Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Dec 1.
Published in final edited form as: Epilepsy Res. 2021 Oct 22;178:106787. doi: 10.1016/j.eplepsyres.2021.106787

Electroencephalography characteristics to predict one-year outcomes in pediatric anti-NMDA receptor encephalitis

Jenny Lin a,b,1, Kathryn Elkins a,b,1, Sonam Bhalla a,c, Satyanarayana Gedela a,c, Ammar Kheder a,c, Guojun Zhang a,c, Leah Loerinc b, Laura Blackwell a,d, Robyn Howarth a,d, Grace Gombolay a,c,*
PMCID: PMC9051600  NIHMSID: NIHMS1800172  PMID: 34715485

Abstract

Background:

Electrographic characteristics (extreme delta brush, posterior dominant rhythm and slow waves) may predict outcomes in anti-NMDA receptor encephalitis (NMDARE). However, whether changes in EEG sleep architecture predict outcomes are unknown. We examine electrophysiological characteristics including sleep architecture in a pediatric NMDARE population and correlate with outcomes at one year.

Methods:

Retrospective chart and EEG review was performed in pediatric NMDARE patients at a single center. Patients with first EEGs available within 48 h of admission, prior to treatment, and one-year follow-up data were included. EEGs were independently reviewed by two epileptologists, and a third when disagreement occurred. Clinical outcomes included modified Rankin scale (mRS) at one year.

Results:

Nine patients (6 females) (range 1.9–16.7 years) were included. Five of nine patients had loss of posterior dominant rhythm (PDR) and three of nine patients had absent sleep architecture. Loss of PDR correlated with a worse mRS score at one year (2.8 versus 0.5, p = 0.038). Loss of PDR and loss of sleep architecture was associated with increased inpatient rehabilitation stay and in higher number of immunotherapy treatments administered. In multivariate analysis, absence of sleep architecture (p = 0.028), absence of PDR (p = 0.041), and epileptiform discharges (p = 0.041) were predictors of mRS at one year.

Conclusions:

Loss of normal PDR, absence of sleep architecture, and epileptiform discharges are associated with worse outcomes at one year which has not been reported before. EEG characteristics may help prognosticate in NMDARE. Larger studies are needed to confirm these findings.

Keywords: Anti-NMDA receptor encephalitis, Delta brush, Electroencephalography, EEG, Pediatric, Sleep

1. Background

Anti-NMDA receptor encephalitis (NMDARE) is an autoimmune encephalitis (AE) caused by antibodies that target the NMDA receptor, resulting in a characteristic neuropsychiatric syndrome. Neuropsychiatric symptoms include seizures, psychosis, movement disorders, speech regression, autonomic instability, and central hypoventilation (Dalmau et al., 2007; Dalmau et al., 2008). NMDARE occurs in all ages, with one-third of patients younger than 18 years of age (Titulaer et al., 2013). Overall, mortality occurs in up to 11% of cases; 20% of patients will not respond to immunotherapy and will be neurologically devastated. Patients who do respond to immunotherapy usually have long term sequelae, including neurocognitive dysfunction (Titulaer et al., 2013; Hacohen et al., 2014; Linnoila et al., 2019; de Bruijn et al., 2018). Thus, identifying features to understand and predict prognosis may help understand the disease better and improve outcomes.

Electrographic findings in NMDARE are variable (Schmitt et al., 2012; Qu et al., 2020; Haberlandt et al., 2017). The most commonly described Electroencephalographic (EEG) features in pediatric NMDARE include background slowing (77–100%), rhythmic delta activity (59–100%), and interictal epileptiform discharges (56–100%) (Haberlandt et al., 2017; Qu et al., 2019; Sands et al., 2015; Gitiaux et al., 2013;Nosadini et al., 2015). Rates of extreme delta brush (EDB) in pediatric NMDARE have been reported as between 6% and 53% (Qu et al., 2020; Haberlandt et al., 2017; Sonderen et al., 2018) and rates of electrographic seizures range from 11% to 52% with status epilepticus in 0–50% (Haberlandt et al., 2017; Qu et al., 2019; Sands et al., 2015; Gitiaux et al., 2013; Nosadini et al., 2015). EEG characteristics may predict outcomes in NMDARE (Schmitt et al., 2012; Sonderen et al., 2018; Mo et al., 2020). While changes in sleep architecture have been reported in autoimmune encephalitis (Blattner et al., 2019), whether electrophysiological changes in sleep architecture are associated with outcomes in NMDARE are unknown. Here we report on electrophysiological characteristics including sleep architecture in a pediatric NMDARE population and correlate with outcomes at 1 year following diagnosis.

2. Design/Methods

Institutional Review Board approval was obtained for this study. Retrospective review of charts and EEGs was performed in pediatric NMDARE patients who presented to Children’s Healthcare of Atlanta between January 1, 2017 and April 1, 2021. NMDARE was diagnosed in patients with positive cerebrospinal fluid (CSF) anti-NMDA receptor antibody testing from Mayo Laboratories and if patients had at least one of six clinical symptoms (seizures, psychosis, movement disorders, speech regression, autonomic instability, and/or central hypoventilation) (Graus et al., 2016). Patients with first EEGs available and 1 year follow up were included. All first EEGs occurred within 48 h of admission to the hospital and prior to treatment. Fourteen NMDARE patients were identified and five patients were excluded due to the following reasons: one had no EEGs available for review; one had EEGs available but the first EEG obtained within 48 h of presentation was unavailable; one had prior herpes simplex viral encephalitis, and two patients did not have one year follow up data available. The first EEG was reviewed, which was 24 h of the full recording in all patients except for one patient with a routine EEG. EEG characteristics included extreme delta brush (EDB), posterior dominant rhythm (PDR), generalized rhythmic delta activity (GRDA), epileptiform discharges, electrographic seizures, and presence of sleep architecture (see Fig. 1). Background was assessed at onset. Sleep was marked as present if N1, N2, N3 or REM was present. N1 was defined if vertex waves were present; N2 was defined if sleep spindles or K complexes were present; N3 was defined if slow waves were identified, and REM was identified if central saw tooth waves were present.

Fig. 1.

Fig. 1.

Open in a new tab

Representative electroencephalography (EEG) characteristics from anti-NMDA receptor encephalitis patients in standard bipolar montage with sensitivity at 7 microvolts and time base of 30 mm per second. A. Depicts sleep stages: N1 with vertex waves, N2 with sleep spindles and vertex waves, N3 with slow wave sleep and REM. B. Example of normal awake background with posterior dominant rhythm of 8–9 Hertz. C. Example of diffuse delta brush, with prominent delta waves with overlying lower amplitude beta activity in the bifrontal leads, highlighted by the red box.

EEGs were independently reviewed by two epileptologists, and a third in the case of disagreement. Clinical outcomes included modified Rankin scale (mRS), a measure of neurological disability (de Haan et al., 1995), at 1 year from diagnosis. A good outcome was defined mRS from 0 to 2 and a poor outcome was defined as mRS 3 or greater. For statistical analysis, Student’s t-test, univariate and multivariate regression was used (STATA 16, 2019) and p value < 0.05 was considered significant.

3. Results

Individual and overall patient characteristics are noted in Table 1. Six girls and three boys were included. Average age at presentation was 11.3 years (range 14 months to 16 years old). Loss of PDR was noted in five of nine patients were noted at the beginning and persistent throughout the 24 h recording. EDB was present in six of nine patients; interestingly, evolution of delta brush (i.e more prominent) was noted over the 24 h in two patients who had poor outcomes but for the rest of the patients, the background did not change much over the 24 h recording. For one patient with just the routine EEG, all normal features were present. Seven of nine patients reported sleep problems, but only three had loss of sleep architecture on EEG. If sleep architecture was present on EEG, most patients had N1, N2, N3 with or without REM with the exception of the one routine EEG that had N1 and N2, but not N3. Six of nine patients reported clinical seizures and three of nine had epileptiform discharges, but no electrographic seizures were captured on first EEG. The clinical seizures occurred prior to the EEG recording. While delta brush, epileptiform discharges, and loss of sleep architecture were observed in three patients, these characteristics were found in different patients. Absence of posterior dominant rhythm (PDR) and presence of generalized rhythmic delta activity (GRDA) were each found in 4 patients, but also in different patients.

Table 1.

Individual and summary clinical and EEG characteristics of pediatric NMDARE patients.

Patient Number Age (years) Sex Hospital LOS (days) Rehab LOS (days) Clinical Seizures Immune tx number Immune tx given Sleep Problems EEG patterns present? (+/−)
 mRS at one year
Delta Brush GRDA Normal PDR Epileptiform discharges Sleep stages
1 8.6 M 15 28 + 4 IVMP, IVIG, PEX, RTX − + + − − N1, N2, N3, REM 3
2 14.1 M 11 13 + 3 IVMP, IVIG, RTX + − − + − N1, N2, N3 0
3 14.9 M 4 0 + 3 IVMP, IVIG, RTX − − − + − N1, N2, N3 0
4 1.9 F 7 8 + 3 IVMP, IVIG, RTX + + − + + N1, N2, N3, REM 0
5 9.0 F 6 12 − 4 IVMP, IVIG, PEX, RTX + + − − − N1, N2, N3 4
6 11.7 F 24 23 + 5 IVMP, IVIG, PEX, RTX, CYC + + + − + − 3
7a 12.7 F 40 19 − 4 IVMP, IVIG, PEX, RTX + − − + − N1, N2b 2
8 16.7 F 22 33 − 4 IVMP, IVIG, PEX, RTX + + + − − − 0
9 16.7 F 169 35 + 5 IVMP, IVIG, PEX, RTX, CYC + + + − + − 4
Total Average (range) 11.3 (1.9–16.7) 3M:6F 34.2 (4–169) 22.8 (0–42) 6 3.5 (3–5) 7 6 4 4 3 6 2.0 (0–4)
Open in a new tab

M = male, F = female, LOS: length of stay, tx: treatment, EEG: electroencephalogram, GRDA: generalized rhythmic delta activity, PDR: posterior dominant rhythm, mRS: modified Rankin score, IVMP: intravenous methylprednisolone, IVIG: intravenous immunoglobulin, PEX: plasmapheresis, RTX: rituximab, CYC: cyclophosphamide

a

Designates routine EEGs.

b

N3 not captured on this routine EEG.

The number of and specific immunotherapies administered, inpatient hospital length of stay (LOS), inpatient rehabilitation LOS, and total hospitalization LOS (inpatient hospital plus inpatient rehabilitation LOS) were examined for each EEG characteristic. Loss of PDR was associated with increased length of inpatient rehabilitation (28.8 versus 10 days, p = 0.026) as did loss of sleep architecture (30.3 versus 13.3 days, p = 0.020). The number of immunotherapy treatments administered was also increased in patients with loss of PDR (4.33 vs 3.25, p = 0.014) and loss of sleep architecture (4.67 versus 3.5, p = 0.045; Fig. 2).

Fig. 2.

Fig. 2.

Open in a new tab

The effects of posterior dominant rhythm (PDR) and sleep architecture on inpatient rehabilitation length of stay (LOS, A) and number of immunotherapies administered (B). *indicates p < 0.05.

One-year outcomes were assessed by EEG characteristic using simple linear regression. Loss of PDR was associated with mRS scores at one year (simple linear regression, p = 0.045) and predicted poor outcomes at one year (Fischer’s exact test, p = 0.048). We then performed a multivariate analysis to determine which EEG characteristic predicted one-year outcomes. We initially included all EEG variables and then removed a variable in a stepwise fashion. Once GRDA was removed, loss of PDR (p = 0.013), loss of sleep architecture (p = 0.041), and epileptiform discharges (p = 0.041) correlated with mRS at one year, with delta brush trending toward significance (p = 0.061).

4. Discussion

Electrographic characteristics may predict outcomes in NMDARE. EDB, an electrophysiologic NMDARE biomarker, is associated with prolonged hospitalization in adults (Schmitt et al., 2012). Additionally, normal posterior dominant rhythm (PDR) has been associated with better outcomes (Sonderen et al., 2018) whereas greater than 50% slow waves have been associated with worse prognosis (Mo et al., 2020). In our cohort, loss of PDR and absent sleep architecture were suggestive of more severe clinical course as evidenced by an increased number of immunotherapies administered and longer inpatient rehabilitation. This study may reflect a more severely impacted patient cohort as our patients received a high number of immunotherapies.

While loss of PDR was the only EEG characteristic associated with higher mRS scores at one year, we found that absent sleep architecture and epileptiform discharges also may correlate with worse outcomes at one year, which to our knowledge has not previously been reported. These characteristics were contributory to one-year outcomes, even when adjusting for PDR. If sleep was present, then the 24 h EEG was mainly helpful in capturing slow wave and/or REM sleep, as the one patient without N3 or REM captured was a routine EEG. In patients in which sleep was absent, even the prolonged recording did not capture sleep. Two patients had more evolving delta brush as the recording continued so the 24 h recording was helpful to detect prominent delta brush but 24 h recordings did not alter the results of the study except to capture sleep. Nevertheless, whether evolution of features associates with prognosis would be interesting to examine in a larger cohort.

The EEG changes are attributable to the pathophysiology of NMDARE. Anti-NMDA receptor antibodies are pathogenic and cause receptor downregulation. This downregulation disrupts multiple networks, including loss of inhibition. This loss of inhibitory synapses results in neuronal excitation and subsequent epileptiform discharges and seizures (Moscato et al., 2014). Additionally, the NMDA receptor is important in sleep circuits (Song et al., 2017) and anti-NMDAR antibodies likely disrupt sleep pathways. Interestingly, different EEG features were observed in different patients, which demonstrates the variable electrophysiological phenotypes.

The limitations of this study include a small number of patients at a single institution. Another limitation is that inpatient rehabilitation length of stay (LOS) may be influenced by other factors such as insurance coverage along with functional outcomes. Despite the small sample, we were able to identify potential EEG characteristics that affect one-year outcomes. Additionally, we examined outcomes at a set time-point of one year, whereas prior studies correlating EEG features with outcomes examined outcomes at variable times after diagnosis, such as greater than 6 months (Sonderen et al., 2018) or at time of discharge (Mo et al., 2020).

In conclusion, loss of normal PDR, absence of sleep architecture, and epileptiform discharges are associated with higher modified Rankin scores at one year in NMDARE. Larger studies are needed to confirm these findings. Understanding the underlying mechanisms in the heterogeneity in clinical presentation and prognosis is important to improve treatments and outcomes in NMDARE.

Funding

This work was funded by a Pediatric Research Alliance Pilot Grant from the Center for Clinical and Translational Research at Emory University and Children’s Healthcare of Atlanta.

Conflicts of interest

J.L., K.E., S.B., S.G., A.K., G.Z., L.L., L.B., and R.H. have nothing to declare.

G.G. receives salary support from the Centers for Disease Control and Prevention for surveillance for acute flaccid myelitis.

Data Availability

Data is available from the corresponding author upon reasonable request.

References

  1. Blattner MS, de Bruin GS, Bucelli RC, Day GS, 2019. Sleep disturbances are common in patients with autoimmune encephalitis. J. Neurol 266, 1007–1015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. de Bruijn M, Aarsen FK, van Oosterhout MP, van der Knoop MM, Catsman-Berrevoets CE, Schreurs MWJ, Bastiaansen D, Sillevis Smitt P, Neuteboom RF, Titulaer MJ, CHANCE Study G, 2018. Long-term neuropsychological outcome following pediatric anti-NMDAR encephalitis. Neurology 90, el997–e2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Dalmau J, Gleichman AJ, Hughes EG, Rossi JE, Peng X, Lai M, Dessain SK, Rosenfeld MR, Balice-Gordon R, Lynch DR, 2008. Anti-NMDA-receptor encephalitis: case series and analysis of the effects of antibodies. Lancet Neurol. 7, 1091–1098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Dalmau J, Tüzün E, Wu HY, Masjuan J, Rossi JE, Voloschin A, Baehring JM, Shimazaki H, Koide R, King D, Mason W, Sansing LH, Dichter MA, Rosenfeld MR, Lynch DR, 2007. Paraneoplastic anti-N-methyl-D-aspartate receptor encephalitis associated with ovarian teratoma. Ann. Neurol 61, 25–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Gitiaux C, Simonnet H, Eisermann M, Leunen D, Dulac O, Nabbout R, Chevignard M, Honnorat J, Gataullina S, Musset L, Scalais E, Gauthier A, Hully M, Boddaert N, Kuchenbuch M, Desguerre I, Kaminska A, 2013. Early electro-clinical features may contribute to diagnosis of the anti-NMDA receptor encephalitis in children. Clin. Neurophysiol 124, 2354–2361. [DOI] [PubMed] [Google Scholar]
  6. Graus F, Titulaer MJ, Balu R, Benseler S, Bien CG, Cellucci T, Cortese I, Dale RC, Gelfand JM, Geschwind M, Glaser CA, Honnorat J, Hoftberger R, Iizuka T, Irani SR, Lancaster E, Leypoldt F, Priiss H, Rae-Grant A, Reindl M, Rosenfeld MR, Rostasy K, Saiz A, Venkatesan A, Vincent A, Wandinger KP, Waters P, Dalmau J, 2016. A clinical approach to diagnosis of autoimmune encephalitis. Lancet Neurol. 15, 391–404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. de Haan R, Limburg M, Bossuyt P, van der Meulen J, Aaronson N, 1995. The clinical meaning of Rankin ‘handicap’ grades after stroke. Stroke 26, 2027–2030. [DOI] [PubMed] [Google Scholar]
  8. Haberlandt E, Ensslen M, Gruber-Sedlmayr U, Plecko B, Brunner-Krainz M, Schimmel M, Schubert-Bast S, Neirich U, Philippi H, Kurleman G, Tardieu M, Wohlrab G, Borggraefe I, Rostasy K, 2017. Epileptic phenotypes, electroclinical features and clinical characteristics in 17 children with anti-NMDAR encephalitis. Eur. J. Paediatr. Neurol 21, 457–464. [DOI] [PubMed] [Google Scholar]
  9. Hacohen Y, Deiva K, Pettingill P, Waters P, Siddiqui A, Chretien P, Menson E, Lin JP, Tardieu M, Vincent A, Lim MJ, 2014. N-methyl-D-aspartate receptor antibodies in post-herpes simplex virus encephalitis neurological relapse. Mov. Disord 29, 90–96. [DOI] [PubMed] [Google Scholar]
  10. Linnoila J, Pulli B, Armangué T, Planagumà J, Narsimhan R, Schob S, Zeller M, Dalmau J, Chen J, 2019. Mouse model of anti-NMDA receptor post-herpes simplex encephalitis. Neurol Neuroimmunol. Neuroinflamm 6, e529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Moscato EH, Peng X, Jain A, Parsons TD, Dalmau J, Balice-Gordon RJ, 2014. Acute mechanisms underlying antibody effects in anti-N-methyl-D-aspartate receptor encephalitis. Ann. Neurol 76, 108–119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Mo Y, Wang L, Zhu L, Li F, Yu G, Luo Y, Ni M, 2020. Analysis of risk factors for a poor prognosis in patients with anti-N-Methyl-D-aspartate receptor encephalitis and construction of a prognostic composite score. J. Clin. Neurol 16, 438–447. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Nosadini M, Boniver C, Zuliani L, de Palma L, Cainelli E, Battistella PA, Toldo I, Suppiej A, Sartori S, 2015. Longitudinal electroencephalographic (EEG) findings in pediatric anti-N-methyl-D-aspartate (anti-NMDA) receptor encephalitis: the Padua experience. J. Child Neurol 30, 238–245. [DOI] [PubMed] [Google Scholar]
  14. Qu XP, Vidaurre J, Peng XL, Jiang L, Zhong M, Hu Y, 2020. Seizure characteristics, outcome, and risk of epilepsy in pediatric anti-N-Methyl-d-aspartate receptor encephalitis. Pediatr. Neurol 105, 35–40. [DOI] [PubMed] [Google Scholar]
  15. Qu XP, Vidaurre J, Peng XL, Jiang L, Zhong M, Hu Y, 2019. Seizure characteristics, outcome, and risk of epilepsy in pediatric anti-N-Methyl-d-aspartate receptor encephalitis. Pediatr. Neurol 105, 35–40. [DOI] [PubMed] [Google Scholar]
  16. Sands TT, Nash K, Tong S, Sullivan J, 2015. Focal seizures in children with anti-NMDA receptor antibody encephalitis. Epilepsy Res. 112, 31–36. [DOI] [PubMed] [Google Scholar]
  17. Schmitt SE, Pargeon K, Frechette ES, Hirsch LJ, Dalmau J, Friedman D, 2012. Extreme delta brush: a unique EEG pattern in adults with anti-NMDA receptor encephalitis. Neurology 79, 1094–1100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Sonderen AV, Arends S, Tavy DLJ, Bastiaansen AEM, Bruijn M, Schreurs MWJ, Sillevis Smitt P, Titulaer MJ, 2018. Predictive value of electroencephalography in anti-NMDA receptor encephalitis. J. Neurol. Neurosurg. Psychiatry 89,1101–1106. [DOI] [PubMed] [Google Scholar]
  19. Song Q, Feng G, Zhang J, Xia X, Ji M, Lv L, Ping Y, 2017. NMDA receptor-mediated Ca2+ influx in the absence of Mg2+ block disrupts rest: activity rhythms in drosophila. Sleep 40, 40. [DOI] [PubMed] [Google Scholar]
  20. Titulaer MJ, McCracken L, Gabilondo I, Armangué T, Glaser C, Iizuka T, Honig LS, Benseler SM, Kawachi I, Martinez-Hernandez E, Aguilar E, Gresa-Arribas N, Ryan-Florance N, Torrents A, Saiz A, Rosenfeld MR, Balice-Gordon R, Graus F, Dalmau J, 2013. Treatment and prognostic factors for long-term outcome in patients with anti-NMDA receptor encephalitis: an observational cohort study. Lancet Neurol. 12, 157–165. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Data is available from the corresponding author upon reasonable request.

RESOURCES