Dietary Therapy in Genetic Epilepsies

Ketogenic diets (KDs) are formulated to consist of high fat intake combined with restricted carbohydrate and adequate protein intake to yield a state of nutritional ketosis. In the strictest form—classic KD—foods are weighed and measured to conform to the prescribed ratio for the individual. The KD ratio refers to the grams of fat in relation to the grams of carbohydrate plus protein of each meal. Therefore, an individual on a 3:1 diet consumes 3 grams fat for every 1 gram of carbohydrate plus protein.¹ Less restrictive diets have been developed that allow individuals to enter the state of nutritional ketosis by restricting carbohydrates to specific daily targets and adding fat to each meal. For example, in a modified Atkins diet, an individual may have a net carbohydrate goal of 20 grams per day.¹ If using a low glycemic index diet, the goal is 40 to 60 grams of carbohydrates per day, with the focus on consuming lower glycemic index carbohydrates.² The glycemic index is a measure of how quickly various carbohydrates are broken down and metabolized, and the effect they have on blood glucose levels² (Figure).

KDs were used widely to treat epilepsy beginning in the early 1920s through the 1930s, prior to FDA approval of phenytoin in 1939.³ Thereafter, medication discovery and approval displaced KDs as a major treatment for epilepsy until KDs returned to public attention through media portrayals of an infant named Charlie who experienced freedom from seizures after treatment with KD at Johns Hopkins Medicine.³ In the past 20 to 30 years, interest in and use of KDs for drug-resistant epilepsy has risen, and KDs have been adapted for broader use outside of the pediatric population. Drug-resistant epilepsy is defined as failure to achieve seizure freedom after adequate trials of at least 2 appropriately chosen antiseizure medications (ASM).⁴ This article reviews the evidence for use of KDs in various genetic conditions, including Angelman syndrome (AS), glucose transporter type 1 deficiency syndrome (GLUT1-DS), tuberous sclerosis complex (TSC), and early-onset epileptic encephalopathy with burst suppression.

Angelman Syndrome

AS, caused by loss or reduction of maternal UBE3A (15q11-q13) function, is characterized by developmental delay, ataxic gait, and limited expressive speech.⁵ More than 80% of individuals with AS develop epilepsy, and the majority of cases (~70%) are medically refractory. The typical onset of seizures is between age 1 and 3 years. Generalized seizure types are the most common, but individuals often have multiple seizure types. EEG tests can have characteristic findings, including notched occipital delta rhythms and rhythmic theta patterns.⁶

A small prospective study of the low glycemic diet included 6 children with AS between ages 1 and 18 who had never been treated with dietary therapy.⁶ A baseline EEG test was obtained, and seizure diaries were kept. Participants were followed up at 1 and 4 months, including a repeat EEG test at 4 months. At baseline, participants had between 0.4 and 30.9 seizures per week. Three participants became seizure-free, 2 had greater than 80% seizure reduction, and 1 had no response. In a retrospective case series of 23 participants with AS treated with the low-glycemic diet,⁷ 22% achieved seizure freedom and 43% were seizure-free except during illness or episodes of nonconvulsive status epilepticus. An additional 30% of participants had decreased seizure frequency, nearly all of whom achieved 50% to 90% reduction in seizures. The average duration of treatment was 3 ± 2.5 years. Most recently, supplementation of exogenous ketones in 13 children with AS was shown to be safe, though study size and duration limited conclusions about efficacy.⁸ A large proportion of patients with Angelman syndrome respond favorably to dietary therapy.

GLUT1-DS

Glucose is transported across the blood–brain barrier by the glucose transporter protein (GLUT1), which is encoded by the SLC2A1 gene on chromosome 1.^9,10 In GLUT1-DS, this transport is faulty, leading to low glucose levels in the cerebrospinal fluid. A diagnosis of classic GLUT1-DS is based on low levels of glucose in the cerebrospinal fluid in the presence of normal serum glucose (hypoglycorrhachia) and/or a pathogenic variant in the SLC2A1 gene; missense, nonsense, and structural variants all presumably causing haploinsufficiency have been described.¹⁰ GLUT1-DS is a heterogeneous disease. In the classic form, individuals present with encephalopathy and seizures. However, individuals also can present with movement disorders (eg, ataxia, chorea, dystonia), varying degrees of intellectual and developmental disability, and eye movement abnormalities, including aberrant gaze saccades.¹⁰ KD can provide an alternate fuel source for the central nervous system in the form of ketone bodies, which do not rely on GLUT1 for movement across the blood–brain barrier; rather, they enter the brain by a monocarboxylic acid transporter.⁹

A retrospective case series¹¹ looked at 20 children with GLUT1-DS who were diagnosed between age 4 weeks and 18 months and treated with classic KD. All had epilepsy and were refractory to multiple ASM. Seizure types included generalized tonic-clonic, absence, focal, myoclonic, and astatic or atonic. All caregivers of these 20 infants and children reported resolution or reduction in seizures on classic KD.

A small prospective study¹² included 15 children prescribed a 3:1 KD. There were between 2 and 5 years of follow-up for each participant. Ten of 15 participants remained seizure-free on KD in monotherapy. Two of 15 participants had seizures recur after 2 years despite adequate levels of ketosis, and these seizures were controlled with ethosuximide. One participant experienced seizure reduction but not seizure freedom. No seizure adverse effects occurred and parental satisfaction with the diet was good. Two participants discontinued the diet.

Information also is available on the long-term safety of KDs in GLUT1-DS. A prospective, multicenter case series looked at the long-term cardiovascular safety of KDs in GLUT1-DS,¹³ including 10 participants followed over 10 years. Outcome measures included body mass index and total cholesterol, high-density lipoprotein, low-density lipoprotein, and triglyceride levels at baseline, 6 months, and 2, 5, and 10 years. There was no significant difference in cardiovascular risk for participants on long-term KDs compared with a reference population.

In 2014, a case series of 3 individuals provided information on long-term bone health and body composition in individuals with GLUT1-DS on KDs. Fourteen participants were adults between age 24 and 25 who had been on a 3:1 KD for more than 5 years. Outcome measures included anthropometric and body composition measurements, bone mineral content, and bone mineral density at baseline and annually. Two individuals had slight fat mass loss; 1 had slight fat mass gain. All 3 had reductions in bone mineral content (–5.5% to –10.8%) at 3 years, which was stable thereafter. Both bone mineral content and bone mineral density were within the normal range at 5 years.

Tuberous Sclerosis Complex

TSC is an inherited multisystem disorder that is typically caused by pathogenic variants in TSC1 or TSC2. It is characterized by pleomorphic features involving several organ systems. Hallmark central nervous system features include the presence of cortical tubers, subependymal nodules, and subependymal giant cell astrocytomas. People with TSC can experience epilepsy, behavioral disorders, and cognitive defects, with epilepsy being the most common, reported in up to 90% of individuals. Many individuals with TSC can go on to develop medically refractory epilepsy.^15,16

A retrospective review¹⁵ of individuals with TSC-associated medically refractory epilepsy treated with KD at a single center over a period of 7 years demonstrated that 10 of 12 participants (83%) experienced 50% or greater reduction in seizures after 3 months of treatment. Four of those individuals (33%) were seizure-free. Two participants discontinued the diet because of worsened seizures. At 6 months, 60% of individuals continuing on the diet were seizure-free, and 80% had greater than 50% reduction in seizures. Six of the initial 12 individuals maintained the diet for more than 12 months, and 4 had a 90% or greater reduction in seizures, with 2 of these individuals being seizure-free.

An observational study¹⁶ of 31 children with TSC-associated epilepsy treated with KDs demonstrated that 45% of participants had greater than 50% reduction in seizures at 18 months, with this percentage dropping to 32% at 24 months. Approximately 20% of individuals were seizure-free after 24 months on KDs. This study also looked at the time to response (defined as greater than 50% reduction in seizures) and found that the majority of participants who would be responders had response within the first month on KDs. No additional individuals obtained greater than 50% reduction in seizures after 3 months on KDs.

Early-Onset Epileptic Encephalopathy With Burst Suppression

Early-onset epileptic encephalopathy with burst suppression, also known as Ohtahara syndrome, is often associated with structural or metabolic etiologies. Varying genetic factors are increasingly being implicated in this disorder. A retrospective chart review¹⁷ included individuals with seizure onset at 3 months or younger, EEG tests showing burst-suppression pattern, a full etiologic workup completed, and follow-up available for at least 1 year. Individuals with hypoxic-ischemic encephalopathy were excluded. Forty-eight individuals were included and a genetic diagnosis was established in 31 (64%). A total of 27% were found to have STXBP1 sequence variations; 10%, KCNQ2 sequence variations; and 10%, SCN2A sequence variations. A variety of other sequence variations accounted for the remaining cases. Among people in whom KD was attempted, all of those with a KCNQ2 sequence variation (n=4) and 7 of 10 with an STXBP1 sequence variation were responders. However, only 1 of 5 with an SCN2A sequence variation had response.

Infantile Spasms

Infantile spasms (IS) are an age-dependent form of epilepsy which usually present between 3 and 10 months of age. The incidence is 2 to 3 per 1000 live births, and IS may be associated with structural, metabolic, or genetic abnormalities. However, in many cases, the etiology is unknown. A systematic review¹⁸ of 13 observational studies examined the efficacy of KDs as an adjunctive therapy in 341 patients with IS and found that ~65% experienced >50% reduction in spasms. Median spasm-free rate was ~35%, with those with IS of unknown etiology having an increased probability of achieving this status.

Conclusion

A variety of dietary therapies are available throughout the lifespan for people with difficult-to-control epilepsy. Although data delineating the efficacy of KDs on the majority of genetic epilepsies are not yet available, certain genetic epilepsies (ie, AS and GLUT1-DS) are suggested to be particularly responsive to dietary therapy. Response to dietary therapy in TSC-associated epilepsy seems to be more modest but is still likely better than the 6-month to 1-year seizure-freedom outcomes reported for additional medication trials beyond 2 failed ASM, which range from 11% to 28%.^19-21 In infants with catastrophic onset of epilepsy in whom genetic etiologies can be identified, those with STXBP1 or KCNQ2 sequence variations should have classic KDs considered early in the course of their treatment. In some individuals, KDs can and should be continued for years, and evidence suggests the overall long-term risks from a cardiovascular and bone health perspective are acceptable.