Sustained rescue of seizure control in patients with highly refractory chronic epilepsy using empiric immunotherapy

Abstract

Objective

Following emerging evidence of autoimmune‐associated seizures in medication‐refractory epilepsy, we began offering a trial of immunotherapy to selected patients. Here, we review this approach's treatment response, predictive clinical features, and utility.

Methods

This was a retrospective single‐center cohort study (2018–2022) of empiric, palliative immunotherapy in 31 adults with highly refractory, highly active epilepsy. Since 2018, in line with the International League Against Epilepsy's addition of “immune” as an etiology in the Classification of Epilepsy, we initiated immunotherapy after comprehensive antiseizure medication failures while at the same time screening for an autoimmune origin. The workup included assessing clinical features, serum autoantibody testing, cerebrospinal fluid testing (where feasible), magnetic resonance imaging (MRI), and electroencephalography. All patients received intravenous methylprednisolone or IV immunoglobulin according to previously published protocols, and follow‐up was for at least 12 months.

Results

Nine patients (29%) in this highly refractory cohort demonstrated a sustained treatment response, measured as a greater than 50% improvement in seizure frequency for at least 12 months. Three patients (10%) became seizure‐free. Six patients (20%) were classified as partial responders and experienced an initial response that was not sustained. Apart from a trend toward a diagnosis of focal epilepsy, we did not identify any specific serological, clinical, electrodiagnostic, or imaging features with statistical significance that were predictive of treatment response.

Significance

This patient group demonstrated a reasonable response rate to an immunotherapy trial. These findings are surprising but support the consideration of an immunotherapy trial in patients with refractory epilepsy. Requirements for repeated courses of immunotherapy differed significantly between patients, and this is an area of interest for further research. The basis for response in this cohort remains unclear; in some cases, antiseizure medication changes may have contributed; however, without any apparent autoimmune features, we consider potential blood–brain barrier repair or a placebo effect as hypothetical alternative mechanisms of action for the response to immunotherapy.

Key points.

• Nine patients (29%) in our single‐center cohort with chronic medication‐refractory epilepsy who received a trial of immunotherapy had a sustained reduction in seizure frequency of more than 50% at the time of follow‐up (minimum 12 months, median 22 months), with three (10%) of them becoming completely seizure‐free.

• Apart from a trend toward a diagnosis of focal epilepsy, we did not identify any specific serological, clinical, electrodiagnostic, or imaging features with statistical significance that were predictive of treatment response.

• Our experience of the use of empiric immunotherapy for refractory epilepsy, demonstrated here, supports a more systematic prospective trial of immunotherapy for patients with chronic medication‐refractory epilepsy.

1. INTRODUCTION

Epilepsy is a common neurological disease affecting ~60 million people worldwide with significant morbidity and mortality, particularly in those with highly frequent refractory seizures. ¹ Approximately one‐third of patients with epilepsy fall into this category, with seizures that do not respond to conventional antiseizure medications. ² These patients experience a worse quality of life, more comorbidity, and an increased risk of sudden unexplained death in epilepsy, placing a substantial burden on the individuals, carers, and society. ³ Aiming for complete seizure freedom, or at least reducing seizure frequency, is an important goal of epilepsy management.

In recent years, it has come to light that certain types of epilepsy may be intrinsically immune driven, presumably related to the ictal potential of neuroinflammation, which has been widely demonstrated. ⁴ These autoimmune‐associated epilepsies were initially thought to be part of the spectrum of autoimmune encephalitides, although these patients show few classic symptoms of autoimmune encephalitis. ⁵ Such patients tend to present with subacute adult‐onset epilepsy with frequent seizures of focal semiology, which are typically refractory to antiseizure medications and often have additional neuropsychiatric symptoms. ⁶ In many patients, neuronal autoantibodies typically directed against synaptic proteins are found to support the diagnosis, but the exact role of these in ictogenesis and epileptogenesis remains to be elucidated. ⁷ , ⁸ Certainly, some “antibody‐negative” patients may respond as well to immunotherapy as those with a proven antibody association. ⁹

In January 2018, after initial positive responses to palliative treatment trials with immunotherapy for patients with highly refractory epilepsy, we developed an immunotherapy treatment pathway in our regional epilepsy service to systematically screen (for autoimmunity) and treat patients with highly refractory epilepsy with either intravenous methylprednisolone or intravenous immunoglobulin (IVIG). In this article, we share our experience developing this clinical pathway and report on patient outcomes. Finally, we highlight areas for further research to help refine treatment options and explore the basis for the observed responses.

2. METHODS

Between 2018 and 2022, following the inclusion of immune‐mediated epilepsies in the International League Against Epilepsy (ILAE) classification of epilepsy and reports of improvements in seizure control in other centers, we adopted a treatment pathway based on existing literature ¹⁰ for patients with highly refractory, highly active (>1 seizure per week for at least 6 months) epilepsy in whom all conventional non‐surgical approaches to seizure control had been employed with limited success. Most of these patients were on the waiting list for surgical evaluation or were not suitable for surgery, and all required a palliative approach to treatment. Using published diagnostic criteria for autoimmune epilepsy, ⁵ , ¹⁰ we developed an immunotherapy protocol utilizing intravenous methylprednisolone, IVIG, oral prednisolone, and steroid‐sparing agents in selected cases.

The diagnostic evaluation included serum and cerebrospinal fluid (CSF) autoantibodies, electroencephalography (EEG), and magnetic resonance imaging (MRI) of the brain. A screening form (Figure 1) was used to identify specific clinical features previously associated with autoimmune epilepsy in prior studies.

FIGURE 1.

Autoimmune epilepsy screening form. An example of our screening form, which was used to capture specific clinical features previously associated with autoimmune epilepsy in prior studies and help identify potential candidates for immunotherapy.

Treatment response was measured by seizure frequency reduction at the first review (8 weeks) and final review (at least 12 months). Clinical and diagnostic features were evaluated for their ability to suggest an autoimmune origin to the epilepsy and a positive response to immunotherapy.

We retrospectively reviewed the clinical, diagnostic features, and treatment responses of 31 such patients who attended our regional epilepsy service in St James's Hospital, Dublin, between January 2018 and August 2022. Demographic, clinical, and diagnostic features were collected, including autoimmune serum and CSF (23/31 patients) markers, MRI of the brain, and EEG. Follow‐up visits were at 8 weeks and 12 months, including patients treated up to July 2022. The screening form and treatment protocol are shown in Figures 1 and 2. Serum autoantibody testing included a vasculitis screen: antinuclear antibody (ANA), anti‐neutrophil cytoplasmic antibody (ANCA), complement, C‐reactive protein (CRP), and erythrocyte sedimentation rate (ESR); thyroid peroxidase antibody (TPO); a neuronal antibody panel, which included the surface‐antibodies: N‐methyl‐d‐aspartate (NMDA), leucine‐rich glioma‐inactivated 1 (LGI1), contactin‐associated protein‐like 2 (CASPR2), α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid (AMPA), glycine, γ‐aminobutyric acid receptor B (GABA B) and dipeptidyl‐peptidase–like protein‐6 (DPPX), as well as intracellular antibodies glutamic acid decarboxylase 65 (GAD65), Hu, Yo, Ri, amphiphysin, CV2/collapsin response mediator protein 5 (CRMP5), Ma‐2, and Zic4.

FIGURE 2.

Diagnostic and treatment protocol (adapted from reference). Shows our current simplified diagnostic and treatment protocol for patients undergoing immunotherapy, which is intended to be followed from top to bottom. MACS, medical ambulatory care suite; OPD, outpatient department.

Patients received immunotherapy following our treatment protocol as closely as possible. However, due to the complexity of the cases, the approach was individualized and adapted to clinical response, and changes in antiseizure medications were allowed when indicated. Response to immunotherapy was evaluated by comparing the seizure frequency at the time of data analysis. The median follow‐up was 22 months, and the seizure frequency was documented before the initiation of immunotherapy. Responders were considered those with a sustained reduction in seizure frequency of greater than 50%, whereas partial responders were those with an initial (2‐month) reduction in seizure frequency of greater than 50% that was not sustained.

This study was a retrospective analysis of existing clinical data, so ethics committee review and patient consent were not required. Data are summarized descriptively. Statistical analysis was conducted using SPSS Version 26.0 for Windows. Cross‐tabulation was used to compare differences between groups for categorical variables. Significant differences were tested using Pearson chi‐square analysis. For a two‐by‐two variable analysis, the Yates Correction of Continuity was used to compensate for the overestimation of the chi‐square analysis. The expected frequency in any cell should be 10 or more, and if this assumption was violated, Fisher's exact probability test was used.

3. RESULTS

Overall, we treated 31 patients with immunotherapy.

3.1. Clinical characteristics

The cohort's clinical and diagnostic characteristics are summarized in Table 1. Of the 31 patients, 22 (77%) were female. The median age was 31 years, and the range was from 17 to 60 years of age. The median duration of epilepsy before treatment was 11.4 years, with a range from 1 to 30 years. The median number of prior antiseizure medications was 5, with a range from 2 to 13. All patients had at least weekly seizures, with 18 (58%) having daily seizures before treatment with immunotherapy. Twenty‐six patients (84%) had a diagnosis of focal epilepsy, whereas only five (16%) had a diagnosis of generalized epilepsy. Of the 23 patients who underwent CSF testing, none were found to have inter‐cellular neuronal or surface antibodies in their CSF.

TABLE 1.

Summary of clinical and diagnostic data.

Characteristics	All (n = 31)	Responders (n = 9)	Non‐responders (n = 22)	p value
Sex
Female	22 (71%)	5 (56%)	17 (77%)	.217
Male	9 (29%)	4 (44%)	5 (23%)
Median age	31 (17–60)	26 (17–55)	33 (18–60)	.189
Median duration of epilepsy diagnosis in years	11.4 (1–30)	8.4 (1–25)	12.6 (2–30)	.19
Seizure frequency at time of enrolment
≥1/week but not daily	13 (42%)	2 (22%)	11 (50%)	.154
≥1/day	18 (58%)	7 (78%)	11 (50%)
Median number of prior ASMs	5.4 (2–13)	6 (2–13)	5.2 (2–12)	.491
Serum status
Negative	27 (87%)	8 (89%)	19 (89%)	.673
Positive	4 (13%)	1 (11%)	3 (14%)
CSF changes	4 (17%)	2 (8%)	2 (8%)	.148
MRI brain
Non‐lesional	22 (71%)	5 (56%)	17 (77%)	.217
Structural abnormality	9 (29%)	4 (44%)	5 (23%)
Screening questions
Encephalopathic features on EEG	4 (13%)	1 (11%)	3 (14%)	.673
Focal‐onset semiology	26 (84%)	9 (100%)	17 (77%)	.155
Generalized‐onset semiology	5 (16%)	0	5 (100%)
Adult‐onset epilepsy	11 (35%)	4 (44%)	7 (32%)	.394
Seizure semiology changing over time	9 (29%)	2 (22%)	7 (32%)	.817
Neuropsychiatric features	18 (58%)	6 (67%)	12 (55%)	.417
No early life risk factors for epilepsy	8 (26%)	2 (22%)	6 (27%)	.576
Onset with status	1 (3%)	0 (0%)	1 (4%)	.71
No EEG correlate for seizures	8 (26%)	1 (11%)	7 (32%)	.235
Other autoimmune conditions	7 (23%)	2 (22%)	5 (23%)	.681

Note: Outlines the clinical characteristics identified at screening among responders and non‐responders. The final column shows the statistical significance of these characteristics in relation to the outcome response.

Abbreviations: ASM, anti‐seizure‐medication; CSF, cerebrospinal fluid; MRI, magnet resonance tomography; EEG, Electroencephalogram

3.2. Autoantibody profiles

Neuronal autoantibodies were identified in only four patients (13%). One patient was positive for GAD65 autoantibodies and responded to immunotherapy but was found to be antibody negative on repeat testing. None of the other patients who responded to Immunotherapy had positive autoantibodies identified. Of the three non‐responders with positive autoantibodies, one patient had weak VGKC (Voltage gated Potassium Channel) positivity at the time of screening but also had cortical dysplasia in her frontal lobe and has since benefited from epilepsy surgery; we assume the VGKC was a non‐specific finding unrelated to her epilepsy. Two patients had autoimmune encephalitis at initial diagnosis many years prior, and both tested GAD positive on one occasion at the time of screening (titers of 245 and 719, respectively). Neither of the patients responded to further immunotherapy. We did not find a significant correlation between autoantibody status and response to immunotherapy in this patient cohort.

3.3. Neuroimaging

The majority (71%) of the patients in our cohort had no identified epileptogenic lesion on their brain MRI using the hospital's standard epilepsy protocol. Of the patients who responded to immunotherapy, one patient had a previous unsuccessful temporal lobectomy, one patient had Sturge–Weber syndrome and had an enlarged right choroid plexus on MRI; in keeping with the diagnosis, this patient also had right temporal lobe and hippocampal atrophy. The third patient had a developmental abnormality of cortical organization of the gray matter in the right hippocampus and occipital and parietal lobes. We did not find a significant correlation between MRI imaging and response to immunotherapy in our patient cohort.

3.4. Response to immunotherapy

Nine patients (29%) had a sustained reduction in seizure frequency of more than 50% at the time of follow‐up, with three (10%) of them becoming completely seizure‐free (Table 2). Six patients (20%) were considered partial responders, having an initial reduction in seizure frequency of more than 50% at the 2‐month follow‐up, which was not sustained even with further immunotherapy for the remaining follow‐up period (Table 2). None of the patients with generalized epilepsy (16%) responded to treatment. None of the other clinical characteristics captured by the screening questionnaire correlated significantly with response to immunotherapy.

TABLE 2.

Clinical details and treatment courses in the responder group.

Patient details	Epilepsy diagnosis	Baseline seizure frequency	Duration of epilepsy at enrolment	MRI brain	Serum autoantibodies	Immunotherapy start date	Immunotherapy	Duration of immunotherapy	Initial response at ~8 weeks	ASM start (mg)	ASM change	Final ASM (mg)	Final seizure frequency
22‐year‐old woman	Focal epilepsy	>15 seizures/day	2 years	Non‐lesional	Negative	Jan 18	Oral steroids 1 mg/kg	12 months	>70% reduction in seizures	LEV 1500 BD, PER 8 OD, ZNS 250 OD	CLB 10 TDS commenced PRN, LEV changed to BRV, PER weaned	BRV 150 BD, CLB 10 OD PRN, ZNS 250 OD	Seizure‐free
17‐year‐old boy	Focal epilepsy	>15 seizures/day	3 years	Sturge–Weber syndrome	Negative	Dec 18	Oral steroids 1 mg/kg	1 month	Seizure freedom	CLB 10 OD PRN, LCM 150 BD, VPA 700 BD	LCM and CLB ceased, LEV changed to BRV	BRV 100 BD	Seizure‐free
27‐year‐old woman	IGE and focal epilepsy	>15 seizures/day	2 years	Non‐lesional	Negative	Jan 19	1*IVMP 5*IVIG Mycophenolate	18 months	Seizure freedom	ESL 800 OD, LEV 1500 BD	LEV weaned off	ESL 800 OD	Seizure‐free
58‐year‐old man	Focal epilepsy	Multiple per week	25 years	Prior left temporal lobectomy	Negative	Mar 19	6*IVMP 18*IVIG Mycophenolate	28 months	3 weeks of seizure freedom	BRV 100 BD, ESL 800 OD, PB 15 30 45 TDS	BRV and ESL increased, CNB trialed, PB weaned off	BRV 150 BD, ESL 1000 OD	Monthly
23‐year‐old woman	Multifocal epilepsy	Multiple per week	8 years	Non‐lesional	Negative	Apr 18	1*IVMP 5*IVIG Mycophenolate	7 months	>70% reduction in seizures	BRV 100 BD, ESL 800 OD, PB 15 30 45 TDS	BRV and ESL reduced	BRV 75 BD, ESL 1600 OD, ZNS 500 OD	Monthly
21‐year‐old man	Focal epilepsy	3–4 per day	10 years	Cortical dysplasia	Negative	Aug 22	6*IVMP	6 months	>50% reduction in seizures	BRV 75 BD, CLB 10 BD, ESL 800 OD, PER 12 OD, VPA 700 BD	BRV and CLB weaned off, PER reduced	ESL 1600 OD, PER 8 OD, VPA 1200 OD	Less than weekly
24‐year‐old man	Focal epilepsy	Multiple per week	6 years	Non‐lesional	Negative	Aug 21	1*IVMP 22*IVIG	24 months	>50% reduction in seizures	CLB 10 PRN, ESL 1600 OD, VPA 700 BD, ZNS 500 OD	No Change	CLB 10 OD PRN, ESL 1600 OD, VPA 700 BD, ZNS 500 OD	Weekly
25‐year‐old woman	Focal epilepsy	>20 seizures/day	12 years	Non‐lesional	GAD positive, negative on repeat	Oct 18	1*IVMP 23*IVIG Mycophenolate	24 months	>70% reduction in seizures	BRV 150 BD, ESL 1800 OD	CNB commenced, LCM trialed	CNB 100 OD, ESL 1800	Less than 10 daily
40‐year‐old woman	Focal epilepsy	Daily	0.5 years	Lesional	Negative	Nov 21	1*IVMP 10*IVIG	12 months	>70% reduction in seizures	BRV 300 OD, CLB 10 OD, ESL 1200 OD	No Change	BRV 300 OD, CLB 10 OD, ESL 1200 OD	Bi‐weekly

Note: This table shows the clinical details, treatment course, antiseizure medications, and magnitude of response among the responder group.

Abbreviations: BRV, brivaracetam; CBD, cannabidiol; CLB, clobazam; CNB, cenobamate; CZP, clonazepam; ESL, eslicarbazepine; LAM, lamotrigine; LCS, lacosamide; LEV, levetiracetam; PB, phenobarbital; PER, perampanel; PHT, phenytoin; RFM, rufinamide; VPA, sodium valproate; ZNS, zonisamide.

Table 2 lists the details of treatment in the responder patients. The frequency and type of immunotherapy required to respond varied between patients. Two patients became seizure‐free with just one course of oral prednisolone, 1 mg/kg, tapered over 6 weeks, whereas a third patient who became seizure‐free during the follow‐up required repeated courses of steroids and IVIG with the eventual introduction of mycophenolate. Those patients who responded to the initial immunotherapy trial were given continued and escalated immunotherapy, including IVIG and mycophenolate, as required. If no sustained response was demonstrated with escalating immunotherapy, the treatment was ceased (partial responders).

Table 3 lists the details of treatment in the non‐responder group.

TABLE 3.

Clinical details and treatment courses in the non‐responder group.

Patient details	Epilepsy diagnosis	Baseline seizure frequency	Duration of epilepsy at enrolment	MRI brain	Serum autoantibodies	Immunotherapy	Duration of immunotherapy	Initial response at ~8 weeks	ASM start (mg)	ASM change	Final ASM (mg)
29‐year‐old woman	Generalized genetic epilepsy	Multiple daily	24 years	Normal	Negative	Oral Steroids 1 mg/kg 1*IVIG	2 months	No change	LAM 300 BD, LEV 2000 BD, PER 12 OD, CLB 5 TDS PRN	LEV reduced	LAM 300 BD, LEV 1500 BD, PER 12 OD, CLB 5 TDS PRN
27‐year‐old woman	Multifocal	Multiple daily	10 years	Normal	Negative	1*IVMP	1 month	No change	BRV 100 BD, ZNS 500 OD	BRV changed to LCM	LCM 200 BD, ZNS 500 OD
43‐year‐old man	Focal epilepsy	Weekly	15 years	Postsurgical change right temporal lobe	Negative	6*IVIG	6 months	No change	ESL 2000 OD, LCM 250 BD, PHT 325 OD, CLB 15 OD PRN	LCM changed to LAM	ESL 2000 OD, LAM 20 BD, PHT 325 OD, CLB 15 OD PRN
34‐year‐old woman	IGE	Daily	16 years	Normal	Negative	1*IVMP 4* IVIG	5 months	No change	VAL 600 BD, LAM 100 BD, LCM 200 BD	VAL ceased, LAM increased, PB and CLB commenced	CLB 5 TDS, LCM 200 BD LAM 250 BD, PB 30 TDS
27‐year‐old woman	Focal epilepsy	Daily	4 years	Normal	Negative	1*IVMP 4* IVIG	5 months	No change	BRV 100 BD, ESL 1200 OD, CLB 10 OD PRN	BRV increased, ESL ceased	BRV 150 BD, CLB 10 OD PRN
24‐year‐old woman	Focal epilepsy left frontal	Multiple daily	8 years	Cortical dysplasia left frontal lobe	Negative	Oral steroids 1 mg/kg	1 month	No change	BRV 125 BD, ESL 2000 OD, PHT 200 OD	BRV, ESL and PHT increased	BRV 150 BD, ESL 2400 OD, PHT 225 OD
22‐year‐old woman	IGE	Multiple daily	8 years	Normal	Negative	1*IVMP	1 month	No change	BRV 150 BD, CLB 20 OD, LAM 225 BD, ZNS 300 OD	ZNS changed to VPA, CLB changed to PRN	BRV 150 BD, LAM 225 BD, VPA 1000 BD, CLB 10 OD PRN
30‐year‐old man	Focal epilepsy, right temporal	Daily	18 years	Focal cortical dysplasia, right temporal lobe	Negative	1*IVMP	1 month	No change	LCM 200 BD, LAM 150 BD, CLB 5 BD, VPA 500 BD	No change	LCM 200 BD, LAM 150 BD, CLB 5 BD, VPA 500 BD
55‐year‐old woman	TLE	Weekly	7 years	Normal	Negative	1*IVMP 1* IVIG	2 month	No change	ESL 400 OD	ESL ceased	Nil
32‐year‐old woman	Focal epilepsy	Multiple daily	12 years	Normal	Negative	1*IVMP	1 month	No change	ESL 1200 OD, LEV 2500 BD, ZNS 500 OD, CZP 0.5 OD PRN	No change	ESL 1200 OD, LEV 2500 BD, ZNS 500 OD, CZP 0.5 OD PRN
49‐year‐old woman	TLE	Daily	10 years	Normal	Negative	1*IVMP	1 month	No change	BRV 100 BD, LCM 150 BD, ZNS 300 OD	ZNS changed to LCM	BRV 100 BD, LCM 150 BD, LAM 200 BD
45‐year‐old man	TLE	Daily	22 years	Right mesial temporal sclerosis	Negative	1*IVMP	1 month	No change	ESL 1600 OD, PER 6 OD, LAM 300 BD, LEV 750 BD	PER and LEV increased	ESL 1600 OD, PER 8 OD, LAM 300 BD, LEV 1500 BD
28‐year‐old man	IGE	Daily	19 years	Normal	Negative	1*IVMP 1* IVIG	2 month	No change	LEV 1500 BD, VPA 1500 BD, ESM 500 BD, CLB 10 OD PRN	No Change	LEV 1500 BD, VPA 1500 BD, ESM 500 BD, CLB 10 OD PRN
44‐year‐old woman	IGE and focal frontal lobe epilepsy	Multiple daily	31 years	Cortical dysplasia left frontal lobe	VGKC‐AB weak pos. in serum	1*IVMP 2*IVIG	3 months	No change	CBZ 400 BD, LEV 1250 BD, CZP 0.5 OD	CBZ increased, LEV changed to BRV, VPA commenced	CBZ 600 BD, BRV 150 BD, VPA 800 BD, CZP 0.5 OD
43‐year‐old woman	TLE	Daily	2 years	B/L mesial temporal sclerosis	GAD +ve (42)	1*IVMP. 4* IVIG	5 months	>50% reduction in seizures	ESL 1200 OD, LCM 200 BD, LEV 1750 BD, CLB 10 BD, PER 8 OD	ESL and PER ceased, CNB commenced	CNB 200 OD, LCM 200 BD, LEV 1750 BD, CLB 10 BD
57‐year‐old woman	TLE	Weekly	12 years	Normal	GAD +ve(721)	1*IVMP	1 month	No change	LAM 250 BD, BRV 150 BD, ZNS 300 OD	No Change	LAM 250 BD, BRV 150 BD, ZNS 300 OD
28‐year‐old woman	Focal epilepsy	Multiple weekly	9 years	Normal	Negative	2*IVMP 2*IVIG	6 months	No change	LCM 100 BD, LAM 250 BD	LCM changed to CBZ, PER trialed, PB commenced	CBZ 400 BD, LAM 250 BD, PER 4 OD, PB 15 TDS
34‐year‐old woman	JME	Weekly	15 years	Normal	Negative	Oral Steroids 1 mg/kg	1 month	>70% reduction in seizures	CLB 10 BD, LCM 200 BD	No Change	CLB 10 BD, LCM 200 BD
61‐year‐old woman	Focal epilepsy	Weekly	4 years	Normal	Negative	1*IVMP	1 month	No change	CLB 5 TDS PRN	ESL commenced	ESL 1200 OD, CLB 5 TDS PRN
36‐year‐old woman	Focal epilepsy	Weekly	7 years	Normal	Negative	7*IVIG Mycophenolate	9 months	No change	ESL 2000 OD, ZNS 200 BD	ESL reduced	ESL 1200 OD, ZNS 200 BD
30‐year‐old woman	Focal epilepsy, right temporal	Multiple daily	10 years	Normal	Negative	1*IVMP 2*IVIG	3 months	No change	CBZ 400 BD, LAM 50 BD	LAM changed to ZNS, CBZ reduced	CBZ 400 mane, 200 nocte, ZNS 500 OD
19‐year‐old man	Focal epilepsy, left temporal	Multiple daily	8 years	Postsurgical change left temporal lobe	Negative	1*IVMP 4*IVIG	5 months	>50% reduction in seizures	PHT 200 BD, LCM 200 BD, RFM 600 BD, CBD 3 mL BD	PER, VPA and CLB PRN commenced, RFM ceased	LCM 200 BD, PER 10 OD, VPA 400 TDS, PHT 400 OD, CLB 10 BD PRN, CBD 4 mL BD

Note: This table shows the clinical details, treatment course, antiseizure medications, and magnitude of response in the non‐responder group.

Abbreviations: BRV, brivaracetam; CBD, cannabidiol; CLB, clobazam; CNB, cenobamate; CZP, clonazepam; ESL, eslicarbazepine; LAM, lamotrigine; LCS, lacosamide; LEV, levetiracetam; PB, phenobarbital; PER, perampanel; PHT, phenytoin; RFM, rufinamide; VPA, sodium valproate; ZNS, zonisamide.

Initially, we treated patients preferably with oral prednisolone, as this was possible in an outpatient setting, but due to the side effects of the treatment, we subsequently used intravenous methylprednisolone.

3.5. Adverse events

Eleven patients (35%) reported side effects; most were mild (see Table S1). Adverse events prompted cessation of immunotherapy in two cases (6%), both with steroid‐induced psychosis. One adverse event (3%) of IVIG‐associated aseptic meningitis prompted immunotherapy cessation.

4. DISCUSSION

In 2018, our epilepsy service, guided by the emerging concept of autoimmune epilepsy and reports of therapeutic responses to immunotherapy, started treating patients with highly refractory chronic epilepsy with intravenous steroids and IVIG as a palliative treatment trial. Here, we discuss the observed outcomes of these patients.

The most striking finding is that more than a quarter of our cohort of patients with highly refractory epilepsy had a significant reduction in seizure frequency of over 50% after immunotherapy, with three patients achieving complete seizure freedom and a relatively low side‐effect profile. We were surprised by these positive responses, considering that all the patients in the cohort had epilepsy for many years and had failed multiple conventional antiseizure medications. The duration and type of immunotherapy required to achieve improved seizure control varied considerably between patients. A positive response to an immunotherapy trial, as described in this study, supports an immune‐mediated etiology and the consideration of long‐term maintenance immunotherapy. Although some patients became seizure‐free with a single course of steroids, others required continuous IVMP (Intravenous Methylprednisolone) or IVIG and occasional addition of mycophenolate, suggesting either varied immune processes contributing to epileptogenesis or other undetermined effects of the treatment.

With this review, we aimed to understand the clinical profile of patients with medication‐refractory epilepsy who respond to immunotherapy. None of the clinical features we recorded differed significantly between responders and non‐responders, and we could not discriminate any clinical features that could help predict treatment response apart from a diagnosis of focal‐onset epilepsy in all responders. This may be due to the small size of our cohort and the less well‐defined phenotype for autoimmunity in epilepsy at the outset of the study period. ¹¹

Furthermore, none of the diagnostic investigations listed correlated with response to treatment, particularly antibody positivity, in our patient group. Positive antibodies have been reported in ~11% of unselected epilepsy patients, ¹² which is similar to the percentage of patients with positive antibodies in our patient group.

Although a positive response to immunotherapy supports an immune‐mediated etiology, ⁴ immunotherapy has also been used successfully in treating intractable epilepsies not proven to be autoimmune, as is the case for infantile spasms. ¹³ The placebo effect may contribute to our findings, given that this is an observation study. However, the magnitude and duration of the effect lead us to consider alternatives such as blood–brain barrier function. Increasing evidence shows that disruption of the blood–brain barrier occurs in patients with refractory seizures, possibly causing local neuroinflammation and peripheral immune cell influx, leading to seizure recurrence. ¹⁴ , ¹⁵ , ¹⁶ The degree of blood–brain barrier disruption strongly correlates with seizure severity. ¹⁷ We hypothesize that the immunotherapy trial in our highly refractory autoantibody‐negative patients may have stabilized the blood–brain barrier in some patients, thereby reducing seizures. Steroids have been described to have a stabilizing structural effect on the blood–brain barrier. ¹⁸ As such, we speculate for at least some patients treated with steroids that improving the structural integrity of the blood–brain barrier may have led to the improvement in seizure control rather than relying on the actual immunomodulatory properties of the steroids. This would explain why some of the patients responded so dramatically to just one or two courses of steroids while others required maintenance immunotherapy to control their seizures.

5. LIMITATIONS

This study is limited by small numbers and retrospective design. Immunotherapy provided to patients was homogenous; although we tried to adhere to the clinical care pathway described, the treatment of patients was individualized due to the severity and complexity of each patient's epilepsy. The seizure quantification methodology available was not ideal despite using a standardized screening form. There may be inter‐clinician variability in assessing patients and documenting seizure frequency. In addition, reporting of seizure frequency is based on patient accounts and, as such, subjective. Due to this, we aimed to be conservative in rating patients as responders. A further likely confounder was the medical need for antiseizure medication changes occurring during the trial period; while attempts were made to lessen this, it could not be controlled, and in some cases, it may have been the primary driver for improvement.

6. CONCLUSIONS

Our experience of the use of empiric immunotherapy for refractory epilepsy, demonstrated here, supports a more systematic prospective trial of immunotherapy for patients with chronic medication‐refractory epilepsy prior to more invasive treatments, such as surgery.

Different guidelines have been proposed to identify and classify central nervous system autoimmunity. Although helpful, these guidelines fail to predict treatment response. In cases of suspected autoimmune epilepsy, response to an immunotherapy trial remains a powerful predictor of outcome, and we suggest these patients should be monitored closely in a structured treatment program. In patients with highly refractory chronic epilepsy, given the relatively few side effects associated with pulse steroid therapy compared to the adverse effects of uncontrolled seizures and the high burden of conventional antiseizure medication side effects, a treatment trial with steroids and, if needed, long‐term immunotherapy may be considered ideally in the setting of a clinical trial. There is a clear need for such prospective clinical trials to better understand the effects of immunotherapy in patients with refractory chronic epilepsy.

AUTHOR CONTRIBUTIONS

Elisabeth Doran, Albert Kelly, and Raluca Stanila collected and analyzed the data and wrote the first draft of the work. Albert Kelly wrote the second and third drafts of the work. Laura Healy completed the statistics for the descriptive analysis of the data. Colin Doherty oversaw the whole project and edited the final draft of the work.

FUNDING INFORMATION

None.

CONFLICT OF INTEREST STATEMENT

No conflict of interest. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Supporting information

Table S1.

Doran E, Kelly A, Stanila R, Healy L, Doherty CP. Sustained rescue of seizure control in patients with highly refractory chronic epilepsy using empiric immunotherapy. Epilepsia. 2025;66:2743–2753. 10.1111/epi.18417

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.