Epilepsy is a common neurological disease with approximately 50 million patients globally. Only fewer than 70% of epileptic patients can achieve seizure control with anti-epileptic drugs [1]. There are certain syndromes in which that percentage is significantly lower, e.g., Lennox-Gastaut syndrome (LGS). LGS is a rare, severe, early-onset developmental epileptic encephalopathy characterized by a triad of multiple drug-resistant seizure types, a specific electroencephalography pattern showing bursts of slow spike-wave complexes or generalized paroxysmal fast activity, and intellectual impairment. Not all patients have all of the core seizure types (e.g., tonic, atonic, and atypical absences), especially at onset [2]. The tonic type of seizure is seen in all patients with LGS, but may not be present at the time of its onset [3]. The incidence of LGS is estimated at 0.1 to 0.28 per 100,000 people per year; the lifetime prevalence at the age of ten years amounts to 0.26 per 1000 children [4]. It is estimated that LGS patients account for 1–10% of cases of childhood epilepsy and 1–2% of all epilepsy patients [2]. There are identifiable and non-identifiable causes of LGS. Approximately, 65 to 75% of patients have an identifiable cause. These include brain damage (e.g., head injury), perinatal complications (e.g., birth asphyxia, intrauterine growth retardation, kernicterus), congenital central nervous system malformations (e.g., tuberous sclerosis), infections (e.g., meningitis, sepsis), or metabolic disorders ^[2][3][2,3]. When LGS has no apparent cause, a genetic predisposition or etiology is probable [2]. The mutation of genes involved in human brain development (e.g., the forkhead box G1 (FOXG1), or chromodomain-helicase-DNA-binding protein 2 (CHD2) genes) and the gene for presynaptic protein dynamin 1 (DNM 1) have been found to be associated with this syndrome. Valproate, lamotrigine, and topiramate are considered the first-choice drugs. According to randomized control trials, other anti-epileptics reported to be effective are clobazam, felbamate, and rufinamide. It is important to start with as few drugs at a time as possible, in the lowest possible doses. If the first drug fails, change to another drug, but if the second drug is also not effective, a second agent can be added to the existing regimen [3]. As available pharmacologic treatments for seizures are limited in their efficacy, several non-pharmacological methods are used, including surgical and dietary interventions. Surgical procedures can lead to control in a select subset of patients suitable for that kind of treatment but still, a significant number of patients are left with uncontrolled seizures. The ketogenic diet and related diets have proven to be useful in drug-resistant childhood epilepsy [5].

2. Ketogenic Dietary Therapies

KDTs are well-established, nonpharmacological treatments used for children and adults with pharmacoresistant epilepsy [6]. Their mechanisms of action rely on putting the organism into a ketogenic state, similar to starvation, which has been reported as a treatment for epilepsy since ancient times [7]. KDT refers to any diet therapy in which dietary composition results in a ketogenic state of human metabolism. The diet generally refers to a high-fat, low-carbohydrate, and moderate-protein diet. Following the development of the classic ketogenic diet (cKD), new diets have been proposed in an attempt to increase retention and savoriness while imitating the effects produced by the original diet. Currently, there are four major KDTs—cKD, the modified Atkins diet (MAD), the medium-chain triglyceride diet (MCT), and the low glycemic index treatment (LGIT). 

The cKD and the MCT diet have been in existence the longest and are typically started in the hospital by a dietitian and neurologist. cKD is the oldest and the most researched diet among dietary therapies for epilepsy. It was designed in the year 1923 by doctor Russell Wilder from the Mayo Clinic and was designed to particularly treat this disease ^[8][9][9,10]. In the cKD, the fat source is largely long-chain triglycerides (LCT), which are gained mainly from standard foods. cKD is typically administered in a 4:1 ratio of fat to carbohydrate and protein, providing 90% of the total calories from the fat. MCT oils provide more ketones per kilocalorie of energy than LCTs. This enlarged ketogenic potential means less total fat is needed in the MCT diet, which allows the inclusion of more carbohydrates and protein and more potential food choices [6]. In both of the above-mentioned KDTs, calculations and arrangements concerning the diet and education of a patient and their family should be conducted by a dietitian ^[6][10][6,11]. As the cKD and MCT can be too restrictive, alternative diets were created. They are more liberal versions of the cKD, which are less restrictive on protein and calorie intake, more affordable for patients, and can be similarly effective in the treatment of epilepsy. The MAD was created at Johns Hopkins Hospital in 2003, primarily for children with behavioral difficulties and adolescents whose parents and neurologists were unwilling to start on the cKD ^[11][12]. The MAD is a high-fat, low-carbohydrate therapy which typically provides approximately a 1:1–1.5:1 ketogenic ratio, but no set ratio is required and some children can achieve a ratio as high as 4:1. The initial daily carbohydrate consumption on the MAD is approximately 10–15 g with a possible increase to 20 g per day after 1–3 months. In addition, there is no restriction on protein, fluids, or calories, making meal planning easier. Detailed calculations are not required [6]. The LGIT was introduced as an elasticized version of the cKD and was first found to be successful in Massachusetts General Hospital in 2005 ^[12][13]. The diet is more liberal and allows 40 to 60 g of carbohydrates per day, but restricts sources of carbohydrates to a glycemic index of less than 50 to prevent postprandial increases in blood glucose. Fats and proteins are unrestricted. This diet is preferred by adolescents because of their difficulty in following the cKD. Admission to hospital is not required to implement the treatment ^[13][8].

3. Mechanisms of KDTs’ Antiseizure Effect

Despite the antiseizure effect of ketosis being known from ancient times [7], the mechanisms of KDTs’ action have not been fully elucidated ^[14][18]. Among proposed theories, the abundance of ketone bodies (KBs) and fatty acids, the decreased glucose supply, and the impact on gut microbiota are mainly considered, inducing the modifications of neuronal metabolism, neurotransmitters, and ion channels ^[15][16][17][19,20,21].

KBs and fatty acids can generate direct changes in organism homeostasis. Acetoacetate ^[18][25] and beta-hydroxybutyrate ^[18][19][25,26] were shown to induce activation of ATP-sensitive potassium (KATP) channels ^[18][19][25,26], while polyunsaturated fatty acids (PUFA) and, presumably, KBs activate two-pore domain potassium (K2P) channels ^[20][27], increasing the seizure threshold via neuronal hyperpolarization ^[18][19][20][25,26,27]. A medium-chain fatty acid—decanoic acids—was shown to independently contribute to seizure control through alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor inhibition ^[21][23].

KDTs conditions also promote neuronal stability via their effects on the neurotransmitters mainly affecting gamma-aminobutyric acid (GABA) ^[15][19]. Upregulation of glutamic acid decarboxylase ^[22][29] and reduction in GABA transaminase activity ^[23][30] were suggested to result in enhanced GABA concentration. Regarding glutamate, the impact of KDTs on its levels is inconstant, and further research in this field is needed. Norepinephrine, dopamine, serotonin, galanine, and neuropeptide Y are highlighted among other neurotransmitters presumably involved in the antiseizure effect of KDTs ^[15][19].

Regarding energy metabolism, prolonged KDT implementation promotes mitochondria synthesis and concentration, increasing adenosine triphosphate (ATP) generation and enhancing the brain’s resistance to metabolic stress ^[15][24][19,28]. Reduced glycolysis presumably contributes to the antiseizure effect, as an intravenous glucose bolus reverses ketonemia and the antiseizure effects of KDTs rapidly ^[25][31]. 

4. Efficacy of cKD

The efficacy of cKD in the treatment of LGS is well-documented in scientific research. A literature review conducted by Lemmon ME et al. ^[26][37], which included 18 papers from the years 1989–2010 concerning applying cKD in, overall, 189 children with LGS, showed that after 3–36 months of using this diet, 47% of investigated children experienced ≥50% seizure reduction, whereas 16% of children achieved seizure-free status. Moreover, it was proven that the cKD causes a reduction in multiple types of epileptic seizures, including tonic, atonic, or atypical absence seizures. No predictive factors were found for the reduction in the number of seizures after using cKD in LGS. Research by Lemmon ME et al. ^[26][37] showed that age, gender, presence of side effects, etiology of LGS, number of previously taken anticonvulsants, and history of seizures during infancy were not predictive of >90% seizure reduction after 12 months of using cKD ^[26][37]. 

5. Efficacy of Alternative Ketogenic Diets

Although alternative ketogenic diets were introduced to provide better tolerability and palatability than cKD, their efficacy is the most important issue. The comparison of the efficacy of MAD and cKD was assessed among the general population of children and adolescents with epilepsy. In a meta-analysis conducted by Razaei et al. ^[27][43], no statistically significant differences in efficacy were reported between MAD and cKD. Regarding LGS patients, a study conducted by Sharma S et al. ^[28][15] comprised of 25 pediatric LGS patients treated with MAD. After three months, nearly half of the patients had a ≥50% reduction in seizure frequency. After six months, 14 children discontinued the diet; however, all of the children who remained on the diet had more than a 50% reduction in seizure frequency both at the 6th and 12th month of the study ^[28][15].

6. Safety and Tolerability of Ketogenic Dietary Therapies

The literature reports from the general epileptic population show that KDTs are generally considered as safe treatment options ^[29][30][31][40,41,45]. They are applied in adults, as well as in children ^[32][46], showing good tolerability for patients as young as six weeks ^[33][47]. Additionally, cKD may be used in groups of LGS patients presenting certain metabolic abnormalities. In the case of LGS, the use of cKD could have certain side effects including gastrointestinal disturbances, such as diarrhea, constipation or vomiting, hypocalcemia, kidney stones, hyperuricemia, and metabolic acidosis ^[34][50]. Patients subjected to cKD also experience problems with the cardiovascular system, kidneys, and skeletal system ^[35][51]. Moreover, research shows that this diet could cause growth retardation in children ^[36][37][52,53]; however, the data regarding growth retardation is mixed [6]. It was proven that patients who experience side effects have a higher risk of non-adherence to this diet ^[38][49]. Compared to cKD, a lower occurrence of side effects is an advantage of LGIT. As a result, this diet is reported as easier to use and continue ^[15][39][19,44]. The side effect among all patients in one of the reported studies was transient diarrhea. In addition, laboratory abnormalities were observed—reduced serum total carbon dioxide, hypercholesterolemia, increased alanine aminotransferase, increased lipase, and increased blood urea nitrogen. However, mentioned conditions did not require additional management or medication ^[40][17].

7. Conclusions

KDTs are effective in LGS treatment. KDTs are reported as safe treatment options and can be administered to different populations of epileptic patients. The most common complications include constipation, diarrhea, and vomiting, while severe adverse effects are rarely reported. Several less-restrictive KDTs, including MCT, MAD, and LGIT, are worth considering, due to their better tolerability, lower treatment costs, and easier administration.