Keywords autoimmune encephalitis - anti-NMDA receptor encephalitis - ovarian teratoma - paraneoplastic
Introduction
Anti-N-methyl-D-aspartate receptor encephalitis (NMDARE) is an autoimmune encephalitis
with neuropsychiatric symptoms caused by anti-NMDA receptor immunoglobulin G antibodies.[1 ] While a paraneoplastic syndrome is most commonly associated with ovarian teratomas,[1 ]
[2 ]
[3 ]
[4 ] the NMDARE rate in ovarian teratomas is unknown. In this retrospective study, we
determine the rate of NMDARE in ovarian teratomas and compare tumor sizes between
NMDARE and non-NMDARE patients.
Methods
Institutional review board approval (STUDY00000209) was obtained for this study. Informed
consent was waived. Patients were identified by the Montage database system of radiological
imaging by the search terms “ovarian” and “teratoma or dermoid” at a single tertiary
pediatric free-standing hospital between January 2011 and December 2019. Chart review
was performed on 163 of 461 patients that had confirmed ovarian teratomas by resection
and pathology. NMDARE was diagnosed if patients met criteria for confirmed NMDARE
defined as having positive anti-NMDA receptor cerebrospinal fluid (CSF) antibodies
and one or more of six neuropsychiatric symptoms: (1) abnormal psychiatric behavior/cognitive
dysfunction; (2) speech dysfunction; (3) seizures; (4) movement disorder, dyskinesias,
or rigidity/abnormal postures; (5) decreased level of consciousness; and (6) autonomic
dysfunction or central hypoventilation[5 ] (see [Fig. 1 ]).
Fig. 1 Flow diagram of participants at each stage of the study. CSF, cerebrospinal fluid;
NMDA, N-methyl-D-aspartate.
Descriptive summary statistics, Student's t -test, Fischer's exact test, and an analysis of variance (ANOVA) on a multiple linear
regression model were performed using the R statistical software.
Results
Clinical characteristics are shown in [Table 1 ]. Initial imaging methods to detect teratomas were ultrasound (70.55%), computed
tomographic scan (26.4%), and magnetic resonance imaging (3.0%). Five of 163 (3.1%)
patients with confirmed ovarian teratomas on pathology were diagnosed with NMDARE
with the presence of at least one neurological symptom along with positive anti-NMDA
receptor antibody detected in the CSF. Student's t -test demonstrated no difference in the ages of NMDARE versus non-NMDARE patients.
All five NMDARE patients and 151/158 non-NMDARE patients had ovarian teratomas with
benign pathology. Three of five (60%) NMDARE had teratomas with identifiable neuroglial
elements as compared with 32/158 (20.3%) non-NMDARE patients, which trended toward
significance (Fisher's exact test, p = 0.067). For neurological symptoms, zero of the non-NMDARE patients reported any
of the neurological symptoms associated with anti-NMDARE. For the NMDARE patients,
all five patients exhibited psychiatric symptoms, speech dysfunction, and a form of
movement disorder. A subset of NMDARE patients had one or more additional neurological
symptoms, that is, seizures, decreased level of consciousness, and autonomic dysfunction/central
hypoventilation ([Table 1 ]).
Table 1
Clinical characteristics in anti-NMDARE as compared with non-NMDARE patients
NMDARE (n = 5)
Non-NMDARE (n = 158)
p -Value
Age, average years (SD)
13.8 (3.90)
12.4 (4.04)
0.5
Race, n (%)
African American
4 (80)
67 (41)
Caucasian
1 (20)
55 (34)
Latino
0
24 (15)
Asian
0
10 (6)
Native American
0
1 (0.6)
Mixed
0
4 (3)
Unknown
0
2 (1)
Ovarian teratoma volume, median cm3 (IQR)
28.3 (431.2)
182.8 (635.0)
0.008
Ovarian teratoma pathology, n (%)
Mature/benign
5 (100)
150 (95)
1.0
Neuroglial element
3 (60)
32 (20)
0.067
Neurological symptom, n (%)
Psychiatric/cognitive decline
5 (100)
0 (0)
Speech dysfunction
5 (100)
0 (0)
Seizures
3 (60)
0 (0)
Movement disorder
5 (100)
0 (0)
Decreased level of consciousness
2 (40)
0 (0)
Autonomic dysfunction or central hypoventilation
3 (60)
0 (0)
Abbreviations: IQR, interquartile range; NMDARE, N-methyl-D-aspartate receptor encephalitis;
SD, standard deviation.
Using ANOVA, we evaluated predictors of ovarian teratoma volumes. Age had a positive
correlation with teratoma size (p = 0.013), whereas NMDARE had an inverse correlation (p = 0.008, adjusted for age and race). Teratoma volumes from NMDARE patients were smaller
than that of non-NMDARE patients (median 28.3 cm3 with interquartile range [IQR] of 431.2 and median 182.8 with IQR of 635.0, respectively;
p = 0.01).
Discussion
In our cohort, 3.07% of patients with ovarian teratomas had NMDARE. Moreover, ovarian
teratoma volumes are smaller in NMDARE compared with non-NMDARE patients and increased
age associates with a larger teratoma size. Additionally, 80% of NMDARE patients were
African American, in comparison to 42.4% of non-NMDARE patients in this study; prior
studies have varied in rates of African American NMDARE patients but another larger
NMDARE cohort is 67%.[6 ]
NMDARE pathogenesis can be a paraneoplastic response to the neural components in ovarian
teratomas.[2 ]
[7 ] Neuroglial elements are possibly more commonly detected in NMDARE versus non-NMDARE
ovarian teratomas, as our data trended toward significance and as discussed in a small
case–control study.[8 ] We also found that teratoma volumes are smaller in NMDARE versus non-NMDARE, which
is most likely due to earlier tumor detection, similar to how neuroblastomas are detected
earlier in opsoclonus myoclonus ataxia syndrome[9 ] due to a paraneoplastic response to the tumor.
Limitations of this study include the retrospective nature of this study at a single
center. Another limitation is that CSF anti-NMDA receptor antibody testing was not
available for all individuals as routine anti-NMDA receptor antibody testing is not
performed in patients with ovarian teratomas. However, none of the symptoms associated
with NMDARE were present in the non-NMDARE group and positive antibody alone would
not be sufficient for diagnosis.[5 ] Serum anti-NMDAR antibody testing is inadequate because anti-NMDAR antibody detection
is more sensitive in CSF[3 ] than serum and serum anti-NMDAR antibody testing can have false positives.[10 ]
In conclusion, rates of NMDARE in ovarian teratomas are low (3.07%) and ovarian teratomas
are smaller in NMDARE than in non-NMDARE. Neuroglial elements may contribute to the
development of NMDARE. Further studies are needed to understand the timing of anti-NMDA
receptor antibodies in teratomas and development of NMDARE.
