Lennox-Gastaut Syndrome (LGS) is a severe developmental and epileptic encephalopathy of childhood, characterized by drug-refractory epilepsy with multiple seizure types, cognitive impairment, and a characteristic interictal electroencephalogram (EEG) pattern of slow spike-wave discharges and generalized paroxysmal fast activity.1 Seizures in LGS are notoriously refractory to treatment and are traditionally managed with polytherapy, commonly involving broad-spectrum antiseizure medications (ASMs) such as valproic acid, lamotrigine, clobazam, topiramate, rufinamide, and felbamate.2,3 Recent clinical trials have demonstrated additional benefit from newer agents including cannabidiol (CBD) and fenfluramine (FFA), which have shown efficacy in reducing both seizure and non-seizure outcomes.4,5 Despite these advances, the challenges of pharmacoresistance and the adverse effects of polytherapy highlight the need for non-pharmacologic interventions.

Non-pharmacologic adjunctive therapies (i.e., dietary therapies, neuromodulation, surgical interventions) are increasingly recognized as important in LGS. Among these, dietary therapies such as the classic ketogenic diet (cKD) and its variants (e.g., modified Atkins diet, low glycemic index treatment, medium-chain triglyceride diet) are the most widely studied. These approaches have demonstrated significant efficacy in reducing seizure burden and improving quality of life in patients with LGS.6, 7, 8, 9, 10, 11 This review aims to provide a comprehensive overview of current dietary therapies for refractory epilepsies, with a particular focus on the efficacy and safety of these interventions in LGS.

The use of dietary modification, specifically fasting, to manage seizures dates back to Hippocrates. However, the concept of the ketogenic diet, characterized by high fat and low carbohydrate intake, was not introduced until 1921, when it was recognized that this dietary approach could mimic the metabolic effects of fasting by inducing ketone production while still providing the necessary calories for growth.12,13 Over the last several decades, the KD has evolved into a well-established non-pharmacologic therapy for epilepsy, expanding beyond its original role as a treatment reserved for refractory cases to become a more mainstream intervention applied for a variety of indications with numerous clinical benefits. In epilepsy, the goals of KD therapies include seizure frequency reduction, reduction or discontinuation of ASMs, improvement of cognitive function and overall, quality of life.14

The classic KD refers to a diet with high-fat, low-carbohydrate, and moderate protein. It remains the most extensively studied and is considered one of the most effective dietary therapies for refractory epilepsies.15 Because of the restrictive nature, several modified approaches aiming at enhancing patient’s adherence while replicating its therapeutic effects have been developed including the modified Atkins diet (MAD), low glycemic index treatment (LGIT), and the medium-chain triglyceride diet (MCT).16

The KD is thought to exert its anticonvulsant effect through mechanisms that influence neuronal excitability, energy metabolism and neuroprotection.17, 18, 19, 20, 21, 22, 23 The KD promotes a metabolic shift that favors the conversion of glutamate to γ-aminobutyric acid (GABA), thereby enhancing inhibitory neurotransmission and decreased neuronal excitability.17 KD also enhances adenosine triphosphate (ATP) production which further contributes to seizure control by supporting ion channel and transporter function, with downstream effects of reduced neuronal excitability through potassium channel activated hyperpolarization.18 KD also induces mitochondrial biogenesis and upregulates antioxidant responses to reactive oxygen species (ROS), strengthening neuronal energy reserves, and improving resilience to seizure-related oxidative injury.19,20 Additionally, ketone body production exerts direct antiepileptic actions through activation of ATP-sensitive potassium channels, decreased neuronal excitation and modulation of pathways involved in stress-response and anti-inflammatory pathways.21,22

The classic ketogenic diet (cKD) is a high-fat, low-carbohydrate, and adequate-protein diet. It is typically administered in a 4:1 ratio of fat to carbohydrate and protein, providing 90 % of the calories from fat.24 It is considered a restrictive diet and requires that all foods are weighed on a gram scale.13 Although the KD can be initiated gradually in the outpatient setting, many centers prefer in-hospital initiation to allow for close patient monitoring and comprehensive caregiver education. In-hospital KD initiation is strongly recommended particularly in patients with high seizure burden, or if the child is under 1 year of age.25 Ketogenic diet can be safely initiated in early infancy, with an all liquid, formula-based KD recommended for infants who have not yet started solids and for those fed enterally,26 and it is even possible to continue breast milk supplementation with KD.27,28

The available literature has demonstrated the efficacy of KD in the management of LGS.6, 7, 8, 9, 10, 11 A retrospective U.S. cohort study of 71 patients with LGS reported that, after 6 months on cKD, 51 % achieved > 50 % seizure reduction, including 23 % with > 90 % reduction. At 12 months, this effect was sustained, and 45 % of patients were able to reduce the number of ASMs. Generalized seizures appeared to respond more favorably to dietary therapy in this cohort.6 Some studies have demonstrated more favorable responses for generalized tonic-clonic seizures and drop seizures, however, this data is variably examined in the literature.29

Similarly, an Argentinian study of 20 children with LGS, 40 % achieved >50 % seizure reduction after 18 months on cKD, with three patients achieving complete seizure freedom.30 A retrospective study of 47 patients with LGS reported that 48.9 % experienced ≥ 50 % seizure reduction after 3 months of KD initiation, with 10 % achieving seizure freedom and 12.5 % ≥ 90 % seizure reduction by 6 months.9 Finally, a comprehensive literature review encompassing 18 studies and outcome data from 189 children with LGS reported that 47 % patients achieved a > 50 % reduction in seizure frequency following 3-36 months of KD.6

Long-term outcomes in a study of 68 patients with LGS, followed for a mean duration of 19 years, found that among 19 patients who underwent dietary therapy, 26.3 % and 0.005 % were seizure free on cKD and modified Atkins diet, respectively.8

KD has demonstrated effects on electroencephalographic (EEG) parameters in patients with refractory epilepsies, including LGS. Several studies have reported improvements in EEG background activity and interictal epileptiform discharges (IEDs) following KD initiation.10,31, 32, 33, 34 While several studies have reported a correlation between EEG improvement and clinical response to KD,10,31 other investigations have failed to establish a consistent association,34 highlighting ongoing debate and the need for further research to identify reliable predictors of treatment response.

KD has also been associated with cognitive benefits.14,35 A systematic review of 33 studies involving both adults and children subjective improvements in alertness, attention, and overall cognitive performance during KD therapy.14 These findings were partially supported by neuropsychological testing, which confirmed enhanced alertness, though no significant changes were observed in global cognitive function. The improvements observed were attributed to both seizure reduction and potential direct effects of KD.

A prospective study of 65 children with developmental and epileptic encephalopathies (DEEs) (including West syndrome, Lennox-Gastaut Syndrome, Doose Syndrome, and Dravet syndrome) found increases in mean developmental quotient, as well as improvements in attention and social functioning, among 34 patients who remained on KD at one-year follow-up. Similarly, another study of 52 children with DEEs, including 7 with LGS, reported cognitive and language improvements in 23 and 7 children, respectively.10

The modified Atkins diet (MAD) is similar to cKD, however, is less restrictive and therefore can be more accessible particularly for teenagers and adults. In MAD, carbohydrates are typically limited to 20 grams or less per day although does not require rigorous meal planning.36 MAD can often be started at home, however in younger children consideration can be given to in-hospital initiation as there can be a risk of hypoglycemia. Children can still produce ketones on MAD and can be as effective as cKD for some children. One prospective study comparing cKD to MAD in 51 children with refractory epilepsy, including 18 with LGS, did not find statistically significant changes in mean percentage of seizures, except in patients less than 2 years of age who achieved significantly higher rates of seizure freedom on the cKD at 3 months compared to those on MAD (53 % vs 20 %, p = 0.047). Mean blood ketones levels were higher in those on KD, but there was not a statistically significant difference between the two groups.37

Specific evidence regarding the efficacy of MAD in LGS is limited. One retrospective study evaluated 25 children with LGS, aged 2 to 14 years, maintained on carbohydrate restriction of 10 grams per day with the primary outcome being percentage seizure reduction. The authors found that at 3 months, 48 % (n = 12) had >50 % seizure reduction included 2 children who were seizure free. However, over half of the children discontinued the diet by 6 months either due to unsatisfactory seizure control or because the diet was too restrictive. Those children who remained on diet continued to experience a > 50 % reduction in seizures. No serious adverse effects were reported by the authors.38

Additional evidence for MAD is based on studies in intractable epilepsy. In a prospective study of 20 children with drug-resistant epilepsy experiencing three or more seizures per week, Kossoff and colleagues evaluated the efficacy and tolerability of MAD over a six-month span. They found that moderate urinary ketones were seen within four days of initiation of the diet. At 6 months, 65 % (n = 13) had >50 % improvement, and of those, seven children (35 %) having >90 % improvement, including four who were seizure-free. Prior to diet initiation, mean seizure frequency was 163 seizures per week, while at 6 months the average seizure frequency was 40 per week (p = 0.005). Efficacy was maintained even in children with lower levels of ketosis.36 In another prospective study of children with intractable epilepsy, MAD was evaluated in 14 children, including 6 with Lennox-Gastaut syndrome. Only half of the children continued the diet at 6 months, with 5 children having > 50 % reduction in seizures, including 3 who became seizure-free. The authors also looked at the effect of beta-hydroxybutyrate levels on outcomes. Interestingly, two patients who did not achieve beta-hydroxybutyrate > 3 millimole/Liter (mmol/L) became seizure-free, while four patients with consistent beta-hydroxybutyrate levels over 3 mmol/L had better outcomes, compared to those who had fluctuations in their beta-hydroxybutyrate levels.39 An open-label randomized study in children with nonsurgical drug-resistant epilepsy evaluated MAD versus levetiracetam. In this study of 101 children, including 47 with LGS, found a significantly higher proportion of children in the MAD arm had > 50 % seizure reduction compared to the levetiracetam arm (52.9 % versus 22 %) and a higher proportion of children with > 90 % seizure reduction and seizure-freedom on MAD versus levetiracetam, although this did not achieve significance.40

Medium chain triglycerides (MCT) diet is, again, similar to cKD, however, MCT oil is one of the sources of fat, compared to long-chain triglycerides (LCT) used in the cKD. MCTs yield more ketones than LCTs, so this diet allows for a greater allowance for carbohydrates and protein, which is felt to improve overall tolerability of the diet. Gastrointestinal side effects can limit use and therefore different percentages of MCT used typically range from 30 to 60 % of the overall diet.41, 42, 43

A randomized controlled trial comparing cKD and MCT diet in 145 children found no statistically significant differences in mean percentage of baseline seizures between the 2 groups at any time point up to 12 months. This study utilized KD ratios of 2:1 to 5:1, with the majority being on the classic 4:1 ratio, and MCT starting at 40-45 % and up to 60 % based on efficacy and tolerability. The authors found that at 3 months of treatment, 26 children on cKD and 27 children on MCT diet could have ASM doses reduced (p > 0.5). Treatment withdrawals and tolerability were similar in both groups with the only significant differences being lack of energy at 3 months and vomiting at 12 months in cKD.44

A prospective long-term study up to 24 months from the Netherlands, looked at MCT-ketogenic diet as add-on therapy in childhood refractory epilepsy in 48 children. This included one patient with LGS. MCT diet was used in 79 % of those followed. The retention rate was 33 % (n = 16) and 23 % (n = 11) at one and two years, respectively, with most discontinuations after three months being related to lack of efficacy, compliance, and tolerability. Of 35 patients with available data, 11 were considered responders with statistically significant decrease in mean seizure frequency seen at all points (p < 0.05) except for the 12-month visit (p = 0.056).45

The low glycemic index diet (LGIT) is a more liberalized diet that allows more carbohydrates, specifically those that will not elevate blood sugar, i.e. the glycemic index (GI), such as whole grains and certain fruits, but limits the intake of high glycemic foods such as baked goods.46 Carbohydrates are typically limited to 40-60 grams per day on LGIT. This diet is felt to help maintain stable blood sugar levels, but the production of ketone bodies is not expected on the LGIT. In an early retrospective study of LGIT for treatment of epilepsy, Pfiefer and Thiele reported on 20 patients undergoing therapy with LGIT, finding that 10 of these patients achieved a > 90 % reduction in seizures.46 A subsequent larger retrospective series from the same institution in 76 children, 89 % of whom had intractable epilepsy failing 3 or more ASMs, demonstrated >50 % seizure reduction in 66 % of patients at 12-months.47

Coppola and colleagues published an Italian retrospective study of LGIT in children and young adults with refractory epilepsy, using the LGIT protocol initially described by Pfiefer and Thiele. In 15 patients, 20 % with LGS, eight patients utilized LGIT as their first dietary therapy, with the remaining patients previously trialed KD. With a median follow-up of 12 months (range 1-60 months), 6 patients (40 %) had 75-90 % seizure reduction, 2 (13.3 %) had 50 % reduction, and the remainder had unchanged seizure frequency.48

A retrospective study of 36 patients with DRE, one-third of whom had LGS, found responder rates with ≥ 50 % seizure reduction of 56 %, 58 %, and 53 % at 3, 6 months respectively, with 2 patients becoming seizure-free at 3 months for up to one year. The greatest response rates were seen in those patients with Dravet Syndrome (4/5, 80 %) and LGS (9/12, 75 %). Seven individuals produced urine ketones on LGIT.49

An open-label randomized controlled study of LGIT in children aged 2-8 years with refractory epilepsy, including 9 children with LGS (45 % of the treatment arm), found that 30 % of children had a > 50 % seizure reduction with one patient with LGS becoming seizure free, compared to no response in the control arm.50 A systematic review and meta-analysis of randomized controlled studies on MAD versus LGIT, which ultimately included only three studies, found comparable rates of seizure reduction, as well as reported improvements in cognition and behavior, between both groups between treatments, with less side effects seen in LGIT.51

Other diets have not been frequently studied for the treatment of refractory epilepsy. A case report of a child with a phenotype consistent with LGS demonstrated improvement with initiation of a gluten-free diet following a diagnosis of celiac disease, confirmed by positive serum testing and jejunal biopsy.52 A systematic review found a bidirectional relationship between epilepsy and celiac disease, with each population at risk of developing the other disorder. This review broadly defined a positive response to a gluten-free diet as reduction in seizure frequency, seizure cessation, ASM reduction or cessation of ASM following diet initiation. They found a positive response in 53 % of 34 patients in included studies.53 These findings could be applicable if there are other signs or symptoms of celiac disease or concerns regarding ASM absorption. It should be noted that a gluten-free diet has not been formally studied in randomized controlled or open-label trials for epilepsy.

Depending on efficacy and tolerability, KD is typically maintained for 3 months to 2 years, with longer duration possible based on continued benefit.26,54 Similar to pharmacologic treatments, KD therapies are associated with adverse effects which tend to be more frequent with stricter therapies such as the cKD and MCTD.55 Commonly reported side effects across all 4 KTDs include gastrointestinal disturbances (i.e., constipation, diarrhea, emesis), weight loss or gain, and less commonly, osteopenia, nephrolithiasis, hypoglycemia, and stunted growth.10,30,36,44,49, 50, 51,55, 56, 57 A rare complication is a selenium-deficiency induced cardiomyopathy.58 Overall, most adverse events can be transient and mitigated through dietary adjustments. To ensure safety, laboratory monitoring is recommended at one month following initiation and subsequently every three months to evaluate for hypoglycemia, excessive ketosis, dyslipidemia, nephrolithiasis, osteopenia, and vitamin or electrolyte abnormalities.25