Glucose Transporter Type 1 Deficiency Syndrome

Clinical characteristics.

The phenotypic spectrum of glucose transporter type 1 deficiency syndrome (Glut1 DS) is now known to be a continuum that includes the classic phenotype as well as paroxysmal exercise-induced dyskinesia and epilepsy (previously known as dystonia 18 [DYT18]) and paroxysmal choreoathetosis with spasticity (previously known as dystonia 9 [DYT9]), atypical childhood absence epilepsy, myoclonic astatic epilepsy, and paroxysmal non-epileptic findings including intermittent ataxia, choreoathetosis, dystonia, and alternating hemiplegia. The classic phenotype is characterized by infantile-onset seizures, delayed neurologic development, acquired microcephaly, and complex movement disorders. Seizures in classic early-onset Glut1 DS begin before age six months. Several seizure types occur: generalized tonic or clonic, focal, myoclonic, atypical absence, atonic, and unclassified. In some infants, apneic episodes and abnormal episodic eye-head movements similar to opsoclonus may precede the onset of seizures. The frequency, severity, and type of seizures vary among affected individuals and are not related to disease severity. Cognitive impairment, ranging from learning disabilities to severe intellectual disability, is typical. The complex movement disorder, characterized by ataxia, dystonia, and chorea, may occur in any combination and may be continuous, paroxysmal, or continual with fluctuations in severity influenced by environmental factors such as fasting or with infectious stress. Symptoms often improve substantially when a ketogenic diet is started.

Diagnosis/testing.

The diagnosis of Glut1 DS is established in a proband with suggestive clinical findings, normal blood glucose concentration, CSF glucose concentration <60 mg/dL, and the identification of a heterozygous pathogenic variant (or rarely, biallelic pathogenic variants) in SLC2A1 by molecular genetic testing. If no pathogenic variant is identified, 3-O-methyl-D-glucose uptake in erythrocytes can be performed; results between 35% and 74% of controls are diagnostic.

Management.

Treatment of manifestations: The ketogenic diet is highly effective in controlling the seizures and improving gait disturbance and is generally well tolerated. Affected individuals treated effectively at a younger age have a better outcome The ketogenic diet is deficient in L-carnitine necessitating dietary supplementation.

Prevention of primary manifestations: Early diagnosis and treatment with a ketogenic diet is associated with improved neurologic outcome.

Prevention of secondary complications: For those on a ketogenic diet: L-carnitine supplementation; proper hydration and avoidance of carbonic anhydrase inhibitors; avoidance of carbohydrate-containing foods, intravenous fluids, and medications that interrupt the state of ketosis; avoidance of valproic acid because it increases the risk of a Reye-like illness.

Surveillance: Periodic measurement of blood ketone concentration with a target beta-hydroxybutyrate concentration of 3-5 mmol/L.

Agents to avoid: Barbiturates (e.g., phenobarbital, the antiepileptic drug most commonly used in treating infants), methylxanthines (e.g., caffeine), valproic acid.

Evaluation of relatives at risk: If the pathogenic variant has been identified in an affected family member, it is appropriate to test at-risk newborns, infants, and other relatives at risk in order to identify as early as possible those who would benefit from initiation of treatment and preventive measures.

Genetic counseling.

Glut1 DS is most commonly inherited in an autosomal dominant (AD) manner. About 90% of individuals with AD Glut1 DS have the disorder as the result of a de novo heterozygous pathogenic variant; about 10% have a clinically affected parent. Parents who are heterozygous for the pathogenic variant may have a mild phenotype or be asymptomatic, findings that can suggest mosaicism in the parent. Offspring of an individual with AD Glut1 DS have a 50% chance of inheriting the pathogenic variant and being clinically affected.

Rarely, Glut1 DS is inherited in an autosomal recessive (AR) manner. Heterozygotes (carriers) in families with AR Glut1 DS are asymptomatic.

Once the SLC2A1 pathogenic variant(s) have been identified in an affected family member, prenatal diagnosis for a pregnancy at increased risk and preimplantation genetic diagnosis for Glut1 DS are possible.

Suggestive Findings

Glucose transporter type 1 deficiency syndrome (Glut1 DS) should be suspected in individuals with one of the following two phenotypes.

Classic Glut1 DS (~90% of affected individuals)

• Seizures with onset:
  • Typically before age two years (most frequently between ages 1 and 6 months)
  • Less often after age two years
• Delayed neurologic development
• Dysarthria
• Acquired microcephaly
• Complex movement disorders including:
  • Ataxia
  • Dystonia
  • Chorea

Non-classic Glut1 DS (~10% of affected individuals)

• No clinical seizures
• Milder phenotype
• Frequent paroxysmal dyskinesias including:
  • Intermittent ataxia
  • Choreoathetosis
  • Dystonia
  • Alternating hemiplegia

Testing

CSF glucose concentration. The single most important laboratory observation in Glut1 DS is hypoglycorrhachia (reduced cerebrospinal fluid [CSF] glucose concentration). Following a four-hour fast, a blood sample is obtained just before performing the lumbar puncture:

• The blood glucose concentration should be normal, ruling out hypoglycemia as the cause of the hypoglycorrhachia.
• The CSF/blood glucose ratio id usually less than 0.4 (range 0.19-0.59) in persons with Glut1 DS; however, this value is less reliable than the absolute CSF glucose value.
• All affected individuals reported to date have had CSF glucose values below 60 mg/dL (range: 16.2-52 mg/dL); in more than 90% it is below 40 mg/dL and in approximately 10% it is 41-52 mg/dL [De Vivo & Wang 2008, Yang et al 2011, Leen et al 2013].

Establishing the Diagnosis

The diagnosis of Glut1 DS is established in a proband with suggestive clinical findings and one or both of the following:

• Normal blood glucose concentration with CSF glucose concentration <60 mg/dL
• Identification of a heterozygous pathogenic variant (or rarely, biallelic pathogenic variants) in SLC2A1 by molecular genetic testing (see Table 1)

Molecular genetic testing approaches can include single-gene testing, use of a multigene panel, and more comprehensive genomic testing:

• Single-gene testing. Sequence analysis of SLC2A1 is performed first and followed by gene-targeted deletion/duplication analysis if no pathogenic variant is found.
• A multigene panel that includes SLC2A1 and other genes of interest (see Differential Diagnosis) may be considered. Note: (1) The genes included in the panel and the diagnostic sensitivity of the testing used for each gene vary by laboratory and are likely to change over time. (2) Some multigene panels may include genes not associated with the condition discussed in this GeneReview; thus, clinicians need to determine which multigene panel is most likely to identify the genetic cause of the condition at the most reasonable cost while limiting identification of variants of uncertain significance and pathogenic variants in genes that do not explain the underlying phenotype. (3) In some laboratories, panel options may include a custom laboratory-designed panel and/or custom phenotype-focused exome analysis that includes genes specified by the clinician. (4) Methods used in a panel may include sequence analysis, deletion/duplication analysis, and/or other non-sequencing-based tests.
  For an introduction to multigene panels click here. More detailed information for clinicians ordering genetic tests can be found here.
• More comprehensive genomic testing (when available) including exome sequencing and genome sequencing may be considered. Such testing may provide or suggest a diagnosis not previously considered (e.g., mutation of a different gene or genes that results in a similar clinical presentation).
  For an introduction to comprehensive genomic testing click here. More detailed information for clinicians ordering genomic testing can be found here.

If molecular genetic testing fails to detect a pathogenic SLC2A1 variant, consider performing erythrocyte glucose uptake assay (when feasible). An abnormally low assay implies an abnormality in SLC2A1.

Erythrocyte 3-O-methyl-D-glucose uptake assay. The uptake assay is a functional measure of glucose transport across the cell membrane. Individuals with Glut1 DS have abnormal values that range from 35% to 74% of controls, with an average reduction of approximately 50% [Yang et al 2011]. As such, it is abnormally low in almost all suspected cases.

• Decreased 3-O-methyl-D-glucose uptake in erythrocytes confirms the diagnosis of SLC2A1-related Glut1 DS.
• Molecular genetic testing detects a pathogenic variant in 95% of persons with abnormally low uptake assay.
• Of note, 3% of persons with Glut1 DS have a normal uptake assay, a finding that correlates with the presence of an SLC2A1 pathogenic missense variant (see Genotype-Phenotype Correlations).
  Note: The 3-O-methyl-D-glucose uptake assay is currently viewed as the diagnostic gold standard for this disease [Yang et al 2011]; however, such testing may not be available on a clinical basis, so it is primarily used if molecular genetic testing fails to detect a pathogenic SLC2A1 variant.

Clinical Description

Glucose transporter type 1 deficiency syndrome (Glut1 DS) usually presents as either classic Glut1 DS (~90% of affected individuals) or, more rarely, non-classic Glut1 DS (~10% of affected individuals), which comprises a broad phenotypic spectrum.

• Infants with the classic phenotype appear normal at birth following an uneventful pregnancy and delivery. Birth weight and Apgar scores are normal. Affected individuals commonly experience infantile-onset epileptic encephalopathy refractory to anticonvulsants and associated with delayed neurologic development; later deceleration of head growth and acquired microcephaly; and ataxia, dystonia, and spasticity [Klepper et al 1999b, De Vivo et al 2002a, De Vivo et al 2002b, Pons et al 2010, Yang et al 2011].
• The phenotypic spectrum designated as non-classic Glut1 DS has expanded over the past few years as more affected individuals have been identified. Paroxysmal non-epileptic manifestations that have been reported include intermittent ataxia, choreoathetosis, dystonia, and alternating hemiplegia. Several disorders including paroxysmal choreoathetosis with spasticity (dystonia 9), paroxysmal exercise-induced dyskinesia and epilepsy (dystonia 18), atypical childhood absence epilepsy, and myoclonic astatic epilepsy are now known to be caused by Glut1 deficiency [Chinnery 2010, Leen et al 2010, Yang et al 2011]. Some findings may show overlap with those seen in classic Glut1 DS.

Seizures. Seizures in classic early-onset Glut1 DS, which usually begin between age one and six months, are often the first clinical indication of brain dysfunction. In some infants, apneic episodes and abnormal episodic eye-head movements similar to opsoclonus may precede the onset of seizures [Pearson et al 2017]. Infantile focal seizures are clinically fragmented (i.e., non-generalized; typical at this age) and may include paroxysmal eye-head movements, cyanotic spells, and complex absence and atonic seizures. The electroencephalogram (EEG) may demonstrate multifocal spike discharges in infancy.

With further brain maturation, the seizures become synchronized and manifest clinically as generalized events associated with 3- to 4-Hz spike and wave discharges. Several seizure types have been described: generalized tonic or clonic, focal, myoclonic, atypical absence, atonic, and unclassified [Leary et al 2003].

The frequency of seizures varies among individuals: some experience daily events; others have only occasional seizures separated by days, weeks, or months. Seizure frequency does not correlate with phenotypic severity.

Some individuals with Glut1 DS never have a clinical seizure [von Moers et al 2002, Leary et al 2003]. About 10%-15% of cases, diagnosed thus far, never have had clinical seizures [Leen et al 2010, Pong et al 2012, Pearson et al 2013].

Speech and language impairment. Varying degrees of speech and language impairment are observed in all affected individuals.

Dysarthria is common and is accompanied by dysfluency (i.e., excessively interrupted speech).

Both receptive and expressive language skills are affected, with expressive language skills being disproportionately affected.

Intellectual disability. Varying degrees of cognitive impairment, ranging from learning disabilities to severe intellectual disability, are observed. Minimally affected individuals have estimated IQ scores in the normal range.

Social adaptive behavior is an exceptional strength. Individuals with Glut1 DS tend to be comfortable in group and school settings and interact well with others. Autistic spectrum disorders appear to be underrepresented in those with Glut1 DS.

Movement disorders. A complex movement disorder is commonly seen and is characterized by ataxia, dystonia, and chorea that may be continuous, paroxysmal, or continual with fluctuations determined by environmental stressors [Leen et al 2010, Pons et al 2010, Pearson et al 2013, Alter et al 2015]. Often, paroxysmal worsening occurs before meals, during fasting, or with infectious stress.

Pons et al [2010] described the frequency of abnormal movements in 57 persons with Glut1 DS. Clinical findings included the following:

• Gait disturbance (89%), the most frequent being ataxia and spasticity together or ataxia alone
• Action limb dystonia (86%)
• Mild chorea (75%)
• Cerebellar action tremor (70%)
• Non-epileptic paroxysmal events (28%)
• Dyspraxia (21%)
• Myoclonus (16%)

The 40 individuals on a ketogenic diet had less severe gait disturbances, but more complex movement disorders than those on a conventional diet [Pons et al 2010], an observation suggesting that the extrapyramidal and cerebellar findings are more apparent in the milder phenotypes.

Paroxysmal movement disorders. Paroxysmal exercise-induced dyskinesia and epilepsy (previously known as dystonia 18 [DYT18] [Suls et al 2008, Weber et al 2008, Zorzi et al 2008, Urbizu et al 2010]) and paroxysmal choreoathetosis with spasticity (previously known as dystonia 9 [DYT9] [Weber et al 2011]) are now recognized to be part of the phenotypic spectrum of Glut1 DS.

• Paroxysmal exercise-induced dyskinesia and epilepsy differs clinically from classic Glut1 DS in that most affected individuals appear to have a normal interictal neurologic examination and a normal head circumference, and experience exercise-induced dyskinesias and later-onset seizures [Suls et al 2008, Weber et al 2008, Zorzi et al 2008, Urbizu et al 2010]. The CSF glucose concentrations tend to be higher (41-52 mg/dL) than those in classic Glut1 DS [De Vivo & Wang 2008].
• The two families with paroxysmal choreoathetosis with spasticity had paroxysmal, mainly exercise-induced dyskinesia with onset between ages one and 15 years [Weber et al 2011] caused by heterozygous SLC2A1 pathogenic variants (p.Arg212Cys and p.Arg126Cys). Dyskinesia triggers included prolonged exercise, anxiety, and emotional stress. The dyskinesias decreased in frequency or stopped later in life. Other associated findings included progressive spastic paraparesis with onset in early adulthood, mild gait ataxia, mild-to-moderate cognitive impairment, and epileptic seizures.

Other paroxysmal events have been reported [Overweg-Plandsoen et al 2003, Pérez-Dueñas et al 2009, Pons et al 2010, Urbizu et al 2010]. It is unclear whether these events represent epileptic or non-epileptic phenomena. These neurologic signs, which generally fluctuate and may be influenced by factors such as fasting or fatigue, include the following:

• Confusion
• Lethargy
• Somnolence
• Recurrent headaches, migraines
• Sleep disturbances
• Hemiparesis
• Total body paralysis
• Intermittent ataxia
• Chorea
• Action limb dystonia
• Cerebellar action tremor
• Writer's cramp
• Dystonic tremor; described as the only finding in a mother and daughter [Roubergue et al 2011]
• Parkinsonism
• Myoclonus
• Dyspraxia
• Nonkinesigenic dyskinesias

Microcephaly. Thirty-two of 58 persons with Glut1 DS had microcephaly ranging from mild (<1 SD below the mean in 14 patients) to moderate (<2 SD below the mean in 10) to severe (<3 SD below the mean in 8).

Additional findings. In addition to the low CSF-to-blood glucose ratio and decreased 3-O-methyl-D-glucose uptake in erythrocytes, the following findings may also be observed in individuals with this disorder:

• Positron emission tomography (PET). Cerebral fluoro-deoxy-glucose PET findings are distinctive with diffuse hypometabolism of the cerebral cortex and regional hypometabolism of the cerebellum and thalamus. Basal ganglia metabolism appears relatively preserved. This distinctive PET signature appears in early infancy and persists into adulthood regardless of disease severity or therapy with a ketogenic diet [Pascual et al 2002, Akman et al 2015]. The sensitivity and specificity of PET in the diagnosis of Glut1 DS have not been established.
• CSF lactate concentration is low-normal or low, often below 1.3 mmol/L or 11.7 mg/dL (range: 5.4-13.5 mg/dL) [De Vivo et al 1991, Wang et al 2005, Yang et al 2011, Leen et al 2013].

Pathophysiology

The disease manifestations can be explained in terms of current understanding of glucose transport in the brain: glucose is the principal fuel source for brain metabolism; the glucose transporter, Glut1 (solute carrier family 2, facilitated glucose transporter 1), the protein product of SLC2A1, is the fundamental vehicle that facilitates glucose entry into the brain. The cerebral metabolic rate for glucose (which is low during fetal development) increases linearly after birth, peaks around age three years, remains high for the remainder of the first decade of life, and declines gradually during the second decade of life to the rate of glucose utilization seen in early adulthood. It thus appears that the risk for clinical manifestations during fetal development is low and then rises throughout infancy and early childhood.

Human and animal data suggest that the margin of safety for glucose transport across the blood-brain barrier to meet the needs of brain metabolism and cerebral function is narrow. A milder clinical phenotype with intermittent symptoms (epilepsy, dyskinesias, and ataxia) may be predicted with 25%-35% reduction in Glut1 transporter function [Rotstein et al 2010]; a more severe phenotype results from greater reductions (perhaps 40%-75%) [Yang et al 2011].

The erythrocyte glucose uptake assay is a functional surrogate measure of residual Glut1 transporter function. Individuals displaying the classic phenotype have on average a 50% uptake assay, resulting from pathogenic loss-of-function variants that result in 50% reduction in Glut1 activity. Absence of Glut1 transporter expression is embryonic lethal [Wang et al 2006].

Genotype-Phenotype Correlations

A correlation between the specific type of SLC2A1 pathogenic variant and the clinical severity has been noted [Leen et al 2010, Yang et al 2011]. A clinical scoring system developed to classify phenotypic severity for mitochondrial diseases and Glut1 DS [Kaufmann et al 2004] has been used to correlate phenotype with other markers including genotype [Levy et al 2010, Rotstein et al 2010, Yang et al 2011].

Fifty-three affected individuals were stratified clinically according to the Columbia Neurological Score (CNS) into four groups [Yang et al 2011]:

• Minimal (CNS 70-76)
• Mild (CNS 60-69)
• Moderate (CNS 50-59)
• Severe (CNS 40-49)

Comparison of the type of SLC2A1 heterozygous pathogenic variants among the four groups revealed the following:

• Missense variants occurred predominantly in the mild and moderate clinical categories.
• Splice site and nonsense variants and insertions, deletions, and exon deletions occurred almost exclusively in the moderate and severe clinical categories.
• Complete gene deletions clustered in the severe clinical category.

A significant inverse correlation (R² = 0.94) was observed between the median values of the erythrocyte 3-O-methyl-D-glucose uptake and the clinical severity as determined by the Columbia Neurological Score. Thus, the erythrocyte glucose uptake is an indication of the functional effect of the pathogenic variant.

Many pathogenic variants have been identified [Wang et al 2005, Pascual et al 2008, Wang et al 2008, Leen et al 2010, Yang et al 2011]; several mutation hot spots and gene regions have been detected:

• Three individuals (each from a different family) have a mild clinical phenotype (as determined by the CNS) and a heterozygous p.Arg333Trp pathogenic missense variant.
• Twelve individuals have a heterozygous pathogenic missense variant at amino acid residue arginine 126.
  • Five had a mild phenotype: three members of one family with a p.Arg126His pathogenic variant [Brockmann et al 2001] and two members of one family with a p.Arg126Cys pathogenic variant [Ho et al 2001a].
  • Five with either the classic phenotype or the non-epileptic phenotype had a p.Arg126Cys pathogenic variant [Zorzi et al 2008, Leen et al 2010].
  • Two monozygotic twin brothers with paroxysmal choreoathetosis with spasticity were heterozygous for the p.Arg126Cys pathogenic variant [Weber et al 2011].
• A significant fraction (5/21) of pathogenic variants is located in a vulnerable region of the Glut 1 protein that involves the fourth transmembrane domain encoded by exon 4, suggesting a critical functional disturbance associated with structural alterations in this region of the protein [Pascual et al 2008].
• One individual had a p.Arg126Leu pathogenic variant in trans configuration with a p.Lys256Val pathogenic variant associated with a severe phenotype [Pascual et al 2008]. The father, who was clinically well, had neither missense variant. The asymptomatic mother was heterozygous for the p.Lys256Val pathogenic variant. Mutagenesis studies in Xenopus oocytes showed that in an uptake study under zero-trans influx conditions the p.Arg126Leu missense variant is more pathogenic than the p.Lys256Val missense variant. Therefore, the asymptomatic mother may have sufficient residual Glut1 activity to allow her to function normally. (Her glucose uptake assay revealed 83% residual activity; values of >74% residual activity correlate with a clinically normal state.) It is unclear whether synergy between the two pathogenic missense variants caused the severe phenotype in the child, a compound heterozygote [Rotstein et al 2010].
• Three individuals heterozygous for the p.Thr295Met pathogenic variant had:
  • A similar early-onset classic phenotype including monthly seizures beginning in infancy, developmental delay, ataxia, microcephaly, and language deficit;
  • Hypoglycorrhachia and low normal CSF lactate concentrations;
  • Normal erythrocyte zero-trans influx of 3-OMG.
    Mutagenesis studies suggest that p.Thr295Met specifically affects glucose flux by perturbing efflux rather than influx. These findings explain the seemingly paradoxic findings of Glut1 DS with hypoglycorrhachia and "normal" erythrocyte glucose uptake [Wang et al 2008, Fujii et al 2011]. Furthermore, this pathogenic missense variant may be a kinetic variant, functioning normally at 4° C (the temperature of the in vitro glucose uptake assay) and abnormally at body temperature [Cunningham & Naftalin 2013].
• A fourth individual heterozygous for the p.Thr295Met pathogenic variant with a similar mild classic phenotype was reported [Leen et al 2010].

Penetrance

Penetrance in Glut1 DS inherited in an autosomal dominant manner is complete.

An asymptomatic parent harboring the pathogenic variant implies a mosaic state.

Fewer pathogenic variants may be transmitted as an autosomal recessive trait; carriers are asymptomatic [Rotstein et al 2010].

Nomenclature

Paroxysmal exercise-induced dyskinesia and epilepsy (previously known as dystonia 18 [DYT18] or DYT-SLC2A1) [Suls et al 2008, Weber et al 2008, Zorzi et al 2008, Urbizu et al 2010] and paroxysmal choreoathetosis with spasticity (DYT9) [Weber et al 2011] are now recognized to be part of the Glut1 DS phenotypic spectrum.

Prevalence

No firm estimates of incidence and prevalence can be made, as cases have been reported worldwide and are biased by physician awareness of the disorder. Authors estimate the incidence/prevalence in Queensland, Australia, at approximately 1:90,000, which is likely a conservative estimate under the conditions of ascertainment [Coman et al 2006]. A more recent study from Scandinavia yielded a similar incidence/prevalence of 1:83,000 [Larsen et al 2015].

Evaluations Following Initial Diagnosis

To establish the extent of disease and needs of an individual diagnosed with glucose transporter type 1 deficiency syndrome (Glut1 DS), the following evaluations are recommended if they have not already been completed:

• EEG (pre-prandial and post-prandial tracings)
• Brain imaging, including FDG-PET in selected individuals
• Neuropsychological assessment
• Consultation with a clinical geneticist and/or genetic counselor

Treatment of Manifestations

Ketogenic diet. Diffusion of ketone bodies across the blood-brain barrier is facilitated by the monocarboxylic transporter 1 (MCT1). The ketogenic diet, introduced as a treatment for Glut1 DS in 1991, primarily provides an alternative fuel for brain metabolism. The ketogenic diet creates chronic ketosis by largely replacing carbohydrates and proteins with lipids in varying ratios.

Experience over the past two decades indicates that the ketogenic diet is well tolerated in most cases and is highly effective in controlling the seizures and improving gait disturbance. Of note, seizures may recur even with good dietary compliance [Klepper et al 2005]. Even when seizures are controlled, affected individuals may continue to have neurobehavioral deficits involving cognition and social adaptive behavior [Klepper et al 2002].

In the authors' experience, the neurologic outcome is influenced by the age at which treatment is initiated: affected individuals treated effectively at a younger age have a better outcome [Alter et al 2015].

Note: Because the ketogenic diet is deficient in L-carnitine, a cofactor that is essential for the metabolism of fats, dietary supplementation with 50 mg/kg/day of L-carnitine is recommended [De Vivo et al 1998].

Affected individuals develop a mild compensated metabolic acidosis when ketotic.

Antiepileptic drugs (AEDs) are generally ineffective or afford only limited improvement in the absence of a ketogenic diet. Certain AEDs may be relatively contraindicated as adjunctive treatment in children on the ketogenic diet (see Agents/Circumstances to Avoid).

• Phenobarbital and valproic acid can inhibit Glut1 transport.
• Valproic acid may partially inhibit beta-oxidation of fatty acids.
• Acetazolamide, topiramate, and zonisamide inhibit carbonic anhydrase and may potentiate the metabolic acidosis. These agents can also cause kidney stones.

Alpha-lipoic acid (thioctic acid) has also been shown to facilitate glucose transport in Glut4-dependent cultured skeletal muscle cells. Preliminary in vitro studies with Glut1 transport systems have shown similar results. Thus, alpha-lipoic acid supplements have been recommended, without supportive clinical evidence, as a treatment for Glut1 DS. Response has been modest at best; however, the dose taken by mouth may be insufficient to approximate experimental conditions [Kulikova-Schupak et al 2001].

Prevention of Primary Manifestations

Clinical observations suggest that early diagnosis and treatment with a ketogenic diet is associated with improved neurologic outcome by nourishing the immature brain during this critical period of growth and development [Alter et al 2015].

Prevention of Secondary Complications

For those who are treated with the ketogenic diet:

• L-carnitine supplementation to avoid carnitine deficiency
• Proper hydration and avoidance of carbonic anhydrase inhibitors to minimize likelihood of kidney stones
• Avoidance of carbohydrate-containing foods, intravenous fluids, and medications that will interrupt the state of ketosis. Family care providers often need to serve as the "watchdogs" to intercept these indiscretions.
• Avoidance of valproic acid treatment, which may be dangerous in individuals on a ketogenic diet because it increases the risk of a Reye-like illness. Valproic acid may also inhibit glucose transport.

Surveillance

Blood ketone concentrations should be monitored daily, weekly, or as needed to document the state of ketosis. A blood beta-hydroxybutyrate concentration of 3-5 mmol/L is recommended to insure a proper ketotic state.

Urinary measurement of ketonuria is only qualitative, and may be falsely reassuring as a strongly positive urine test for ketones may correlate with hypoketonemia. Blood measurement of ketone concentration is the preferred method.

Agents/Circumstances to Avoid

Barbiturates. Generally, children with infantile-onset seizures are treated with phenobarbital, the most commonly used antiepileptic drug in this age group. In vitro studies indicate that barbiturates aggravate the Glut1 transport defect in erythrocytes of individuals with Glut1 DS [Klepper et al 1999a]. On occasion, parents have reported that phenobarbital did not improve their child's seizure control or may have worsened their child's clinical condition.

Methylxanthines (e.g., caffeine), which are known to inhibit transport of glucose by Glut1 [Ho et al 2001b], have also been reported to worsen the clinical state of individuals with Glut1 DS [Brockmann et al 2001]. Thus, it is advisable for affected individuals to avoid coffee and other caffeinated beverages.

Valproic acid. The following studies suggest that valproic acid effects in vitro are mixed and the clinical consequences of valproic acid usage in patients with Glut1 DS cannot be predicted.

• Valproic acid inhibited Glut1 transport activity in normal and Glut1-deficient erythrocytes by 20%-30%. In primary astrocytes as well as in normal and Glut1-deficient fibroblasts, sodium valproate inhibited glucose transport by 20%-40%, accompanied by an up to 60% downregulation of GLUT1 mRNA expression [Wong et al 2005].
• A study using cultured astrocytes from the Glut1 DS mouse model showed an upregulation of Glut1 activity at lower valproic acid concentrations presumably from the valproic acid-associated inhibition of histone deacetylase activity [Kim et al 2013].

Evaluation of Relatives at Risk

It is appropriate to evaluate at-risk newborns, infants, and other relatives in order to identify as early as possible those who would benefit from initiation of treatment and preventive measures; early initiation of the ketogenic diet, ideally in infancy, results in better seizure control and improves long-term neurologic outcome. Molecular genetic testing can be used to clarify the genetic status of at-risk relatives if the pathogenic variant in the family is known.

See Genetic Counseling for issues related to testing of at-risk relatives for genetic counseling purposes.

Therapies Under Investigation

Triheptanoin is a specially designed synthetic triglyceride compound of three seven-carbon (C7) fatty acids intended to provide affected individuals with the medium-length, odd-chain, fatty acid heptanoate. Triheptanoin is metabolized rapidly in the liver to form a series of energy-containing metabolites, including heptanoate, which is further metabolized to 4-carbon (C4) and 5-carbon (C5) ketone bodies. Ketone bodies cross the blood-brain barrier via the monocarboxylate transporter, bypassing the deficient Glut1 transporter and providing alternative energy sources to the brain. The metabolites also have the ability to resupply intermediates of the tricarboxylic acid (TCA) cycle (i.e., anaplerosis). A pilot study of the effects of ingestion of triheptanoin as a dietary supplement in individuals with Glut1 deficiency was conducted and the preliminary results of this trial showed increased oxygen cerebral metabolic rate (CMRO2), decreased seizures, and improved neuropsychological performance [Pascual et al 2014]. A Phase II randomized double-blind placebo-controlled parallel-group study of the triheptanoin effects in those with Glut1 DS is currently enrolling affected persons who are not fully compliant with the ketogenic diet and continue to have seizures. The primary efficacy objective is reduction in seizure frequency [Mochel 2017].

Gene therapy. Murine Glut1 was packaged into adeno-associated virus 9 (AAV9) and systemically introduced into a neonatal mouse model of Glut1 deficiency. AAV9-Glut1-treated mice and relevant controls were assessed during adult life. In AAV9-Glut1-treated mice, Glut1 RNA and protein levels rose, CSF glucose levels were restored when the mice were treated early in development (but not later), brain size was normalized, and motor defects corrected [Monani et al 2014, Tang et al 2017]. Glut1 DS caused an arrest in cerebral angiogenesis that could be avoided if gene therapy was performed before the age of two weeks.

These results provide important proof-of-concept data of the therapeutic effects of restoring Glut1 protein function early in the life of patients with Glut1 DS and represent an important step toward finding a disease-modifying treatment for the human disease. These findings also indicate the need for newborn screening to treat individuals with genetically confirmed Glut1 DS presymptomatically in infancy.

Search ClinicalTrials.gov in the US and EU Clinical Trials Register in Europe for information on clinical studies for a wide range of diseases and conditions.