Diagnostic utility of magnetic resonance imaging in autoimmune encephalitis: prognostic implications

Abstract

Background

The role of imaging in autoimmune encephalitis (AIE) remains unclear, and there are limited data on the utility of magnetic resonance imaging (MRI) to diagnose, treat, or prognosticate AIE.

Purpose

To evaluate whether MRI is a diagnostic and prognostic marker for AIE and assess its efficacy in distinguishing between various AIE subtypes.

Material and Methods

We analyzed data from 96 AIE patients from our prospective autoimmune registry. MRI sequences examined were FLAIR, diffusion, SWI, T2WI, ASL, and contrast enhancement. Short-term outcomes were measured using the Modified Rankin Scale (mRS) at discharge; long-term outcomes were assessed with the Functional Independence Measure (FIM) at 6 months.

Results

MRI confirmed AIE in cases of new-onset seizures (82.1%, P < 0.001) and dementia (100%, P = 0.02). Antibody-negative AIE exhibited significant multifocal FLAIR abnormalities compared to antibody-positive cases (P = 0.002). LGI1 and CASPR2 encephalitis frequently involved the mesial temporal region (P = 0.004), while ASL revealed hyperperfusion of the contralateral basal ganglia in faciobrachial dystonic seizures (P = 0.016). GAD65 encephalitis predominantly affected the cerebellum (P = 0.002), and NMDA encephalitis showed contrast enhancement in five cases (P = 0.045). MRI was not useful for predicting short-term outcomes but was associated with long-term outcomes; specifically, a normal MRI was linked to a better long-term outcome in 47.8% of patients (P = 0.035), and resolution of abnormalities correlated with a favorable FIM score (>54) in 76.7% (P = 0.016).

Conclusion

MRI is valuable for early detection of seizures or dementia as initial manifestations of AIE and for differentiating AIE subtypes. Follow-up MRI is significant in predicting long-term outcomes.

Keywords

Magnetic resonance imaging autoimmune encephalitis utility prognosis

Introduction

In the past decade, there has been a dramatic rise in the identification and characterization of autoimmune encephalitis, driven by heightened awareness and advances in research. The California Encephalitis Project revealed that anti-NMDAR (N-methyl D-aspartate receptor) encephalitis is four times more prevalent than encephalitis caused by Herpes simplex virus 1, West Nile virus, or Varicella Zoster virus (1,2).

Autoimmune encephalitis (AIE) needs to be considered in any patient with a subacute onset of rapidly progressive (within 3 months) working memory impairment, altered mentation, or psychiatric symptoms, particularly if accompanied by at least one of the following: new focal neurological deficits; unexplained seizures; cerebrospinal fluid (CSF) pleocytosis; or magnetic resonance imaging (MRI) features indicative of encephalitis, after excluding alternative causes (3). Although detecting autoantibodies through cell-based immunoassays or immunoblotting is the most specific diagnostic tool, many autoimmune encephalitis cases are antibody-negative (4). Previous studies on imaging findings in AIE associated with surface antigens have shown that MRI findings may vary widely, ranging from normal to increased T2 signal intensities, often in the mesial temporal region (5). However, there has also been reports of involvement beyond the limbic system, including other lobes, the basal ganglia, cerebellum, brainstem, spinal cord, and multifocal abnormalities (6,7). The utility of MRI for diagnosing, treating, or prognosticating AIE has not been critically analyzed so far.

The aim of the present study was to describe MRI findings in patients with antibody-positive (non-paraneoplastic) and antibody-negative AIE at a tertiary care center in India. In doing so, we sought to address the following research questions, which remain unresolved: the utility of one or multiple MRI scans for diagnosing and differentiating between the various subtypes of AIE; and in management, the role of MRI in predicting short- and long-term outcomes for patients with AIE.

Material and Methods

We collected data from patients enrolled in our prospective AIE registry, which has been operational since 2013. The registry includes cases of antibody-positive and antibody-negative AIE diagnosed between 2013 and 2018. Patient selection was based on the International Encephalitis Consortium Diagnostic Criteria for Encephalitis and Encephalopathy of Presumed Infectious or Autoimmune Etiology (8) Approval was obtained from the ethics committee of Sree Chitra Tirunal Institute for Medical Sciences and Technology (SCT/IEC/714). The procedures used in this study adhere to the tenets of the Declaration of Helsinki.

The exclusion criteria were infectious causes of encephalitis and other differential diagnoses, such as primary central nervous system angiitis, Bickerstaff's brainstem encephalitis, acute disseminated encephalomyelitis, Rasmussen's encephalitis, underlying malignancy, septic encephalopathy, metabolic encephalopathy, Hashimoto's encephalopathy, drug toxicity, cerebrovascular disease, neoplastic disorders, Creutzfeldt-Jakob disease, rheumatologic disorders (e.g. lupus, sarcoidosis), Reye’s syndrome (in children), mitochondrial diseases, and inborn errors of metabolism. These exclusion criteria were based on clinical findings and appropriate investigations.

As part of the protocol for our prospective registry, all patients underwent testing for serum and cerebrospinal fluid (CSF) antibodies to neuronal surface antigens and non-paraneoplastic antibodies. The autoimmune panel included a combination of NMDA, LGI1, CASPR2, GABABR1/B2, and AMPA1&2. For autoimmune encephalitis antibody testing, we used a fixed cell-based assay with HEK293 transfected cells (Euroimmun, Lübeck, Germany). Serum samples were diluted to 1:10, and CSF samples were tested neat (undiluted).

Our paraneoplastic panel included amphiphysin, CV2, PNMA2 (Ma2/Ta), Ri, Yo, Hu, SOX1, Recoverin, Titin, Zic4, Tr, and GAD65. Paraneoplastic antibodies were detected in serum using a commercially available immunoblot kit (Euroimmun, Lübeck, Germany) at a 1:101 dilution.

Patients underwent 3-T MRI scans (Discovery MR 750W; GE Healthcare, Milwaukie, USA), which were reviewed by two experienced neuroradiologists. The MRI sequences analyzed included T1-weighted (T1W), T2-weighted (T2W), 3D fluid attenuation inversion recovery (FLAIR), diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC), susceptibility-weighted imaging (SWI), magnetic resonance spectroscopy (MRS), arterial spin labeling (ASL), and T1 volumetric interpolated breath-hold examination (VIBE)/T1 blood-rapid volumetric acquisition (BRAVO) (post-contrast gradient sequences). We evaluated the follow-up MRI for resolution, volume loss, and cortical laminar necrosis.

Short-term outcomes were assessed using the modified Rankin Scale (mRS) at discharge, and long-term outcomes were assessed using the Functional Independence Measure (FIM) (9) at a 6-month follow-up. The follow-up period was uniform for all patients up to 6 months. An mRS score >3 and FIM score <54 was considered to have a poor outcome.

Statistical analysis

Demographic data and variables were analyzed; proportions were used for categorical variables and means with standard deviation for continuous variables. Bivariate analysis was performed using the Chi-square test for proportions. Unpaired t-tests and Mann–Whitney U-tests were used to analyze continuous variables, and the Shapiro–Wilk test was used to assess normality. The association between MRI and outcome was analyzed, and multivariate regression was carried out to evaluate the role of MRI in predicting outcomes. Variables with a P value < 0.1 were included in the multivariate regression model. A P value < 0.05 was considered statistically significant. The statistical analysis was performed using SPSS Statistics for Windows Statistics version 21.0 (IBM Corp., Armonk, NY, USA).

Results

Clinical characteristics

Of the 5459 patients admitted to the ward and intensive care unit during the study period, 96 (1.7%) were diagnosed with AIE, of which 59 (61.5%) patients were antibody positive. The mean age of the cohort was 28 ± 10 years (age range = 4–79 years). Among the antibody-positive cases, the majority were positive for anti-NMDA antibodies followed by anti CASPR2, LGI1, GAD65, and anti-glycine in that order. Seizures were the most common clinical presentation (including status epilepticus), followed by neuropsychiatric manifestations, involuntary movements, dementia, and Isaac's syndrome (Table 1).

Table 1.

Demographic, clinical, and imaging variables.

Variable	Total (n = 96)
Mean age (years)	28 ± 10
Men	45 (47)
Women	51 (53)
Antibody positive	59 (61.5)
Antibody negative	37 (38.5)
Antibody
Anti NMDA	38 (64.4)
Anti CASPR2	9 (15.3)
Anti LGI1	7 (11.9)
Anti GAD65	4 (6.8)
Anti-Glycine	1 (1.7)
Clinical presentation
Seizures	45 (47.9)
Status epilepticus	17 (17.7)
Neuropsychiatric manifestations	32 (33.3)
Involuntary movements	11 (11.5)
Dementia	6 (6.25)
Isaac's syndrome	2 (2.1)
MRI findings
FLAIR abnormalities	61 (63.5)
1. Unilateral	48 (78.6)
2. Bilateral	13 (21.4)
3. Multiple lobes	32 (54.1)
4. Mesial temporal involvement	20 (32.8)
5. Cerebellar	4 (6.6)
6. Insula	2 (3.3)
7. Basal ganglia	2 (3.3)
8. Parietal	1 (1.6)
Diffusion restriction	1 (1.6)
T2 shine-through	17 (17.7)
Contrast enhancement	18 (18.8)
ASL hyperperfusion	9 (9.4)
1. Basal ganglia	3 (3.1)
2. Mesial temporal	4 (4.2)
3. Multifocal abnormalities	2 (2.09)
Follow-up MRI	45 (46.8)
Resolution	29 (64.4)
Volume loss	28 (62.2)
Cortical laminar necrosis	3 (6.7)

Values are given as n (%) or mean ± SD.

ASL, arterial spin labeling; FLAIR, fluid attenuation inversion recovery; MRI, magnetic resonance imaging.

Diagnostic role of MRI in AIE

MRI was found to be the primary diagnostic tool for patients with AIE with initial clinical presentation of new-onset seizures or rapidly progressive dementia. However, in patients who presented with status epilepticus or neuropsychiatric manifestations, MRI served only as an ancillary investigation. In these cases, AIE was diagnosed and treatment was initiated before MRI (Table 2).

Table 2.

Role of MRI in diagnosis of AIE.

Clinical presentation of AIE	AIE diagnosed		P value*
Clinical presentation of AIE	After MRI	Before MRI	P value*
New onset seizures	23	5	<0.001
Status epilepticus	5	12	0.012
Neuropsychiatric manifestations	11	21	0.003
Dementia	6	0	0.01
Involuntary movements	7	4	0.669

AIE, autoimmune encephalitis; MRI, magnetic resonance imaging.

*Chi-square test.

MRI findings in AIE

MRI was carried out in all 96 patients, of which 61 (63.5%) had FLAIR abnormalities. Among these, majority patients showed bilateral involvement. The distribution of FLAIR abnormalities was predominantly involvement of multiple lobes, followed by involvement of mesial temporal, cerebellar, insula, basal ganglia, and parietal lobe, in that order (Table 1).

Diffusion restriction was observed in 17 (17.7%) patients, T2 shine-through in the ADC was seen in 18 (18.8%), and contrast enhancement was noted in 9 (9.4%) patients. SWI sequences did not show blooming. ASL hyperperfusion was noted in basal ganglia, mesial temporal region, and was also multifocal (Table 1). Follow-up MRI was available for 45 (46.8%) patients. Of these, resolution was noted in the majority, followed by volume loss and cortical laminar necrosis (Table 2).

MRI in antibody-positive versus antibody-negative cases

On comparing imaging findings between antibody-positive and antibody-negative cases, we noted a significantly higher prevalence of multifocal FLAIR abnormalities (Fig. 1) in antibody-negative cases (n = 19, 79.2%) than antibody-positive cases (n = 13, 37.8%) (P = 0.002). Follow-up MRI showed resolution in 15 (83.3%) antibody-positive cases compared to 14 (51.9%) antibody-negative cases (P = 0.015). No significant differences were found between the two groups with regard to other MRI parameters (Table 3).

Fig. 1.

Antibody negative encephalitis: (a) Axial FLAIR images; (b, c) axial DWI and ADC shows multiple gyriform swelling in the form of cortical and subcortical white matter hyperintensity in bilateral insula and frontotemporal region. These areas appear to be ADC hyperintense consistent with facilitated diffusion or edema. (d) T2* MR perfusion image shows high perfusion in left insulotemporal region suggestive of more cerebral blood flow due to active inflammation. ADC, apparent diffusion coefficient; DWI, diffusion-weighted imaging; FLAIR, fluid attenuation inversion recovery; MR, magnetic resonance.

Table 3.

Comparison of MRI characteristics in antibody positive and antibody negative patients.

	Total (n = 96)	Antibody positive (n = 59)	Antibody negative (n = 37)	P value*	Anti NMDA (n = 38)	Anti LGI1 & CASPR (n = 16)	Anti GAD65 (n = 4)	P value*
MRI abnormal	61 (63.5)	37 (62.7)	24 (64.9)	0.416	21 (57.9)	12 (75)	4 (100)	0.114
Laterality
Unilateral	13 (21.4)	6 (16.2)	7 (29.2)	0.114	4 (18.2)	1 (8.3)	2 (50)	0.177
Bilateral	48 (78.6)	31 (83.8)	17 (70.8)		17 (81.8)	11 (91.7)	2 (50)
FLAIR Abnormalities	61 (63.5)	37 (62.7)	24 (64.9)	0.415	21 (57.9)	12 (75)	4 (100)	0.114
FLAIR abnormalities – site
Mesial temporal lobe	20 (32.8)	15 (40.5)	5 (20.8)	0.081	6 (27.3)	9 (75)	0	0.004
Cerebellum	4 (4.9)	4 (8.1)	0	0.053	2 (9.1)	0	2 (50)	0.002
Parietal	1 (1.6)	1 (2.7)	0	0.213	1 (4.5)	0	0	0.765
Insula	2 (3.3)	2 (5.4)	0	0.129	2 (9.1)	0	0	0.538
Basal ganglia	2 (3.3)	2 (5.4)	0	0.129	0	2 (16.7)	0	0.066
Multiple	32 (54.1)	13 (37.8)	19 (79.2)	0.002	10 (50)	1 (8.3)	2 (50)	0.106
Diffusion restriction	17 (17.7)	8 (13.6)	9 (24.3)	0.089	6 (15.8)	2 (12.5)	0	0.674
T2 shine-through	18 (18.8)	4 (23.7)	4 (10.8)	0.057	5 (13.2)	7 (43.8)	2 (50)	0.026
Contrast enhancement
Present	9 (9.4)	5 (8.6)	4 (10.5)	0.351	5 (13.2)	0	0	0.045
Gyriform	2	1	1		1
Sulcal/Pial	7	4	3		4
Follow-up MRI (n = 45)
Normal	15 (15.6)	7 (38.8)	8 (29.6)	0.259	4 (33.3)	1 (25)	2 (100)	0.163
Abnormal	30 (31.3)	11 (61.1)	19 (70.4)		8 (66.7)	3 (75)	0
Resolution on follow-up MRI	29 (64.4)	15 (83.3)	14 (51.9)	0.015	8 (66.7)	4 (100)	2 (100)	0.277
Volume loss	28 (62.2)	12 (66.7)	16 (59.3)	0.308	8 (66.7)	3 (75)	1 (50)	0.829
Cortical laminar necrosis	3 (6.7)	1 (5.6)	2 (7.4)	0.404	1 (8.3)	0	0	0.767

Values are given as n (%).

*Chi-square test.

FLAIR, fluid attenuation inversion recovery; MRI, magnetic resonance imaging.

MRI findings relevant to various antibody-positive subtypes

MRI findings relevant to the various antibody subtypes are summarized in Table 3. NMDA encephalitis (see Fig. 2) showed contrast enhancement in 5 (13.2%) cases (P = 0.045). In addition, the T2 shine-through effect was less evident in patients with NMDA encephalitis (n = 5, 13.2%; P = 0.026) when compared to other etiologies.

Fig. 2.

NMDA encephalitis: (a, b) axial FLAIR and T2 images show gyral swelling and hyperintensity of right cerebral hemisphere, left posterior temporal, and perirolandic regions; (c) diffusion restriction in the above regions; (d) post-contrast leptomeningeal enhancement in the affected areas. FLAIR, fluid attenuation inversion recovery.

Patients with CASPR2 and LGI1 encephalitis (Fig. 3) had a higher prevalence of mesial temporal hyperintensities (n = 9, 75%; P = 0.004), which were bilateral and symmetric in the majority (n = 5, 56%; P = 0.019), compared to other conditions. ASL hyperperfusion of the contralateral basal ganglia (Fig. 4) was noted in three cases of LGI1 and CASPR2 encephalitis with prototype faciobrachial dystonic seizures (FBDS) (P = 0.016) (Fig. 5).

Fig. 3.

LGI1 encephalitis: subacute phase (3-month follow-up); (a) axial FLAIR images and (b) post-contrast T1 images showing symmetric FLAIR hyperintensity with atrophy of bilateral limbic region with no contrast enhancement. FLAIR, fluid attenuation inversion recovery.

Fig. 4.

CASPR2 encephalitis: (a) axial T2W image shows asymmetric hyperintensity involving bilateral caudate and lentiform nucleus with patchy involvement of cingulate and left posterior temporal lobe; (b) DWI and (c) ADC show restricted diffusion of caudate and lentiform nucleus; (d) ASL shows asymmetric hyperperfusion of left caudate and insular-perisylvian region in a patient presenting with faciobrachial dystonic seizures. ADC, apparent diffusion coefficient; ASL, arterial spin labeling; DWI, diffusion-weighted imaging; T2W, T2-weighted.

Fig. 5.

Representative MR images of a case of LGI1 positive AIE at baseline and 6-month follow-up. Baseline: (a) axial FLAIR depicting hyperintensities in bilateral anteromedial temporal lobes, bilateral hippocampi, basifrontal regions, anterolateral basal ganglia; (b) axial T2W depicting hyperintensities in bilateral anteromedial temporal lobes, insular cortices, cingulate, bilateral hippocampi, anterolateral basal ganglia; (c) axial DWI showing restriction of rim of insular cortices, bilateral cingulate; (d) ASL showed no perfusion abnormalities; (e) axial T1 contrast-enhanced images showed no enhancement. Follow-up at 6 months: (f) axial FLAIR image shows follow-up atrophy of medial temporal lobes, decrease in hyperintensities in other regions; (g) axial T2 image shows a decrease in temporal lobe hyperintensities in bilateral anteromedial temporal lobes, insular cortices, and cingulate; (h) axial DWI depicting decreased diffusion restriction of rim of insular cortices and cingulate region on follow-up; (i) axial T1 contrast-enhanced images showed no enhancement. ADC, apparent diffusion coefficient; AIE, autoimmune encephalitis; ASL, arterial spin labeling; DWI, diffusion-weighted imaging; FLAIR, fluid attenuation inversion recovery; MR, magnetic resonance.

GAD65 antibody-mediated encephalitis had a propensity for cerebellar involvement (Fig. 6), with FLAIR hyperintensities in the cerebellum evident in 2 of 4 (50%) cases (P = 0.002). The single patient with anti-glycine antibody-mediated encephalitis showed multifocal hyperintensities on MRI.

Fig. 6.

GAD encephalitis: (a) axial and coronal FLAIR images shows patchy areas of hyperintensity in bilateral cerebellar hemispheres and vermis showing GAD cerebellitis; (b) representative images of another patient – axial FLAIR images depicting the bilateral hippocampal hyperintensities and gliotic changes in left parietal operculum and left inferior parietal lobule. FLAIR, fluid attenuation inversion recovery.

Short- and long-term outcomes in patients with AIE

At discharge, 71 (74%) patients had a mRS score of 0–2, while 25 (26%) had a mRS score of >3. At the last follow-up, 64 (67%) patients had a FIM score of ≥54 and 32 (33%) had a FIM score of <54.

MRI as a prognostic marker in AIE

On assessing the MRI parameters for their utility in predicting short-term outcome, no statistical significance was noted. However, in evaluating MRI and its association with long-term outcome (Table 4), the presence of a normal MRI in 11 (47.8%) patients (P = 0.035) and resolution of MRI abnormalities in 23 (76.7%) patients with a FIM score >54 (P = 0.016) were noted to be statistically significant. Multivariate regression analysis also indicated that a normal MRI (P = 0.001) and resolution of MRI changes (P = 0.02), particularly in antibody-positive patients (P = 0.037), were associated with a better outcome (FIM >54), suggesting a definitive role for MRI as a prognostic marker.

Table 4.

MRI as a predictor of short- and long-term outcomes.

	mRS 0–2*	mRS > 3*	P value^#	FIM > 54^§	FIM <54^§	P value^#
MRI abnormal	46 (64.8)	15 (60)	0.334	43 (67.2)	18 (56.2)	0.147
FLAIR abnormalities	46 (64.8)	15 (60)	0.334	43 (67.2)	18 (56.2)	0.147
Diffusion restriction
Present	15 (21.1)	2 (8)	0.070	13 (20.3)	4 (12.5)	0.172
T2 shine-through	15 (21.1)	3 (12)	0.157	14 (21.9)	4 (12.5)	0.134
Contrast enhancement	5 (7)	4 (16)	0.093	6 (9.4)	3 (9.4)	0.500
MRI on follow-up
Abnormal				12 (52.2)	18 (81.8)	0.018
Resolution on follow-up				23 (76.7)	6 (40)	0.008

Values are given as n (%).

*Modified Rankin Scale.

Chi-square test.

^§

Functional independence measure.

Discussion

Our study involved a large cohort of patients with both antibody-positive and antibody-negative autoimmune encephalitis (AIE). The clinical presentations in these patients were varied, and a broad range of differential diagnoses were considered. We investigated the role of MRI as a diagnostic and prognostic marker in AIE and analyzed its ability to differentiate between various subtypes. In resource-limited settings, MRI can be expensive and labor-intensive, particularly for patients with AIE with acute presentation as recurrent seizures and behavioral abnormalities. However, the Antibody Prevalence in Epilepsy (APE) and Response to Immunotherapy in Epilepsy (RITE) scores, which include MRI findings as one of the components, highlight the role of MRI in the diagnosis of autoimmune epilepsy and evaluating response to immunotherapy even before antibody testing or in antibody-negative patients (10). This reiterates the importance of MRI in the diagnosis, management, and the prognosis of AIE.

We found that MRI plays an essential diagnostic role in patients with AIE presenting with new-onset seizures and rapidly progressive dementia. For new-onset seizures, in spite of similar MRI changes in post-ictal states (e.g. T2 signal abnormalities or diffusion restriction in gray or white matter) (11) and can mimic changes in infectious encephalitis, low-grade tumors, drug-induced toxicity, autoimmune encephalitis should still be considered if MRI findings include focal (more common in antibody-positive cases) or multifocal (more common in antibody-negative cases) FLAIR hyperintensities restricted to gray matter of limbic or extra-limbic regions, often with gyral swelling and without diffusion restriction in most cases. In rapidly progressive dementia, MRI findings, such as propensity for the limbic system and multifocal hyperintensities without diffusion restriction, suggest AIE and help exclude other causes like Creutzfeldt-Jakob disease, which has characteristic findings like cortical ribboning (12).

We identified specific MRI findings associated with each AIE subtype. Leptomeningeal enhancement was a prominent feature for NMDA encephalitis (Fig. 2). Previous studies showed that NMDA encephalitis often has a para-infectious trigger and the severity of neuropsychiatric manifestations associated with anti-NMDA receptor autoantibodies depends on blood–brain barrier integrity (13–15). A substantial proportion (42.1%) of patients with anti-NMDA positive encephalitis had either normal MRI or non-specific findings involving multiple regions, such as the mesial temporal lobe, cerebellum, insula, and parietal lobe. This is consistent with prior studies where MRI was normal in 21%–89% of patients (16–18). In addition, the T2 shine-through effect was less prevalent in NMDA encephalitis than in other etiologies. There are reports of posterior reversible encephalopathy syndrome due to autoimmune etiology associated with CASPR2 and LGI1 antibodies, the resultant vasogenic edema produces a T2 shine-through effect on MRI (19). In LGI1 and CASPR2-related encephalitis (Fig. 3), bilateral symmetric mesial temporal hyperintensities were observed, consistent with previous findings (20,21). Both LGI1 and CASPR2 are expressed in the hippocampus and limbic system, and thereby explain the predilection of these areas (22). ASL hyperperfusion of the contralateral basal ganglia (Fig. 4) was noted in three cases with LGI1-related encephalitis presenting with FBDS, which aligns with prior reports of perfusion, diffusion, or FLAIR abnormalities in the basal ganglia (23–26). This suggests a network hypothesis involving cortical and subcortical areas, with seizures typically originating in the temporal lobes and potentially spreading to the basal ganglia.

Anti-GAD65 antibody-positive cases (Fig. 6) showed a predilection for cerebellar involvement. Prior studies showed association of anti-GAD65 positivity with limbic encephalitis involving the mesial temporal lobes (27). Cerebellar atrophy was found in 8 of 14 patients with cerebellar ataxia who were anti-GAD65 positive earlier (28). Our study's novel finding of cerebellar involvement can be attributed to the high density of GAD-containing axon terminals in the cerebellum (29).

Our study also addressed antibody-negative autoimmune encephalitis, a diagnostic dilemma with scarce data. In patients with encephalitis of non-infectious etiology who test antibody-negative but are steroid-responsive, the presence of bilateral mesial temporal hyperintensities on T2-weighted/FLAIR MRI is highly suggestive of an immune-mediated disorder (4). Our findings showed a significantly higher prevalence of multifocal FLAIR abnormalities in antibody-negative cases and less resolution on follow-up MRI, indicating worse outcome when compared to antibody-positive cases. This expands the diagnostic armamentarium for detecting antibody-negative autoimmune encephalitis. Based on our data, the presence of normal MRI with resolution of FLAIR abnormalities on follow-up is a potential prognostic marker for long-term outcome. The greater resolution observed in antibody-positive cases compared to antibody-negative cases may be ascribed to the delay in diagnosis and initiation of immunomodulatory therapy in antibody-negative patients.

The present study has some limitations. These include the retrospective nature of the study and the need for high patient cooperation to minimize movement artefacts, resulting in ASL being performed in only 23 patients. The number of cases with anti-GAD (n = 4) and anti-glycine (n = 1) mediated encephalitis was limited, and follow-up MRI was available for only 45 patients. The follow-up MRI was also not obtained at a uniform time for all the patients.

Despite these limitations, our study highlights the role of MRI as both a diagnostic and prognostic marker in AIE. We developed an algorithm to assist in identifying specific AIE subtypes based on characteristic MRI findings (Fig. 7).

Fig. 7.

Utility of MRI in diagnosis of AIE. Alternate etiologies: primary central nervous system angiitis, Bickerstaff's brainstem encephalitis, acute disseminated encephalomyelitis, Rasmussen's encephalitis, underlying malignancy, septic encephalopathy, metabolic encephalopathy, Hashimoto's encephalopathy, drug toxicity, cerebrovascular disease, neoplastic disorders, Creutzfeldt-Jakob disease, rheumatologic disorders (e.g. lupus, sarcoidosis), Reye's syndrome (in children), mitochondrial diseases, and inborn errors of metabolism.

In conclusion, autoimmune encephalitis is primarily diagnosed based on clinical suspicion, with MRI providing crucial diagnostic clues to exclude alternative causes. MRI aids in the early detection of AIE, particularly in patients with seizures or dementia. The location of MRI abnormalities can be useful in differentiating between various AIE subtypes despite significant overlap. The resolution of MRI changes on follow-up is associated with a favorable long-term outcome.

Footnotes

Acknowledgments

We express our profound gratitude to Chandrasekharan Kesavadas, Neuroradiologist, and Rajalakshmi Poyuran, Neuropathologist, for their invaluable assistance in the methodology. Their expertise and guidance have been instrumental in the successful completion of this work

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors

ORCID iDs

KY Manisha

S Vinayagamani

References

Lancaster

. The diagnosis and treatment of autoimmune encephalitis. J Clin Neurol 2016;12:1.

Glaser

Gilliam

Schnurr

, et al. In search of encephalitis etiologies: diagnostic challenges in the California encephalitis project, 1998–2000. Clin Infect Dis 2003;36:731–742.

Gable

Sheriff

Dalmau

, et al. The frequency of autoimmune N-methyl-D-aspartate receptor encephalitis surpasses that of individual viral etiologies in young individuals enrolled in the California Encephalitis Project. Clin Infect Dis 2012;54:899–904.

Graus

Titulaer

Balu

, et al. A clinical approach to diagnosis of autoimmune encephalitis. Lancet Neurol 2016;15:391–404.

Lee

. The laboratory diagnosis of autoimmune encephalitis. J Epilepsy Res 2016;6:45–52.

Kelley

Patel

Marin

, et al. Autoimmune encephalitis: pathophysiology and imaging review of an overlooked diagnosis. AJNR Am J Neuroradiol 2017;38:1070–1078.

Heine

Prüss

Bartsch

, et al. Imaging of autoimmune encephalitis–relevance for clinical practice and hippocampal function. Neuroscience 2015;309:68–83.

Venkatesan

Tunkel

Bloch

, et al. Case definitions, diagnostic algorithms, and priorities in encephalitis: consensus statement of the international encephalitis consortium. Clin Infect Dis 2013;57:1114–1128.

Kidd

Stewart

Baldry

, et al. The functional independence measure: a comparative validity and reliability study. Disabil Rehabil 1995;17:10–14.

10.

Dubey

Singh

Britton

, et al. Predictive models in the diagnosis and treatment of autoimmune epilepsy. Epilepsia 2017;58:1181–1189.

11.

Cianfoni

Caulo

Cerase

, et al. Seizure-induced brain lesions: a wide spectrum of variably reversible MRI abnormalities. Eur J Radiol 2013;82:1964–1972.

12.

Fragoso

Gonçalves Filho

AdM

Pacheco

, et al. Imaging of creutzfeldt-jakob disease: imaging patterns and their differential diagnosis. Radiographics 2017;37:234–257.

13.

Prüss

Finke

Höltje

, et al. N-methyl-D-aspartate receptor antibodies in herpes simplex encephalitis. Ann Neurol 2012;72:902–911.

14.

Schäbitz

Rogalewski

Hagemeister

, et al. VZV brainstem encephalitis triggers NMDA receptor immunoreaction. Neurology 2014;83:2309–2311.

15.

Hammer

Stepniak

Schneider

, et al. Neuropsychiatric disease relevance of circulating anti-NMDA receptor autoantibodies depends on blood-brain barrier integrity. Mol Psychiatry 2014;19:1143–1149.

16.

Irani

Bera

Waters

, et al. N-methyl-D-aspartate antibody encephalitis: temporal progression of clinical and paraclinical observations in a predominantly non-paraneoplastic disorder of both sexes. Brain 2010;133:1655–1667.

17.

Florance

Davis

Lam

, et al. Anti–N-methyl-D-aspartate receptor (NMDAR) encephalitis in children and adolescents. Ann Neurol 2009;66:11–18.

18.

Titulaer

McCracken

Gabilondo

, et al. Treatment and prognostic factors for long-term outcome in patients with anti-NMDA receptor encephalitis: an observational cohort study. Lancet Neurol 2013;12:157–165.

19.

Hadad

Billingsley

. Posterior reversible encephalopathy syndrome (PRES) associated with ovarian cancer and voltage-gated potassium channel antibodies: a case report. Gynecol Oncol Rep 2017;20:67–69.

20.

Kotsenas

Watson

Pittock

, et al. MRI Findings in autoimmune voltage-gated potassium channel complex encephalitis with seizures: one potential etiology for mesial temporal sclerosis. AJNR Am J Neuroradiol 2014;35:84–89.

21.

Vincent

Buckley

Schott

, et al. Potassium channel antibody-associated encephalopathy: a potentially immunotherapy-responsive form of limbic encephalitis. Brain 2004;127:701–712.

22.

Binks

SNM

Klein

Waters

, et al. LGI1, CASPR2 and related antibodies: a molecular evolution of the phenotypes. J Neurol Neurosurg Psychiatry 2018;89:526–534.

23.

Irani

Stagg

Schott

, et al. Faciobrachial dystonic seizures: the influence of immunotherapy on seizure control and prevention of cognitive impairment in a broadening phenotype. Brain 2013;136:3151–3162.

24.

Plantone

Renna

Grossi

, et al. Teaching NeuroImages: basal ganglia involvement in facio-brachial dystonic seizures associated with LGI1 antibodies. Neurology 2013;80:e183–e184.

25.

López Chiriboga

Siegel

Tatum

, et al. Striking basal ganglia imaging abnormalities in LGI1 ab faciobrachial dystonic seizures. Neurol Neuroimmunol Neuroinflamm 2017;4:e336.

26.

Iyer

Ramakrishnan

TCR

Karunakaran , et al. Faciobrachial dystonic seizures result from fronto–temporo–basalganglial network involvement. Epilepsy Behav Case Rep 2017;8:47–50.

27.

Malter

Helmstaedter

Urbach

, et al. Antibodies to glutamic acid decarboxylase define a form of limbic encephalitis. Ann Neurol 2010;67:470–478.

28.

Honnorat

Saiz

Giometto

, et al. Cerebellar ataxia with anti-glutamic acid decarboxylase antibodies: study of 14 patients. Arch Neurol 2001;58:225–230.

29.

Esclapez

Tillakaratne

Kaufman

, et al. Comparative localization of two forms of glutamic acid decarboxylase and their mRNAs in rat brain supports the concept of functional differences between the forms. J Neurosci 1994;14:1834–1855.