Beruflich Dokumente
Kultur Dokumente
In its early stages, anti-N-methyl-D-aspartate receptor disproportionate cognitive disturbance (p,0.005) with high
(NMDAR) encephalitis is often characterized by prominent negative symptom scores and excitability. Those present-
psychiatric manifestations that can lead to delays in di- ing with psychotic symptoms were significantly younger
agnosis and treatment. The authors aimed to address this than those without (p,0.005). Patients with anti-NMDAR
problem by providing a detailed description of the psychi- encephalitis present with a somewhat distinct cluster of
atric phenotype and demographic features that may influ- psychiatric symptoms not commonly seen in functional
ence presentation. Eighty-six patients with positive serum psychoses. When encountered, this atypical pattern should
NMDAR antibodies were identified, 22 of whom met di- warrant further investigation and a high index of suspicion
agnostic criteria for anti-NMDAR encephalitis. Medical notes for anti-NMDAR encephalitis. The more prominent psy-
were reviewed retrospectively to rate psychiatric symptoms chotic features in younger adults may reflect greater sus-
using standardized scales. Clinical and demographic char- ceptibility of the young brain to exogenous psychosis.
acteristics were compared for patients with and without
psychosis. Patients with psychosis exhibited severe psy-
chopathology with a characteristic phenotype: severe and JNCN in Advance (doi: 10.1176/appi.neuropsych.17120343)
Anti-N-methyl-D-aspartate receptor (NMDAR) encephalitis, behavioral abnormalities.7,8 Age-related differences are also
the most common of the autoimmune encephalitis condi- seen in functional psychoses, where psychosis occurs without
tions, is a progressive and potentially fatal condition.1 Initial a clearly defined organic cause; onset is rare in childhood,
cognitive deficits, psychiatric symptoms, and speech disin- with the median age of first-episode psychosis in the early
tegration are often followed by depression of consciousness 20s, often slightly later in women.9 Where schizophrenia-like
and dystonic movements. However, the condition is het- psychosis does present in later life, there are marked differ-
erogeneous and can occur across the life span, with young ences in presentation with paranoia prominent, whereas neg-
women, often with ovarian teratomas, most commonly af- ative symptoms and thought disorder are infrequent.10 No
fected.2 The existence of isolated psychiatric presentations study to date has explored how the described demographic
that respond to immunotherapy raises concerns of missed differences in functional psychoses might be reflected in the
diagnoses.3 Indeed, prominent psychiatric symptoms in psychiatric presentation of anti-NMDAR encephalitis.8
anti-NMDAR encephalitis result in up to 60% of patients The present study was designed to address these ques-
being initially admitted to psychiatric units.4 This can delay tions by examining whether the phenomenology and pattern
diagnosis and appropriate treatment significantly. While of psychotic symptoms might allow the identification of
there is hope that with increasing recognition of the disease distinct encephalitic subgroups and also discrimination from
this number is decreasing, it remains an important differ- functional or idiopathic psychosis.
ential for patients under psychiatric care.5 Autonomic in-
stability and neurological symptoms, such as seizures, have
METHODS
been highlighted as important signs to aid clinicians in
identifying such patients; however, no defining psychiatric This study was approved by the institutional review board of
features have so far been found.5,6 South London and Maudsley NHS Foundation Trust.
Anti-NMDAR encephalitis can affect men and women
of any age, but demographic differences may influence the Patients
clinical phenotype. Both children and men present more In this observational retrospective cohort study, 86 pa-
commonly with seizures and less prominent psychiatric and tients were identified with positive serum NMDAR antibodies
FIGURE 1. Flow Chart of Patient Selection patients were excluded because, despite
having positive NMDAR antibodies and in
some cases receiving a contemporaneous
86
diagnosis of anti-NMDAR encephalitis
anti-NMDAR +ve from their treating clinician, they did not
30
satisfy the Graus et al. criteria. Children
<18 years
were excluded due to the anticipated dif-
56 ficulty in making meaningful compari-
9 sons of psychiatric symptoms with adults
47 >18 years
No accessible notes
(Figure 1).
Notes available
25 22 Inclusion Criteria
NMDA encephalitis unlikely NMDA encephalitis probable Patients over 18 years old, with a serum
from available data positive NMDAR antibody result, clini-
cal records available at Kings College
12 10 Hospital, and symptomatic features for
Psychosis present Psychosis absent probable anti-NMDAR encephalitis were
included for analysis. Also included was a
patient (#27) who had the clinical fea-
between July 2010 and May 1, 2017, at Kings Health Partners, tures required but did not undergo reported CSF or EEG
London, United Kingdom. Testing was performed at Oxford testing, as she was admitted to a psychiatric unit. After several
Neuroimmunology Laboratory with cell-based assays for se- months, she improved spontaneously, but the treating psy-
rum samples. A live cell-based assay was used for all patients chiatrist acknowledged that it was likely that NMDAR en-
except two, whose serum was tested after December 2015 cephalitis contributed to her presentation and arranged future
when testing methodology changed to a fixed cell-based outpatient testing, which was not available at the time of this
assay. Only two of the patients underwent CSF testing of study. Statistical analyses were performed with and without
NMDAR antibodies, as NMDAR antibody testing of CSF this individual without any change in significant findings.
was not common practice in the United Kingdom during Twenty-one other patients were identified with rapid onset of
the period of this study. at least 4/6 groups of symptoms and abnormal CSF or EEG
The international consensus criteria in the position pa- testing (or 3/6 symptoms alone in the presence of a teratoma)
per by Graus et al.11 were used to identify those patients (Table 1).
with probable NMDAR encephalitis with a positive serum
NMDAR antibody result and clinical features of anti- Data Collection
NMDAR encephalitis. Briefly, the criteria include rapid on- Data were collected by retrospectively reviewing a combi-
set; at least four out of six major groups of clinical symptoms nation of electronic patient medical and psychiatric records.
(unless accompanied by a teratoma, when only three are Records were used to identify demographic information,
required); an abnormal EEG and/or CSF; and reasonable symptomatic timing onset and nadir, clinical features in-
exclusion of other disorders.11 cluding psychotic symptoms, and results of investigations
These criteria were used to retrospectively identify pa- performed, such as other serum antibodies, EEG, CSF, and
tients with probable NMDAR encephalitis, as all but two of MRI.
our patients were diagnosed and treated prior to the pub-
lication of these criteria. These diagnostic criteria for proba- Measures
ble anti-NMDAR encephalitis are designed for use in the Psychotic symptoms were assessed using definitions from
absence of a serum or CSF NMDAR antibody result. Our the Positive and Negative Syndrome Scale (PANSS), sup-
patients both had a positive serum result and met the di- plemented by the Scale for the Assessment of Positive
agnostic criteria, leading to greater certainty of their anti- Symptoms (SAPS) and Scale for the Assessment of Negative
NMDAR encephalitis diagnosis. The high sensitivity and Symptoms (SANS) criteria because of their finer grained
specificity of these criteria in correctly identifying patients coverage of certain psychotic symptoms, such as bizarre
with anti-NMDAR encephalitis have since been affirmed.12,13 behavior and avolition.14 The quality of the patient notes
varied; items were rated on the basis of the greatest intensity
Exclusion Criteria of symptoms documented, from initial onset to nadir. The
Patients were excluded if they were under 18 years, had PANSS is validated for use in first-episode psychosis and
insufficient clinical information available to make a di- schizophrenia and has been used in anti-NMDAR enceph-
agnostic assessment, or did not meet diagnostic criteria (see alitis case studies.15–18 Items are scored between 1 and 7 with
Figure 1 for flow chart). Nine patients were excluded due increasing severity. Factor analyses of symptoms suggest
to insufficient available clinical information. Twenty-five that a five-factor model provides the best fit for PANSS
JNCN in Advance
4, Female, 76 Nonpsychotic + – + + – – Dominant theta NMDAR + Normal Mild cerebral small
background; VGKC – vessel disease
bursts of
bitemporal slow
waves, left
predominant
1, Male, 77 Nonpsychotic + + + + + – Periodic temporal NMDAR+ Normal Diffuse T2
sharp waves and VGKC 101 hyperintensity and
polyspikes. swelling; left
Diffuse low hippocampus,
amplitude delta cortical/
activity with subcortical inferior
intermixed fast frontal and
activity temporal, insula,
pulvinar; mature
left middle
cerebral artery
territory infarct
2, Male, 40 Nonpsychotic + – + + + – Slow background NMDAR + WCC 25, 94% lymph Diffuse T2
activity with VGKC – hyperintensity of
frequent VGCaC – amygdala,
bifrontal sharp Ro/SSA + hippocampus,
delta waves. TPO – splenium of
Triphasic at corpus callosum
times and periventri-
cular white matter
bilaterally
13, Female, 27 Nonpsychotic + + + – – – Runs of theta NMDAR + Normal Normal
slowing and ANA –
sharp waves left
temporal
16, Female, 71 Nonpsychotic + + + + + – Nearly continuous NMDAR + OCB matched T2 hyperintensity
periodic LGI1 – cortical/subcortical
sharpened slow CASPR2 – in posterior
waves (PLEDs) TPO 61 temporal and
left temporal, ANA – parietal lobes, left
parietal and ds DNA – temporal lobe most
occipital extensive
9, Female, 54 Nonpsychotic – + + + – – Mild, diffuse NMDAR+ N/A Nonspecific
background VGKC 277 scattered foci of
slowing Hu – T2 high signal in
Yo – cerebral white
matter bilaterally
neuro.psychiatryonline.org
continued
GIBSON ET AL.
3
4
TABLE 1, continued
Patient Auto-
Number, Speech Move- Reduced nomic EEG Tested
Gender, Age Distur- ment Con- Dys- (Abbreviated Serum MRI (Abbreviated
(years) Psychosis bance Seizure Disorder sciousness function Teratoma Reports) Antibody CSF Reports)
23, Male, 42 Nonpsychotic + + + – – – Intermittent NMDAR+ OCB unmatched Resection cavity in
bitemporal GAD + right temporal
slowing with lobe; Known
spikes and temporal lobe
sharpened glioma resection
neuro.psychiatryonline.org
slow waves $20years prior
24, Female, 69 Nonpsychotic + – – + + – Frequent slow/ NMDAR + N/A Generalized volume
sharp waves, Hu – loss.
bifronto- Yo –
temporal, left
emphasis
15, Male, 72 Nonpsychotic + – + + – – Diffusely slow NMDAR + Normal Extensive T2
background, VGKC – hyperintensity in
left temporal insula, parietal and
focal slowing lateral temporal
lobe
49, Female, 44 Nonpsychotic + – + + + – Bifrontal and left NMDAR+ Normal T2/flair
PSYCHIATRIC PHENOTYPE OF ANTI-NMDA RECEPTOR ENCEPHALITIS
JNCN in Advance
TABLE 1, continued
Patient Auto-
Number, Speech Move- Reduced nomic EEG Tested
Gender, Age Distur- ment Con- Dys- (Abbreviated Serum MRI (Abbreviated
(years) Psychosis bance Seizure Disorder sciousness function Teratoma Reports) Antibody CSF Reports)
12, Female, 22 Psychotic + + + + + + Diffuse bilateral NMDAR + WCC 30 Diffuse cerebral
JNCN in Advance
left predominant ANCA – edema and mild
fronto-temporal dsDNA – bilateral uncal
slow wave with GAD – herniation; diffuse
sharpened theta leptomeningeal
activity enhancement
14, Female, 45 Psychotic + + + + – – Theta and delta NMDAR + IgG 37 Normal
transients
mainly left
temporal
17, Female, 22 Psychotic + + + + + + Right fronto- NMDAR + Protein 0.47g/dL Normal
temporal sharp VGKC –
and slow ANCA –
waves and dsDNA –
independent left GAD –
sharp transients ganglioside –
18, Female, 23 Psychotic + – + + + – Diffusely slow NMDAR + Normal (no OCB, Normal
background (CSF and serum) WCC,5)
with frontal VGKC – NMDA +
bilateral high
amplitude delta
slow waves
19, Female, 32 Psychotic + + + – – – Abnormal (report NMDAR + OCB unmatched Left mesial temporal
unavailable) (CSF and NMDA+ sclerosis
serum)
VGKC –
5, Female, 52 Psychotic + – + – + – Sharp spikes in NMDAR + Normal Mild small vessel
Ovarian temporal area VGKC – disease cerebral
cyst Hu – hemispheres
Yo –
11, Female, 19 Psychotic + – + – + + Normal NMDAR + Normal Normal
VGKC –
Hu –
Yo –
20, Male, 56 Psychotic + + – – + – Focal slowing over NMDAR+ Normal Normal
left temporal VGKC 193
with sharp Hu –
waves seen Yo –
intermixed and
independently
27, Female, 33 Psychotic + – + – + – Not available NMDAR + N/A N/A
VGKC –
TPO –
neuro.psychiatryonline.org
continued
5
GIBSON ET AL.
PSYCHIATRIC PHENOTYPE OF ANTI-NMDA RECEPTOR ENCEPHALITIS
The + symbol denotes the presence of the symptom or the antibody indicated; the – symbol denotes the absence of the symptom or the antibody indicated. ANCA=antineutrophil cytoplasmic antibodies; anti-
GAD=glutamic acid decarboxylase antibody; anti-NMDAR=N-methyl-D-aspartate receptor antibody; ANA=anti-nuclear antibody; anti-TPO=thyroid peroxidase antibody; anti-VGCaC=voltage gated calcium
TABLE 2. Demographic and Clinical Characteristics of Patients
channel antibodies; anti-VGKC=voltage gated potassium channel antibodies; CASPR2=contactin-associated protein 2; ds DNA=double stranded DNA; LGI1=leucine-rich glioma inactivated 1; HSV=herpes
T2 hyperintensity and
MRI (Abbreviated
hippocampus;
With Without
Reports)
encephalitis
recent HSV
Psychosis Psychosis
Characteristic (N=12) (N=10) p
N % N %
Female 10/12 83 6/10 60 0.348
Psychiatric admission 7/12 58 0/10 0 0.005
Psychiatric review 12/12 100 4/10 40 0.003
Presence of teratoma 5/12 42 0/10 0 0.040
OCB unmatched
Mean SD Mean SD
Age (years) 33.8 12.0 57.2 18.0 0.003
MRS at nadir 4.0 0.95 3.9 1.4 0.842
a
Statistically significant data are indicated in bold. MRS=Modified Rankin
Scale.
temporal delta
(Abbreviated
brush activity
Frequent right
Reports)
EEG
Dys-
SAPS/SANS.19
Statistical Analyses
+
48, Male, 37
(years)
files for the PANSS, SAPS and SANS were compared with
within-subjects analysis of variance (ANOVA). These
a
FIGURE 2. Mean Scores for the Positive and Negative Symptom Scale (PANSS) and the Scale for the Assessment of Negative
Symptoms (SANS)a
A
Positive
7
6
5
4
3
Depressed 2 Negative
1
0
Excited Cognitive
B Hallucinations C Affect
5 5
4 4
3 3
2 2
Attention Alogia
1
Formal 1
thought 0 Delusions
disorder 0
Anhedonia Avolition
Bizarre behavior
a
Panel A shows the mean scores on the Wallwork domains of PANSS. Panel B shows the mean global symptom scores on the Scale for the
Assessment of Positive Symptoms. Panel C shows the mean global symptom scores for SANS.
analyses are based on the mean aggregate score of the two presented with dyskinesia alone (67% of which were oro-
independent rating clinicians. facial in nature), one with hypertonicity, four with an-
other movement disorder such as dystonia, and nine with
more than one of these three symptom groups.
RESULTS
Descriptive Data Investigations. Ninety percent (N=19) had an abnormal EEG;
Just under half of the adults (47%, N=22) with serum 42% (N=8) had CSF showing oligoclonal bands or pleo-
NMDAR antibodies satisfied the clinical criteria for proba- cytosis; 33% (N=7) had MRI brain changes suggestive of
ble NMDAR antibody encephalitis and were included in the encephalitis; 23% (N=5) had an ovarian teratoma. Details
subsequent analyses. Seventy-three percent (N=16) were of these clinical symptoms and investigations are shown in
female; the mean age was 44 (range, 19–77 years); and 55% Table 1.
(N=12) were acutely psychotic based on the presence of
hallucinations or delusions. Thirty-two percent (N=7) had an Psychosis
admission to a psychiatric unit prior to a diagnosis of anti- Patients presenting with psychotic symptoms were signifi-
NMDAR encephalitis. cantly younger than those without (33.8 years versus 57.2
years, t=3.50, df=15.2, p,0.005). There was no significant
Clinical symptoms. Fifty-five percent (N=12) presented with gender difference in patients with or without psychosis
seizures: two patients with generalized seizures, seven with (p=0.4); however, patients with psychosis were significantly
partial seizures (most commonly complex partial), and three more likely to have an ovarian teratoma (p=0.040). Patients
with both generalized and partial seizures. Ninety-one per- with psychosis were more likely to have a psychiatric in-
cent (N=20) presented with movement disorder; these patient admission (p=0.005) and psychiatric review while
were classified into dyskinesia, rigidity/abnormal posture, on a medical ward (p,0.005). There was no difference in
or movement disorder as per Graus et al. 11 Six patients overall clinical severity at nadir between patients with
psychosis and patients without psychosis (based on MRS, positive symptoms such as bizarre behavior, hallucinations,
t(20)=20.201, p=0.8). See Table 2 for figures. and delusions. However, the detailed description of psy-
chopathology in our patient group has revealed for the first
Psychosis subgroup: PANSS symptom profile. For the 12 pa- time, in addition to positive and classic encephalitic symp-
tients with psychosis according to our definition, PANSS toms marked negative symptoms as commonly described in
scores indicated a severe degree of psychopathology; mean functional psychoses such as schizophrenia. This is con-
score 135.2 (range, 86–169).21 The highest mean scores were gruent with an animal model of anti-NMDAR encephalitis
on items of conceptual disorganization (6.4), poor abstract that showed profound memory deficits and anhedonia.22
thinking (6.1), attention (6.1), and poor rapport (6.2). Low However, this does little to distinguish the psychiatric fea-
scores were seen on items measuring guilt (1.5) and somatic tures of anti-NMDAR encephalitis from other psychiatric
concern (1.8). populations.
PANSS five-factor analysis showed some symptomatic Apparent differences did emerge when considering the
domains to be disproportionately affected (Figure 2A). phenomenology of the psychosis seen in anti-NMDAR
Within-subjects ANOVA showed a significant difference encephalitis. In particular, patients had severe cognitive
between the means of the five domains (F=19.3, df=4, 44, disturbance relative to other domains; marked cognitive
p,0.005), with post hoc tests indicating that patients had impairment is a clear phenotypic red flag for anti-NMDA
more significant cognitive symptoms than depressive (p,0.005) encephalitis. There was a tendency for patients to be more
or positive features (p=0.003). Patients exhibited high nega- disturbed by thought disorder and bizarre behavior as op-
tive symptom scores (Figure 2A). posed to delusions and hallucinations, in contrast with the
typical presentation in first-episode psychosis and schizo-
Psychosis subgroup: SAPS and SANS profile. A significant phrenia populations.23 Furthermore, clinicians should not
effect for SAPS subscale on ANOVA (F=20.469, df=3, 33, let the presence of negative symptoms distract from the
p,0.005) indicated a higher degree of thought disorder possibility of anti-NMDAR encephalitis; indeed, particular
and bizarre behavior than other SAPS factors (Figure 2B). suspicion should be aroused in cases of prominent alogia and
Post hoc analyses on the SAPS showed that formal thought distractibility, less common in functional psychoses.23 Given
disorder was significantly more severe than delusional the difficulty in diagnosing anti-NMDAR encephalitis, such
(p=0.002) or hallucinatory behavior (p,0.005). Similarly, unusual presentations within psychiatric inpatients should
patients were more affected by bizarre behavior than hal- warrant further investigation and antibody testing.
lucinations (p,0.005) or delusions (p=0.002). A qualitative This pattern of symptoms mirrors those seen in phar-
review of the notes revealed that delusions were often poorly macological models of NMDAR hypofunction. Prominent
formed and unstable. negative symptoms, cognitive impairment, and thought
On the SANS, the observed differences between the disorder with relatively less severe hallucinations and de-
means were also significant (F=9.07, df=1.81, 20.0, p=0.002), lusions are characteristic in ketamine-induced models of
with speech production and attention most profoundly im- psychosis.24,25 While in anti-NMDAR encephalitis the down-
paired (Figure 2C). regulation of receptors is antibody mediated, ketamine non-
competitively antagonizes NMDAR.25,26 Given that NMDAR
Nonpsychosis subgroup: symptom profile. Ten patients did hypofunction is common to both, the degree of similarity in
not have psychotic symptoms, 80% (N=8) presented with symptom profile is perhaps unsurprising. This lends support
confusion and disorientation, 60% (N=6) had memory defi- to suggestions of a unique psychopathological profile arising
cits, 30% (N=3) were aphasic, and a further 40% (N=4) from glutamatergic psychosis, as well as a potential mecha-
presented with partial language disturbance: one predom- nism underpinning the dimensions of psychosis seen in anti-
inantly expressive, one receptive, and two with receptive NMDAR encephalitis.
and expressive components. These distinctions were docu- Those encephalitis patients with prominent psychotic
mented by the treating clinician. Four were unable to re- features were significantly younger than those without. This,
liably follow commands, and three had symptoms of low without any other differences seen in symptoms or investi-
mood. In one patient, this was particularly severe, with quiet gations, is consistent with greater susceptibility of the young
speech and psychomotor retardation. brain to psychosis. Indeed, in general psychiatry, the higher
Finally, a contrast between older and younger patients incidence and greater severity of psychosis in younger
based on a cut-off of 40 years revealed no significant dif- (versus older) adults is well recognized.27
ferences in gender, symptom groups other than psychosis,
or abnormalities on clinical investigations. Limitations
Few of our patients underwent CSF antibody testing and no
confirmatory immunoassays were included, as these had not
DISCUSSION
been commonly used in U.K. clinical practice at the time
Much of the literature to date on the psychiatric features of the study. For a diagnosis of definite anti-NMDAR en-
of anti-NMDAR encephalitis has highlighted prominent cephalitis, Graus et al. require either positive CSF NMDAR
17. van de Riet EH, Esseveld MM, Cuypers L, et al: Anti-NMDAR 23. Shtasel DL, Gur RE, Gallacher F, et al: Phenomenology and func-
encephalitis: a new, severe and challenging enduring entity. Eur tioning in first-episode schizophrenia. Schizophr Bull 1992; 18:
Child Adolesc Psychiatry 2013; 22:319–323 449–462
18. Zandi MS, Deakin JB, Morris K, et al: Immunotherapy for patients 24. Krystal JH, Perry EB Jr, Gueorguieva R, et al: Comparative and inter-
with acute psychosis and serum N-Methyl D-Aspartate receptor active human psychopharmacologic effects of ketamine and amphet-
(NMDAR) antibodies: a description of a treated case series. Schizo- amine: implications for glutamatergic and dopaminergic model psychoses
phr Res 2014; 160:193–195 and cognitive function. Arch Gen Psychiatry 2005; 62:985–994
19. Landis JR, Koch GG: The measurement of observer agreement for 25. Carhart-Harris RL, Brugger S, Nutt DJ, et al: Psychiatry’s next top
categorical data. Biometrics 1977; 33:159–174 model: cause for a re-think on drug models of psychosis and other
20. Bonita R, Beaglehole R: Recovery of motor function after stroke. psychiatric disorders. J Psychopharmacol 2013; 27:771–778
Stroke 1988; 19:1497–1500 26. Moscato EH, Peng X, Jain A, et al: Acute mechanisms underlying
21. Leucht S, Kane JM, Kissling W, et al: What does the PANSS mean? antibody effects in anti-N-methyl-D-aspartate receptor encepha-
Schizophr Res 2005; 79:231–238 litis. Ann Neurol 2014; 76:108–119
22. Planagumà J, Leypoldt F, Mannara F, et al: Human N-methyl 27. Howard R, Rabins PV, Seeman MV, et al: Late-onset schizophrenia
D-aspartate receptor antibodies alter memory and behaviour in and very-late-onset schizophrenia-like psychosis: an international
mice. Brain 2015; 138:94–109 consensus. Am J Psychiatry 2000; 157:172–178