Clinical heterogeneity and immunotherapy outcomes in anti-GAD65 antibody-associated autoimmune encephalitis: a retrospective study

Abstract

Anti-glutamic acid decarboxylase 65 (GAD65) antibody-associated autoimmune encephalitis (AE) is characterized by diverse neurological manifestations and systemic autoimmune comorbidities, posing diagnostic and therapeutic challenges. This study aimed to investigate the clinical heterogeneity, immunological characteristics, neuropathological features, and therapeutic outcomes in patients with anti-GAD65 AE. A retrospective study was performed in six patients with anti-GAD65 AE diagnosed at Guangdong Provincial People’s Hospital. Collected data included demographic, clinical manifestations, cerebrospinal fluid (CSF) profiles, neuroimaging findings, antibody titers, immunotherapy regimens, and histopathology. Initial presentations included epilepsy (50%), hyperthyroidism (33%), and stiff-person syndrome (17%). Concurrent autoimmune diseases (type 1 diabetes, autoimmune thyroiditis, and polyendocrine syndrome) were present in half of the patients. All patients exhibited elevated anti-GAD65 antibody titers in serum and CSF. CSF analysis revealed lymphocytic pleocytosis in one patient and oligoclonal bands in four tested cases. Acute immunotherapy involving corticosteroids and intravenous immunoglobulin achieved seizure control in 83%. Maintenance therapy with mycophenolate mofetil or biologics (e.g., ofatumumab, telitacicept) elicited heterogeneous serological responses, including antibody titer rebound despite clinical stability. Notably, Case 6 represented the first reported case worldwide of AE with triple positivity (anti-GAD65, anti-GABA_ARα1, and anti-GQ1b IgG antibodies), presenting with rapidly progressive cognitive impairment, extensive parenchymal lesions, and extremely high anti-GAD65 titers, yet exhibited significant improvement post-immunotherapy. Histopathology revealed CD3⁺/CD8⁺ T-cell infiltrates in perivascular, meningeal, and parenchymal regions, accompanied by aggregates of CD20⁺ B cells, CD38⁺ plasma cells, GrB⁺ non-T cells (e.g., NK cells), and activated CD68⁺ macrophages/microglia. Anti-GAD65 AE exhibits marked clinical and immunological heterogeneity, involving both humoral and T-cell-mediated pathogenic mechanisms. Coexisting antibodies may exacerbate neuronal injury through mechanisms such as antigen unmasking and antibody-dependent cellular cytotoxicity. Persistent antibody elevation and subclinical immune activity underscore the necessity of integrating multimodal biomarkers and developing personalized treatment strategies.

Keywords

anti-GAD65 antibodies autoimmune encephalitis clinical heterogeneity immunotherapy

Introduction

Autoimmune encephalitis (AE) is an increasingly recognized inflammatory disorder of the central nervous system (CNS), for which diagnosis largely relies on the sensitivity and specificity of neuronal antibody testing. Among the subtypes of AE, those associated with antibodies against glutamic acid decarboxylase 65 (GAD65) have attracted considerable attention owing to their heterogeneous clinical presentations and frequent coexistence with multiple autoimmune syndromes. GAD65 is an intracellular enzyme that catalyzes the synthesis of gamma-aminobutyric acid (GABA), the principal inhibitory neurotransmitter in the CNS. GAD65 antibodies are associated with a spectrum of neurological disorders, including limbic encephalitis, cerebellar ataxia, and stiff-person syndrome (SPS).^1–4 Furthermore, GAD65-associated AE may coexist with other autoimmune diseases, including autoimmune thyroiditis (AIT), type 1 diabetes mellitus (T1DM), and autoimmune polyendocrine syndrome type II (APS II).^5–7

The broad clinical spectrum of GAD65-associated disorders complicates diagnosis, affects treatment strategies, and determines prognosis. Early diagnosis and timely immunotherapy are essential for optimizing clinical outcomes. By synthesizing evidence from recent cases and clinical studies, we aim to enhance the understanding of this complex autoimmune disorder and to inform future research directions and clinical management.

Materials and methods

Study participants

This retrospective study was conducted in the Department of Neurology of Guangdong Provincial People’s Hospital. We systematically reviewed the medical records of patients diagnosed with anti-GAD65 antibody-positive AE between January 2020 and December 2024.

Inclusion criteria were as follows: (1) diagnosis of anti-GAD65 AE according to the criteria proposed in the “Chinese Expert Consensus on the Diagnosis and Treatment of Autoimmune Encephalitis (2022 Edition),” in conjunction with the diagnostic framework described in the Lancet publication entitled “clinical approach to diagnosis of autoimmune encephalitis”⁸; (2) positive anti-GAD65 antibodies in serum or cerebrospinal fluid (CSF); and (3) complete clinical and paraclinical data.

Exclusion criteria were as follows: (1) isolated serum or CSF GAD65 positivity without neurological manifestations; (2) negative GAD65 antibodies in both serum and CSF; and (3) incomplete clinical data or poor treatment compliance.

All data were retrieved from electronic medical records following acquisition of written informed consent from the patients.

Research methods

Peripheral blood and CSF samples were collected from all enrolled patients. Antibody testing was performed on both CSF and serum samples employing a cell-based assay to detect antibodies targeting neuronal cell-surface and synaptic proteins, including the N-methyl-D-aspartate receptor, GABA receptor, leucine-rich glioma-inactivated 1 protein, GAD65, etc.

For each patient, the following clinical and laboratory data were collected: general demographics characteristics and clinical symptomatology, including gender, age at onset, clinical manifestations, coexistence of other autoimmune disorders, immunosuppressive therapies administered, and clinical outcomes; laboratory assessments included complete blood count, liver and renal function, fasting blood glucose, glycated hemoglobin (HbA₁c), autoimmune diabetes-associated antibodies, thyroid function, and thyroiditis antibodies; lumbar puncture was performed for CSF analysis, including routine examination, biochemistry, microbiology, and immunofixation electrophoresis; tumor screening included tumor markers, chest computed tomography, and abdominal ultrasonography; neuroimaging and monitoring evaluation included magnetic resonance imaging (MRI) and 24-h video electroencephalogram (EEG) monitoring.

Case 6, which was initially misdiagnosed as a low-grade glioma, subsequently underwent surgical resection. Formalin-fixed, paraffin-embedded tissue specimens were sectioned at a thickness of 4 μm. Routine hematoxylin and eosin (H&E) staining was performed for initial histological evaluation. Immunohistochemical (IHC) staining was performed to qualitatively evaluate immune responses within the tissue using markers specific for inflammatory cells, including T cells, B cells, plasma cells, macrophages, and microglia.

Statistical analysis

Given the retrospective case series design with a small sample size (n = 6), only descriptive statistics were applied. Categorical variables are reported as frequencies (percentages), and continuous variables as mean ± standard deviation or median (range), as appropriate. No inferential statistical tests were performed.

Reporting guideline

This case series has been reported in accordance with the CARE (CAse REport) guidelines. The completed CARE checklist is provided as Supplementary Material.

Results

Demographic and clinical characteristics

The study cohort consisted of six patients diagnosed with anti-GAD65 antibody-positive AE (4 females; mean age: 32.8 years, range: 22–36). Initial clinical manifestations included epilepsy (n = 3), hyperthyroidism (n = 2), and axial muscle rigidity (n = 1). Autoimmune comorbidities were identified in three patients, including autoimmune diabetes mellitus (Case 3), APS II accompanied by anemia (Case 4), and AIT (Case 5). Islet cell antibody positivity in the absence of diabetes was observed in Cases 2 and 5. Case 5 fulfilled the diagnostic criteria for SPS and exhibited emotion-triggered rigidity. No history of antecedent infections or familial genetic disorders was reported in any patient (Tables 1 and 2).

Table 1.

Clinical data of six AE patients with anti-glutamic acid decarboxylase antibodies.

Patient No.	Age at onset (years)	Sex (F/M)	Onset symptoms	Clinical manifestations		Serum/CSF GAD65 antibodies	MR	EEG		Therapeutic regimen		Prognosis
Patient No.	Age at onset (years)	Sex (F/M)	Onset symptoms	Prodromal infection	Clinical phenotype	Serum/CSF GAD65 antibodies	MR	EEG background	Epileptiform discharges	Acute treatment	Maintenance immunotherapy	Prognosis
Case 1	27	F	Epilepsy	No	AE	1:100/1:100	Normal	Intermittent θ waves	Lt. temporal, Rt. frontotemporal	High-dose corticosteroids and IVIG pulses	Prednisone	Serum GAD65 titer 1:300; Recurrent generalized seizures
Case 1	27	F	Epilepsy	No	AE	1:100/1:100	Normal	Intermittent θ waves	Lt. temporal, Rt. frontotemporal	High-dose corticosteroids and IVIG pulses	Prednisone + Ofatumumab	Serum GAD65 titer 1:100; Intermittent complex partial seizures
Case 2	27	F	Hyperthyroidism	No	AE	1:100/1:100	Normal	Scattered θ waves	Absence of epileptiform activity	High-dose corticosteroids and IVIG pulses	MMF	Serum GAD65 titer 1:320; no seizures
Case 2	27	F	Hyperthyroidism	No	AE	1:100/1:100	Normal	Scattered θ waves	Absence of epileptiform activity	High-dose corticosteroids and IVIG pulses	Switched to Taisemp	Serum GAD65 titer 1:10; Seizure-free, with normalized thyroid function
Case 3	22	M	Epilepsy	No	AE and ADM	1:100/1:100	Normal	Scattered θ waves	Rt. frontopolar + ant./mid-temporal	High-dose corticosteroids and IVIG pulses	MMF	Serum GAD65 titer 1:1000; no seizures, with blood glucose control
Case 4	36	F	Hyperthyroidism	No	AE with APS II and anemia	1:100/1:100	Normal	Scattered θ waves	Rt. frontopolar + ant./mid-temporal	IVIG pulses + Initial IVMP 40 mg	MMF	Seizures uncontrolled
Case 5	33	M	Muscle rigidity	No	AE with SPS and AIT	1:100/1:10	Bil. frontoparietal punctate: ↓T1, ↑T2/FLAIR, no enh.	Scattered θ waves	Rt. anterior/mid-temporal	High-dose corticosteroids and IVIG pulses	MMF	Serum GAD65 titer 1:10,000; no seizures or muscle rigidity.
Case 6	36	F	Epilepsy	No	AE	1:10,000/1:320	Bil. frontotemporal + Lt. CC genu: ↓T1, ↑T2/DWI (no ↓ADC), meningeal enh. ↓NAA (Lt. Cho/NAA = 1.58; Rt. 0.95), post-immunotherapy reduction.	Bil. ant./mid- temporal + Bil. frontopolar + Lt. post. occipitotemporal: Bursts δ-θ waves	Absence of epileptiform activity	High-dose corticosteroids and IVIG pulses	MMF	Serum GAD65 negative, no seizures, MRI lesions reduced

ADC, apparent diffusion coefficient; ADM, autoimmune diabetes mellitus; AE, autoimmune encephalitis; AIT, autoimmune thyroiditis; Ant, anterior; APS II, autoimmune polyendocrine syndrome type II; Bil, bilateral; CC genu, corpus callosum genu; CSF, cerebrospinal fluid; DWI, diffusion-weighted imaging, EEG, electroencephalogram; enh., enhancement; GAD, anti-glutamic acid decarboxylase; IVIG, intravenous immunoglobulin; Lt, left; Mid, middle; MMF, mycophenolate mofetil; NAA, N-acetylaspartate; Post, posterior; Rt, right; SPS, stiff-person syndrome.

Table 2.

Autoimmune diabetes and thyroiditis in autoimmune encephalitis patients with GADAs: serological and metabolic biomarkers.

Patient No.	FBG (mmol/L)	HbA₁c (%)	ADAb				TF	TAB				Tumor screening
Patient No.	FBG (mmol/L)	HbA₁c (%)	IAA (RU/mL)	ICA (IU/mL)	A-IA2 (IU/mL)	GADA (IU/mL)	TF	TG (ng/mL)	TgAb (IU/ML)	TPOAb (IU/mL)	TRAb (IU/L)	Tumor screening
Case 1	4.92	4.8	Not performed	Not performed	Not performed	Not performed	T4 170.93↑ nmol/L	Not performed	Not performed	Not performed	Not performed	Negative
Case 2	4.91	5.5	6.31	258.41↑	<5.00	221.34 ↑	T4 238.51↑ nmol/L, FT4 41.24↑ pmol/L, TSH <0.015 ↓μIU/mL	28.39	1047.00 ↑	442.20 ↑	11.44 ↑	Negative
Case 3	7.89↑	6.3 ↑	2.91	>400.00	<5.00	20.40 ↑	Normal	Not performed	Not performed	Not performed	Not performed	Negative
Case 4	3.94	7.9 ↑	15.14	>400.00	<5.00	11.49↑	Normal	1.11↓	172.00↑	89.02↑	12.59↑	Negative
Case 5	4.55	5.2	<2.00	>400.00	<5.00	3.19	Normal					Negative
								1.04↓	120.00↑	208.00↑	1.34
Case 6	3.84	4.9	Not performed	Not performed	Not performed	Not performed	TSH 0.451↓ μIU/mL	0.93↓	44.10	26.56	1.27	Negative

ADAb, autoimmune diabetes antibodies; A-IA2, anti-IA-2 antibody; FBG, fasting blood glucose; GADA, anti-glutamic acid decarboxylase antibody; HbA₁c, glycated hemoglobin; IAA, insulin autoantibody; ICA, islet cell autoantibody; TAB, thyroiditis antibodies; TF, thyroid function; TG, thyroglobulin; TgAb, thyroglobulin antibodies; TPOAb, thyroid peroxidase antibodies; TRAb, thyroid stimulating hormone receptor antibody.

“↓” indicates below the normal rang; “↑” indicates above the normal range

Laboratory and immunological profiles

CSF analysis revealed elevated lymphocytic pleocytosis (34 cells/μL; 98% lymphocytes) exclusively in Case 1, whereas all other patients exhibited normal leukocyte counts. Mild elevation of CSF protein was observed in 50% of patients (3/6), while glucose and chloride levels remained within normal ranges. Oligoclonal IgG bands restricted to CSF were detected in 3/4 tested cases. Case 3 demonstrated parallel oligoclonal bands in both CSF and serum, indicative of systemic humoral immune activation (Table 3).

Table 3.

The CSF results of six cases.

Patient No.	Appearance	Pressure (mmH₂O)	Cell count/cell differential	Biochemistry			Microbiology	Immunofixation electrophoresis
Patient No.	Appearance	Pressure (mmH₂O)	Cell count/cell differential	Chloride (mmol/L)	Glucose (mmol/L)	Total protein (mg/L)	Microbiology	IgG index	Oligoclonal bands
Case 1	Lucid	325	34.00 × 10⁶/L ↑ (Lymphocytes 98%)	124.6	3.86	520↑	Negative	0.66	Oligoclonal IgG bands in CSF only
Case 2	Lucid	120	3.00 × 10⁶/L	129.5	3.67	298	Negative	0.65	Oligoclonal IgG bands in CSF only
Case 3	Lucid	330	8.00 × 10⁶/L	125.2	4.69	507↑	Negative	0.66	Oligoclonal bands in CSF: unique bands + shared IgG bands
Case 4	Lucid	205	2.00 × 10⁶/L	126.2	5.00	283	Negative	0.53	Oligoclonal IgG bands in CSF only
Case 5	Lucid	160	2 × 10⁶/L	125	3.38	376	Negative	NTR	NTR
Case 6	Lucid	NTR	6 × 10⁶/L	122.1	3.41	595↑	Negative	NTR	NTR

CSF, cerebrospinal fluid; NTR, not traceable record.

“↑” indicates above the normal range.

All patients exhibited elevated anti-GAD65 antibody titers in both serum and CSF, with the highest titers observed in Case 6 (serum 1:10,000; CSF 1:320). In addition, Case 6 exhibited elevated anti-GABA_ARα1 antibodies in CSF (titer 1:3.2) and positive serum anti-GQ1b IgG antibodies, rendering the diagnosis controversial. Malignancy screening yielded negative results in all patients (Tables 1 and 2).

Neuroimaging, electrophysiological, and pathological findings

Case 6 was a 36-year-old woman who presented with recurrent focal seizures for a duration of 3 months. Initial brain MRI demonstrated a patchy abnormal signal in the right frontal lobe, appearing hypointense on T1-weighted imaging, hyperintense on T2-weighted and diffusion-weighted imaging (DWI), without restricted diffusion on apparent diffusion coefficient (ADC) maps, suggestive of a low-grade glioma (Figure 1(a1)–(a4)). Histopathological examination of the surgically resected brain tissue revealed vascular dilation and congestion, accompanied by scattered perivascular lymphocytic infiltrates on H&E staining (Figure 2(a)). IHC analysis demonstrated distinct patterns of inflammatory infiltration within various brain regions. CD3⁺ T cells and CD8⁺ T cells were observed infiltrating the meninges, brain parenchyma, and perivascular regions, with the most prominent accumulation localized around blood vessels (Figures 2(b) and (c) and 3(c), (d), (g), and (h)). Within the meninges, aggregates of CD20⁺ B cells and CD38⁺ plasma cells were identified (Figure 3(a) and (b)). In the brain parenchyma, scattered granzyme B-positive (GrB⁺) cells, identified as CD3⁻CD8⁻ non-T cells, were observed in close apposition to neurons (Figure 2(d) and (h)). GrB-positive granules were identified within the cytoplasm of some neurons (Figure 2(d) and (h)). In addition, CD68⁺ activated macrophages/microglia were also observed (Figure 2(g)). By contrast, CD20 and CD38 immunostaining revealed scarcely any detectable B cells within the parenchyma (Figure 2(e) and (f)). In perivascular regions, T-cell infiltration was prominent, whereas B cells were not detected (Figure 3(f)–(h)). GrB immunostaining in the meninges yielded negative results (Figure 3(e)). Postoperatively, the patient developed speech difficulties, memory impairment, and recurrent seizures. Follow-up MRI demonstrated progressive enlargement of the lesions, involving the bilateral frontal and temporal lobes, as well as the left genu of the corpus callosum. Contrast-enhanced MRI revealed linear enhancement along the lesion margins and the adjacent meninges. Magnetic resonance spectroscopy (MRS) of the bilateral frontal regions revealed reduced N-acetylaspartate (NAA) peaks in both the left (Cho/NAA ratio: 1.58) and right lesion (Cho/NAA ratio: 0.95), with mildly elevated choline (Cho) peaks in the left lesion (Figure 1(b1)–(b6)). Following immunotherapy, follow-up MRI revealed marked lesion reduction. By contrast, MRI scans of the remaining five patients revealed no structural abnormalities or pathological signal intensities.

Figure 1.

Evolution of brain lesions on MRI in response to immunotherapy. (a1–a4) Baseline MRI (Month 0): A right frontal lesion exhibited T2 hyperintensity (a1), diffusion restriction on DWI (a2) without corresponding ADC hypointensity (a3), and T1 hypointensity (a4). (b1–b6) Pre-treatment MRI (Month 2): Disease progression with new lesions in the bilateral frontal and temporal lobes and corpus callosum (b1–b4). Post-contrast T1-weighted imaging (b4) revealed linear leptomeningeal enhancement. MRS of a left frontal lesion (b5) showed an elevated choline level (Cho/NAA = 1.58), while a right frontal lesion (b6) exhibited a Cho/NAA ratio of 0.95. (c1–c4) Post-immunotherapy MRI (Month 4): Significant regression of the lesions was observed following immunotherapy.

Figure 2.

Pathological features of the brain parenchyma. (a) H&E staining. (b) CD3 and (c) CD8 immunostaining revealed T-cell infiltration within the parenchyma. (d, h) GrB staining identified GrB⁺ cells (CD3⁻CD8⁻ non-T cells) in the parenchyma adjacent to neurons. GrB granules were identified within the cytoplasm of some neurons. (g) CD68⁺ staining indicated activated macrophages/microglia. (e) CD20 and (f) CD38 staining showed scant detectable B cells. Scale bars: (a–g) 200 μm, (h) 60 μm.

Figure 3.

Pathological features of the meninges and perivascular regions. Immunohistochemical staining was performed on serial sections: panels (a–e) and (f–h) represented two independent sets of consecutive sections, respectively. In the meninges, (a) CD20, (b) CD38, (c) CD3, and (d) CD8 staining demonstrated B-cell and T-cell infiltration. (e) GrB staining was negative. In the perivascular regions, (f) CD20, (g) CD3, and (h) CD8 staining indicate T-cell infiltration, with no detectable B cells. Scale bars: (a–e) 200 μm; (f–h) 60 μm.

Abnormalities in background EEG activity were universally present (6/6, 100%), predominantly manifesting as intermittent or scattered theta waves. Epileptiform discharges were observed in four patients (66.7%), localized to frontotemporal regions. Case 6 exhibited bursts of delta-theta waves involving bilateral frontopolar, anterior/middle temporal, and left posterior occipitotemporal regions, in the absence of epileptiform discharges (Table 1).

Treatment and clinical outcomes

Acute phase treatment: five patients received intravenous methylprednisolone (IVMP; 1000 mg/day for 3 days) combined with intravenous immunoglobulin (IVIG; 0.4 g/kg/day for 5 days). IVMP dose was subsequently reduced by half every 3 days until reaching 60 mg/day, converted to oral administration, and then tapered by 4 mg every 1–2 weeks. Exceptionally, Case 4, with hepatitis B virus comorbidity, initiated IVMP at 40 mg alongside IVIG pulse therapy prior to tapering.

Maintenance therapy: Mycophenolate mofetil (MMF) 1000 mg/d was administered to five patients (Cases 2–6), yielding variable therapeutic responses. In Case 2, serum anti-GAD65 titers increased to 1:320 despite the absence of seizures. Subsequent treatment with telitacicept (a BLyS/APRIL inhibitor) reduced the titers to 1:10.

Clinical stability was achieved in Case 3 (with seizure control, glucose normalization) and Case 5 (with the absence of seizures and rigidity). However, concurrent elevation of anti-GAD65 titers (1:1000 in Case 3 and 1:10,000 in Case 5) in these patients indicated a dissociation between serological and clinical responses suggestive of subclinical immune activity. Case 4 experienced recurrent seizures. Case 6 exhibited seroconversion of serum GAD65 antibody from 1:10,000 to negative, coinciding with seizure remission and marked regression of MRI lesions (Figure 1(c1)–(c4)), whereas serum anti-GQ1b IgG remained persistently positive.

In Case 1, prednisone monotherapy failed to prevent recurrent generalized seizures, with titers rising from 1:100 to 1:300). The addition of ofatumumab (an anti-CD20 monoclonal antibody) reduced titers to 1:100 and attenuated seizure frequency to intermittent complex partial episodes.

Discussion

The role of GAD65 antibodies in the pathogenesis of neurological disorders, particularly AE, continues to be a topic of considerable interest and debate. In this study, we delineate the clinical and immunological heterogeneity of anti-GAD65-associated AE in a cohort of six patients, emphasizing cross-tissue autoimmune comorbidities and divergent therapeutic responses.

Clinical heterogeneity and autoimmune comorbidities

Our cohort exhibited a heterogeneous spectrum of clinical manifestations, with autoimmune comorbidities (diabetes mellitus, thyroiditis, APS II) in 50% of patients, consistent with prior reports.⁹ Anti-GAD65 antibodies are associated with a diverse array of neurological syndromes, including SPS, limbic encephalitis, cerebellar ataxia, and epilepsy.^10–12 GAD65 comprises three functional domains: N-terminal, catalytic middle, and C-terminal.¹³ Although SPS and T1DM exhibit distinct epitope specificities (linear vs conformational),¹⁴ epitope mapping within GAD-associated neurological disorders reveals a consistent pattern. Burbelo et al.¹⁵ showed that all SPS sera strongly recognized the central fragment corresponding to the catalytic domain, whereas the C-terminal fragment yielded weak signals only in a subset. Fouka et al.¹⁶ extended these findings to 27 patients across diverse phenotypes (SPS, SPS with ataxia, epilepsy), confirming exclusive reactivity against the catalytic core with no phenotype-specific differences. Thus, epitope diversity does not explain clinical heterogeneity. Instead, coexisting neuronal surface autoantibodies may contribute, as three GAD-positive epilepsy patients harbored unidentified antibodies binding to live hippocampal neurons.¹⁶ These findings underscore the need for comprehensive autoantibody screening beyond GAD65 alone. Given the clinical heterogeneity, systematic surveillance for subclinical autoimmune comorbidities is key to early detection and interrupting metabolic-neurological deterioration.

Pathogenicity of anti-GAD65 antibodies: Beyond GABAergic dysregulation

Anti-GAD65 antibodies inhibit GABA synthesis in vitro,¹⁷ but their intracellular target and failed passive transfer challenge direct pathogenicity.^18,19 Nevertheless, functional evidence supports GABAergic impairment: MRS revealed significantly reduced sensorimotor GABA levels in SPS patients,^20,21 correlating with clinical severity²¹; transcranial magnetic stimulation demonstrated motor cortex hyperexcitability with reduced intracortical inhibition and enhanced facilitation, confirming cortical GABAergic dysfunction²²; and a subset of SPS patients with cerebellar ataxia (SPS-Cer) suggests GABAergic dysfunction can affect distinct neural circuits.²³

Beyond humoral mechanisms, T-cell-mediated pathology has emerged as a critical contributor.²⁴ Burton et al.²⁵ demonstrated that GAD65-specific CD4⁺ T cells can infiltrate the CNS and induce neuronal injury in murine models. Bien et al.²⁶ reported granzyme-positive T cells apposing neurons without IgG or complement deposition in GAD65 encephalitis. In GAD-associated temporal lobe epilepsy, early-stage brain tissue shows dense CD8⁺ cytotoxic T-cell infiltration and microglial nodules without complement membrane attack complex deposition.²⁷ In GAD65 limbic encephalitis, activated CD8⁺ T cells are elevated in blood and CSF, and their intrathecal frequency inversely correlates with hippocampal volume and memory performance, with parenchymal CD8⁺ T cells expressing perforin.²⁸ Our histopathology (Case 6) revealed T-cell-dominated infiltrates (CD3⁺/CD8⁺) with meningeal B/plasma cells. Collectively, these observations strongly support a “dual-hit” model: antibodies may initiate synaptic dysfunction, while T-cell cytotoxicity perpetuates injury.²⁹

Immunotherapy responses and novel therapeutic targets

Current immunotherapeutic strategies target both GABAergic dysfunction and autoimmunity.^24,30 IVIg remains the only randomized controlled trials (RCT)-proven immunotherapy for SPS, improving stiffness and quality of life.³¹ Long-term data from 36 GAD antibody-positive SPS patients confirmed that approximately two-thirds maintain benefit with monthly IVIg maintenance therapy over a median of 3.3 years, though some exhibit diminishing efficacy, highlighting the need for more durable therapeutic options.³²

MMF suppresses B-cell/T-cell proliferation via inosine monophosphate dehydrogenase inhibition, impairing antibody-secreting cell (ASC) function.^11,29,33 In our cohort, it effectively reduced autoantibody titers in most patients (e.g., Cases 3 and 6), but Cases 2 and 5 showed paradoxical serological rebounds. Other immunotherapies, including conventional immunosuppressants, corticosteroids, plasmapheresis, and autologous stem cell transplantation, have shown limited or inconsistent efficacy.³⁰

B-cell depletion strategies remain controversial. The largest rituximab RCT (n = 24) found no significant difference from placebo at 6 months in SPS; strong placebo effects and scale insensitivity likely contributed. Yet, 4/12 (33%) rituximab-treated patients showed meaningful, video-documented clinical improvement.³⁴ In SPS-associated autoimmune retinopathy, rituximab stabilized systemic symptoms but failed to halt retinal degeneration.³⁵ By contrast, ofatumumab, a high-affinity anti-CD20 monoclonal antibody with broader lymphoid tissue penetration, benefited Case 1, reducing antibody titers and improving seizure control. CD20-targeted antibodies deplete mature B cells but spare autoantibody-producing plasma cells, and T-cell-mediated cytotoxicity limits efficacy, underscoring the need for strategies targeting both ASCs and T cells.

Emerging strategies directly target ASCs via dual BLyS/APRIL inhibition. Telitacicept, a BLyS/APRIL dual inhibitor, has been validated as effective and well-tolerated in autoimmune diseases.³⁶ In Case 2, it reduced GAD65 antibody titers and stabilized symptoms. Notably, CD19 CAR-T cells target pathogenic ASCs while sparing protective long-lived plasma cells and may also modulate cytotoxic CD8⁺ T cells,³⁷ aligning with the “dual-hit” model and offering a unified strategy for refractory GAD-associated syndromes. The first case of GAD65 cerebellar ataxia failing rituximab and cyclophosphamide achieved a 95% titer reduction, CSF seroconversion, and sustained improvement with only grade 1 cytokine release syndrome.³⁸ For severe refractory cases, tocilizumab (IL-6 receptor blockade) has shown efficacy in anti-GAD65-associated epilepsy by suppressing Th17-mediated inflammation and inhibiting B-cell differentiation into ASCs, thus also targeting both T-cell and humoral pathways.³⁹

Antibody titers and emerging biomarkers

Anti-GAD65 titers poorly reflect disease activity. High titers (>10,000 IU/mL) are associated with classic phenotypes but not severity.¹⁰ Immunotherapy may reduce titers, yet complete seronegativity is rare, and paradoxical rebounds occur.^29,40 In our cohort, Case 1 showed rising titers during steroid tapering with seizure relapse; B-cell depletion reduced both. Cases 3 and 5 had titer fluctuations without clinical change, suggesting intrathecal synthesis or persistent ASCs decouple serology from clinical activity.

Neurofilament light chain (NfL) may offer complementary prognostic information. In anti-GAD65 AE secondary to checkpoint inhibitors, elevated NfL correlated with progression despite falling GAD65 titers and normal MRI,⁴¹ consistent with NfL as a sensitive axonal injury marker.^42–44 Integrating multiple biomarkers—GAD65 titers for initial response, NfL for axonal integrity—may improve disease monitoring and treatment stratification.

Multi-antibody synergism in AE: Distinctive pathogenicity and therapeutic implications

The coexistence of multiple neural autoantibodies in AE constitutes a complex immunopathological phenomenon with significant implications for both diagnosis and treatment. This first-reported case of triple antibody positivity (anti-GAD65, anti-GABA_ARα1, and anti-GQ1b IgG) expands the clinical spectrum of multi-antibody syndromes. Only three prior cases of coexisting anti-GAD and anti-GABA_AR antibodies exist.^45–47 Multi-antibody AE typically presents with greater clinical severity compared to single-antibody disease.⁴⁸ This case differed from the other five and exhibited rapidly progressive cognitive decline, extensive parenchymal lesions, and exceptionally high GAD65 titers, suggesting synergistic pathogenicity mechanisms.

Neuroimaging revealed atypical bifrontotemporal and callosal involvement, diverging from the classic limbic-predominant pattern of anti-GAD65 encephalitis^1,3,5,9,49 and the brainstem/cerebellar of anti-GQ1b-related disorders,^50,51 while aligning with the multifocal pattern of anti-GABA_AR encephalitis.^52,53 Immunotherapy response with anti-GAD65 seroconversion supports its primary role; persistent anti-GQ1b without relapse suggests an epiphenomenon. Histopathology revealed T-cell-dominated infiltrates (CD3⁺/CD8⁺) in perivascular, meningeal, and parenchymal regions, with meningeal CD20⁺ B cells and CD38⁺ plasma cells indicating robust intrathecal antibody production—supporting a “dual-hit” model. Notably, GrB⁺/CD3⁻/CD8⁻ cells (likely NK cells) and intraneuronal GrB granules were observed, along with CD68⁺ activated macrophages/microglia. We propose that low-titer anti-GABA_ARα1 antibodies may enhance neuronal vulnerability via membrane disruption, exposing intracellular GAD65 (“antigen unmasking”). This facilitates antibody-dependent cellular cytotoxicity (NK cell-mediated) and T-cell amplification (via macrophage antigen presentation), driving synergistic neuronal injury. This model explains both disease severity and the favorable response to immunotherapy, as surface-directed antibodies enabled effective humoral immunotherapies.

Limitations

Limitations include retrospective design, small sample size, lack of neuronal co-labeling for GrB localization, and no immunofluorescence analysis for immune cell–neuron interactions.

Conclusion

Anti-GAD65-associated neurological disorders exemplify humoral–cellular interplay. The coexistence of multiple neural autoantibodies may amplify clinical severity through synergistic effects, underscoring the importance of comprehensive antibody profiling at diagnosis. Given clinical diversity, heterogeneous treatment responses, and prognostic challenges, personalized therapy is essential. Early diagnosis, integration of multimodal biomarkers, and strategic immunotherapy targeting distinct pathogenic pathways are cornerstone strategies for optimizing outcomes. Future research should focus on elucidating immunopathological mechanisms and conducting RCT to establish evidence-based treatments.

Supplemental Material

sj-docx-1-tan-10.1177_17562864261456519 – Supplemental material for Clinical heterogeneity and immunotherapy outcomes in anti-GAD65 antibody-associated autoimmune encephalitis: a retrospective study

Supplemental material, sj-docx-1-tan-10.1177_17562864261456519 for Clinical heterogeneity and immunotherapy outcomes in anti-GAD65 antibody-associated autoimmune encephalitis: a retrospective study by Yihui Qiu, Zhiheng Huang, Tengyun Ma, Lihuan Lan, Guixian Ma, Kun Nie, Yuhu Zhang, Lijuan Wang and Jiehao Zhao in Therapeutic Advances in Neurological Disorders

Footnotes

Acknowledgements

The authors would like to express their sincere gratitude to all the participants and clinical staff for their invaluable contributions to this study.

Declarations

ORCID iDs

Yihui Qiu

Yuhu Zhang

Supplemental material

Supplemental material for this article is available online.

References

Chengyu

Weixiong

Chao

, et al. Clinical features and immunotherapy outcomes of anti-glutamic acid decarboxylase 65 antibody-associated neurological disorders. J Neuroimmunol 2020; 345: 577289.

Zhang

Lin

, et al. Clinical characteristics and outcomes of autoimmune encephalitis patients associated with anti-glutamate decarboxylase antibody 65. Clin Neurol Neurosurg 2020; 196: 106082.

Zhang

Dai

Han

, et al. A case of anti-GAD 65 autoimmune encephalitis associated with focal segmental Stiff-Person syndrome. Brain Sci 2023; 13: 369–376.

Raju

Foote

Banga

, et al. Analysis of GAD65 autoantibodies in Stiff-Person syndrome patients. J Immunol 2005; 175: 7755–7762.

Sapana

Tian

, et al. A case report of anti-GAD65 antibody-positive autoimmune encephalitis in children associated with autoimmune polyendocrine syndrome type-II and literature review. Front Immunol 2023; 14: 1274672.

Bach

Ananth

Thind

, et al. Co-occurrence of anti-GAD65 syndrome, type 1 diabetes mellitus, and focal seizures with impaired awareness. Cureus 2023; 15: e40611.

Fattahzadeh Ardalani

Samady Khanghah

Jahanpanah

, et al. Successfully treated anti-GAD limbic encephalitis in a 15-year-old diabetic boy with intravenous immunoglobulin: case report. Ann Med Surg (Lond) 2024; 86: 1173–1181.

Graus

Titulaer

Balu

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

Moura

Sambayeta

Sousa

, et al. Characterization of anti-GAD65-associated neurological syndromes: clinical features and antibody titers. NeuroSci 2024; 5: 201–208.

10.

Dalakas

MC.

Stiff-person syndrome and GAD antibody-spectrum disorders: GABAergic neuronal excitability, immunopathogenesis and update on antibody therapies. Neurotherapeutics 2022; 19: 832–847.

11.

Tsiortou

Alexopoulos

Dalakas

MC.

GAD antibody-spectrum disorders: progress in clinical phenotypes, immunopathogenesis and therapeutic interventions. Ther Adv Neurol Disord 2021; 14: 17562864211003486.

12.

Dalakas

Fujii

, et al. The clinical spectrum of anti-GAD antibody-positive patients with Stiff-Person syndrome. Neurology 2000; 55: 1531–1535.

13.

Fenalti

Buckle

AM.

Structural biology of the GAD autoantigen. Autoimmun Rev 2010; 9: 148–152.

14.

Kim

Namchuk

Bugawan

, et al. Higher autoantibody levels and recognition of a linear NH2-terminal epitope in the autoantigen GAD65, distinguish stiff-man syndrome from insulin-dependent diabetes mellitus. J Exp Med 1994; 180: 595–606.

15.

Burbelo

Groot

Dalakas

, et al. High definition profiling of autoantibodies to glutamic acid decarboxylases GAD65/GAD67 in stiff-person syndrome. Biochem Biophys Res Commun 2008; 366: 1–7.

16.

Fouka

Alexopoulos

Akrivou

, et al. GAD65 epitope mapping and search for novel autoantibodies in GAD-associated neurological disorders. J Neuroimmunol 2015; 281: 73–77.

17.

Manto

Laute

Aguera

, et al. Effects of anti-glutamic acid decarboxylase antibodies associated with neurological diseases. Ann Neurol 2007; 61: 544–551.

18.

Chang

Alexopoulos

McMenamin

, et al. Neuronal surface and glutamic acid decarboxylase autoantibodies in nonparaneoplastic Stiff Person syndrome. JAMA Neurol 2013; 70: 1140–1149.

19.

Dalakas

MC.

Progress and stiff challenges in understanding the role of GAD-antibodies in stiff-person syndrome. Exp Neurol 2013; 247: 303–307.

20.

Levy

Dalakas

Floeter

The Stiff-Person syndrome: an autoimmune disorder affecting neurotransmission of gamma-aminobutyric acid. Ann Intern Med 1999;131: 522–530.

21.

Levy

Levy-Reis

Fujii

, et al. Brain gamma-aminobutyric acid changes in stiff-person syndrome. Arch Neurol 2005; 62: 970–974.

22.

Sandbrink

Syed

Fujii

, et al. Motor cortex excitability in stiff-person syndrome. Brain 2000; 123: 2231–2239.

23.

Rakocevic

Raju

Semino-Mora

, et al. Stiff person syndrome with cerebellar disease and high-titer anti-GAD antibodies. Neurology 2006; 67: 1068–1070.

24.

Dalakas

MC.

Stiff-person syndrome and related disorders—diagnosis, mechanisms and therapies. Nat Rev Neurol 2024; 20: 587–601.

25.

Burton

Baquet

Eisenbarth

, et al. Central nervous system destruction mediated by glutamic acid decarboxylase-specific CD4⁺ T cells. J Immunol 2010; 184: 4863–4870.

26.

Bien

Vincent

Barnett

, et al. Immunopathology of autoantibody-associated encephalitides: clues for pathogenesis. Brain 2012; 135: 1622–1638.

27.

Tröscher

Mair

Verdú

Juan

, et al. Temporal lobe epilepsy with GAD antibodies: neurons killed by T cells not by complement membrane attack complex. Brain 2023; 146: 1436–1452.

28.

Dik

Widman

Schulte-Mecklenbeck

, et al. Impact of T cells on neurodegeneration in anti-GAD65 limbic encephalitis. Ann Clin Transl Neurol 2021; 8: 2289–2301.

29.

Muñoz-Lopetegi

de Bruijn

MAAM

Boukhrissi

, et al. Neurologic syndromes related to anti-GAD65: clinical and serologic response to treatment. Neurol Neuroimmunol Neuroinflamm 2020; 7: e696.

30.

Dalakas

MC.

Therapies in Stiff-Person syndrome: advances and future prospects based on disease pathophysiology. Neurol Neuroimmunol Neuroinflamm 2023; 10: e200109.

31.

Dalakas

Fujii

, et al. High-dose intravenous immune globulin for stiff-person syndrome. N Engl J Med 2001; 345: 1870–1876.

32.

Dalakas

MC.

Long-term effectiveness of IVIg maintenance therapy in 36 patients with GAD antibody-positive Stiff-Person syndrome. Neurol Neuroimmunol Neuroinflamm 2022; 9: e200011.

33.

Allison

Mechanisms of action of mycophenolate mofetil. Lupus 2005; 14: s2–s8.

34.

Dalakas

Rakocevic

Dambrosia

, et al. A double-blind, placebo-controlled study of rituximab in patients with stiff person syndrome. Ann Neurol 2017; 82: 271–277.

35.

Nipp

Woodward

Gross

, et al. Evolution of autoimmune retinopathy in Stiff Person syndrome: a case report. Retin Cases Brief Rep 2025; 19: 297–303.

36.

Liu

Zhao

Liu

, et al. The latest progress in the application of telitacicept in autoimmune diseases. Drug Des Devel Ther 2024; 18: 5811–5825.

37.

Dalakas

MC.

Promising effects of CAR T-cell therapy in refractory Stiff Person syndrome and a hopeful future for all neuroautoimmunities. Neurol Neuroimmunol Neuroinflamm 2026; 13: e200511.

38.

Vaisvilas

Cernauskiene

Petrosian

, et al. Successful fresh formulation CD19 CAR-T cell therapy for GAD65 antibody-mediated cerebellar ataxia. A case report. Front Immunol 2026; 17: 1755797.

39.

Chen

Kessi

Tan

, et al. Anti-GAD65 antibodies related refractory epilepsy successfully treated with tocilizumab: a case report and systematic literature review. Immunotargets Ther 2025; 14: 491–500.

40.

Budhram

Sechi

Flanagan

, et al. Clinical spectrum of high-titre GAD65 antibodies. J Neurol Neurosurg Psychiatry 2021; 92: 645–654.

41.

Piepgras

Müller

Steffen

, et al. Neurofilament light chain levels reflect outcome in a patient with glutamic acid decarboxylase 65 antibody-positive autoimmune encephalitis under immune checkpoint inhibitor therapy. Eur J Neurol 2021; 28: 1086–1089.

42.

Syk

Tornvind

Gallwitz

, et al. An exploratory study of the damage markers NfL, GFAP, and t-Tau, in cerebrospinal fluid and other findings from a patient cohort enriched for suspected autoimmune psychiatric disease. Transl Psychiatry 2024; 14: 304–314.

43.

Khalil

Teunissen

Lehmann

, et al. Neurofilaments as biomarkers in neurological disorders—towards clinical application. Nat Rev Neurol 2024; 20: 269–287.

44.

Falk

Cabrera

Junker

, et al. Serum NfL predicts outcome and secondary autoimmunity in herpes-simplex encephalitis. J Neuroimmunol 2025; 400: 578528.

45.

Tan

, et al. Coexistence of anti-GAD and anti-GABAAR antibodies in an autoimmune encephalitis patient: a case report. Int Med Case Rep J 2025; 18: 105–109.

46.

Gagnon

Savard

Mourabit Amari

Refractory status epilepticus and autoimmune encephalitis with GABA A R and GAD65 antibodies: a case report. Seizure 2016; 37: 25–27.

47.

Nakano

Chihara

Matoba

, et al. Immunotherapy-responsive non-paraneoplastic encephalitis with antibodies against GAD, LGI1, and GABAA receptor. Intern Med 2022; 61: 419–423.

48.

Zhou

Chen

Zhu

, et al. Clinical characteristics of autoimmune encephalitis with co-existence of multiple anti-neuronal antibodies. BMC Neurol 2024; 24: 1–13.

49.

Malter

Helmstaedter

Urbach

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

50.

Donadon

ALF

da Silva

de Mendonça

, et al. Atypical diffuse white matter changes in a patient with anti-GQ1B spectrum disorder: a case report. Radiol Case Rep 2025; 20: 2365–2367.

51.

Chataway

Larner

Kapoor

Anti-GQ1b antibody status, magnetic resonance imaging, and the nosology of Bickerstaff’s brainstem encephalitis. Eur J Neurol 2001; 8: 355–357.

52.

Deng

Cai

Qiu

, et al. MRI characteristics of autoimmune encephalitis with autoantibodies to GABAA receptor. Neurol Neuroimmunol Neuroinflamm 2022; 9: e1158.

53.

Spatola

Petit-Pedrol

Simabukuro

, et al. Investigations in GABAA receptor antibody-associated encephalitis. Neurology 2017; 88: 1012–1020.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB