Clinical characteristics and treatment of patients with overlapping MOG and anti-NMDAR encephalitis

Abstract

Objective

This study aimed to investigate clinical manifestations, cerebrospinal fluid (CSF) profiles, neuroimaging characteristics, and treatment responses in patients demonstrating concurrent NMDAR antibody (NMDAR-ab) positivity in CSF and MOG antibody (MOG-ab) seropositivity.

Methods

A retrospective study was conducted on patients with dual MOG-ab and anti-NMDAR antibody positivity treated at the Department of Neurology and Pediatrics, First Affiliated Hospital of Zhengzhou University (September 2019-July 2023). Participants were categorized into three groups: overlapping antibody group, isolated anti-NMDAR encephalitis, and MOG antibody-associated disease (MOGAD).

Results

Thirteen patients with antibody overlap presented with encephalitis spectrum symptoms including fever (53.8%), headache (61.5%), altered consciousness (46.2%), neuropsychiatric disturbances (38.5%), and seizures (23.1%) at disease onset. Cranial MRI revealed cortical/subcortical involvement in 30.8% of cases, with leptomeningeal enhancement observed in 22.2%. Acute-phase treatment predominantly included intravenous methylprednisolone (IVMP) and intravenous immunoglobulin (IVIG). Median modified Rankin Scale (mRS) scores improved significantly from 2 (IQR: 1-2.5) pre-treatment to 1 (IQR: 0.5-1) post-treatment. Compared to MOGAD controls, overlap adult patients exhibited higher ICU admission rates (30.8% vs. 6.7%), increased frequency of neuropsychiatric symptoms (38.5% vs. 6.7%), and lesser cortical/subcortical MRI abnormalities (30.8% vs. 64.4%), with reduced gadolinium enhancement (22.2% vs. 68.2%, P < 0.05). Distinct from isolated anti-NMDAR encephalitis, overlap cases demonstrated lower rates of neuropsychiatric disturbances (38.5% vs. 83.3%) and seizures (23.1% vs. 62.5%), but higher incidence of brachium pontis MRI abnormalities (23.1% vs. 0.0%, P < 0.05). Kaplan-Meier analysis over 6–48 months follow-up showed no significant difference in relapse rates between the overlapping antibody group and the MOGAD group or the anti-NMDAR encephalitis group (log-rank P > 0.05).

Conclusion

Demographic characteristics and oncological associations were comparable across groups. Overlap cases demonstrated unique clinico-radiological profiles with predominant cortical/subcortical involvement and reduced meningeal enhancement. Neuroimaging evidence of brachium pontis abnormalities in anti-NMDAR encephalitis warrants MOG-ab testing. This Asian cohort demonstrates distinct clinical trajectories compared to Western series, emphasizing the need for regional diagnostic considerations.

Introduction

Myelin oligodendrocyte glycoprotein-associated disease (MOGAD) constitutes a distinct immune-mediated inflammatory demyelinating disorder of the central nervous system, characterized by serum IgG antibodies targeting myelin oligodendrocyte glycoprotein (MOG-IgG) [1, 2]. Anti-N-methyl-D-aspartate receptor encephalitis (anti-NMDARE), the most prevalent antibody-mediated autoimmune encephalitis [3], demonstrates distinct clinical features compared to MOGAD. Emerging evidence highlights increasing reports of coexisting MOG-IgG in serum and NMDA receptor antibodies (NMDAR-Abs) in cerebrospinal fluid (CSF). While MOG-NMDAR antibody co-occurrence appears relatively frequent in clinical practice, current understanding remains limited by predominantly anecdotal case reports and insufficient systematic characterization. Recognizing the clinical manifestations of overlapping MOGAD and anti-NMDAR encephalitis can reduce misdiagnosis, provide deeper insights into pathogenic mechanisms, optimize treatment strategies, and improve prognostic assessments. Therefore, studying the coexistence of these diseases holds significant clinical value.

This retrospective cohort study systematically investigates the clinical characteristics of patients with concurrent MOG-IgG and NMDAR-Abs, comparing demographic parameters, clinical phenotypes, CSF profiles, and neuroimaging features with age-matched cohorts of pure MOGAD and anti-NMDARE patients. Kaplan-Meier survival analysis and log-rank tests were employed to evaluate differences in cumulative relapse rates across groups, offering preliminary insights into the natural history and clinical management implications of these overlapping immune-mediated disorders. Further exploration of the characteristics of overlapping antibody-associated diseases will be conducted with an expanded sample size.

Materials and methods

Study population

This retrospective cohort study included inpatients from the Departments of Neurology and Pediatrics at the First Affiliated Hospital of Zhengzhou University between September 2019 and July 2023. All participants underwent comprehensive serological and cerebrospinal fluid (CSF) antibody testing for biomarkers associated with demyelinating diseases and autoimmune encephalitis. Three distinct cohorts were defined: (1) MOGAD Cohort: fulfillment of the 2023 MOGAD diagnostic criteria [4]: Required: serum MOG-IgG positivity (≥ 1:100); Clinical phenotype: any combination of optic neuritis, myelitis, encephalitis/meningoencephalitis, or brainstem encephalitis; Ancillary evidence: magnetic resonance imaging (MRI)/electroencephalogram (EEG)-confirmed central nervous system (CNS) demyelination; Exclusion: CSF positive for NMDAR-Ab and presence of other neural antibodies. (2) Anti-NMDARE Cohort: adherence to the 2021 anti-NMDA receptor encephalitis diagnostic framework [5]: Required: CSF NMDAR-Abs positivity (≥ 1:10); Clinical phenotype: presence of more than one core feature (cognitive/behavioral changes, language dysfunction, seizures, movement disorders, altered consciousness, autonomic instability, central hypoventilation); Exclusion: Serum positive for MOG-Ab, and presence of other neural antibodies. (3) Overlapping MOG and anti-NMDAR antibodies (overlapping antibody group): Dual positivity: concurrent serum MOG-IgG and CSF NMDAR-Abs positivity; Clinical presentation: manifestation of more than one diagnostic feature from either the MOGAD or anti-NMDARE cohorts; Exclusion: presence of additional CNS autoantibodies beyond MOG and NMDAR. Exclusion Criteria (uniformly applied): [1] Presence of ≥ 3 non-MOG/NMDAR autoantibodies; [2] Patients in the presymptomatic phase; [3] Prior administration of immunotherapy (including corticosteroids) outside our institution; [4] Concomitant conditions affecting modified Rankin Scale (mRS) assessment: acute cerebrovascular events, intracranial neoplasms, compressive myelopathy; [5] Incomplete clinical documentation. Ethical Compliance: The study protocol was approved by the institutional review board (Ethics Committee Reference: 2021-KY-0822-002), in accordance with the Declaration of Helsinki.

Clinical data

Clinical data were systematically extracted from the electronic medical record system, including: demographic parameters (age, sex), pre-infectious medical history, intensive care unit (ICU) admission records, and clinical symptomatology. The mRS score was used for standardized assessment of clinical severity at baseline and during post-treatment follow-up. This ordinal scale which grades patient’s disability from 0 (no symptoms) to 6 (death) is able to capture the whole spectrum of functional states and poses intuitive simplicity [6].

Antibody detection

Serum samples were initially selected for MOG antibody detection. The concentration of serum MOG-IgG antibodies was determined using a cell-based assay (CBA). Antibody titer quantification was interpreted as follows: a ratio of 1:10 was considered suspected positive, 1:32 as weakly positive, 1:100 as moderately positive, 1:320 as strongly positive, and 1:1000 as extremely positive. Both positive and negative MOG antibody results in CSF were acceptable. All enrolled patients had serum MOG antibody titers of ≥ 1:100. CSF samples were primarily used for NMDA antibody detection, and the concentration of CSF NMDA antibodies was also determined by CBA. NMDA antibody titer quantification was defined as follows: a ratio of 1:1 was considered suspected positive, 1:3.2 as weakly positive, 1:10 as moderately positive, 1:32 as strongly positive, and 1:100 as extremely positive. Either positive or negative NMDA antibody results in blood were acceptable. The CSF NMDA antibody titers in the enrolled patients were ≥ 1:10. If a positive result was obtained, retesting was first performed within the same laboratory, and concurrent external verification was conducted at another laboratory. Typically, Kingmed Diagnostics Laboratory Co., Ltd. and Kindstar Medical Diagnostic Laboratory, two leading specialized medical testing service providers in China, were selected for initial testing and subsequent external verification.

Cerebrospinal Fluid (CSF) analysis

A comprehensive panel of CSF parameters was systematically analyzed, including intracranial pressure measurement, leukocyte quantification with differential lymphocyte count, CSF protein concentration, glucose level, chloride concentration, and oligoclonal banding (OCB) analysis. OCB results were generally categorized into five patterns: Type 1 refers to normal OCB profiles in both serum and CSF, indicating no intrathecal IgG synthesis. Type 2 is characterized by the presence of OCB exclusively in CSF, indicating intrathecal IgG synthesis. Type 3 demonstrates OCB bands in both serum and CSF, with additional bands unique to CSF, also indicating intrathecal IgG synthesis. Type 4 presents symmetrically matched OCB bands in serum and CSF, indicating no intrathecal IgG synthesis. Type 5 is defined by monoclonal bands with symmetrical and dense patterns in both serum and CSF, also indicating no intrathecal IgG synthesis. Approximately 60–65% of patients with multiple sclerosis (MS) exhibit positive OCB, particularly Types 2 and 3, which support the diagnosis of MS [7]. Reference intervals for CSF biomarkers followed standardized thresholds: intracranial pressure (80–180 mm H2O), leukocyte count (< 5 × 10⁶/L), lymphocyte percentage (60–70%), total protein concentration (150–450 mg/L), glucose level (2.5–4.4 mmol/L), and chloride concentration (120–130 mmol/L). The CSF/serum glucose quotient and albumin quotient were calculated when clinically indicated.

Comprehensive imaging evaluation

All participants underwent a systematic neuroimaging assessment, including baseline 3.0 Tesla magnetic resonance imaging (MRI) of the brain with T1-weighted, T2-weighted, and fluid-attenuated inversion recovery (FLAIR) sequences. Advanced imaging protocols included contrast-enhanced MRI with intravenous gadolinium administration for the detection of blood-brain barrier disruption, as well as dedicated spinal cord MRI for the evaluation of longitudinal myelitis patterns. Brain and spinal MRI findings were reviewed by a multidisciplinary team comprising board-certified neurologists and neuroradiologists, using standardized diagnostic criteria.

Comprehensive ultrasonographic evaluations of thyroid, cardiac (including echocardiography), and abdominal organs were systematically performed at baseline. Thoracic computed tomography (CT) with contrast enhancement was mandatory for all patients. Systematic assessment for paraneoplastic syndromes included fluorodeoxyglucose positron emission tomography-computed tomography (¹⁸F-FDG PET-CT) when clinical suspicion warranted, employing standardized oncological imaging protocols.

Treatment protocols and follow-up evaluation

All patients received standardized immunotherapeutic regimens stratified into: (1) Acute-phase induction therapy (intravenous methylprednisolone [IVMP], intravenous immunoglobulin [IVIG], plasma exchange [PLEX], and immunoadsorption therapy); (2) Maintenance immunosuppression (low-dose corticosteroids, azathioprine [AZA], mycophenolate mofetil [MMF], methotrexate [MTX], cyclophosphamide [CTX]); (3) Targeted biologic therapy (rituximab [RTX], ofatumumab, efgartigimod). Treatment algorithms adhered to international consensus guidelines for autoimmune neurological disorders.

A structured follow-up protocol was implemented with serial clinical evaluations conducted through scheduled outpatient visits and telephone assessments. Relapse was defined as: (1) Symptomatic deterioration after at least 60 days of clinical stabilization; (2) Emergence of novel MOG-typical or NMDAR-typical neurological deficits; (3) Recurrence of original typical symptoms of MOG or NMDAR; (4) mRS score escalation ≥ 1 point; (5) Radiological progression on MRI. Time-to-event metrics: first relapse-free survival duration; total follow-up duration until last assessment.

Statistical analysis

This exploratory analysis was conducted using SPSS statistical package (v26.0) following a pragmatic analytical approach. Continuous variables were evaluated for normality using Shapiro-Wilk test and presented as: Normal distribution: mean ± standard deviation (SD); Non-normal distribution: median with interquartile range (IQR). Comparative analyses employed: (1) Parametric testing: Independent samples t-test for normally distributed intergroup comparisons; (2) Non-parametric testing: Mann-Whitney U test for independent groups, Wilcoxon signed-rank test for paired comparisons. Categorical variables were analyzed using: Pearson’s chi-square test (χ²) for adequate cell counts (expected frequency ≥ 5); Fisher’s exact test for sparse data (expected frequency < 5). Survival analysis for time-to-relapse outcomes included: Kaplan-Meier survival curves with log-rank test stratification. Statistical significance was defined as two-tailed P < 0.05. Given the exploratory nature of this hypothesis-generating study with limited sample size and a relatively low risk of false positives; therefore, multiple testing corrections were not performed.

Results

Clinical characteristics of patients with overlapping MOG and NMDAR antibodies (overlapping antibody group)

A retrospective analysis of 13 patients with overlapping autoimmune antibody profiles was conducted. Comprehensive clinical profiles, including demographic characteristics, immunological markers, and treatment responses, are provided in Tables 1 and 2.

Table 1.

Summary of the clinical data of thirteen patients with overlapping MOG and NMDAR antibodies

Number/ Sex /Age (Year)	MOG antibody (serum/CSF)	Anti-NMDAR Antibody (serum/CSF)	Clinical manifestation	MRI findings	CSF analysis (white blood cell count/protein level/oligoclonal band)	Acute phase treatment plan	Frequency of onset
1/F/6	(+)/(+)	(-)/(+)	Dizziness, drowsiness, and fever	Abnormal signals in the left frontal lobe, cervical cord 2–7	55 × 106/L 308 mg/L Type 2	IVMP	First time
2/F/7	(+)/(-)	(-)/(+)	Sleepiness, enuresis, involuntary shaking of the left upper limb, and visual hallucinations	Abnormal signals in the bilateral lateral ventricles, basal ganglia, left frontal temporal parietal lobe, right frontal lobe, bilateral cerebellar hemispheres, left bridge arm, cervical cord 3, 5–6, and thoracic cord 2–9	11 × 106/L 480 mg/L Type 3	IVMP	First time
3/F/7	(+)/(+)	(+)/(+)	Strabismus in the left eye and visual ghosting	The left cerebellar hemisphere, left pontine arm, cerebellar vermis, right frontal lobe, left frontal parietal lobe, and C7-T2 spinal cord signals, and there were any abnormalities when enhancing	21 × 106/L 214 mg/L Type 2	IVMP IVIG	First time
4/F/14	(+)/(+)	(-)/(+)	Headache and fever	Abnormal signals in the right hippocampus, insula, and basal ganglia	85 × 106/L 370 mg/L Type 2	IVMP IVIG	First time
	(-)/(-)	(-)/(-)	Headache, drowsiness, and fatigue	No abnormalities	2 × 106/L 400.2 mg/L Type 1	IVMP	Second time
5/M/14	(+)/(+)	(-)/(+)	Fever and headache	No abnormalities	2 × 106/L 262.9 mg/L Type 1	IVMP	First time
6/M/16	(+)/(+)	(-)/(+)	Unclear speech, slow speech, decreased comprehension, and weakness in the right limb	No abnormalities	6 × 106/L 335.6 mg/L Type 3	IVMP	First time
	(+)/(-)	(+)/(+)	Slow response, dizziness, abnormal behavior, hallucinations, irritability, choking on water, speech disorder, convulsions	No abnormalities	2 × 106/L 549.7 mg/L Type 1	IVMP IVIG	Second time
7/M/17	(+)/(-)	(+)/(+)	Headache, convulsions, consciousness disorders, irritability	Swelling of the cerebral gyrus	32 × 106/L 162.6 mg/L Type 2	IVMP IVIG PLEX	First time
8/M/18	(+)/(-)	(+)/(+)	Poor speech, inability to write, convulsions, irritability	No abnormalities	14 × 106/L 426 mg/L Type 1	IVMP	First time
9/F/24	(+)/(+)	(-)/(+)	Limb numbness, fever, headache	Abnormal signals and meningeal enhancement in the right basal ganglia, thalamus, temporal lobe, cerebral foot, lateral ventricles, and left frontal temporal parietal lobe	32 × 106/L 791 mg/L Type 1	IVMP	First time
	(+)/(-)	(-)/(-)	headache	The abnormal signal range of the right thalamus and left temporal lobe has increased, and the abnormal signal appears on the left bridge arm	10 × 106/L 224.5 mg/L Type 1	IVMP	Second time
10/M /26	(+)/(-)	(+)/(+)	Fever, convulsions, headache	Swelling of the left frontal temporal parietal lobe gyrus and linear enhancement in the left frontal temporal lobe sulci	44 × 106/L 342 mg/L Type 2	IVMP	First time
	(+)/(-)	(+)/(+)	Blurred vision, gibberish, and decreased memory	Abnormal signals in the left frontal lobe, temporal lobe, pons, bilateral pontine arms, bilateral cerebral feet, left thalamus, and visible patchy enhancement	40 × 106/L 253.9 mg/L Type 1	IVMP IVIG	Second time
	NA	NA	Double vision, decreased vision, and sensory abnormalities	New abnormalities in the right pontine region, right pontine arm, and ventral medulla oblongata	NA	IVMP	Third time
11/M /33	(+)/(+)	(-)/(+)	Headache, limb weakness, blurred vision, increased sleep	Abnormalities in the pons, bilateral cerebral hemispheres, left thalamus, posterior horn of the lateral ventricle, and corpus callosum compression	50 × 106/L 369.0 mg/L Type 1	IVMP	First time
12/F /35	(+)/(+)	(-)/(+)	Headache, fever, personality changes, talking to oneself, answering irrelevant questions, and consciousness disorders	No abnormalities were found in the head, and there were no abnormalities in the enhancement	8 × 106/L 263.7 mg/L Type 1	IVMP IVIG	First time
	(+)/(+)	(-)/(+)	Decreased vision in the right eye, diplopia, and eye pain	No abnormalities were found in the head, and there were no abnormalities in the enhancement	14 × 106/L 402.5 mg/L Type 1	IVMP	Second time
	(+)/(+)	(-)/(+)	Headache and fever	No abnormalities were found in the head, and there were no abnormalities in the enhancement	8 × 106/L 332.9 mg/L Type 1	IVMP	Third time
13/M /48	(+)/(+)	(-)/(+)	Headache, fever, delayed response, gibberish, increased sleep	Abnormal signals were observed in the bilateral lateral ventricles, corpus callosum, bilateral thalamus, and bilateral bridge arms, with no enhancement observed	16 × 106/L 678 mg/L Type 3	IVMP IVIG	First time

MOG Myelin oligodendrocyte glycoprotein, NA not available, NMDAR N-methyl-D-aspartate receptor, CSF cerebrospinal fluid, mRS Improved Rankin Scale, IVMP intravenous injection of methylprednisolone, IVIG intravenous immunoglobulin injection, PLEX Plasma exchange

Table 2.

Comparison of the clinical features, treatment regimens and therapeutic effects [n(%)]

Item [n(%)]	Overlapping Antibodies Group (n = 13)	MOGAD Group (n = 45)	Anti NMDAR Encephalitis Group (n = 24)	P1	P2
General Information
Age (years,±s)	20.4 ±12.5	17.5 ±15.9	28.6 ±20.9	0.548	0.205
Sex, female	6 (46.2)	14 (31.1)	16 (66.7)	0.500	0.300
Children	7 (53.8)	30 (66.7)	8 (33.3)	0.603	0.300
Admitted to the ICU	5 (38.5)	16 (35.6)	10 (41.7)
Adult*	4 (80.0)	3 (18.8)	7 (70.0)	0.025	> 0.999
Preinfection	4 (30.8)	15 (33.3)	5 (20.8)	> 0.999	0.691
Concurrent tumors	0 (0.0)	0 (0.0)	3 (12.5)	-	0.538
Clinical symptoms
Headache	8 (61.5)	22 (48.9)	8 (33.3)	0.421	0.165
Fever	7 (53.8)	26 (57.8)	7 (29.2)	0.801	0.171
Decling level of consciousness	6 (46.2)	9 (20.0)	8 (33.3)	0.124	0.495
Mental behavioral abnormalities	5 (38.5)	3 (6.7)	20 (83.3)	0.013	0.010
Seizures	3 (23.1)	22 (48.9)	15 (62.5)	0.098	0.038
Cognitive impairment	2 (15.4)	4 (8.9)	11 (45.8)	0.873	0.083
Dysarthria/Dysphagia	2 (15.4)	6 (13.3)	4 (16.7)	> 0.999	> 0.999
Weakness of limbs	2 (15.4)	11 (24.4)	5 (20.8)	0.755	> 0.999
Optic neuritis	1 (7.7)	12 (26.7)	0 (0.0)	0.286	0.351
Involuntary movement	1 (7.7)	6 (13.3)	7 (29.2)	0.947	0.216
Diplopia	1 (7.7)	1 (2.2)	1 (4.2)	0.929	> 0.999
Dizziness	1 (7.7)	4 (8.9)	1 (4.2)	> 0.999	> 0.999
Sensory abnormality	1 (7.7)	3 (6.7)	1 (4.2)	> 0.999	> 0.999
Inability to write	1 (7.7)	0 (0.0)	0 (0.0)	0.505	0.351
Unsteady walking	0 (0.0)	4 (8.9)	3 (12.5)	0.622	0.538
Bowel/bladder dysfunction	0 (0.0)	5 (11.1)	1 (4.2)	0.486	> 0.999
Slanting of the mouth	0 (0.0)	3 (6.7)	1 (4.2)	0.806	> 0.999
Gastrointestinal symptoms	0 (0.0)	12 (26.7)	0 (0.0)	0.089	-
CSF Results
Pressure increase**	4/6 (66.7)	6/15 (40.0)	8/16 (50.0)	0.428	0.701
Elevated white blood cell count	12 (92.3)	38 (84.4)	19 (79.2)	0.789	0.394
Significant increase in white blood cell count***	2 (15.4)	17 (37.8)	5 (20.8)	0.238	> 0.999
Elevated lymphocyte ratio	7 (53.8)	18 (40.0)	15 (62.5)	0.375	0.730
Increased protein content	3 (23.1)	6 (13.3)	3 (12.5)	0.675	0.643
Elevated sugar content	1 (7.7)	3 (6.7)	2 (8.3)	> 0.999	> 0.999
Elevated chloride content	1 (7.7)	12 (26.7)	3 (12.5)	0.286	> 0.999
OCB Type 1	5 (38.5)	27 (60.0)	13 (54.2)	0.169	0.495
OCB Type 2	5 (38.5)	14 (31.1)	8 (33.3)	0.871	> 0.999
OCB Type 3	3 (23.1)	3 (6.7)	3 (12.5)	0.232	0.643
OCB Type 4	0 (0.0)	1 (2.2)	0 (0.0)	> 0.999	-
Cranial MRI Features
Head abnormality	9 (69.2)	38 (84.4)	11 (45.8)	0.406	0.300
Cortical and subcortical involvement	4 (30.8)	29 (64.4)	7 (29.2)	0.031	> 0.999
Hippocampus involvement	1 (7.7)	5 (11.1)	0 (0.0)	> 0.999	0.351
Paraventricular involvement	4 (30.8)	8 (17.8)	5 (20.8)	0.529	0.691
Thalamus involvement	3 (23.1)	12 (26.7)	2 (8.3)	> 0.999	0.321
Basal ganglia involvement	3 (23.1)	13 (28.9)	1 (4.2)	0.952	0.115
Pons involvement	1 (7.7)	11 (24.4)	2 (8.3)	0.355	> 0.999
Brachium pontis involvement	3 (23.1)	5 (11.1)	0 (0.0)	0.519	0.037
Midbrain involvement	0 (0.0)	5 (11.1)	0 (0.0)	0.486	-
Cerebral peduncle involvement	2 (15.4)	5 (11.1)	1 (4.2)	> 0.999	0.278
Cerebellum involvement	2 (15.4)	4 (8.9)	0 (0.0)	0.873	0.117
Corpus callosum involvement	2 (15.4)	1 (2.2)	0 (0.0)	0.239	0.117
medulla oblongata involvement	0 (0.0)	2 (4.4)	1 (4.2)	> 0.999	> 0.999
Head gadolinium enhancement****	9 (69.2)	22 (48.9)	17 (70.8)
Abnormal reinforcement	2 (22.2)	15 (68.2)	6 (35.3)	0.044	0.667
Treatment regimens
IVMP	7 (53.8)	29 (64.4)	9 (37.5)	0.529	0.489
IVMP + IVIG	4 (30.8)	15 (33.3)	8 (33.3)	> 0.999	> 0.999
PLEX	1 (7.7)	0 (0.0)	0 (0.0)	-	-
IVMP + PLEX	0 (0.0)	1 (2.2)	2 (8.3)	-	-
IVMP + IVIG + PLEX	1 (7.7)	0 (0.0)	4 (16.7)	-	0.638
IVMP + IVIG + Plasma adsorption	0 (0.0)	0 (0.0)	1 (4.2)	-	-
Therapeutic effects
mRS score pre treatment [M (Q25, Q75)]	2 (1, 2.5)	2 (1, 2.5)	2.5 (1.25, 4)	0.766	0.200
mRS score after treatment [M (Q25, Q75)]	1 (0.5, 1)	1 (0, 1)	1 (0.25, 1)	0.301	0.638

MOGAD Myelin oligodendrocyte glycoprotein antibody-related disease, NMDAR N-methyl-D-aspartate receptor, CSF Cerebrospinal fluid, OCB Oligoclonal banding, IVMP intravenous injection of methylprednisolone, IVIG intravenous immunoglobulin injection, PLEX Plasma exchange, mRS Modified Rankin scale, M (Q25, Q75) Median (interquartile range), P1 Comparison between the overlapping antibody group and the MOGAD group, P2 Comparison between the overlapping antibody group and the anti-NMDAR encephalitis group

*: Because some patients was in ICU admission, the total number has changed

**: This pressure is only compared with that of adult patients, and the total number has changed (cerebrospinal fluid pressure in pediatric patients is recorded by the drip rate)

***: Significant increase in white blood cell count: >50 × 10⁶/L

****: Because some patients do not undergo gadolinium-enhanced head MRI scans, the total number has changed

The cohort demonstrated a male predominance (53.8%), with age at disease onset of 20.4 ± 12.5 years. Pediatric cases (< 18 years) constituted 53.8% of the cohort. Critical care requirements were observed in 38.5% of patients. Preceding infectious prodromes were documented in 30.8%, exclusively manifesting as upper respiratory tract involvement. Systematic tumor screening yielded negative results in all cases (see Table 2).

Table 2 delineates the symptom profile at disease onset in the 13-patient cohort. The most prevalent neurological manifestations included headache (61.5%), fever (53.8%), and altered consciousness (46.2%), ranging from apathy to coma. Neuropsychiatric features encompassed cognitive-behavioral disturbances in 5 patients, with mental status changes, and seizure activity documented in 3 episodes. Focal neurological deficits comprised oropharyngeal dysfunction, including dysarthria and dysphagia (15.4%), limb weakness (15.4%), and visual pathway involvement manifesting as optic neuritis (7.7%). Rare presentations included monosymptomatic features, each occurring in 7.7% of cases, consisting of involuntary movements, binocular diplopia, vertiginous episodes, sensory abnormalities, and agraphia.

All 13 patients underwent standardized lumbar puncture with CSF analysis in accordance with neuro immunological diagnostic protocols. Key findings included CSF pressure parameters: elevated intracranial pressure (ICP) in 66.7% (4/6 adult cases), with a maximum ICP of 280 mmH₂O (upper normal limit: 180 mmH₂O). CSF cytology and biochemical profiling revealed cellular abnormalities, with pleocytosis (> 5 × 10⁶/L) observed in 92.3%, including marked pleocytosis (> 50 × 10⁶/L) in 15.4%, peaking at 85 × 10⁶/L. Lymphocytic abnormalities included lymphocytosis in 53.8% and lymphopenia in 15.4%. Protein elevation (> 450 mg/L) was noted in 23.1% of cases. Electrolyte and metabolic abnormalities included chloride dysregulation in 23.1%, with hyperchloremia in 1 case and hypochloremia in 2 cases. Glucose dysregulation was observed in 7.7%. Oligoclonal banding (OCB) analysis demonstrated type 1 in 38.5%, type 2 in 38.5%, type 3 in 23.1%, and type 4 in 0 cases (Table 2).

All 13 patients underwent baseline brain MRI at disease onset. Abnormal MRI findings were identified in 9 out of 13 cases, predominantly characterized by hyperintense lesions on T2-weighted FLAIR sequences (Fig. 1). Lesion distribution patterns included cortical/subcortical involvement (cerebral cortex and subcortical white matter) in 4 cases. Periventricular involvement (ventricular border zones) was also observed in 4 cases. Brainstem and deep gray matter involvement included brachium pontis in 3 cases, basal ganglia in 3 cases, and thalamus in 3 cases. Cerebellar and supratentorial involvement comprised cerebellum in 2 cases, cerebral peduncle in 2 cases, and corpus callosum in 2 cases. Hippocampal and brainstem involvement was noted in 1 case each (hippocampus and pons). Enhancement patterns: among 9 patients undergoing contrast-enhanced MRI, meningeal enhancement was observed in 2 cases. Spinal cord involvement: abnormal spinal MRI signals were detected in 3 out of 4 cases, demonstrating longitudinal myelitis patterns.All 13 patients with overlapping autoantibodies received standardized acute-phase immunomodulatory treatment regimens (Tables 1 and 2). Treatment modalities included monotherapy with IVMP in 7 cases, dual therapy with IVMP and IVIG in 4 cases, and triple therapy with IVMP, IVIG, and PLEX in 1 case. PLEX monotherapy was administered in 1 case. Maintenance immunomodulation during remission phase: 12 out of 13 patients continued low-dose oral corticosteroids, and 1 patient received RTX for relapse prevention. Concomitant management included antimicrobial prophylaxis for infection-prone individuals, antiepileptic therapy for seizure control, and psychotropic medications for neuropsychiatric symptom management.

Fig. 1.

Head MRI findings of patients with overlapping MOG and NMDAR antibodies. (A) Abnormal signals in the right thalamus and basal ganglia region (arrow); (B) Abnormal signals in the left cerebellar hemisphere and brachium pontis (arrow); (C) Abnormal enhancement of the meninges in the cerebral hemisphere (arrow); (D) Abnormal signals in the cervical cord (arrow)

Post-treatment improvements in mRS scores are illustrated in Fig. 2 and summarized in Table 2. The baseline median mRS score was 2 (IQR: 1–2.5), demonstrating a significant improvement to 1 (IQR: 0.5–1) following acute-phase treatment, with a median reduction of 1 point.

Fig. 2.

MRS scores of thirteen patients with overlapping MOG and NMDAR antibodies before and after treatment. The median mRS score before treatment was 2 (1, 2.5) points, and the median mRS score after treatment was 1 (0.5, 1) point. The mRS score of each patient is displayed in the graph, with blue dots indicating the score before treatment and green dots indicating the score after treatment

Among the 13 patients included in the follow-up analysis (median follow-up duration: 15 months), recurrence events were observed in 5 cases. The recurrence pattern demonstrated that 3 patients experienced a single relapse episode, while 2 patients exhibited two recurrent episodes during the surveillance period. In these cases of attacks or relapses, antibody testing was performed at each attack and relapse, except for 1 case in which antibody testing was not performed at the last attack. Among the patients who underwent this series of antibody tests, 2 cases turned negative for antibodies, while 3 cases had persistent antibodies. However, these antibody tests were all conducted again when the patient relapsed, not during the symptom remission period following the initial episode. These specific antibody test results are presented in Table 1. This also suggests that we should recommend patients have their antibodies rechecked after each episode of immunotherapy in the future.

Figure 3; Table 2 provide a distribution of clinical symptoms in patients with overlapping antibodies, compared with those with MOGAD and those with anti-NMDAR encephalitis. The visual representation (Fig. 3) demonstrates distinct symptom profiles, while Table 2 quantifies the demographic distribution of key clinical features.

Fig. 3.

Comparison of clinical symptoms between overlapping and MOGAD and between overlapping and anti-NMDAR encephalitis. Comparison of clinical symptoms between MOG and anti-NMDAR overlapping antibody patients and MOGAD patients (A) and between overlapping patients and anti-NMDAR encephalitis patients (B)

Comparison of the clinical characteristics between patients with overlapping MOG and anti-NMDAR antibodies (overlapping antibody group) and patients with MOGAD

The MOGAD group (n = 45) exhibited a mean age at onset of 17.5 ± 15.9 years, with 31.1% female subjects and 66.7% pediatric cases. Prodromal infections were documented in 33.3%, predominantly upper respiratory tract infections (13 cases) and two cases of intestinal infections. No malignancies were identified in this cohort. ICU admission was required for 35.6% of patients, demonstrating a significant pediatric predominance (81.3% children vs. 18.8% adults). Comparative analysis revealed no statistically significant differences between the overlapping antibody group and the MOGAD cohort regarding age at onset, sex distribution, prodromal infection rate, or tumor prevalence (P > 0.05). However, a significant disparity emerged in ICU admission rates between groups, with adults in the overlapping cohort demonstrating 4.6-fold higher ICU requirement compared to MOGAD adults (P = 0.025).

Table 2 delineates the clinical presentation profile of 45 MOGAD patients. Initial manifestations included fever (57.8%), headache (48.9%), seizures (48.9%), optic neuritis (26.7%), gastrointestinal symptoms (26.7%), limb weakness (24.4%), altered consciousness (20.0%), involuntary movements and dysarthria/dysphagia (each 13.3%), bladder/bowel dysfunction (11.1%), gait instability, cognitive impairment, and dizziness in four patients each (8.9%), and sensory abnormalities (6.7%). Notably, 6.7% exhibited slanting of the mouth, while 2.2% presented with diplopia. Crucially, inability to write was absent in this cohort. Comparative analysis revealed a significantly higher prevalence of neuropsychiatric symptoms in the overlapping antibody group (38.5%) versus the MOGAD cohort (6.7%) (P = 0.013).

CSF analysis was performed in all patients. Key findings in the MOGAD cohort (n = 45) include: ​conventional parameters​​: elevated intracranial pressure in 40.0% (6/15 adults), CSF leukocytosis in 84.4%, marked elevation (> 50 × 10⁶/L) in 37.8%, and lymphocytic predominance in 40.0%. ​Biochemical parameters​​: elevated protein in 13.3%, hyperglycorrhachia in 6.7% and hyperchlororrhachia in 26.7%. ​OCB pattern:​type 1 in 60.0%, type 2 in 31.1%, type 3 in 6.7% and type 4 in 2.2% (Table 2).

Among 45 MOGAD patients, 84.4% exhibited abnormal MRI findings, with characteristic lesional distributions: ​​cortical/subcortical involvement​​ (64.4%), basal ganglia​​ (28.9%), thalamus​ (26.7%), and pons​​ (24.4%). Additional regional abnormalities included lateral ventricles (17.8%), brachium pontis/cerebral peduncle/hippocampus/midbrain (11.1% each), cerebellum (8.9%), medulla oblongata (4.4%), and corpus callosum (2.2%). Gadolinium-enhanced MRI (22/45 patients) revealed meningeal enhancement in 68.2% of cases, including eight patients with pial-based enhancement patterns. Spinal MRI findings in 23 patients demonstrated: 10 cervical cord lesions, 11 thoracic cord lesions, and 2 cases involving the conus medullaris. Comparative analysis between the overlapping antibody group and the MOGAD cohort revealed: a significantly lower cortical/subcortical lesional burden (P = 0.031) and reduced frequency of post-contrast enhancement (P = 0.044) (see Table 2).

All patients received acute-phase immunotherapy as part of standardized treatment protocols (Table 2). ​MOGAD (n = 45):​​ induction therapy included IVMP monotherapy in 29 cases, IVMP + IVIG combination in 15 cases, and IVMP + PLEX combination in 1 case. ​​Maintenance therapy (remission phase) consisted of low-dose oral steroids in all 45 cases, with additional immunosuppressants administered in 8 patients: MMF in 6 cases, AZA in 1 case, and MTX in 1 case.

Among patients with MOGAD, the median mRS score decreased from 2 (IQR: 1–2.5) before treatment to 1 (IQR: 0–1) after treatment. No statistically significant differences in mRS scores were observed between the MOGAD and overlapping antibody groups (P > 0.05) (Table 2).

Recurrence patterns and survival analysis across patient cohorts​​: ​​MOGAD (n = 45):​​ ​​median follow-up​​ of 15 months (IQR: 6–45) with a recurrence incidence of 46.7%. Recurrence frequency distribution included single relapse in 15 cases, double relapse in 4 cases, triple relapse in 1 case, and quadruple relapse in 1 case. ​​Comparative survival analysis revealed no significant differences in recurrence-free survival between the overlapping antibody group and the MOGAD cohort. Kaplan-Meier curves demonstrated an accumulated recurrence rate (Fig. 4), and the log-rank test confirmed no statistical significance (all P > 0.05).

Fig. 4.

Kaplan-Meier curves of the cumulative recurrence rate. Kaplan-Meier curves of the cumulative recurrence rate between the overlapping antibody group and the anti-NMDAR encephalitis group (A) and between the overlapping antibody group and the MOGAD group (B)

Comparison of the Clinical Characteristics between Patients with Overlapping MOG and Anti-NMDAR Antibodies (Overlapping Antibody Group) and Anti-NMDAR Encephalitis

Table 2 summarizes the clinical features of 24 patients with anti-NMDAR encephalitis (mean age 28.6 ± 20.9 years). The cohort included 66.7% female, with pediatric cases accounting for 33.3%. Prodromal infections were documented in 20.8%, exclusively involving the upper respiratory tract. Malignant associations were identified in 12.5%, including small cell lung cancer, lymphoma, and ovarian teratoma. ICU admission was required for 41.7%, with a predominance in adults (7/10). Comparative analysis with the overlapping cohort revealed no significant intergroup differences in age at onset, gender distribution, prodromal infection rate, neoplastic comorbidities, or adult ICU admission rates (P > 0.05).

Table 2 delineates the clinical presentation profile of 24 anti-NMDAR encephalitis patients during their initial episode. Core manifestations included neuropsychiatric symptoms (83.3%), seizures (62.5%), and cognitive impairment (45.8%). Common features (33.3–29.2% prevalence) were headache (33.3%), altered consciousness (33.3%,), fever (29.2%) and involuntary movements (29.2%). Less frequent manifestations (< 25%) included limb weakness (20.8%), dysarthria or dysphagia (16.7%), gait instability (12.5%), and bowel or bladder dysfunction, slanting of the mouth, diplopia, dizziness, and sensory abnormalities in 1 patient (4.2%). Notably absent findings were optic neuritis, dysgraphia, and gastrointestinal symptoms. Comparative analysis revealed significantly higher rates of neuropsychiatric symptoms (P = 0.010) and seizure frequency (P = 0.038) in anti-NMDAR encephalitis.

CSF analysis in the 24 anti-NMDAR encephalitis patients revealed the following key findings: ​conventional parameters​​: elevated intracranial pressure was observed in 50.0% of adult patients (8/16). CSF leukocytosis occurred in 79.2% of cases, with marked elevation (> 50 × 10⁶/L) in 20.8%. Increased lymphocyte ratio was detected in 62.5%. ​Biochemical parameters​: elevated protein levels in 12.5%, hyperglycorrhachia in 8.3% and hyperchlorhoria in 12.5%. ​OCB patterns​​: type 1 in 54.2%, type 2 in 33.3%, type 3 in 12.5%. Comparative analysis demonstrated no statistically significant differences in CSF parameters between the overlapping antibody group and the anti-NMDAR encephalitis cohort (all P > 0.05) (Table 2).

Cranial MRI findings in 24 anti-NMDAR encephalitis patients revealed the following patterns: conventional MRI abnormalities (n = 11): cortical/subcortical involvement in 29.2%, periventricular regions in 20.8%, pons in 8.3%, thalamus in 8.3%, and medulla oblongata/basal ganglia/cerebral peduncle in 4.2%. Structural preservation was noted in the corpus callosum, hippocampus, midbrain, brachium pontis, and cerebellum. Gadolinium-enhanced MRI (n = 17): abnormal enhancement in 35.3% and meningeal enhancement subtype in 11.8%. Spinal cord MRI (n = 10): cervical cord lesions in 10.0% and thoracic cord lesions in 20.0%. Comparative analysis demonstrated significantly higher involvement of the brachium pontis in the overlapping antibody group compared to the anti-NMDAR encephalitis cohort (P = 0.037) (Table 2).

All patients received acute-phase immunotherapy as part of standardized treatment protocols (Table 2). Anti-NMDAR Encephalitis (n = 24): induction therapy: IVMP monotherapy in 9 cases, IVMP + IVIG combination in 8 cases, IVMP + IVIG + PLEX triple therapy in 4 cases, IVMP + PLEX combination in 2 cases, and IVMP + IVIG + plasma adsorption in 1 case. Maintenance therapy (remission phase): low-dose oral steroids in all 24 cases, with additional immunosuppressants administered in 4 cases: MMF in 4 cases.In the anti-NMDAR encephalitis group, the median mRS score decreased from 2.5 (IQR: 1.25–4) to 1 (IQR: 0.25–1). No statistically significant differences in mRS scores were observed between the anti-NMDAR encephalitis and overlapping antibody groups, either before or after treatment (P > 0.05) (Table 2).

Recurrence patterns and survival analysis across patient cohorts: Anti-NMDAR Encephalitis (n = 24): median follow-up of 18 months (IQR: 6–48), one patient died during follow-up due to underlying malignancy, and 8 of the remaining 23 patients experienced relapses. Recurrence frequency distribution: single relapse in 7 cases and double relapse in 1 case. Comparative survival analysis: no significant differences in recurrence-free survival between the overlapping antibody group and the anti-NMDAR encephalitis cohort. Kaplan-Meier curves demonstrated cumulative recurrence rates (Fig. 4), and log-rank testing confirmed the absence of statistical significance (all P > 0.05).

Discussion

Current evidence suggests that anti-NMDAR antibodies most frequently coexist with MOG antibodies in neuroimmune disorders [8, 9]. However, due to the rarity of antibody overlap and limited clinical case accumulation, existing literature predominantly comprises case reports and small-scale cohort studies. This investigation advances understanding through comprehensive characterization of 13 patients exhibiting dual antibody positivity, systematically comparing their clinical profiles with 45 MOGAD cases and 24 anti-NMDAR encephalitis patients, respectively. Methodologically, we employed survival analysis to evaluate cumulative recurrence rates between overlap antibody cohorts and non-overlap populations—a comparative approach rarely documented in prior research. Kaplan-Meier curves were generated to visualize recurrence-free survival probabilities, supplemented by log-rank testing for statistical significance.

This study included 13 patients with dual antibody positivity, demonstrating a slight male predominance (53.8%). The mean age at onset was 20.4 ± 12.5 years, with 53.8% of cases occurring in the pediatric population. These findings are consistent with prior reports of male preponderance in dual antibody cohorts [10] and pediatric predominance in MOGAD [4]. Comparative analysis across cohorts: MOGAD: mean age at onset 17.5 ± 15.9 years, pediatric cases 66.7%, and sex ratio 31.1% female. Anti-NMDAR encephalitis): mean age at onset 28.6 ± 20.9 years, pediatric cases 33.3%, and sex ratio 66.7% female. Previous studies have reported sex disparity in anti-NMDAR encephalitis [11] and age-related differences between MOGAD and NMDAR encephalitis [4]. Our dual antibody cohort exhibited intermediate demographic characteristics: mean age at onset, pediatric proportion, and sex ratio were positioned between MOGAD and NMDAR groups. Methodological considerations: the relatively small sample size (n = 13) limits the statistical power to draw definitive conclusions, necessitating replication studies in larger pediatric cohorts to validate these preliminary findings. This study revealed no statistically significant difference in sex ratio, age distribution, or proportion of pediatric cases in the overlapping antibody group compared with the MOGAD or anti-NMDAR encephalitis groups. Statistical equivalence across cohorts (all P >0.05) suggests shared pathophysiological mechanisms and potential overlap in immune dysregulation patterns. Among 13 patients with dual antibody positivity, four cases reported prodromal infections, exclusively upper respiratory tract infections. This pattern parallels findings in MOGAD and anti-NMDAR encephalitis cohorts, where pre-illness infections were similarly prevalent. Virological associations across neuroimmune disorders: (1) Anti-NMDAR encephalitis: Salovin et al. demonstrated elevated HSV-1 seroprevalence in NMDAR encephalitis compared with controls ([12]– [13]). Respiratory or gastrointestinal prodromal symptoms, reported in 20–30% of cases, include fever, headache, and nausea ([13]– [14]). A variety of viral agents, including West Nile virus (WNV), varicella zoster virus (VZV), and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), have been implicated in the development of autoimmune encephalitis (AE). Since the global emergence of the post-COVID-19 era, there has been an increase in reports linking SARS-CoV-2 to AE pathogenesis [15–18]. (2) MOGAD: MOGAD is often associated with postinfectious or persistent immune activation. A cohort study of 39 MOGAD cases revealed preceding infections, including respiratory infections (46.2%), gastrointestinal infections (7.7%), urinary tract infections (2.6%), and recent immunizations documented in 12.8% of cases. Fever was reported in 31% of cases. Mycoplasma IgM was elevated in four patients, and evidence of recent Epstein-Barr virus infection was present in two patients. Prodromal symptoms, including fever, malaise, or upper respiratory tract infection, are reported in 30–50% of cases ([18]– [19]). Since the global outbreak of COVID-19, several studies have reported cases of SARS-CoV-2-associated MOGAD, which has become one of the major infectious triggers in current MOGAD-related prodromal infections [20–22]. Li et al. ([22]– [23]) reported a case involving human herpesvirus type 7 infection concurrent with anti-NMDAR encephalitis and MOG-associated inflammatory demyelinating disease, suggesting a potential role of viral infections in the pathogenesis of these conditions. Pathogenetic implications include the possibility that infectious agents may act as immune triggers through mechanisms such as molecular mimicry or bystander activation; overlapping antibody cohorts exhibit intermediate infection prevalence (30.8%) between MOGAD (33.3%) and anti-NMDAR encephalitis (20.8%). However, the limited sample size (n = 13) restricts statistical power for subgroup comparisons and further conclusions will require analysis of additional cases. No neoplastic lesions were detected in the overlapping antibody cohort (n = 13), consistent with established literature demonstrating no significant tumor association in this subgroup [10, 24]. This contrasts with anti-NMDAR encephalitis, where ovarian teratomas are reported in 20–30% of cases [25]; our study identified three malignancies (12.5% incidence), including one ovarian teratoma. In contrast, MOGAD exhibited a complete absence of tumors, consistent with prior reports of low malignancy prevalence in this population [8]. Therefore, tumor screening is not recommended for patients with overlapping antibodies unless clinically indicated. A total of 31 patients required intensive care across the three cohorts: dual antibody-positive cohort, 5 cases; MOGAD, 16 cases; and anti-NMDAR encephalitis, 10 cases. Although anti-NMDAR encephalitis exhibited the highest crude ICU admission rate (41.7%), this difference did not reach statistical significance (P >0.05). The observed trend aligns with prior pediatric studies demonstrating elevated critical care requirements in NMDAR encephalitis compared to MOGAD and dual antibody cohorts [26]. Age distribution analysis across ICU admissions revealed distinct cohort-specific patterns: (1) the dual antibody cohort demonstrated adult predominance (80.0% adults vs. 20.0% children); (2) in contrast, the MOGAD cohort exhibited marked pediatric predominance (81.3% children vs. 18.8% adults); (3) while the anti-NMDAR cohort showed intermediate demographics, with 70.0% adults and 30.0% children. Clinical correlates suggest that MOGAD’s pediatric predominance in ICU admissions likely reflects increased susceptibility to acute disseminated encephalomyelitis (ADEM)-like presentations [27]; anti-NMDAR encephalitis ICU admissions in adults predominantly involved severe neurological deterioration (e.g., status epilepticus, coma); the dual antibody cohort demonstrated bimodal age distribution, with adults exhibiting higher severity scores at ICU admission. This epidemiological profile suggests that immune-mediated central nervous system disorders exhibit distinct critical care requirements based on antibody profile and age.​​Patients with overlapping antibodies exhibit complex clinical profiles characterized by anti-NMDAR encephalitis-like features, including psychobehavioral disturbances, cognitive impairment, and seizure activity; demyelinating disease manifestations include optic neuritis, myelitis, and brainstem encephalitis. This clinical heterogeneity significantly challenges diagnostic accuracy, with initial presentations often mimicking general encephalitis (e.g., headache and fever). A study by Nan et al. [9] systematically characterized this overlap cohort, identifying seizures and cognitive dysfunction as the most prevalent disease-defining features. The coexistence of limbic encephalitis and demyelinating features complicates differential diagnosis, necessitating combined evaluation with serum/CSF antibody testing (anti-MOG/NMDAR), neuroimaging findings (MRI evidence of demyelination), and EEG patterns (generalized slowing with epileptiform discharges).

In this study, patients with MOGAD predominantly exhibited core symptomatology: limb weakness, optic neuritis, gastrointestinal dysfunction, headache, and fever. Previous research involving children with overlapping antibodies has demonstrated that, compared to those with MOGAD, these patients exhibit a higher incidence of psychological and behavioral disturbances as well as epilepsy [28]. Additionally, other studies have reported that sleep disorders, along with seizures and psychological or behavioral symptoms, occur significantly more frequently in MOGAD patients [26]. However, some studies have also suggested that there are minimal differences in clinical manifestations between the two groups [9]. In our study, compared with the MOGAD group, the overlapping antibody group was significantly more likely to experience psychological and behavioral abnormalities (P = 0.013), consistent with findings from previous studies. Previous studies also revealed that the overlapping antibody group exhibited more severe clinical symptoms [9]. Therefore, when patients with MOGAD present with psychiatric and behavioral disturbances, testing for anti-NMDAR antibodies is recommended. Notably, MOGAD patients exhibited higher absolute epilepsy rates (48.9% vs. 23.1%). Previous studies have also reported cases of unilateral cortical encephalitis with epileptic seizures in patients testing positive for MOG antibodies. Dual antibody cohorts demonstrated distinct epileptiform patterns: FLAMES syndrome (FLAIR-hyperintense lesions in anti-MOG encephalitis with seizures), characterized by unilateral cortical hyperintensities on FLAIR imaging, seizure frequency ≥ 3/month in 78% of cases, and coexistence with anti-NMDAR antibodies [29]. This may be attributed to the substantial number of patients in this category included in our study. FLAMES may also coexist with anti-NMDAR antibodies, and distinguishing patients with FLAMES from those with overlapping syndromes based solely on clinical characteristics is considered challenging [30]. Our findings further suggest that in MOGAD patients presenting with seizures, clinicians should not only consider the potential presence of concomitant anti-NMDAR antibodies but also the rare clinical phenotype of FLAMES. For such cases, comprehensive testing for antibodies associated with autoimmune encephalitis is warranted, as this increases the likelihood of detecting coexisting anti-NMDAR and MOG antibodies. Clinical Implications display (1) MOGAD patients with seizures warrant extended autoimmune encephalitis panel testing (including NMDAR antibodies) and FLAIR MRI sequence analysis for unilateral cortical lesions. (2) Dual antibody cohorts require early electroclinical correlation for FLAMES identification and combined immunotherapy targeting MOG/NMDAR pathways. Methodological Considerations show sample size limitations (n = 45 MOGAD) may underestimate comorbidity rate; FLAMES prevalence is likely underreported due to routine FLAIR sequence omission in 35% of MOGAD imaging protocols.

While Ding et al. [31] reported no significant clinical differences between dual antibody-positive and pure anti-NMDAR encephalitis cohorts, our retrospective analysis revealed critical distinctions: psychiatric/behavioral disturbances occurred in 83.3% of anti-NMDAR encephalitis patients versus 38.5% in the dual antibody cohort (P = 0.010); epilepsy prevalence was 62.5% vs. 23.1%, respectively (P = 0.038). Prior research has indicated that patients with overlapping antibodies tend to exhibit milder symptoms—such as psychiatric disturbances and cognitive impairment—compared to those with isolated anti-NMDAR encephalitis [32, 33]. Given the resemblance in clinical presentation, differentiating between anti-NMDAR encephalitis and overlapping antibody syndromes based solely on symptomatology remains challenging.

​​CSF analysis from initial lumbar punctures in 13 patients with overlapping antibodies demonstrated characteristic patterns: elevated intracranial pressure in 66.7% of adults, pleocytosis in 92.3%, and a higher proportion of lymphocytes in >50% of cases. Biochemical analyses revealed slight elevations in CSF protein, glucose, and chloride levels. Isoelectric focusing revealed OCB in these patients, a finding classically associated with multiple sclerosis (MS) [34], but also observed in other immune-mediated neurological disorders. Comparative analysis between the overlapping antibody cohort and separate groups with anti-NMDAR encephalitis or MOGAD revealed comparable biochemical profiles, consistent with prior reports.

In 24 anti-NMDAR encephalitis patients, transient OCB positivity was accompanied by mild protein elevation, aligning with the transient nature of intrathecal immunoglobulin synthesis in this condition [35]. Among overlapping antibody patients experiencing relapse (n = 3), OCB retesting showed persistent positive bands in three cases, while two were type 2 and one was type 3. The current study suggests that the relationship between OCB and MS is relatively certain, while although OCB type 2 or 3 has been detected in other diseases, it only suggests the presence of intrathecal synthesis, and whether it has other clinical significance remains to be determined.

Neuroimaging abnormalities were observed in 69.2% of patients with overlapping autoantibodies, predominantly involving cortical and subcortical regions (Fig. 1A). Secondary involvement occurred in the lateral ventricles (30.8%), brachium pontis (23.1%), basal ganglia (23.1%), thalamus (23.1%), and other supratentorial regions. This pattern aligns with prior descriptions of overlapping antibody syndromes showing preferential supratentorial lesions [36], while infratentorial and spinal cord involvement occurs less frequently. Additionally, leptomeningeal enhancement has been observed in some patients with overlapping antibodies [37]. Comparative analysis in our study revealed distinct patterns across antibody groups: (1) Overlapping antibodies​​: cortical/subcortical involvement predominated, with leptomeningeal enhancement observed in 2/9 gadolinium-enhanced MRIs. (2) Compared with patients with overlapping antibodies, MOGAD patients exhibited more frequent cortical and subcortical involvement. Gadolinium-enhanced imaging revealed a significantly higher incidence of abnormal enhancement in MOGAD patients. In contrast, lesions in MOGAD patients commonly affected the cortex and subcortical white matter, thalamus, brainstem, cerebellum, and other regions, with longitudinally extensive myelitis and optic neuritis also frequently observed [38]. Our MOGAD cohort similarly showed predominant involvement of the cortex, subcortical regions, basal ganglia, thalamus, pons, and lateral ventricles, consistent with previous findings. The emergence of a novel cerebral cortical encephalitis phenotype in MOGAD patients [39] —characterized by frequent seizures (48.9% in our cohort) and pial enhancement (68.2%)—may explain this discrepancy. The greater representation of this phenotype in our MOGAD patients, or a potential association between MOG antibodies and meningeal enhancement, may explain the lower incidence of such enhancement in patients with overlapping antibodies. While gadolinium enhancement patterns showed statistical differences between groups, our findings corroborate emerging evidence that MOG antibody-associated cortical encephalitis represents a distinct clinico-radiological entity with unique enhancement characteristics compared to overlapping syndromes.

In patients with anti-NMDAR encephalitis, characteristic MRI abnormalities predominantly involve the cerebral cortex, subcortical structures (particularly the medial temporal lobes and limbic system), and lateral ventricles. Our cohort demonstrated MRI abnormalities in 45.8% of cases, aligning with established literature reporting a 40—50% prevalence of MRI changes in this condition [40]. Notably, comparative analysis revealed that anti-NMDAR encephalitis patients exhibited significantly lower rates of brachium pontis involvement compared to individuals with overlapping autoantibody (p = 0.037). Enhanced MRI evaluation through gadolinium administration was performed in 17 cases, revealing abnormal meningeal or parenchymal enhancement patterns in 6 patients (35.3%), including two cases demonstrating leptomeningeal contrast enhancement. However, these enhancement patterns showed no statistically significant difference between our anti-NMDAR cohort and patients with overlapping antibody profiles. The absence of spinal cord imaging data and limited sample size underscore the need for larger prospective studies to better characterize central nervous system involvement patterns. Current evidence suggests that while conventional brain MRI remains a critical diagnostic tool, its sensitivity may be suboptimal for early diagnosis. Importantly, the presence of pontine lesions should prompt clinicians to consider differential diagnoses involving MOG antibodies, as this finding may indicate overlapping autoimmune pathology.

The management of patients exhibiting overlapping autoantibody profiles presents unique clinical challenges due to the paucity of evidence-based guidelines. In our cohort of 13 cases, all patients received first-line induction therapy during acute episodes, consisting of IVMP, IVIG, and PLEX. Notably, 7.7% of cases required escalation to RTX. During disease recurrence, treatment strategies demonstrated heterogeneity: with CTX utilized in 20% of cases, MMF in 20%, and combination AZA/MMF in 20%. These observations suggest that conventional immunosuppressive agents beyond first-line therapies may be necessary for refractory cases. Established therapeutic paradigms for autoimmune encephalopathies provide relevant comparative frameworks. Anti-NMDAR encephalitis protocols typically escalate from first-line therapies (IVMP/IVIG/PLEX) to second-line agents such as RTX or CTX in treatment-resistant cases [5]. Similarly, in the management of MOGAD, maintenance immunosuppression with mycophenolate or azathioprine is emphasized following acute-phase induction [1]. Consequently, therapeutic approaches for patients with overlapping antibodies may be guided by treatment protocols established for both anti-NMDAR encephalitis and MOGAD. This dual-tiered approach informs the management of overlapping antibody syndromes, where individualized treatment regimens should incorporate: (1) Early introduction of second-line biologic agents in poor responders; (2) Transition to disease-modifying therapies during remission phases; (3) Active monitoring for antibody-mediated relapse patterns. This comparative analysis underscores the necessity for standardized treatment algorithms integrating biomarker-guided therapeutic escalation (e.g., CSF NMDAR-IgG titers), sequential immunosuppression protocols and long-term maintenance strategies tailored to specific antibody profiles.

Clinical outcome analysis revealed distinct disability trajectories across antibody-defined cohorts. Baseline mRS scores indicated moderate disability among patients with overlapping autoantibody (median = 2, IQR = 1–2.5), which improved to a median mRS score of 1 (IQR = 0.5–1) following treatment. While the anti-NMDAR encephalitis cohort presented with marginally higher baseline disability (median mRS = 2.5, IQR = 1.25–4), this difference did not achieve statistical significance. Post-treatment improvement trends were comparable across groups (p > 0.05 for intergroup comparisons), supporting similar therapeutic responsiveness. Nevertheless, the higher pre-treatment median mRS score observed in patients with anti-NMDAR encephalitis appears to corroborate previous observations suggesting that individuals with overlapping antibodies may exhibit comparatively milder clinical manifestations relative to those with NMDAR encephalitis. Across all three groups, reductions in mRS scores following treatment indicate that acute-phase immunotherapy may effectively mitigate disease severity.

Recurrence dynamics revealed autoantibody -specific patterns (Fig. 4): overlapping autoantibody 38.5%, MOGAD 46.7%, anti-NMDAR encephalitis 34.8%. No statistically significant differences were observed among the groups. The data indicated that the recurrence rate of anti-NMDAR encephalitis is 20–30% [41], consistent with the findings of our study. The recurrence rate of MOGAD has been reported as 40–60% [42], whereas our study observed a rate of 46.7%, with no significant difference. Although overlapping autoantibody demonstrated a numerically lower recurrence rate compared with MOGAD, this contrasts with a meta-analysis reporting a 63.4% recurrence rate in overlapping cohorts [31]. Our study found a lower recurrence rate of 38.5%, which may be attributable to the small sample size. A pediatric study indicated that 72.7% of children with overlapping antibodies experienced recurrence, a figure more closely aligned with the recurrence rate observed in MOG antibody-associated disease. Based on these observations, the authors proposed that MOG antibodies may play a predominant role in the recurrence of overlapping syndromes in pediatric populations [12]. The observed disparity may reflect selection bias in pediatric cohorts, where a 72.7% recurrence rate in children with overlapping antibodies [12] aligns more closely with the natural history of MOGAD. This age-dependent recurrence pattern suggests that MOG antibody-mediated mechanisms may drive relapse propensity in younger patients. The 20–30% recurrence rate in monospecific anti-NMDAR encephalitis [41] contrasts with our overlapping cohort’s rate of 38.5%, potentially indicating an additive risk conferred by multiple autoantibodies. This hypothesis is supported by studies linking antibody heterogeneity to greater severity of immune dysregulation [43].

Given that the recurrence rate in patients with overlapping antibodies may exceed that in patients with single antibody-associated diseases, continuation of immunotherapy beyond the acute treatment phase may be considered as a strategy to reduce recurrence risk. To determine whether antibody coexistence contributes to an increased recurrence rate and to identify which specific antibodies exert the greatest influence on recurrence, long-term follow-up studies are essential. We followed three groups of patients and recorded the time from disease onset to first recurrence. During the follow-up period, Kaplan-Meier curves were constructed based on the time to first recurrence in the three patient groups. No statistically significant differences were observed between the overlapping antibodies group and either the MOGAD group or the anti-NMDAR encephalitis group. Due to the limited follow-up duration and small sample size, the presence of significant differences in cumulative recurrence rates among the groups requires further investigation through long-term follow-up studies. These findings have important therapeutic implications: (1) Prolonged immunosuppression may be warranted in overlapping antibodies given their intermediate recurrence risk; (2) MOG antibody titers should be monitored serially as potential biomarkers of relapse; (3) Age-adjusted recurrence prevention protocols are needed for pediatric populations. Future research directions proposed by this study include: prospective validation of correlations between antibody load and recurrence; extended follow-up (≥ 5 years) to capture late recurrences; and functional imaging studies to elucidate the mechanisms underlying relapse.

For patients with overlapping antibodies, identifying the predominant pathogenic antibodies is crucial for guiding targeted therapeutic interventions. Several observations support the hypothesis that MOG antibodies may represent the primary pathogenic antibodies in overlapping antibody. First, the low frequency of concurrent neoplasms in these patients aligns with the infrequent tumor association typically observed in MOG antibody-associated disease (MOGAD). Clinically, manifestations such as visual impairment and limb weakness, in conjunction with characteristic MRI abnormalities affecting the optic nerve and spinal cord, are more consistent with MOGAD. Additionally, in certain patients, serum MOG antibody titers were substantially higher than cerebrospinal fluid (CSF) NMDAR antibody titers, suggesting a predominant MOG antibody response. In cases where MOG antibodies were detected prior to the emergence of NMDAR antibodies, the latter may represent secondary or concomitant rather than primary pathogenic antibodies.

Evidence supports the pathogenic role of NMDAR antibodies, as clinical features such as psychiatric and behavioral disturbances observed in patients with overlapping antibodies are hallmark manifestations of NMDAR encephalitis. In some instances, CSF NMDAR antibody titers were significantly higher than serum MOG antibody titers [44]. When clinical and MRI features characteristic of both disorders coexist, this mixed phenotype implies the simultaneous activation of two distinct immune mechanisms, suggesting that both MOG and NMDAR antibodies may function as pathogenic antibodies in such cases.

This study has several methodological limitations. Due to geographical recruitment constraints, the single-center design inherently carries a risk of selection bias. However, the following strategies were implemented to mitigate this limitation: [1] Control of selection bias: The case samples were drawn from a wide geographical catchment area, covering the entire province, thereby enhancing sample representativeness. Despite the retrospective design, strict inclusion and exclusion criteria were applied (see Methods section for details) to minimize selection bias. Baseline characteristics were well balanced between comparison groups (Table 2), indicating good sample homogeneity [2]. Standardization of diagnostic procedures: All diagnostic evaluations were conducted according to standardized protocols and international guidelines. To address heterogeneity, cases with incomplete diagnostic data were excluded. For cases with diagnostic uncertainty, two independent physicians performed blinded re-evaluations [3]. A prospective multi-center validation study is currently in preparation. Due to the rarity of overlapping autoantibody, the limited sample size constrained statistical power, precluding the application of multiple testing corrections. Given the relative rarity of diseases with this dual antibody coexistence, even with a limited sample size, highlighting clinically observed trends holds significant value. These findings can provide valuable insights and references to enhance diagnostic accuracy and therapeutic precision in clinical practice.

Future Research Directions: Prospective Cohort Expansion: multi-center studies incorporating standardized treatment protocols are warranted to validate syndrome-specific CSF biomarker panels (e.g., GFAP/NfL ratios) and to develop dynamic threshold models for predicting relapse risk.

Conclusion

The overlapping antibody cohort demonstrated demographic homogeneity with the anti-NMDAR encephalitis and MOGAD groups, with no statistically significant differences in sex ratio, age distribution, or proportion of pediatric cases. While paraneoplastic associations remained infrequent, syndrome-specific clinical profiles were observed. Compared with the MOGAD group, the overlapping antibody group exhibited a higher incidence of psychiatric and behavioral disturbances, a greater proportion of adult patients requiring ICU admission, and reduced gadolinium enhancement on cranial MRI. Conversely, when compared with the anti-NMDAR encephalitis group, the overlapping antibody group showed lower frequencies of psychiatric and behavioral disturbances and seizures. Notably, in cases of anti-NMDAR encephalitis with cranial MRI lesions involving the brachium pontis, MOG antibody testing is recommended. No significant difference was observed in cumulative recurrence rates among the overlapping antibody group, the MOGAD group, and the anti-NMDAR encephalitis group.

Abbreviations

MOG

Myelin oligodendrocyte glycoprotein

MOGAD

Myelin oligodendrocyte glycoprotein-associated disease

anti-NMDARE

Anti-N-methyl-D-aspartate receptor encephalitis

CSF

Cerebrospinal fluid

MRI

Magnetic resonance imaging

EEG

Electroencephalogram

CNS

Central nervous system

AQP4

Aquaporin-4

GFAP

Glial Fibrillary Acidic Protein

mRS

Modified Rankin Scale

ICU

Intensive care unit

CBA

Cell-based immunofluorescence assays

FLAIR

Fluid-attenuated inversion recovery

CT

Computed tomography

FDG PET-CT

Fluorodeoxyglucose positron emission tomography-computed tomography

IVMP

Intravenous methylprednisolone

IVIG

Intravenous immunoglobulin

PLEX

Plasma exchange

AZA

Azathioprine

MMF

Mycophenolate mofetil

MTX

Methotrexate

CTX

Cyclophosphamide

RTX

Rituximab

SD

Standard deviation

IQR

Interquartile range

OCB

Oligoclonal banding

WNV

West Nile virus

VZV

Varicella zoster virus

SARS-CoV-2

Severe acute respiratory syndrome coronavirus 2

ADEM

Acute disseminated encephalomyelitis

FLAMES

FLAIR-hyperintense lesions in anti-MOG encephalitis with seizures

CASPR2

Contact-like protein 2

Authors’ contributions

YTL, JJW and MYS contributed equally to this work. YTL and LMW designed the study. YTL and MYS extracted, collected, and analyzed the data. JJW contributed to data analysis and interpretation. YTL and YL prepared tables and figures. YTL and LMW reviewed the results, interpreted the data, and assisted in writing the manuscript. All authors made intellectual contributions to the manuscript and approved the final submission.

Funding

This study was supported by the Science and Technology Department of Henan Province (Social Development, Grant No. 212102310825) and Natural Science Foundation of Henan Province (Grant No. 242300420072).

Data availability

The data that support the findings of this study are available from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the Clinical Research Ethics Committee of the First Affiliated Hospital of Zhengzhou University (Ethics Committee Reference: 2021-KY-0822-002). Written informed consent was obtained from all participating patients. For participants under the age of 16, consent was provided by their parents or legal guardians.

Consent for publication

Written informed consent for publication was obtained from all participants included in the study. For individuals under the age of 18, consent was provided by their parents or legal guardians.

Competing interests

The authors declare no competing interests.