Targeted Therapy of GRIA1-Related Epilepsy and Intellectual Disability With Perampanel

Abstract

Objectives

This report details a patient with a GRIA1 pathogenic variant presenting with intellectual disability (ID) and epilepsy. We describe clinical features, genetic findings, a personalized treatment approach, and a literature review of GRIA1-related disorders.

Methods

We describe clinical presentation, neuropsychological assessment, and genetic analysis. We conducted a literature review of published GRIA1-related disorders using PubMed, Simons Foundation Autism Research Initiative (SFARI) Gene, and ClinVar databases.

Results

An 8-year-old girl with ID, focal-to-bilateral tonic clonic seizure since age 5, and later atypical absences was diagnosed with a novel, de novo GRIA1 c.2530T > G, p.Leu844Val pathogenic variant. After genetic diagnosis, she was titrated to 4 mg of perampanel, an α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor antagonist, which led to seizure control and improvements in cognition and school performance. Literature review identified 31 patients carrying 15 different pathogenic variants. The c.1906G > A, p.Ala636Thr variant was recurrent in 17 individuals. Intellectual disability and autism spectrum disorder were common while epilepsy was reported in approximately a quarter of patients. Two patients with gain-of-function missense variants in GRIA1 and GRIA2, successfully treated with perampanel, have also been reported.

Discussion

This case emphasizes the role of targeted interventions in the management of rare genetic disorders and underscores the potential of precision medicine in addressing GRIA1-related symptoms.

Introduction

Many neurodevelopmental disorders (NDDs) have monogenic causes, particularly affecting proteins involved in excitatory neurotransmission such as the GRIA1–4 genes.^1,2 Over 100 pathogenic GRIA1–GRIA4 gene variants have been linked to NDDs, including epilepsy.² This report focuses on the GRIA1 gene, which encodes the GluA1 subunit of AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid) receptors.³ These receptors are vital for excitatory synaptic transmission and brain development.³

Several whole-exome sequencing (WES) and whole-genome sequencing studies in cohorts of individuals with NDDs and single case reports identified GRIA1 as a potential candidate for intellectual disability (ID).^4-10,e1

Perampanel (PER) is an AMPA receptor antagonist known for its antiseizure efficacy, approved as adjunctive therapy for focal seizures and generalized tonic-clonic seizures (TCSs) in the United States and European Union.¹¹ Perampanel improved outcomes in a patient with a recurrent GRIA1 gain-of-function (GOF) pathogenic variant (p.Ala636Thr), as well in a patient with GRIA2 GOF variant.^9,12

We present a novel de novo heterozygous missense GRIA1 variant in an 8-year-old girl with childhood epilepsy including focal and generalized seizures and mild ID. Treatment with PER led to seizure control and clear improvements across cognitive, academic, and neurologic domains.

Case Report

The girl was born at 35 weeks and birth was normal, although she presented with hypotonia and congenital left clubfoot. There was no family history for neurologic disorders. She began walking independently at 14 months. Over time, she displayed mild ID and learning difficulties during school years.

At age 5, she had a 20-minute seizure with loss of consciousness, facial expression changes, and vomiting. A postictal EEG revealed right frontocentral spikes. Two additional focal seizures followed over 2 months. At 5.6 years, a focal seizure progressing to bilateral tonic-clonic¹³ prompted initiation of carbamazepine. Four months later, an interictal EEG showed worsening bilateral epileptiform activity with very frequent bilateral discharges, leading to a switch to lamotrigine (LTG). She remained seizure free for over a year until a febrile seizure occurred. Repeated video-EEG revealed episodes of partial loss of consciousness coinciding with irregular spike-and-wave discharges classified as atypical absences¹³ (Figure 1).

Figure 1. Video-EEG Polygraphic Recording During Wakefulness.

(A) Bilateral, irregular, frontally predominant spike-polyspike-wave discharges. (B) These discharges are evoked by hyperventilation, during which the patient exhibits facial expression changes, partial loss of awareness, and inability to follow a given command.

Neuropsychological assessment at 6.9 years revealed impairments in fluid reasoning, processing speed, and sustained attention. On the Wechsler Preschool and Primary Scale of Intelligence, she obtained a Full Scale IQ of 64. Her Verbal Scale score was 86, Performance Scale 49, and Processing Speed 55, indicating a mild, disharmonic IQ profile. Language showed difficulties with consonant articulation, occasionally rendering speech unintelligible. Academic and gross motor skills were delayed, with impaired constructional praxis. Social-emotional immaturity and attentional deficits were noted, with clinically notable elevations (T score >64) on the Child Behavior Checklist (CBCL 6–18). She received specialized teacher support at school.

At age 7 years, a WES of the trio revealed a heterozygous, de novo, GRIA1 (NM_001114183.1) c.2530T > G, p.Leu844Val variant classified as likely pathogenic.^e2 Genetic testing confirmed maternity and paternity and did not reveal additional pathogenic variants. The GRIA1 variant results in the substitution of leucine (Leu) with valine (Val) at position 844 in the protein, is absent in the gnomAD reference population database (n > 120,000 exomes and >15,000 genomes), and is predicted as deleterious by all in silico tools used (POLYPHEN, SIFT, and MUTTASTER).

After genetic diagnosis, PER was titrated to 4 mg daily (body weight 32 kg) while LTG was discontinued. EEG revealed reduced epileptiform abnormalities. Six months from PER initiation, teachers and parents reported improved academic performance and attention, corroborated by a reduced score on the attention scale of the CBCL 6–18 (T score = 60). A cognitive reassessment at 8.3 years—14 months into PER treatment—demonstrated clear improvement in perceptual reasoning and processing speed. On the Wechsler Intelligence Scale for Children-IV, her Full Scale IQ was 69. The Verbal Comprehension Index was 92, Perceptual Reasoning Index was 69, Working Memory Index was 76, and Processing Speed Index was 65. Verbal reasoning fell within the average range. Pretreatment and post-treatment profiles are shown in Figure 2.

Figure 2. Test Results.

The y-axis displays standard scores (mean: 100, standard deviations: 15) for the Wechsler Preschool and Primary Scale of Intelligence (WPPSI-III), used at first assessment (child's age: 6 years 9 months), and the Wechsler Intelligence Scale for Children-IV (WISC-IV), used at retest (child's age: 8 years 3 months). These scales assess different age groups (WPPSI-III: from 2 years 6 months to 7 years 7 months; WISC-IV: from 6 years to 16 years 11 months). The WPPSI-III has fewer subtests and composite areas compared with the WISC-IV. Scores from the 2 tests cannot be directly compared because they are standardized on different age groups and use distinct normative samples.

Written informed consent for the publication of anonymous data was obtained from the patient and her parents. The study, based on retrospective analysis of standard clinical care data collection, anonymized for publication, and involving no experimental interventions, was conducted in accordance with international research ethics guidelines.

Literature Review

We performed a GRIA1 gene variant search using PubMed, SFARI Gene, and ClinVar databases. We identified 31 patients carrying 15 different pathogenic variants. Among these, the p.Ala636Thr variant is recurrent, being present in 17 of 31 probands (54.8%).^8-10,e3-e6 Heterozygous GRIA1 variants occurred de novo in 30 probands, and only 1 was inherited (Table).

Table.

Summary of Genetic and Clinical Data of All Patients Carrying GRIA1

No. of pts	Sex	cDNA variant	AA variant	Type of variant	Variant protein localization	Functional alteration	ID/ASD	ID level	Epilepsy	Reference
1	aNA	c.329C>T	p.Thr110Met	Het. Miss	NTD	Not studied	From ASD cohort	NA	NA	e7
1	aNA	c.603C>A	p.Asp201Glu	Het. Miss	LBD	Not studied	ASD	NA	NA	e8
1	M	c.623G>A	p.Arg208His	Het. Miss	LBD	Not studied	ASD	NA	NR	4
1	aNA	c.653G>A	p.Arg218his	Het. Miss	LBD	Not studied	From ASD cohort	NA	NR	e9
1	aNA	c.772G>A	p.Arg258Thr	Het. Miss	LBD	Not studied	From ASD cohort	NA	NA	e10
1	NA	c.1034G>A	p.Arg345Gln	Het. Miss	NTD	Studied, no alteration of functional properties	ID	NR	NR	8
1	F	c.1129C>T	p.Arg377Ter	Het. Miss	NTD	Destroys GluA1-containing AMPAR expression	ID	Severe	Onset 13 m; LTG, ESM	8
1	F	c.1526 G> A	p.Arg509Gln	Het. Miss	LBD	Not studied	ID, ASD	NA	NR	6
1	NA	c.1568 G> A	p.Gly523Glu	Het. Miss	LBD	Not studied	ID	NA	Yes	e6
1	M	c.1880T>C	p.Ile627Thr	Het. Miss	LBD	Loss-of-function	ID	NR	NR	8
1	NA	c.1906G>A	p.Ala636Thr	Het. Miss	LBD	Gain-of-Function	ID	NA	NR	e3
3	2F+1M	c.1906G>A	p.Ala636Thr	Het. Miss	LBD	Gain-of-Function	ID (3); ASD + ADHD (1)	Severe/moderate	NR	8
1	aNA	c.1906G>A	p.Ala636Thr	Het. Miss	LBD	Gain-of-Function	ID	Severe	NR	e4
1	aNA	c.1906G>A	p.Ala636Thr	Het. Miss	LBD	Gain-of-Function	ASD	Severe	NR	e5
5	4F+1M	c.1906G>A	p.Ala636Thr	Het. Miss	LBD	Gain-of-Function	ID (4); ASD (4); speech problems (3)	Mild/moderate	1 patient with sz at 2–5 y	10
1	NA	c.1906G>A	p.Ala636Thr	Het. Miss	LBD	Gain-of-Function	ID, ASD, hypotonia		NR	e6
5	3M+2F	c.1906G>A	p.Ala636Thr	Het. Miss	LBD	Gain-of-Function	ID; behavioral issues and ASD (5 pts); hyp (2)	2 severe, 2 mod.; 1 mild	3 pts with epilepsy, onset: 2, 6, and 10 y	9
1	aNA	c.1988C> T	p.Thr663Met	Het. Miss	LBD	Not studied	NA	NA	NR	e6
1	aNA	c.1994G>A	p.Gly665Asp	Het. Miss	LBD	Not studied	Bipolar disorder	NA	NR	7
1	F	c.2234G>A	p.Gly745Asp	Het. Miss	TD	Loss-of-function	ID, ASD	Severe	NR	8
1	NA	c.2645 G> C	p.Ser882Thr	Het. Miss	TD	Not studied	ID	NA	Yes	e6
1	F	c.2530T>G	p.Leu844Val	Het. Miss	TD	Not studied	ID	Mild	Yes	This study

Abbreviations: AA = amino acid; AMPARs = α-amino-3-hydroxy-5-methyl-4-isoxazole propionate receptors; ASD = autism spectrum disorder; ESM = ethosuximide; F = female; Hyp = hypotonia; Het. Miss = heterozygous missense; ID = intellectual disability; LBD = ligand-binding domain; LTG = lamotrigine; M = male; m = months; mod = moderate; NTD = N-terminal domain; NA = not available; NR = not reported; pts = patients; Sz = seizures; TD = transmembrane domain; yrs = years.

^aCohort study; °maternal.

All but one of the variants are heterozygous missense variants; the only exception is a homozygous nonsense variant, c.1129C > T; p.Arg377Ter, identified in a single patient (Table).

Functional characterization of 5 variants⁸ showed the following: the p.Arg345Gln variant has no significant effect on functional properties; p.Ala636Thr causes a GOF; the p.Ile627Thr and p.Gly745Asp variants lead to a loss of function, and the p.Arg377Ter variant completely disrupts GluA1-containing α-amino-3-hydroxy-5-methyl-4-isoxazole propionate receptor expression.

All patients exhibit ID, ranging from mild to severe; autism spectrum disorder is present in 16 probands (51.6%); and other reported features include attention-deficit/hyperactivity disorder, hypotonia, behavioral problems, and speech difficulties. Epilepsy has been reported only in 8 of 31 patients (25, 8%). A summary of genetic and clinical data, including this case, is provided in the Table.

Data Availability

Anonymized data not included in this article will be made available on request to qualified investigators.

Discussion

We report a child with a novel de novo GRIA1 c.2530T > G, p.Leu844Val missense variant, presenting with ID and childhood-onset epilepsy including both focal and generalized seizures not classifiable under a defined epilepsy syndrome. The relatively mild phenotype observed in our patient aligns with existing evidence suggesting that heterozygous GRIA1 variants result in less severe clinical manifestations. By contrast, homozygous nonsense GRIA1 variants lead to a complete loss of function and are associated with more severe outcomes, including early-onset seizures and profound speech and language delay.⁸ Epilepsy has been documented in approximately 25% of patients.^8-10,e6 However, given that the average age at seizure onset is around 6 years, this percentage may be underestimated because some individuals may develop epilepsy later in life.

Perampanel has demonstrated efficacy for both focal and generalized seizures in adults and children aged 4 years and older, in both clinical trials and real-world settings.¹⁴ Functional studies have demonstrated that PER can fully block GluA2 currents, indicating its potential therapeutic benefit in disorders involving GRIA1–4 variants.¹² This experimental evidence led to its translational application in patients with GRIA1 and GRIA2 pathogenic variants.^9,12 The patient with the GRIA2 variant began PER treatment at 27 months alongside a ketogenic diet.¹² Over a 7-month period, the patient experienced a significant reduction in seizure frequency and showed developmental and cognitive improvements. Similarly, a patient carrying the recurrent GRIA1 p.(Ala636Thr) variant achieved seizure freedom for 6 months after PER initiation.⁹

Our patient was also treated with PER, which led to a reduction in seizures and noticeable cognitive improvements. These clinical benefits suggest that PER effectively counteracts the presumed neuronal hyperexcitability associated with the GRIA1 p.Leu844Val variant. The positive treatment response further supports a GOF effect of the variant, mitigated by PER-AMPA receptor inhibition.

Our case resembles a previously reported patient carrying the GRIA1 GOF variant, also treated with PER and presenting with mild/moderate ID, language impairment, and childhood-onset epilepsy.⁹ Both patients have focal-to-bilateral TCSs, yet our patient later also manifested atypical absences.

As observed in other epilepsy/NDD genes, it can be hypothesized that seizures are the core feature of GOF variants, whereas loss of function (LOF) variants may be more closely linked to NDDs without epilepsy. However, current data do not support this theory because several patients with presumed GOF variants do not exhibit seizures while at least 1 patient with a nonsense (LOF) variant exhibits epilepsy⁸ (Table).

Notably, as with the GRIA2-treated case,¹² our patient also exhibited improved alertness, mood, and behavior after PER initiation—findings contrary to the more commonly reported side effects of PER, such as mood disturbances and aggression. This case, the third reported instance of a GRIA disorder treated with PER, underscores the therapeutic potential of targeted modulation of AMPA receptors in rare genetic epilepsy syndromes.

Future Perspectives

Although our case adds to growing evidence supporting PER as a precision treatment for GRIA1-related epilepsy, more studies are needed. Owing to the rarity of these conditions, randomized controlled trials are unlikely. Instead, N-of-1 trials may provide valuable insights into individual responses.¹⁵ As a selective AMPA receptor negative allosteric modulator, PER holds promise for treating GOF GRIA1 variants. This case supports its broader potential in precision medicine approaches for rare monogenic epilepsies.

Author Contributions

E. Cesaroni: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. C. Passamonti: analysis or interpretation of data. C. Marini: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data.

Study Funding

The authors report no targeted funding.

Disclosure

The authors report no relevant disclosures. Go to Neurology.org/NG for full disclosures.