using the EEG to identify the appearance abnormalities [18] .

Seizures are a common clinical manifestation in ANMDARE along with psychiatric manifestations and movement disorders. According to a systematic review by Gillinder The types of seizures are variable and patients may have generalized, partial, or mixed type seizures [16] . In the study by Zhang et al., who compared seizures in ANMDARE patients with and without teratomas, complex partial seizures were more common in the teratoma group [20] . There are no systematic reviews that include pooled data of EDB. In our study, it was present in 30.89% (55/178) patients with seizures.

The etiology of the EDB pattern remains mainly unknown. Additionally, defining the pattern of ictal vs interictal is also not known. The EDB pattern is a mix of beta waves with superimposed delta activity. The beta activity tends to occur in a burst, synchronized, and diffuse fashion, while the delta activity tends to be more localized frontally [14] . Schmit et al proposed that there is a disruption of the rhythmic neuronal activity in ANMDAR encephalitis. When antibodies block/target NMDAR, the rhythmic neuronal activity is disrupted which could lead to the unique EDB pattern [20] .

The term EDB is derived from its resemblance to the delta brush pattern seen in premature infants, also known as beta-delta complexes [20] . The typical neonatal delta brushes are a combination of delta frequency transients with superimposed 8−20 Hz fast activity. However, there are differences between the two. Neonatal delta brushes are usually symmetric but not synchronous and are seen less commonly in the frontal regions of the head. In contrast, the EDB pattern is often symmetric and synchronous and is typically seen broadly across all head regions [20] . Another theory proposed that delta activity is caused by focal abnormalities in the brain, and the superimposition of the beta waves is related to the alterations of the NMDA receptors [21] .

Jeanin-Mayer et al. explained additional EEG patterns, excess beta activity (EBA), and generalized rhythmic delta activity (GRDA). EBA is a pattern ranged 14-20 hertz that is also found in benzodiazepine and barbiturate use, while GRDA is present in several comatose patients. All three patterns were found frequently in ANMDARE patients. A clear chronological organization in the course of the disease was evidenced. They found that EBA appears first, followed by EDB and then GRDA with a respective median time of appearance of 10, 16.5, and 21.5 days [17] .

In all our studies reviewed, the combined prevalence of the EDB pattern was 22.08% (55/249) in all patients and 30.89% (55/178) in patients with seizures. Overall, two studies concluded worse short-term outcomes in patients with EDB but the other two studies displayed a lack of such evidence.

In the Veciana et al study, patients who suffered status epilepticus and had EDB on EEG tended to have fewer abnormal findings on MRI than patients without EDB. Additionally, patients with EDB patterns did not have a worse long-term outcome, but they needed more aggressive treatment with anesthetic coma and spent more days either in the hospital or the intensive care unit setting, thus confirming worse short-term outcomes [10] . According to Schmit et al, patients with EDB were hospitalized and monitored with cEEG longer and demonstrated a trend toward a worse outcome. These findings suggest that the EDB pattern may be a marker of more severe disease and perhaps worse outcomes [14] . On the other hand, Aungsumart et al concluded that abnormal EEG patterns were not associated with poorer outcomes. However, this study did make a direct comparison of patients with and without EDB and clinical outcomes [4] . Similarly, Jeanin-Mayer and Viswanath stated that the EDB pattern was not related to more seizures or status epilepticus, and did not correlate with worse outcomes [17] .

Both Schmit and Veciana supported the notion that patients who had EDB tended to have fewer abnormal findings on MRI than patients without EDB [10, 14] .

In terms of seizure-related outcomes, Alvarez et al. suggested that the need for treatment of seizures and movement disorders was predictive of unfavorable early outcomes. However, this was not confirmed by other studies [21] . Viswanathan LG et al. said that there was no significant difference in outcomes such as seizure recurrence, modified Rankin score (mRS) at follow-up/discharge, or relapse between groups of patients who had EEG abnormalities in the first EEG and with those who did not. According to Zhang et al., the prevalence of seizures was higher in patients under 6 years of age. The most common type was focal seizures and patients with ANMDARE tended to have repetitive seizures as compared to single seizures or status epilepticus [16] . Most patients achieved seizure freedom, so long-term use of anti-epileptic drugs may not be necessary for the pediatric population. According to the literature, adults tend to present with focal seizures and children tend to present with generalized seizures. In the Zhang study, the EDB was only seen in 11.1% of patients, which is lower than the other studies. The author suggested that the difference could be related to the timing of the EEG recording [16] .

From the other EEG patterns studied by Veciana (EBA and GRDA), neither was associated with seizures, but GRDA was associated with abnormal movements [17] .

In the Spianola et al study, neither seizures nor the EDB was related to catatonia. However, patients with Catatonia had more status epilepticus [13] .

Zhang Y et al. studied the efficacy of a newer modality, Therapeutic Plasma Exchange (TPE) in patients with severe ANMDARE (mRS score of 4-5). They found that patients in the TPE group showed greater clinical improvement at 1 and 2 months after treatment, compared to the non-TPE group. There was no significant improvement at 3, 6, and 12-month follow-ups. It also helped to decrease the dose of required AEDs [22] .

In the study of Wang et al., there was a statistically significant difference between patients who receive only first-line treatments (steroids, intravenous immunoglobulin, plasmapheresis) vs first-line plus second-line therapy (rituximab, cyclophosphamide) in the relationship of seizures or EDB. Overall, using second-line therapy significantly improved verbal episodic memory outcomes in patients with anti-NMDAR encephalitis [15] .

A variety of anti-epileptics have been used for the management of seizures in ANMDARE. These include leviteracetam, lacosamide, phenytoin, oxcarbazepine and topiramate. In certain cases, patients required pharmacologic coma with phenobarbital [10] . Patients usually (85%) attain freedom from seizures and do not need AEDs after 1 year [16] . In the study by Zhang, seizures persisted in only 2/34 (11.1%) patients at the last follow-up at 20 months [16] . This pattern may be found when the patients first present to the hospital and usually disappears after treatment with anti-epileptics, but in some cases, it may persist for several months [14] .

The etiology of the EDB remains essentially unknown. However, it has been postulated that in ANMDAR encephalitis, there is a disruption of the rhythmic neuronal activity. When antibodies block/target NMDAR, the rhythmic neuronal activity is disrupted, leading to the unique EDB pattern. Another theory suggests that delta activity is caused because of focal abnormalities in the brain, and the superimposition of the beta waves is related to the alterations of the NMDA receptors.

There was wide variability in the prevalence of EDB (6%-58%) in our sample which could be related to the timing of the EEG recording. The pooled prevalence was 30.89% (55/178) in patients with seizures. Overall, two studies concluded worse short-term outcomes in patients with EDB but the other two studies displayed a lack of such evidence. Some patients with EDB had prolonged hospital stays, increased ICU admission, and a higher frequency of status epilepticus. These findings suggest that the EDB pattern may be a marker of more severe disease and worse short-term outcomes, while long-term outcomes are not affected. Since treatment with TPE in severe anti-NMDAR encephalitis can improve short-term clinical outcomes including epilepsy, we suggest that every patient with EDB pattern may benefit from this treatment.

In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.

Spectrum of Anti-NMDA Receptor Antibody Encephalitis: Clinical Profile, Management and Outcomes

Autoimmune encephalitis: NMDA receptor encephalitis as an example of translational neuroscience

Assessment of care transitions and caregiver burden in anti-NMDA receptor encephalitis

Abnormal level of consciousness predicts outcomes of patients with anti-NMDA encephalitis

Anti-NMDA Receptor Encephalitis In Children

Autoantibodies to synaptic receptors and neuronal cell surface proteins in autoimmune diseases of the central nervous system

Analysing triggers for anti-NMDA-receptor encephalitis including herpes simplex virus encephalitis and ovarian teratoma: results from the Queensland autoimmune encephalitis cohort

Pediatric anti-NMDA receptor encephalitis associated with COVID-19

Extreme delta brush in anti-NMDAR encephalitis correlates with poor functional outcome and death. Front Neurol

EEG extreme delta brush: an ictal pattern in patients with anti-NMDA receptor encephalitis

Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group

ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions

Catatonia in patients with anti-NMDA receptor encephalitis

Extreme delta brush: a unique EEG pattern in adults with anti-NMDA receptor encephalitis

Early second-line therapy is associated with improved episodic memory in anti-NMDA receptor encephalitis

Clinical features, treatment, and outcomes among Chinese children with antimethyl-D-aspartate receptor (Anti-NMDAR) encephalitis. Front Neurol

EEG analysis in anti-NMDA receptor encephalitis: description of typical patterns

Spectrum and evolution of EEG changes in anti-NMDAR encephalitis

EEG findings in NMDA encephalitis -a systematic review

Anti-N-methyl-D-aspartate receptor encephalitis with accompanying ovarian teratoma in female patients from East China: Clinical features, treatment, and prognostic outcomes

Pharmacologic treatment and early rehabilitation outcomes in pediatric patients with anti-NMDA receptor encephalitis

Efficacy of Therapeutic Plasma Exchange in Patients with Severe Refractory Anti-NMDA Receptor Encephalitis