Share

July 2023; 10 (4) Research ArticleOpen Access

Mesiotemporal Volumetry, Cortical Thickness, and Neuropsychological Deficits in the Long-term Course of Limbic Encephalitis

Antonia Harms, Tobias Bauer, Juri-Alexander Witt, Tobias Baumgartner, Randi von Wrede, Attila Racz, Leon Ernst, Albert J. Becker, Christoph Helmstaedter, Rainer Surges, Theodor Rüber
First published May 25, 2023, DOI: https://doi.org/10.1212/NXI.0000000000200125
Full PDF
Citation
Permissions

Make Comment

See Comments

Downloads
0

Share

Abstract

Background and Objectives Limbic encephalitis (LE) is an autoimmune disease often associated with temporal lobe epilepsy and subacute memory deficits. It is categorized into serologic subgroups, which differ in clinical progress, therapy response, and prognosis. Using longitudinal MRI analysis, we hypothesized that mesiotemporal and cortical atrophy rates would reveal serotype-specific patterns and reflect disease severity.

Methods In this longitudinal case-control study, all individuals with antibody-positive (glutamic acid decarboxylase 65 [GAD], leucine-rich glioma-inactivated protein 1 [LGI1], contactin-associated protein 2 [CASPR2], and N-methyl-d-aspartate receptor [NMDAR]) nonparaneoplastic LE according to Graus' diagnostic criteria treated between 2005 and 2019 at the University Hospital Bonn were enrolled. A longitudinal healthy cohort was included as the control group. Subcortical segmentation and cortical reconstruction of T1-weighted MRI were performed using the longitudinal framework in FreeSurfer. We applied linear mixed models to examine mesiotemporal volumes and cortical thickness longitudinally.

Results Two hundred fifty-seven MRI scans from 59 individuals with LE (34 female, age at disease onset [mean ± SD] 42.5 ± 20.4 years; GAD: n = 30, 135 scans; LGI1: n = 15, 55 scans; CASPR2: n = 9, 37 scans; and NMDAR: n = 5, 30 scans) were included. The healthy control group consisted of 128 scans from 41 individuals (22 female, age at first scan [mean ± SD] 37.7 ± 14.6 years). The amygdalar volume at disease onset was significantly higher in individuals with LE (p ≤ 0.048 for all antibody subgroups) compared with that in healthy controls and decreased over time in all antibody subgroups, except in the GAD subgroup. We observed a significantly higher hippocampal atrophy rate in all antibody subgroups compared with that in healthy controls (all p ≤ 0.002), except in the GAD subgroup. Cortical atrophy rates exceeded normal aging in individuals with impaired verbal memory, while those who were not impaired did not differ significantly from healthy controls.

Discussion Our data depict higher mesiotemporal volumes in the early disease stage, most likely due to edematous swelling, followed by volume regression and atrophy/hippocampal sclerosis in the late disease stage. Our study reveals a continuous and pathophysiologically meaningful trajectory of mesiotemporal volumetry across all serogroups and provides evidence that LE should be considered a network disorder in which extratemporal involvement is an important determinant of disease severity.

Glossary

AIC=
Akaike information criterion;
AMPAR=
α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor;
CASPR2=
contactin-associated protein-like 2;
FDR=
false discovery rate;
GABAB=
gamma-aminobutyric acid B;
GAD=
glutamic acid decarboxylase 65;
HEK293=
Human Embryonic Kidney 293;
LE=
limbic encephalitis;
LGI1=
leucine-rich glioma-inactivated 1;
LLR=
likelihood ratio;
NMDAR=
N-methyl-d-aspartate receptor

Limbic encephalitis (LE) is a form of autoimmune encephalitis defined by subacute onset of typical symptoms such as seizures, memory impairment, or psychiatric abnormalities, combined with bilateral MRI abnormalities in the mesial temporal lobes, pleocytosis in the cerebrospinal fluid, and epileptiform EEG activity involving the temporal lobes.1,-,5 Although the diagnosis of LE can be made without the presence of specific antibodies, the detection of antibodies is required when not all diagnostic criteria are met, for instance, when MRI abnormalities are not found bilaterally in the medial temporal lobe. Most common antibodies associated with nonparaneoplastic LE target against glutamic acid decarboxylase 65 (GAD), leucine-rich glioma-inactivated 1 (LGI1), and contactin-associated protein-like 2 (CASPR2).6 While antibodies against N-methyl-d-aspartate receptor (NMDAR) are commonly associated with autoimmune encephalitis involving the entire brain, in rare cases, antibodies against NMDAR can be found in cases with predominant involvement of the limbic system that fulfill diagnostic criteria for typical LE.1,3,7,8 In addition to clinical evaluation, cerebrospinal fluid analysis, and EEG, MRI is not only an integral part of the diagnostic criteria but increasingly plays an important role in the disease monitoring of LE.3 Especially, abnormalities in the mesial temporal lobe have been consistently reported in previous research, such as increased signal in T2-weighted fluid-attenuated inversion recovery imaging3 or volumetric alterations of amygdala and hippocampus.9,-,11 Previous imaging studies in people with LE, however, are limited in 2 regards: first, in a number of studies, multiple data points per subject are usually sampled to discrete and often arbitrarily defined disease stages (e.g., early vs late or acute vs chronic),9,-,13 which is an oversimplified approach unable to capture the by nature continuous dynamics of individual disease trajectories. Second, while the term limbic encephalitis suggests a pathology limited to limbic brain structures,14 disrupted white matter integrity,12,15 network topology,14,16,17 and altered gray matter metabolism18,19 make it seem likely that neocortical gray matter is involved during disease progression.

To overcome these issues and to allow a meaningful longitudinal view on both subcortical and neocortical gray matter, we included all available MRI data from people with antibodies treated at our department together with a longitudinal healthy control group and analyzed the resulting dataset by means of linear mixed-effects models. This type of model allows to include a different number of data points per subject. It has proven to be useful in imaging studies on neurodegenerative diseases20,21 and in the modeling of progressive neurodegenerative processes in epilepsy.22,23 However, despite the compelling advantages of linear mixed-effects models, individual outliers and possible circadian or nonlinear courses are still being overseen on a group level, yet may contribute to a deeper understanding of disease trajectories in LE. For this reason, we arranged all raw data that were included in our statistical analyses in easily readable charts and presented them in this study as a complement to the results of the linear mixed-effects models (Figure 5, eAppendix 1, links.lww.com/NXI/A862).

We hypothesized that linear mixed-effects models would establish sero-specific continuous relations between mesiotemporal inflammation (e.g., amygdalar swelling) usually seen at earlier disease stages and mesiotemporal sclerosis in later stages, differentiate between disease-specific neocortical degeneration and normal aging, and link atrophy patterns to disease severity.

Methods

Study Group

Clinical and MRI data from people with LE who were treated at the Department of Epileptology at the University Hospital in Bonn between 2005 and 2019 and met the diagnostic criteria by Graus et al.3 including serologically proven antibodies were retrospectively ascertained. We aimed to conduct our analyses with a study cohort as homogeneous as possible. For this reason, individuals with paraneoplastic and individuals with antibody-negative LE were excluded. Disease onset was defined as the first occurrence of LE-related symptoms. Clinical information to each scan time point consisting of the patient's neuropsychological performance, type of immunomodulatory treatment, and number of antiseizure medication were retrieved from the hospital records. We defined one predominantly affected hemisphere according to a 2-step classification based on the interictal EEG and mesiotemporal volumetry, as performed earlier.9,15 Because it is subject to an ongoing debate whether LE is a unihemispheric process,9,10,15 we referred to the opposite hemisphere as contralateral hemisphere. As a control group, we selected 41 individuals with no history of psychiatric or neurologic disorders with more than 1 MRI scan available. A maximum of 8 MRI scans were included per control (median = 3). Table 1 lists an overview over the study group. All participants had to be older than 18 years at MRI examination.

View this table:
Table 1

Subjects Overview

Standard Protocol Approvals, Registrations, and Patient Consents

Written informed consent was obtained from all participants in the study. The study was approved by the internal review board of the University Hospital Bonn (ethics vote no. 136/19).

Serologic Analysis

Antibody screening was performed at the Department of Neuropathology at the University Hospital in Bonn, as outlined in detail earlier.24 Before 2014, the detection of GAD antibodies in serum was performed by using an anti125 I-GAD radioimmunoprecipitation assay (normal values ≤1 U/mL; Weatherall Institute, Oxford, UK, or Euroimmun, Lübeck, Germany). Antibodies against LGI1 and CASPR2 were examined by indirect immunofluorescence using formalin-fixed Human Embryonic Kidney 293 (HEK293) cells containing membrane-bound LGI1 or CASPR2 (normal values <1: 10; all tests performed by Euroimmun). After 2014, screening for onconeuronal antibodies was performed using semiquantitative immunoblots (EUROLINE PNS 12; Euroimmun, DL 1111–1601-7 G) coated with recombinant antigen or antigen fragments (dilution: serum 1:100, cerebrospinal fluid: 1:1). Moreover, immunocytochemistry was performed using HEK293 cells with expression of antigens on the cell surface (IIFT: Autoimmune-Encephalitis-Mosaik1, Euroimmun, FA 1120–1005-1; GAD65-IIFT, Euroimmun, FA 1022–1005-50) for NMDAR, CASPR, LGI1, GABAB, AMPAR, and GAD65 antibodies (dilution: serum 1:10, cerebrospinal fluid: 1:1).9,10,15,16 Thirty people were identified with GAD-LE (30/30 > 1,000 U/mL, 19/30 > 2,000 U/mL, 5/30 > 10,000 U/mL), 15 with LGI1-LE, 9 with CASPR2-LE, and 5 with NMDAR-LE.

Neuropsychological Testing

Psychometric testing was available for 227 time points for verbal memory performance and 223 time points for figural memory. Verbal memory performance was assessed by the verbal learning and memory test.25 Abilities in figural learning were tested by the revised Diagnosticum fuer Cerebralschaedigung.26 Memory parameters were standardized according to a conormalization sample of 488 healthy volunteers (mean = 100, SD = 10), applying a correction for age.15,16 A person was categorized as impaired if the individual memory performance was in at least half of all available time points 2 SDs below the mean of the normative sample.

Image Acquisition

All participants underwent T1-weighted structural imaging at the University Hospital in Bonn using a 3T MRI scanner at the Life & Brain Center (Magnetom Trio; Siemens Healthineers, Erlangen, Germany). Due to a scanner hardware update in 2014, images acquired after this update were acquired with slightly adjusted sequences parameters. Before the update: voxel size = 1 × 1 × 1 mm3, repetition time = 1,570 ms, echo time = 3.42 ms, flip angle = 15°, matrix = 256 × 256 pixel. After the update: voxel size = 0.8 × 0.8 × 0.8 mm3, repetition time = 1,660 ms, echo time = 2.54 ms, flip angle = 9°, matrix = 320 × 320 pixel.

Image Preprocessing and Segmentation

Cortical reconstruction and volumetric segmentation of T1 structural images were performed using the longitudinal framework in FreeSurfer (version 6.0).27 It creates a robust, within-subject template for each subject and therefore increases the accuracy of the automatic surface reconstruction and segmentation of brain MRI of multiple time points compared with the conventional processing stream.28 The surface reconstruction results were visually checked for accuracy. In case of segmentation defects, errors were corrected by manually adding control points and erasing voxels of nongray or nonwhite matter on the within-subject template and subsequently rerunning parts of the FreeSurfer pipeline followed by a final visual recheck of the results.

Volumetric Analysis of Amygdala and Hippocampus

Volumes of amygdala and hippocampus were specified in relation to the total brain volume without ventricles. To avoid bias due to their natural left/right volumetric asymmetry,29 we calculated a left/right correction factor based on volumetric measurements within the healthy controls for amygdala and hippocampus, respectively, and applied it to volumes of structures in the left hemisphere of all participants. Finally, we pooled left and right structures of healthy controls, which resulted in 3 categories for all amygdalar and hippocampal volumes: affected, contralateral, and healthy controls (Figure 1).

Figure 1
Figure 1 Analysis of Subcortical Volumes and Cortical Thickness

(A) After segmentation of T1 structural images, healthy amygdalar and hippocampal volumes were examined for left/right asymmetry. Subsequently, a left/right asymmetry factor of the healthy structures was applied to all left amygdalar and hippocampal volumes. Left and right amygdalar and hippocampal volumes of people with antibodies were reordered into affected and contralateral, according to a published9,15 2-step classification scheme and compared with the corresponding pooled subcortical structures of healthy controls. (B) Cortical thickness was computed as the minimum distance between the gray/white matter boundary and the pial surface. Parcellation of the cortical surface was conducted using the Desikan-Killiany cortical atlas,31 resulting in 68 anatomical regions across both hemispheres. Cortical thickness of people with antibodies was compared regionwise with cortical thickness of healthy controls.

Cortical Thickness Analysis

Thickness of cortical gray matter was computed as the minimum distance between the gray/white matter boundary and the pial surface.30 Cortical surface parcellation into standard gyral-based neuroanatomical regions was automatically conducted as previously described,31 using the Desikan-Killiany cortical atlas.32 For further analysis, we calculated the average cortical thickness of each of the resulting 68 anatomical regions across both hemispheres (Figure 1). In cortical analyses, we used the native left/right orientation due to natural asymmetry of the human cortex regarding both morphometric measures and functional specialization.

Statistical Analysis

Statistical analyses were performed with the freely available Python modules statsmodels33 (version 0.13.0) and scikit-posthocs34 (version 0.6.4).

We chose a linear mixed-effects model for longitudinal data21,35,36 to compare subcortical volumes and cortical thickness over time between people with LE and healthy controls. For amygdala and hippocampus, we defined our model as

Embedded Image

where Yij is amygdalar or hippocampal volume of individual i at scan time point j, β0, − β4 are fixed-effects regression coefficients, and b is a random-effects regression coefficient. The random effects enable modeling of individual-specific intercepts.36 We defined Ai as the age at baseline of participant i, which we consider the age at disease onset for people with antibodies or age at first scan for healthy controls. Xij denotes the time since baseline of participant i at scan time point j. We denoted Gi as the categorial group affiliation of individual i and included an interaction term of group affiliation Gi with time since baseline Xij to investigate differences in longitudinal volume changes between groups. We applied 2 paradigms to assign individuals to groups: First, we performed the analysis with participants categorized by the underlying antibody, resulting in 5 groups: GAD, LGI1, CASPR2, NMDAR, and healthy controls. Second, we classified people with antibodies by verbal and by figural memory performance, each resulting in 3 groups: impaired, unimpaired, and healthy controls.

To analyze cortical thickness changes over time in people with antibodies when compared with those in healthy controls, we used a similar model as described earlier (equation 1), with the thickness Tij at a certain cortical region of participant i at scan time point j as

Embedded Image

Again, we denote Xij as time since baseline of participant i at time point j. Because we investigated a total of 68 anatomical regions per hemisphere, we applied the Benjamini-Hochberg procedure to control for the false discovery rate (FDR) of multiple comparisons with an FDR threshold of 0.05.33,37 In all regression models, we accounted for the 2 different MRI sequences by introducing the variable sequence and included sex as another fixed effect. In the model for thickness changes (equation 2), moreover, we added the total brain volume without ventricles as a fixed effect. We tested for significance of our models by comparing them with a null model containing only the random effects individual-specific intercepts. For this purpose, we applied likelihood ratio (LLR) tests and considered the Akaike information criterion (AIC) as an estimator for information loss. In this study, p values less than 0.05 were considered significant.

Data Availability

Data that support the findings of this study are available on request from the corresponding author. Data are not publicly available because they contain information that could compromise the privacy of research participants.

Results

Clinical Group Characteristics

Two hundred fifty-seven MR images from 59 people with antibodies including follow-up MRI scans up to 10 years after disease onset (maximum number of MRI scans per participant = 13; median = 4) were included. An overview of the distribution of age, sex, and memory performance in the psychometric testing is summarized in Table 1. All people with antibodies received a pulse steroid therapy. Intravenous immunoglobulin or plasmapheresis was applied to 13 people with GAD-LE (43%), 7 with LGI-LE (46%), 3 with CASPR2-LE (33%), and 3 with NMDAR-LE (60%). Oral immunosuppressants were taken by 11 people with GAD-LE (37%), 4 with LGI-LE (27%), 2 with CASPR2-LE (22%), and 1 with NMDAR-LE (20%).

Volumetric Analysis of Amygdala and Hippocampus

Antibody-Specific Subgroups

AIC and an LLR test revealed significant superiority of our model estimating amygdalar (AICfull_model = −11,734.623, AIC difference to null model [AICnull – AICfull_model] = 112.563, LLR = 152.563, p < 0.001) and hippocampal (AICfull_model = −11,033.764, AICnull – AICfull_model = 167.437, LLR = 207.437, p < 0.001) volume changes against the null model (equation 1). The amygdalar volume at disease onset was significantly higher than in controls in the affected and contralateral hemisphere of all antibody subgroups (all p < 0.048), except for the contralateral side in NMDAR-LE. Over time, we observed a significantly stronger decrease relative to controls in the affected hemisphere in CASPR2-LE (p < 0.001) and both hemispheres in LGI1-LE (p < 0.001) and NMDAR-LE (p < 0.007). The hippocampal volume at disease onset was significantly higher than in controls in the affected hemisphere only in CASPR2-LE (p = 0.003) and GAD-LE (p = 0.014). Regarding hippocampal atrophy, we observed a significantly stronger volume decrease relative to controls in the affected and contralateral hemisphere of all antibody subgroups (all p < 0.017), except for the contralateral side in GAD-LE. All results of the mixed-effects regression model are visualized in Figure 2. Regression coefficients of the estimated group intercept at baseline (β3) and the estimated volume change over time (β4) are listed in Table 2.

Figure 2
Figure 2 Amygdalar and Hippocampal Volumes Longitudinally Categorized by Antibodies

Intercepts of straight lines represent the estimated volume at baseline (β3), and slopes show the estimated volume change over time (β4). Volumes were specified in relation to the total brain volume without ventricles. Left and right amygdalar/hippocampal volumes of people with antibodies were reordered into affected and contralateral, according to a 2-step classification scheme as performed earlier and compared with the corresponding pooled subcortical structures of healthy controls (bold gray line). On the left, group estimated volumes are displayed as percentage in relation to the estimated volume of pooled healthy controls at baseline (defined as disease onset of people with antibodies/first scan of healthy controls); on the right, 120 months after baseline (after 10 years). CASPR2 = contactin-associated protein-like 2; GAD = glutamic acid decarboxylase 65; LGI1 = leucine-rich glioma-inactivated 1; NMDAR = N-methyl-d-aspartate receptor.

View this table:
Table 2

Mixed-Effects Linear Regression Results of Amygdalar and Hippocampal Volumes for Serological Subgroups and Subgroups of Impaired and Unimpaired Memory

Subgroups of Impaired and Unimpaired Memory

AIC and the LLR test revealed significant superiority of the models estimating amygdalar (verbal memory: AICfull_model = −11,709.390, AICnull – AICfull_model = 87.329, LLR = 111.329, p < 0.001; figural memory: AICfull_model = −11,702.716, AICnull – AICfull_model = 80.655, LLR = 104.655, p < 0.001) and hippocampal (verbal memory: AICfull_model = −11,006.293, AICnull – AICfull_model = 139.966, LLR = 163.966, p < 0.001; figural memory: AICfull_model = −11017.933, AICnull – AICfull_model = 151.606, LLR = 175.606, p < 0.001) volume changes when tested against the null model, with people with LE categorized by verbal memory performance and by figural memory performance (equation 1). Regarding the atrophy over time, we observed a significant volume decrease of hippocampus and amygdala relative to controls in individuals both in the affected and contralateral hemispheres (all p < 0.002). In individuals with unimpaired verbal memory, we found a significant volume decrease relative to controls of the affected and contralateral hippocampus (both p < 0.001), and the contralateral amygdala (p = 0.012). In cases with unimpaired figural memory, we found only a significant volume decrease relative to controls of the affected hippocampus (p = 0.002) and amygdala (p = 0.032). Results of the memory performance categorization are visualized in Figure 3. Regression coefficients (β3: intercept – estimated volume at baseline; β4: slope – estimated volume change over time) are listed in Table 2.

Figure 3
Figure 3 Amygdalar and Hippocampal Volumes Longitudinally Categorized by Memory Performance

Volumes were specified in relation to the total brain volume without ventricles. Left and right amygdalar/hippocampal volumes of people with antibodies were reordered into affected and contralateral, according to a 2-step classification scheme as published earlier and compared with the corresponding pooled subcortical structures of healthy controls (bold gray line). On the left, estimated volumes are displayed as percentage of the estimated volume of the corresponding structure of pooled healthy controls at baseline (defined as disease onset of people with antibodies/first scan of healthy controls); on the right, 120 months after baseline (after 10 years).

Cortical Thickness Analysis

Antibody-Specific Subgroups

All 68 mixed-effect regression models for all 68 cortical regions of interest (equation 2) were significantly superior to a null model containing only the individual-specific intercepts. In people with GAD-LE, cortical atrophy did not differ from healthy controls after FDR correction at any cortical region. In people with LGI1-LE, we observed a significantly increased atrophy in the right entorhinal cortex compared with that in healthy controls. In people with NMDAR-LE, we found only a diminished atrophy rate in the right caudal anterior cingulate cortex. In people with CASPR2-LE, we found significantly diminished atrophy rates compared with healthy controls in the right caudal and rostral anterior cingulate cortex and a significantly increased atrophy for numerous cortical regions. See Figure 4 for a graphical representation of the regression results. Interactive 3D visualizations of all cortical thickness results, allowing a better view on all cortical regions, can be found in the online supplementary material (eAppendix 2, links.lww.com/NXI/A863). A full list of all regression results is summarized in eTable 1 (links.lww.com/NXI/A864) in the online supplement.

Figure 4
Figure 4 Cortical Thickness Development Over Time

Right and left hemisphere cortex overviews showing cortical thickness development differences between people with antibodies and healthy controls. Regions are marked with an arrow, where cortical atrophy rates of people with antibodies are significantly (p < 0.05) increased compared with healthy controls. People with antibodies were categorized by serology (A), verbal memory performance (B), and figural memory performance (C) and compared with healthy controls. Red: lower cortical atrophy rate of people with antibodies when compared with healthy controls; Blue: higher cortical atrophy rate of people with antibodies when compared with healthy controls. CASPR2 = contactin-associated protein-like 2; GAD = glutamic acid decarboxylase 65; LGI1 = leucine-rich glioma-inactivated 1; NMDAR = N-methyl-d-aspartate receptor.

Subgroups of Impaired and Unimpaired Memory

The group of people with antibodies and impaired verbal memory performance showed multiple cortical regions with a faster atrophy than healthy controls. Atrophy was diminished compared with that in healthy controls at the anterior cingulum (eTable 2, links.lww.com/NXI/A865). People with antibodies and unimpaired verbal memory showed no regions with significantly different atrophy compared with healthy controls. People with antibodies and impaired figural memory performance had a significantly slower cortical thickness reduction than healthy controls in the right caudal anterior cingulate cortex. The unimpaired group showed a significantly stronger atrophy than healthy controls in the right and left entorhinal cortex and the right temporal pole. See Figure 4 and the online supplementary material for a visualization of the regression results (eAppendix 2, links.lww.com/NXI/A863). A full list of all regression results is listed in eTable 2.

Individual Disease Courses

A comprehensive graphical representation of each individual clinical disease trajectory of all people with antibodies can be found in the supplementary material (eAppendix 1, links.lww.com/NXI/A862). In these synoptic representations (see 2 representative examples in Figure 5), we have included verbal and figural memory performance, the number of antiseizure medication prescribed during scan, immunomodulatory therapy, and amygdalar and hippocampal volumes.

Figure 5
Figure 5 Examples of Individual Clinical Disease Trajectories

Blue area in the background represents number of antiseizure medication prescribed. A full overview of all people with antibodies and healthy controls can be found in the online supplement. GAD = glutamic acid decarboxylase 65; LGI1 = leucine-rich glioma-inactivated 1.

Discussion

Clinical Group Characteristics

In our study cohort, GAD-LE and NMDAR-LE were significantly younger than LGI1-LE and CASPR2-LE at disease onset, which corresponds to other cohorts reported in the literature.1,3,38,39

Verbal or figural memory impairment could not be observed in 1 serogroup particularly but was generally found in all groups. This might be explained by the fact that people who recuperate also improve in memory performance and drop out of our retrospective cohort. Given that most of the cases with LE included in this study showed a left hemispheric disease focus that is typically linked to verbal memory impairment,40 it seems surprising that in our cohort, figural memory deficits are as frequent as verbal memory deficits. Nevertheless, this observation is concordant with previous studies showing that often both verbal and figural memory are compromised in LE, irrespective of the lateralization of the disease focus in MRI.13,17,41,42 This may be explained by more recent research suggesting that LE involves memory networks as a whole rather than single nodes of it.16,17,43

Furthermore, all people with LE received at least once a pulse steroid therapy. Intravenous immunoglobulin therapy, plasmapheresis, and oral immunosuppressants were applied in all our groups equally. The lack of clear differences in therapy between serogroups can be explained by the fact that over the years a therapy scheme has been developed and applied in most cases of our cohort.

Mesiotemporal Volumetry Reflects Anticipated Course of LE: Inflammation, Remission, and Atrophy

Our findings on the group level reflect the notion that LE progresses in 3 stages: at first, an increased amygdalar volume at disease onset is consistently found across all serogroups and corresponds to results of previous cross-sectional studies in the acute stage. This is most likely due to neuroinflammation, which is known to present with hyperintense signal in T2-weighted MRI3 and volume increase.9,11 Second, a subsequent reduction of the amygdalar volume throughout disease progression was observed in LGI1-LE, CASPR2-LE, and NMDAR-LE and may be interpreted as the remission of neuroinflammation after immunotherapy. In GAD-LE, by contrast, the absence of volume reduction may be due to persisting inflammation, leading further to a poorer clinical prognosis and lower rates of recovery, as reported in the literature.39 Third, we observed that the volume reduction fell even further below the volume of healthy controls, possibly mirroring neuronal cell loss of the predamaged structure. Of note, we observed that more severe neuropsychological impairment was associated with greater swelling of the amygdala at disease onset and greater subsequent atrophy rates, underscoring the functional relevance of the imaging findings discussed earlier.

Furthermore, we observed decreasing hippocampal volume over time in all serogroups, most impressively in the LGI1 group. Indeed, hippocampal sclerosis is known to be a common long-term consequence of LE.38,44 Previous studies associated diminishing hippocampal volume with neuronal cell loss and impaired memory performance,42 which explains the coobservation of neuropsychological impairment and hippocampal atrophy over time.

Bilateral vs Unilateral Mesiotemporal Involvement

Whether LE is a bilateral disease or indeed has a unihemispheric focus is subject to an ongoing debate.9,10,15,16 In our data, the volume of the primarily affected amygdala, as determined by EEG focus, was higher than on the contralateral side, although increased amygdala volume at disease onset was found bilaterally in all serogroups. Moreover, increased hippocampal volume was found only in the primarily affected hemisphere of people with CASPR2-LE and GAD-LE, and atrophy rates of the hippocampus in the primarily affected hemisphere were greater than in the contralateral hemisphere in all serogroups. Overall, our results show bilateral involvement, but still support the hypothesis that the disease progression is more pronounced in 1 hemisphere, which in the case of LGI1-LE is supported by previous work.18

Neocortical Degeneration Varies Between Serogroups and Determines Disease Severity

There is growing evidence that cerebral inflammatory diseases, not least LE, should rather be considered as network disorders.14,16,17,43 Accordingly, we observed distinct gray matter atrophy of the neocortex in most LE subgroups. Neocortical involvement at disease onset is anecdotally described in the literature45,-,47; however, parietal neocortical involvement is documented in the radiologic reports in only 1 case with GAD-LE from our cohort. In all other cases, radiologic reports at disease onset only state mesiotemporal abnormalities.

In LGI1-LE, we observed increased cortical atrophy when compared with aging in healthy controls in the entorhinal cortex, which lies in anatomical proximity to the inflammatory focus located in hippocampus and amygdala. In CASPR2-LE, we found widespread higher atrophy rates compared with aging in healthy controls over the entire cortex, among others the motor cortex and its associated cortical areas.

In GAD-LE, we did not find any cortical area with increased atrophy when compared with aging in healthy controls. This was contrary to our expectations because people with GAD-LE often have a severe disease course with less respondence to immunotherapy, which should necessarily lead to cortical neurodegeneration.12 We may, however, not see this effect due to the particularly individual and highly variable disease trajectories in this serogroup, which may hamper group-level statistical inference. Furthermore, cortical atrophy may be overseen in GAD-LE for developmental reasons because people with LGI1-LE and CASPR2-LE are on average older at disease onset than people with GAD-LE and are therefore by nature more susceptible to cortical degeneration. This result may be further biased by the fact that the healthy control cohort included in this study is on average younger at baseline than the LGI1-LE and CASPR2-LE subgroups.

Of interest, people with antibodies who are impaired in verbal memory performance present with a higher gray matter atrophy rate than healthy controls, distributed over the entire cortex, while those who are less affected regarding verbal memory do not show significant differences relative to healthy controls. Verbal memory performance is one of the most relevant cognitive deficits in LE and is as such a fairly well-established proxy of disease burden. In this regard, our data support the concept outlined in previous studies that it is not the structural damage to the hippocampus alone but rather the involvement of further anatomical structures beyond it that determines disease severity.14,16,17

Sero-specific Patterns and Immunotherapy Response in Individual Clinical Disease Trajectories

By visualizing the individual volume development of amygdala and hippocampus of both hemispheres, we observed similar patterns within each serogroup. In LGI1-LE, we found either initially increased volumes, which then subside (e.g., participants 2, 31, and 43), or no substantial alteration in the volume of amygdala or hippocampus (e.g., participants 24 and 12) over time. In 4 of the 6 people with CASPR2-LE with more than 2 scan time points, we observed a rather monophasic decrease in the volume of all structures (participants 13, 18, 20, and 35) as seen in most people with LGI1-LE. People with GAD-LE, by contrast, present with a heterogeneous picture of volume changes of amygdala and hippocampus throughout disease progression. The most common pattern in this group, however, is the one in which the volume does not change much over time (e.g., participants 10, 14, 17, 19, and 44).

Regarding therapeutic interventions, previous studies have shown a favorable clinical response to immunotherapy, especially in LGI1-LE.48,-,50 This is supported by our data: mesiotemporal volumes in LGI-LE mostly decrease over time, and neuropsychological performance often recovers under therapy. We found the same response also in fewer people with different antibodies (GAD-LE: participants 22, 38, and 34; CASPR2-LE: participant 20).

Despite all the advantages of discrete statistical models, linear mixed-effects models still represent a linear approximation. As a result, nonlinear or cyclical patterns are not uncovered on group level and can only be surmised from the raw data. Another limitation lies in the retrospective design of our study. The availability of follow-up may be biased in that people with antibodies are more likely to attend follow-up appointments if they continue to experience symptoms due to a more severe disease course, whereas fully recovered people with antibodies are more likely not to attend these appointments and therefore dropout of our cohort earlier. Conversely, people with LE and less obvious symptoms at disease onset may present to a specialized center such as our department only after a certain delay. Our longitudinal study design is also challenged by therapeutic interventions. These are mostly based on the individual disease courses and include both immunotherapeutic and antiseizure medication. Both may potentially interfere with the variables included in our analyses; however, from a statistical perspective, it remains practically impossible to control for this effect in the statistical analyses due to the high variability. Regarding the generalizability of our results, it should be noted that cases with paraneoplastic LE, antibody-negative LE, and LE associated with other less common antibodies such as GABAB or AMPAR were not included in our study. With its focus on LE, this study also ignores most individuals with NMDAR antibody encephalitis who do not fulfill diagnostic criteria for LE (whereas NMDAR antibodies are far more common among “nonlimbic” autoimmune encephalitis). Finally, it should be noted that all individuals with LE received therapy, and for this obvious reason, our data do not depict the natural course of LE.

Our retrospective study puts mesiotemporal volumetry and neocortical thickness of a rather large LE cohort into a meaningful longitudinal statistical model. Regarding mesiotemporal structures, our model reflects the common pathophysiologic course of LE across all observed serogroups: Volume increase, most likely due to edematous swelling in the early disease stage, is followed by a remission with volume normalization and finally atrophy in late disease stages. While seen in both hemispheres, this pattern is more pronounced in the primarily affected hemisphere pointing toward a lateralized pathology.

Neocortical atrophy shows serotype-dependent regional patterns. Furthermore, impaired verbal memory and thus more severe disease courses in people with antibodies could be linked to higher neocortical atrophy.

Taken together, our study reveals a common, continuous, and pathophysiologically meaningful pattern of mesiotemporal volumetry across all serogroups and provides further evidence that LE should be considered a network disorder in which extratemporal involvement is an important factor regarding disease severity.

Study Funding

A. Harms and T. Bauer received support from the BonnNI Promotionskolleg Neuroimmunology of the University of Bonn and the Else-Kröner-Fresenius Stiftung (grants 2020-S1-01, 2018-S2-01). T.B. received support from the BONFOR research commission of the medical faculty of the University of Bonn (grant 2019-4-07). This work was supported by the Verein zur Förderung der Epilepsieforschung.

Disclosure

J.A. Witt reports personal fees from Eisai. These activities were not related to the content of this article. R. Surges has received fees as speaker or served on the advisory board of Angelini, Arvelle, Bial, Desitin, Eisai, Livanova, Janssen-Cilag GmbH, Novartis, Precisis GmbH, UCB Pharma, UnEEG, and Zogenix. These activities were not related to the content of this article. Go to Neurology.org/NN for full disclosures.

Appendix Authors

Table

Footnotes

  • Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

  • The Article Processing Charge was funded by the authors.

  • Submitted and externally peer reviewed. The handling editor was Josep O. Dalmau, MD, PhD, FAAN.

  • Received November 19, 2022.
  • Accepted in final form March 30, 2023.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

References

  1. 1.↵
    1. Dalmau J,
    2. Graus F
    . Antibody-mediated encephalitis. N Engl J Med. 2018;378(9):840-851. doi:10.1056/NEJMra1708712
    OpenUrlCrossRefPubMed
  2. 2.↵
    1. Bien CG
    . Limbic encephalitis. Handb Clin Neurol. 2022;187:467-487. doi:10.1016/B978-0-12-823493-8.00024-9
    OpenUrlCrossRef
  3. 3.↵
    1. Graus F,
    2. Titulaer MJ,
    3. Balu R, et al
    . A clinical approach to diagnosis of autoimmune encephalitis. Lancet Neurol. 2016;15(4):391-404. doi:10.1016/S1474-4422(15)00401-9
    OpenUrlCrossRefPubMed
  4. 4.↵
    1. Linnoila J
    . Autoimmune encephalitis: new hammers in the toolbox. J Neurol Neurosurg Psychiatry. 2021;92(7):686. doi:10.1136/jnnp-2021-326096
    OpenUrlFREE Full Text
  5. 5.↵
    1. Witt JA,
    2. Helmstaedter C
    . Neuropsychological evaluations in limbic encephalitis. Brain Sci. 2021;11(5):576. doi:10.3390/brainsci11050576
    OpenUrlCrossRef
  6. 6.↵
    1. Dubey D,
    2. Pittock SJ,
    3. Kelly CR, et al
    . Autoimmune encephalitis epidemiology and a comparison to infectious encephalitis. Ann Neurol. 2018;83(1):166-177. doi:10.1002/ana.25131
    OpenUrlCrossRefPubMed
  7. 7.↵
    1. Graus F,
    2. Dalmau J
    , editorsOverview. In: Autoimmune Encephalitis and Related Disorders of the Nervous System. Cambridge University Press; 2022:1-106.
  8. 8.↵
    1. Abboud H,
    2. Probasco JC,
    3. Irani S, et al
    . Autoimmune encephalitis: proposed best practice recommendations for diagnosis and acute management. J Neurol Neurosurg Psychiatry. 2021;92(7):757-768. doi:10.1136/jnnp-2020-325300
    OpenUrlAbstract/FREE Full Text
  9. 9.↵
    1. Ernst L,
    2. David B,
    3. Gaubatz J, et al
    . Volumetry of mesiotemporal structures reflects serostatus in patients with limbic encephalitis. Am J Neuroradiol. 2019;40(12):2081-2089. doi:10.3174/ajnr.A6289
    OpenUrlAbstract/FREE Full Text
  10. 10.↵
    1. Wagner J,
    2. Weber B,
    3. Elger CE
    . Early and chronic gray matter volume changes in limbic encephalitis revealed by voxel-based morphometry. Epilepsia. 2015;56(5):754-761. doi:10.1111/epi.12968
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Malter MP,
    2. Widman G,
    3. Galldiks N, et al
    . Suspected new-onset autoimmune temporal lobe epilepsy with amygdala enlargement. Epilepsia. 2016;57(9):1485-1494. doi:10.1111/epi.13471
    OpenUrlCrossRef
  12. 12.↵
    1. Wagner J,
    2. Schoene-Bake JC,
    3. Witt JA, et al
    . Distinct white matter integrity in glutamic acid decarboxylase and voltage-gated potassium channel-complex antibody-associated limbic encephalitis. Epilepsia. 2016;57(3):475-483. doi:10.1111/epi.13297
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Frisch C,
    2. Malter MP,
    3. Elger CE,
    4. Helmstaedter C
    . Neuropsychological course of voltage-gated potassium channel and glutamic acid decarboxylase antibody related limbic encephalitis. Eur J Neurol. 2013;20(9):1297-1304. doi:10.1111/ene.12186
    OpenUrlCrossRefPubMed
  14. 14.↵
    1. Heine J,
    2. Prüss H,
    3. Kopp UA, et al
    . Beyond the limbic system: disruption and functional compensation of large-scale brain networks in patients with anti-LGI1 encephalitis. J Neurol Neurosurg Psychiatry. 2018;89(11):1191-1199. doi:10.1136/jnnp-2017-317780
    OpenUrlAbstract/FREE Full Text
  15. 15.↵
    1. Bauer T,
    2. Ernst L,
    3. David B, et al
    . Fixel-based analysis links white matter characteristics, serostatus and clinical features in limbic encephalitis. NeuroImage Clin. 2020;27:102289. doi:10.1016/j.nicl.2020.102289
    OpenUrlCrossRef
  16. 16.↵
    1. Bauer T,
    2. David B,
    3. Ernst L, et al
    . Structural network topology in limbic encephalitis is associated with amygdala enlargement, memory performance and serostatus. Epilepsia. 2020;61(10):e140-e145. doi:10.1111/epi.16691
    OpenUrlCrossRef
  17. 17.↵
    1. Loane C,
    2. Argyropoulos GPD,
    3. Roca-Fernández A, et al
    . Hippocampal network abnormalities explain amnesia after VGKCC-Ab related autoimmune limbic encephalitis. J Neurol Neurosurg Psychiatry. 2019;90(9):965-974. doi:10.1136/jnnp-2018-320168
    OpenUrlAbstract/FREE Full Text
  18. 18.↵
    1. Navarro V,
    2. Kas A,
    3. Apartis E, et al
    . Motor cortex and hippocampus are the two main cortical targets in LGI1-antibody encephalitis. Brain. 2016;139(4):1079-1093. doi:10.1093/brain/aww012
    OpenUrlCrossRefPubMed
  19. 19.↵
    1. Huang G,
    2. Xin M,
    3. Hao Y,
    4. Bai S,
    5. Liu J,
    6. Zhang C
    . Cerebral metabolic network in patients with anti-N-Methyl-D-aspartate receptor encephalitis on 18F-FDG PET imaging. Front Neurosci. 2022;16:885425. doi:10.3389/fnins.2022.885425
    OpenUrlCrossRef
  20. 20.↵
    1. Thompson WK,
    2. Hallmayer J,
    3. O'Hara R
    , the Alzheimer’s Disease Neuroimaging Initiative. Design considerations for characterizing psychiatric trajectories across the lifespan: application to effects of APOE-ε4 on cerebral cortical thickness in Alzheimer's disease. Am J Psychiatry. 2011;168(9):894-903. doi:10.1176/appi.ajp.2011.10111690
    OpenUrlCrossRefPubMed
  21. 21.↵
    1. Bernal-Rusiel JL,
    2. Greve DN,
    3. Reuter M,
    4. Fischl B,
    5. Sabuncu MR
    . Statistical analysis of longitudinal neuroimage data with Linear Mixed Effects models. NeuroImage. 2013;66:249-260. doi:10.1016/j.neuroimage.2012.10.065
    OpenUrlCrossRefPubMed
  22. 22.↵
    1. Bernhardt BC,
    2. Worsley KJ,
    3. Kim H,
    4. Evans AC,
    5. Bernasconi A,
    6. Bernasconi N
    . Longitudinal and cross-sectional analysis of atrophy in pharmacoresistant temporal lobe epilepsy. Neurology. 2009;72(20):1747-1754. doi:10.1212/01.wnl.0000345969.57574.f5
    OpenUrlCrossRefPubMed
  23. 23.↵
    1. Galovic M,
    2. van Dooren VQH,
    3. Postma TS, et al
    . Progressive cortical thinning in patients with focal epilepsy. JAMA Neurol. 2019;76(10):1230. doi:10.1001/jamaneurol.2019.1708
    OpenUrlCrossRef
  24. 24.↵
    1. Kuehn JC,
    2. Meschede C,
    3. Helmstaedter C, et al
    . Adult-onset temporal lobe epilepsy suspicious for autoimmune pathogenesis: autoantibody prevalence and clinical correlates. PLoS One. 2020;15:e0241289. doi:10.1371/journal.pone.0241289
    OpenUrlCrossRef
  25. 25.↵
    1. Helmstaedter C,
    2. Lendt M,
    3. Lux S
    . VLMT Verbaler Lern- Und Merkfähigkeitstest. 1. Auflage. Beltz Test GmbH; 2001.
  26. 26.↵
    1. Helmstaedter C,
    2. Pohl C,
    3. Hufnagel A,
    4. Elger CE
    . Visual learning deficits in nonresected patients with right temporal lobe epilepsy. Cortex. 1991;27(4):547-555. doi:10.1016/S0010-9452(13)80004-4
    OpenUrlCrossRefPubMed
  27. 27.↵
    1. Reuter M,
    2. Schmansky NJ,
    3. Rosas HD,
    4. Fischl B
    . Within-subject template estimation for unbiased longitudinal image analysis. NeuroImage. 2012;61(4):1402-1418. doi:10.1016/j.neuroimage.2012.02.084
    OpenUrlCrossRefPubMed
  28. 28.↵
    1. Reuter M,
    2. Fischl B
    . Avoiding asymmetry-induced bias in longitudinal image processing. NeuroImage. 2011;57(1):19-21. doi:10.1016/j.neuroimage.2011.02.076
    OpenUrlCrossRefPubMed
  29. 29.↵
    1. Kong XZ,
    2. Postema MC,
    3. Guadalupe T, et al
    . Mapping brain asymmetry in health and disease through the ENIGMA consortium. Hum Brain Mapp. 2020;43(1):167-181. doi:10.1002/hbm.25033
    OpenUrlCrossRef
  30. 30.↵
    1. Fischl B,
    2. Dale AM
    . Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proc Natl Acad Sci U S A. 2000;97(20):11050-11055. doi:10.1073/pnas.200033797
    OpenUrlAbstract/FREE Full Text
  31. 31.↵
    1. Fischl B
    . Automatically parcellating the human cerebral cortex. Cereb Cortex. 2004;14(1):11-22. doi:10.1093/cercor/bhg087
    OpenUrlCrossRefPubMed
  32. 32.↵
    1. Desikan RS,
    2. Ségonne F,
    3. Fischl B, et al
    . An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. NeuroImage. 2006;31(3):968-980. doi:10.1016/j.neuroimage.2006.01.021
    OpenUrlCrossRefPubMed
  33. 33.↵
    1. Seabold S,
    2. Perktold J
    . Statsmodels: Econometric and Statistical Modeling with Python. 2010:92-96. doi:10.25080/Majora-92bf1922-011
    OpenUrlCrossRef
  34. 34.↵
    1. Terpilowski M
    . scikit-posthocs: pairwise multiple comparison tests in Python. J Open Source Softw. 2019;4(36):1169. doi:10.21105/joss.01169
    OpenUrlCrossRef
  35. 35.↵
    1. Verbeke G,
    2. Molenberghs G
    . Linear Mixed Models For Longitudinal Data; Springer, 2000. doi:10.1007/978-1-4419-0300-6
    OpenUrlCrossRef
  36. 36.↵
    1. Wachinger C,
    2. Salat DH,
    3. Weiner M,
    4. Reuter M
    , Alzheimer’s Disease Neuroimaging Initiative. Whole-brain analysis reveals increased neuroanatomical asymmetries in dementia for hippocampus and amygdala. Brain. 2016;139(12):3253-3266. doi:10.1093/brain/aww243
    OpenUrlCrossRefPubMed
  37. 37.↵
    1. Benjamini Y,
    2. Hochberg Y
    . Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B. 1995;57(1):289-300. doi:10.1111/j.2517-6161.1995.tb02031.x
    OpenUrlCrossRefPubMed
  38. 38.↵
    1. Bien CG,
    2. Urbach H,
    3. Schramm J, et al
    . Limbic encephalitis as a precipitating event in adult-onset temporal lobe epilepsy. Neurology. 2007;69(12):1236-1244. doi:10.1212/01.wnl.0000276946.08412.ef
    OpenUrlAbstract/FREE Full Text
  39. 39.↵
    1. Malter MP,
    2. Helmstaedter C,
    3. Urbach H,
    4. Vincent A,
    5. Bien CG
    . Antibodies to glutamic acid decarboxylase define a form of limbic encephalitis. Ann Neurol. 2010;67(4):470-478. doi:10.1002/ana.21917
    OpenUrlCrossRefPubMed
  40. 40.↵
    1. Helmstaedter C,
    2. Grunwald T,
    3. Lehnertz K,
    4. Gleißner U,
    5. Elger CE
    . Differential involvement of left temporolateral and temporomesial structures in verbal declarative learning and memory: evidence from temporal lobe epilepsy. Brain Cogn. 1997;35(1):110-131. doi:10.1006/brcg.1997.0930
    OpenUrlCrossRefPubMed
  41. 41.↵
    1. Hansen N,
    2. Widman G,
    3. Witt JA, et al
    . Seizure control and cognitive improvement via immunotherapy in late onset epilepsy patients with paraneoplastic versus GAD65 autoantibody-associated limbic encephalitis. Epilepsy Behav. 2016;65:18-24. doi:10.1016/j.yebeh.2016.10.016
    OpenUrlCrossRef
  42. 42.↵
    1. Finke C,
    2. Prüss H,
    3. Heine J, et al
    . Evaluation of cognitive deficits and structural hippocampal damage in encephalitis with leucine-rich, glioma-inactivated 1 antibodies. JAMA Neurol. 2017;74(1):50. doi:10.1001/jamaneurol.2016.4226
    OpenUrlCrossRefPubMed
  43. 43.↵
    1. Argyropoulos GP,
    2. Loane C,
    3. Roca-Fernandez A, et al
    . Network-wide abnormalities explain memory variability in hippocampal amnesia. eLife. 2019;8:e46156. doi:10.7554/eLife.46156
    OpenUrlCrossRef
  44. 44.↵
    1. Kotsenas AL,
    2. Watson RE,
    3. Pittock SJ, et al
    . MRI findings in autoimmune voltage-gated potassium channel complex encephalitis with seizures: one potential etiology for mesial temporal sclerosis. Am J Neuroradiol. 2014;35(1):84-89. doi:10.3174/ajnr.A3633
    OpenUrlAbstract/FREE Full Text
  45. 45.↵
    1. Ibi K,
    2. Fujii K,
    3. Kobayashi H,
    4. Senda M,
    5. Kitazawa K,
    6. Honda A
    . Anterior cingulate cortex involvement in non-paraneoplastic limbic encephalitis. Brain Dev. 2019;41(8):735-739. doi:10.1016/j.braindev.2019.04.006
    OpenUrlCrossRef
  46. 46.↵
    1. Bose G,
    2. Zwicker JC,
    3. Sitwell LD,
    4. Osman N,
    5. Fantaneanu TA
    . Anti-LGI1 limbic encephalitis presenting as an expanding insular lesion. Can J Neurol Sci. 2019;46(6):770-772. doi:10.1017/cjn.2019.247
    OpenUrlCrossRef
  47. 47.↵
    1. Papiri G,
    2. Puca E,
    3. Marcucci M,
    4. Paci C,
    5. Cagnetti C
    . A case of anti-leucine-rich glioma-inactivated protein 1 (Anti-LGI1) encephalitis with an unusual frontomesial motor cortex T2 MRI hyperintensity. Cureus. 2022;14(10):e30480. doi:10.7759/cureus.30480
    OpenUrlCrossRef
  48. 48.↵
    1. Toledano M,
    2. Britton JW,
    3. McKeon A, et al
    . Utility of an immunotherapy trial in evaluating patients with presumed autoimmune epilepsy. Neurology. 2014;82(18):1578-1586. doi:10.1212/WNL.0000000000000383
    OpenUrlAbstract/FREE Full Text
  49. 49.↵
    1. Shin YW,
    2. Lee ST,
    3. Shin JW, et al
    . VGKC-complex/LGI1-antibody encephalitis: clinical manifestations and response to immunotherapy. J Neuroimmunol. 2013;265(1-2):75-81. doi:10.1016/j.jneuroim.2013.10.005
    OpenUrlCrossRefPubMed
  50. 50.↵
    1. Irani SR,
    2. Stagg CJ,
    3. Schott JM, et al
    . Faciobrachial dystonic seizures: the influence of immunotherapy on seizure control and prevention of cognitive impairment in a broadening phenotype. Brain. 2013;136(10):3151-3162. doi:10.1093/brain/awt212
    OpenUrlCrossRefPubMed

Letters: Rapid online correspondence

No comments have been published for this article.
Comment

REQUIREMENTS

You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.

Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.

If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.

Submission specifications:

  • Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
  • Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
  • Submit only on articles published within 6 months of issue date.
  • Do not be redundant. Read any comments already posted on the article prior to submission.
  • Submitted comments are subject to editing and editor review prior to posting.

More guidelines and information on Disputes & Debates

Compose Comment

More information about text formats

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Author Information
NOTE: The first author must also be the corresponding author of the comment.
First or given name, e.g. 'Peter'.
Your last, or family, name, e.g. 'MacMoody'.
Your email address, e.g. higgs-boson@gmail.com
Your role and/or occupation, e.g. 'Orthopedic Surgeon'.
Your organization or institution (if applicable), e.g. 'Royal Free Hospital'.
Publishing Agreement
NOTE: All authors, besides the first/corresponding author, must complete a separate Publishing Agreement Form and provide via email to the editorial office before comments can be posted.
CAPTCHA
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.

Vertical Tabs

You May Also be Interested in

Back to top
Advertisement

Related Articles

  • No related articles found.

Topics Discussed

Alert Me