- Email: [email protected]
Therapeutic effects of the ketogenic diet in children with Lennox-Gastaut syndrome
Epilepsy Research 128 (2016) 176–180
Contents lists available at www.sciencedirect.com
Epilepsy Research journal homepage: www.elsevier.com/locate/epilepsyres
Therapeutic effects of the ketogenic diet in children with Lennox-Gastaut syndrome Yunjian Zhang 1 , Yi Wang 1 , Yuanfeng Zhou, Linmei Zhang, Lifei Yu, Shuizhen Zhou ∗ Department of Pediatric Neurology, Children’s Hospital of Fudan University, China
a r t i c l e
i n f o
Article history: Received 21 May 2016 Received in revised form 18 October 2016 Accepted 6 November 2016 Available online 9 November 2016 Keywords: Ketogenic diet Lennox-Gastaut syndrome EEG Children
a b s t r a c t Objective: The aim of this study was to evaluate the efficacy of the ketogenic diet (KD) on the clinical and electroencephalographic (EEG) features of Lennox-Gastaut syndrome (LGS) and explore the relationships between EEG changes and clinical efficacy. Methods: We retrospectively studied 47 patients with LGS who accepted KD therapy between May 2011 and May 2015. Clinical efficacy and EEG features such as background activity, abnormal interictal epileptic discharges (IEDs) and the discharge location were evaluated prior to and at 3 and 6 months after therapy. Responders were defined as ≥50% seizure reduction. Results: At 3 months of treatment, 23 patients (48.9%) had ≥50% seizure reduction. Seven patients (14.9%) discontinued treatment between 3 and 6 months because of lack of efficacy or inability to adhere to the diet. At 6 months of treatment, 4 patients (10%) were seizure free, 5 (12.5%) had ≥90% seizure reduction, 12 (30%) had a reduction of 50–89%, and 19 (47.5%) had <50% reduction. Patients with improved EEG background and reduced IEDs had an improved seizure reduction rate compared with patients without change in EEG background or IEDs (p < 0.01). Conclusions: The results show that the KD is effective in LGS. It can control seizures and improve EEG abnormalities. Early improvement in the EEG background and a reduction in IEDs may be predictors of a patient’s response to diet. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Lennox–Gastaut syndrome (LGS) is a severe childhood epileptic encephalopathy characterised by multiple seizure types, intellectual disability and the presence of diffuse fast activity and slow (<3 Hz) spike-wave complexes on electroencephalogram (EEG) (AlBanji et al., 2015). A literature review of LGS found that because a precise definition of this syndrome is lacking, the literature contains contradictory data on many aspects (van Rijckevorsel, 2008). It is a rare disease, with an estimated incidence of 0.28 per 1000 live births (Rantala and Putkonen, 1999). The prevalence may be higher because of its refractory characteristics, and it perhaps accounts for about 1–10% of all cases of epilepsy in children (Camfield, 2011). The aetiology of LGS is extensive and diverse. About 30% of LGS patients are cryptogenic. Perinatal brain damage, central nervous system infection sequelae, and malformations of cortical development are some of the most common causes, and probably account for 70% of all cases (van Rijckevorsel, 2008). There is only lim-
∗ Corresponding author at: 399 Wanyuan Road, Minhang District, Shanghai, China. E-mail address: [email protected] (S. Zhou). 1 Co-first author. http://dx.doi.org/10.1016/j.eplepsyres.2016.11.003 0920-1211/© 2016 Elsevier B.V. All rights reserved.
ited evidence for a genetic component. Half of West syndrome cases and other infantile epileptic encephalopathies will progress to LGS, and 20–30% of LGS cases develop from West syndrome or other epileptic encephalopathies (Trevathan, 2002; You et al., 2009). Intractable seizures of LGS include tonic, atypical absence, atonic and myoclonic seizures. All seizure types are resistant to antiepileptic drugs (AEDs) (van Rijckevorsel, 2008). An alternative therapeutic option is the ketogenic diet (KD), which may be useful, and according to some authors should be used as early as possible (Caraballo et al., 2014; Kim et al., 2015). The KD is a high-fat, low carbohydrate diet with 90% of calories derived from fat. It has been used in the treatment of intractable childhood epilepsy since the 1920s (Winesett et al., 2015). The KD has been shown to be beneficial for the treatment of drug-resistant epilepsy. It has been reported that more than approximately 50% of intractable epileptic children had at least a ≥50% seizure reduction after KD therapy (Henderson et al., 2006). Some patients achieved >90% reduction (Hallböök et al., 2015). It has been observed that the efficacy of the KD differs among different seizure types or epileptic syndromes (Caraballo et al., 2011). According to the International Ketogenic Diet Study Group recommendations, the KD may be particularly effective in Dravet syndrome, Doose syndrome, and some hereditary epilepsies, with
Y. Zhang et al. / Epilepsy Research 128 (2016) 176–180
larger than expected rates of seizure freedom in these patients (Kossoff et al., 2009). However, there is a paucity of published data on the efficacy of the KD in LGS. In a retrospective study, the records of 71 children with LGS who were treated with the KD were reviewed (Lemmon et al., 2012). Intent-to-treat analysis was used, and at the end of 6 months there were 36 (51%) children who had >50% reduction in seizures and 16 (23%) who had >90% reduction in seizures. One child was seizure free. In a prospective study, 20 children with LGS received KD therapy (Caraballo et al., 2014). Fifteen of the children continued to receive the KD for at least 18 months. Eight of these 15 children had a ≥50% decrease in seizures, including three who became seizure free. Although a number of published studies have attempted to identify predictors of the response to the KD, the studies often used a group of medically refractory epilepsy patients (including patients with LGS), and therefore the results obtained may not be specific to LGS. In particular, the relationship between EEG features and the clinical efficacy of the KD is unknown in patients with LGS. Therefore, the purpose of this study was to examine the effects of the KD on clinical and EEG features in children with LGS and to evaluate the correlation between them. 2. Methods 2.1. Participants A total of 47 LGS patients who experienced daily or weekly seizures despite therapy with at least two appropriate AEDs were enrolled in our study and received KD therapy between May 2011 and May 2015. To be enrolled, patients had to meet the following criteria: (1) multiple seizure types, delay of intelligence development, and an abnormal EEG pattern of fast activity and slow spike-wave complexes; (2) have tonic seizures during sleep; (3) intractable epilepsy defined as two or more drugs used without achieving control; and (4) signed informed consent provided by the parents or legal guardians. Patients were excluded if they were not suitable for the KD; had any deficiencies of the transportation or oxidation of fatty acids or ketobodies; had deficiency of pyruvate carboxylase; had a mitochondrial disease; or their parents did not adhere to providing the diet. A paediatric epileptologist and a dietician managed all patients on the KD. The patients were started on the classic 4:1 ratio KD (fat:protein plus carbohydrate). KD therapy was given for 6 months. All patients continued to take the same drugs they were taking before KD therapy during the 6-month period; no changes in drug prescription were permitted. The patients were given potassium citrate supplementation to prevent hypokalemia or lithiasis. Calcium and vitamin D were also given. Patients stayed in the hospital for 1 week and were closely monitored for efficacy and potential side effects. Seizure frequency was documented daily using seizure diaries kept by the families, which were reviewed at each clinic visit. After discharge, the children were reviewed as outpatients monthly for an initial 3 months and then followed up once every 3 months. Adherence to the KD was evaluated during these visits. The efficacy, side effects, laboratory tests and video EEG recordings were assessed at regular intervals. 2.2. MRI and EEG All patients underwent magnetic resonance imaging (MRI). Imaging abnormalities were classified as lesional (e.g., tuberous sclerosis, subdural effusion, hippocampal atrophy) or other (e.g., atrophy, developmental abnormalities). Electrographic patterns were measured through digital EEG monitoring. EEG was performed when patients were both awake
177
and asleep. Electrocardiography (ECG) and electromyography (EMG) were performed during the awake, drowsy and sleep states. The results from each patient were reviewed by two of the authors; if no consensus was reached, a third author reviewed the results and made the final decision. Hyperventilation and photic stimulation were performed in patients who were cooperative. The EEG markers such as background activity (rhythm, frequency) (Dressler et al., 2010), abnormal interictal epileptic discharges (IEDs), and discharge location were recorded. Improvement of EEG background included an increase of wave frequencies in the occipital region or amelioration of disorganisation. More specifically, after therapy, regular rhythm was present or the rhythm was increased in frequency, that is, the frequency of the background wave on the EEG increased by at least 1 Hz compared with before therapy. The spike-wave index of IEDs was expressed as the number of seconds containing spikes or sharp waves divided by the total number of seconds times 100. Kessler et al. (2011) reported that those patients with an individual IED decline of ≥10% from the baseline level were more likely to be KD responders. Therefore, we adopted ≥10% as the IED reduction criteria for this study. A total of three 5-min wakeful EEG samples without artefacts were selected for evaluation, and the mean spike-wave index was calculated as described previously. 2.3. Seizure frequency Seizure frequency was expressed as the percentage change compared with the baseline level after 1, 3 and 6 months of treatment. All types of seizures were included. The outcome was classified as <50%, 50–89%, 90–99%, or seizure free. Treatment response was defined as ≥50% reduction in seizure frequency. A seizure frequency reduction between 90% and 99% was considered to be highly effective therapy. Patients who had a <50% seizure frequency reduction were defined as non-responders. EEG features were examined prior to and at 3 and 6 months after the initiation of KD therapy, and the EEG characteristics and clinical efficacy were evaluated. 2.4. Statistical analysis General characteristics of patients were represented as mean ± standard deviation (SD) with range (minimum to maximum) for continuous data and n (%) for categorical data. Binary logistic regression was applied to identify the association of successful seizure reduction rate (≥50%) at 6 months after treatment with consideration of patients’ characteristics and EEG characteristics. Results are shown as odds ratio (OR) with corresponding 95% confidence intervals (95%CI). Variables with significant association in univariate analysis (p < 0.05) were selected and put in a multivariate logistic regression model. A bar graph was also used to represent the distribution of patients with regard to seizure reduction for given follow-up times after treatment. All statistical assessments were two-tailed and considered significant at p < 0.05. All statistical analyses were carried out with IBM SPSS statistical software version 22 for Windows (IBM Corp., Armonk, NY, USA). 3. Results 3.1. General characteristics of patients A total of 47 patients (30 males, 17 females) with LGS were treated with the KD between May 2011 and May 2015 and monitored at the Children’s Hospital of Fudan University. The mean age at the initiation of the KD was 4.4 years (range: 1–16 years). The time interval of treatment delay from seizure onset to KD therapy ranged from 0.5 to 15 years. The seizure types other than tonic seizures that were most frequently detected were atypical absence, myoclonic, epileptic spasms and atonic. All patients had received
178
Y. Zhang et al. / Epilepsy Research 128 (2016) 176–180
Table 1 Patients’ demographics and baseline characteristics (n = 47). Characteristics
(n = 47)
Age at initiation of ketogenic diet, years Age at seizure onset, years Time delay to ketogenic diet therapy, years
4.4 ± 3.2 (1–16) 1.9 ± 2.0 (0.08–10) 2.5 ± 2.4 (0.5–15)
Aetiology Cerebral dysplasia/atrophy Post-infectious sequelae Hypoxic Other structural Unknown
17 (36.2) 7 (14.9) 2 (4.2) 8 (17.0) 13 (27.7)
Seizure types Tonic Atypical absence Myoclonic Epileptic spasms Atonic Focal GTCS (Generalised tonic–clonic)
47 (100) 35 (74.5) 17 (36) 15 (32) 12 (26) 8 (17) 5 (10.6)
Number of antiepileptic drugs used before diet Two Three Four
18 (38.3) 22 (46.8) 7 (14.9)
Antiepileptic drugs used before diet (only top five listed) Valproate Topiramate Levetiracetam Clonazepam Lamotrigine History of infantile spasms
37 (78.7) 30 (63.8) 27(57.4) 24 (51.1) 6 (12.8) 18 (38)
MRI abnormality Lesional Developmental Normal
17 (36.2) 15 (31.9) 15 (31.9)
EEG characteristics Discharge location at baseline Generalised Multifocal
9 (19.1) 38 (80.9)
Table 2 Patients’ EEG characteristics, long-term outcomes and side effects after treatment. (n = 47). Characteristics
(n = 47)
EEG background after treatment (3 months later) No change Improved
25 (53.2) 22 (46.8)
IEDs change after treatment (3 months later) No change Decreased Duration of continuation of the diet on the last follow-up of this study after the 6 months plan, months
26 (55.3) 21 (44.7) 9.45 ± 6.71 (3–24)
Effects evaluation <50% seizure reduction ≥50% seizure reduction Early termination
19 (40.4) 21 (44.7) 7 (14.9)
Side effects Hyperlipidaemia Gastrointestinal Fatigue Drowsiness
15 (31.9) 8 (17.0) 7 (14.9) 5 (10.6)
Data are represented as mean ± SD (range: min. to max.) for continuous data and n (%) for categorical data.
Data are represented as mean ± SD (range: min. to max.) for continuous data and n (%) for categorical data.
at least two types of AEDs before being given the KD. The most commonly used drugs were valproate, topiramate, levetiracetam and clonazepam. Eighteen patients (38%) had a history of infantile spasms. Thirty-two patients (68.1%) had abnormalities noted on MRI scans; 17 had lesional and 15 developmental abnormalities. All the patients’ characteristics and post-treatment outcomes are summarised in Tables 1 and 2. The long-term treatment duration was recorded as a mean of 9.45 months (range: 3–24 months), and the therapy was effective for 21 patients (44.7%) and not effective for 19 (40.4%). Seven patients (14.9%) had early termination of therapy before the 6-month follow-up after the start of treatment. Among these seven patients, two had successful seizure reduction (≥50%), but the patients refused to continue the therapy. Five patients did not have effective reduction of seizures, and two of these five patients refused to continue the therapy and the other three stopped the KD because the therapy was ineffective. After 3 months of treatment, 22 (46.8%) of the 47 patients had an improvement in EEG background (increased background frequency), and 21 (44.7%) had decreased IEDs. The side effects that were noted were hyperlipidaemia (n = 15), gastrointestinal (n = 8), fatigue (n = 7) and drowsiness (n = 5) (Table 2). 3.2. Clinical efficacy of ketogenic diet therapy The clinical efficacies at 1, 3 and 6 months after the initiation of KD therapy are shown in Fig. 1. Seven patients who left the
Fig 1. The distribution of patients according to the stratification of seizure percent reduction rate at the first, third and sixth month after treatment.
study were not included for the efficacy evaluation at the sixth month after start of treatment. The proportion of patients who were seizure free at the first, third and sixth month was 4.3%, 4.3% and 10%, respectively. The seizure frequency reduction rate was as follows: at 1 month, 63.8% of patients had <50% reduction in seizure frequency, 25.5% had 50% to 89%, and 6.4% of patients had 90% to 99% reduction in seizure frequency; at 3 months, 51.1% of patients had <50% reduction in seizure frequency, 36.2% had 50% to 89%, and 8.5% of patients had 90% to 99% reduction in seizure frequency; and at 6 months, 47.5% of patients with a reduction rate <50%, 30% from 50% to 89%, 12.5% from 90% to 99%. There was no significant difference in the seizure reduction rate among the time points according to the McNemar test (Fig. 1). The aetiologies of LGS in the 47 patients are presented in Table 1. Table 3 shows the number of patients in seizure reduction groups at the sixth month after treatment with regard to aetiology. Table 4 shows that there might be an association between successful seizure reduction (≥50%) and change in improved EEG background (p = 0.007) or IEDs reduction (p = 0.025) after 3 months of treatment. Fig. 2 illustrates the distribution of patients with successful seizure reduction rate at the first, third and sixth month after
Y. Zhang et al. / Epilepsy Research 128 (2016) 176–180
179
Fig. 2. The distribution of patients with successful seizure reduction rate (≥50%) at the first, third and sixth month after treatment with combining the change of EEG background and IED reduction at the third month after treatment. EEG (−) and (+) mean no change and improved EEG background, and IED (−) and (+) mean no change and reduced IEDs. Table 3 Summary of the patients for given seizure reduction at sixth month after treatment according to the aetiology. Seizure reduction at sixth month after treatment Aetiology Cerebral dysplasia/atrophy Post-infectious sequelae Hypoxic Other structural Unknown
Seizure free 2 (15.4%) 0 (0%) 0 (0%) 0 (0%) 2 (15.4%)
≥90% 1 (7.7%) 0 (0%) 0 (0%) 0 (0%) 4 (30.8%)
50%–89% 4 (30.8%) 2 (40%) 0 (0%) 3 (42.9%) 3 (23.1%)
<50% 6 (46.2%) 3 (60%) 2 (100%) 4 (57.1%) 4 (30.8%)
Data were summarised as n (%).
treatment combined with the change in EEG background or IEDs reduction at the third month after treatment. The combined effect also showed that patients with both improved EEG background and reduced IEDs had an improved seizure reduction rate ≥50% compared with patients without improvement in EEG background and IEDs (OR = 98.0, 95%CI = 9.75–1208.5, p < 0.001) (results not shown in tables). 4. Discussion Our results suggest that the KD is effective for the treatment of LGS. We found that 52% of the children had ≥50% seizure reduction 6 months after KD therapy compared to the baseline level, and four children achieved complete freedom from seizures. Most of
the responders to the diet exhibited a marked reduction in seizure frequency during the first 3 months of KD therapy. Our results also suggest that improved EEG background change and reduction in IEDs are associated with a better seizure reduction rate, and therefore the EEG might be useful as a predictor of response. Our findings on clinical efficacy of the KD are similar to those of two previous studies of children with LGS (Caraballo et al., 2014; Lemmon et al., 2012). The KD has also been shown to be effective in children with various types of epileptic conditions. In a large study of 315 children with epilepsy resistant to drug therapy, Hallböök et al. found that based on intention-to-treat analysis >50% seizure reduction at 6, 12 and 24 months was achieved in 50%, 46% and 28% of the children, respectively (Hallböök et al., 2015). In another large study of 216 children with refractory epilepsy who received KD therapy, Caraballo et al. reported that 56.5% of the patients had >75% seizure reduction by the end of the study period (Caraballo et al., 2011). It was observed that LGS was one of three epileptic conditions in which the KD produced the best results. None of the routine clinical variables including gender, age at seizure onset, age at initiation of the KD, or MRI abnormalities have been found to be predictors of the response to KD therapy (Lemmon et al., 2012). It has been reported that the treatment delay in responders was significantly shorter than in non-responders (Dressler et al., 2010), but in our study, no significant difference between responders and non-responders was found. Thirty-eight percent of our patients had a history of infantile spasms, but it was
Table 4 Association of the patients with successful seizure reduction rate (≥50%) at sixth month after treatment considering patients’ characteristics and EEG characteristics. ≥50% seizure reduction (n = 21)
OR (95%CI)
EEG background after treatment (3 months later) 3 (15) Improved No change 16 (80)
17 (85) 4 (20)
11.820 (1.969, 70.955)* Reference
IEDs change after treatment (3 months later) 3 (15.8) Decreased No change 16 (76.2)
16 (84.2) 5 (23.8)
7.986 (1.305, 48.883)* Reference
Characteristics
<50% seizure reduction (n = 19)
Data are represented as n (%) for patients with and without poor seizure reduction rate (<50% and ≥50%) at sixth month after treatment for given patients’ characteristics and EEG characteristics. Binary logistic regression was applied to identify the association of successful seizure reduction rate (≥50%) at sixth month after treatment considering patients’ characteristics and EEG characteristics. Variables with significant association in univariate analysis (p < 0.05) were selected and put in multivariate logistic regression model. Results are shown as OR with corresponding 95%CI in multivariate logistic regression model. * p < 0.05.
180
Y. Zhang et al. / Epilepsy Research 128 (2016) 176–180
not predictive of a response to the KD. In our sample, differences in aetiologies between responders and non-responders were not statistically significant. Few studies have examined EEG characteristics between KD responders and non-responders. A study of 37 patients undergoing KD therapy who underwent routine EEGs prior to, 1 month, and 3 months after starting the diet showed that patients with ≥10% improvement in IED frequency at 1 month were more than six times as likely to be KD responders (Kessler et al., 2011). In our retrospective study, there were significant differences between responders and non-responders with respect to the characteristics of the EEG. For 21 patients, a significant reduction of IEDs after 3 months of treatment was identified. Sixteen of these patients were clinical responders. After 3 months of treatment, the EEG background of 22 patients was improved; 17 of those were responders. Four patients were seizure free, and three of these patients had a normal EEG. Patients who experienced a reduction in IEDs and improvement in EEG background after KD initiation were more likely to be KD responders. The EEG background rhythms of some patients before therapy were difficult to differentiate and were disordered. After therapy, regular rhythms were present. The rhythms of other patients increased in frequency after therapy, which meant that the frequency of the background wave on the EEG increased by at least 1 Hz compared with before therapy. Thus, EEG background and IEDs can be predictors of the KD clinical outcome. The findings in the present study will help in the recognition of early clinical improvement according to EEG characteristics. For infants who showed seizure reduction after 3 months of treatment, the improvement of EEG also supported the therapeutic effect of the KD. For some infants who had no significant clinical improvement after 3 months of therapy, the improvement shown by EEG could encourage KD therapy to be continued. For some patients who had no improvement clinically or on EEG, the long-term efficacy of KD would be predicted to be not good, and therefore other therapies should be recommended. Thus, the EEG examination at 3 months of KD therapy may provide guidance on continuation of diet therapy. Few side effects were seen in our study population. None of the side effects led to cessation of KD therapy. There were some limitations in this study. Data on seizure frequency was based only on seizures diaries. The EEG reviewers were not blinded to case disposition. The diagnosis of LGS lacked specificity and therefore a mixture of aetiologies was represented in the study group. There was no comparison or control group and no randomisation. Also, data on ketone levels were not available for comparison between responders and non-responders. Of course, more studies, especially multicenter studies with large sample size, are required to confirm our findings because the sample size was small in this study. 5. Conclusion In general, the KD is an effective and safe treatment for children
with LGS. It can not only effectively control clinical seizures, but it can also improve abnormal EEG background activity and reduce IEDs. Early improvement in the EEG background and a reduction in IEDs may be predictors of a patient’s response to diet. Conflict of interest statement None of the authors has any conflicts of interest. References Al-Banji, M.H., Zahr, D.K., Jan, M.M., 2015. Lennox-Gastaut syndrome. Management update. Neurosciences (Riyadh) 20, 207–212. Camfield, P.R., 2011. Definition and natural history of Lennox-Gastaut syndrome. Epilepsia 52 (Suppl. 5), 3–9. Caraballo, R., Vaccarezza, M., Cersosimo, R., Rios, V., Soraru, A., Arroyo, H., Agosta, G., Escobal, N., Demartini, M., Maxit, C., Cresta, A., Marchione, D., Carniello, M., Panico, L., 2011. Long-term follow-up of the ketogenic diet for refractory epilepsy: multicenter Argentinean experience in 216 pediatric patients. Seizure 20, 640–645. Caraballo, R.H., Fortini, S., Fresler, S., Armeno, M., Ariela, A., Cresta, A., Mestre, G., Escobal, N., 2014. Ketogenic diet in patients with Lennox-Gastaut syndrome. Seizure 23, 751–755. Dressler, A., Stocklin, B., Reithofer, E., Benninger, F., Freilinger, M., Hauser, E., Reiter-Fink, E., Seidl, R., Trimmel-Schwahofer, P., Feucht, M., 2010. Long-term outcome and tolerability of the ketogenic diet in drug-resistant childhood epilepsy—the Austrian experience. Seizure 19, 404–408. Hallböök, T., Sjolander, A., Amark, P., Miranda, M., Bjurulf, B., Dahlin, M., 2015. Effectiveness of the ketogenic diet used to treat resistant childhood epilepsy in Scandinavia. Eur. J. Paediatr. Neurol. 19, 29–36. Henderson, C.B., Filloux, F.M., Alder, S.C., Lyon, J.L., Caplin, D.A., 2006. Efficacy of the ketogenic diet as a treatment option for epilepsy: meta-analysis. J. Child Neurol. 21, 193–198. Kessler, S.K., Gallagher, P.R., Shellhaas, R.A., Clancy, R.R., Bergqvist, A.G., 2011. Early EEG improvement after ketogenic diet initiation. Epilepsy Res. 94, 94–101. Kim, H.J., Kim, H.D., Lee, J.S., Heo, K., Kim, D.S., Kang, H.C., 2015. Long-term prognosis of patients with Lennox–Gastaut syndrome in recent decades. Epilepsy Res. 110, 10–19. Kossoff, E.H., Zupec-Kania, B.A., Amark, P.E., Ballaban-Gil, K.R., Christina Bergqvist, A.G., Blackford, R., Buchhalter, J.R., Caraballo, R.H., Helen Cross, J., Dahlin, M.G., Donner, E.J., Klepper, J., Jehle, R.S., Kim, H.D., Christiana Liu, Y.M., Nation, J., Nordli Jr., D.R., Pfeifer, H.H., Rho, J.M., Stafstrom, C.E., Thiele, E.A., Turner, Z., Wirrell, E.C., Wheless, J.W., Veggiotti, P., Vining, E.P., 2009. Optimal clinical management of children receiving the ketogenic diet: recommendations of the International Ketogenic Diet Study Group. Epilepsia 50, 304–317. Lemmon, M.E., Terao, N.N., Ng, Y.T., Reisig, W., Rubenstein, J.E., Kossoff, E.H., 2012. Efficacy of the ketogenic diet in Lennox-Gastaut syndrome: a retrospective review of one institution’s experience and summary of the literature. Dev. Med. Child Neurol. 54, 464–468. Rantala, H., Putkonen, T., 1999. Occurrence, outcome, and prognostic factors of infantile spasms and Lennox-Gastaut syndrome. Epilepsia 40, 286–289. Trevathan, E., 2002. Infantile spasms and Lennox-Gastaut syndrome. J. Child Neurol. 17 (Suppl. 2), 2S9–2S22. Winesett, S.P., Bessone, S.K., Kossoff, E.H., 2015. The ketogenic diet in pharmacoresistant childhood epilepsy. Expert Rev. Neurother. 15, 621–628. You, S.J., Kim, H.D., Kang, H.C., 2009. Factors influencing the evolution of West syndrome to Lennox-Gastaut syndrome. Pediatr. Neurol. 41, 111–113. van Rijckevorsel, K., 2008. Treatment of Lennox-Gastaut syndrome: overview and recent findings. Neuropsychiatr. Dis. Treat. 4, 1001–1019.
Contents lists available at www.sciencedirect.com
Epilepsy Research journal homepage: www.elsevier.com/locate/epilepsyres
Therapeutic effects of the ketogenic diet in children with Lennox-Gastaut syndrome Yunjian Zhang 1 , Yi Wang 1 , Yuanfeng Zhou, Linmei Zhang, Lifei Yu, Shuizhen Zhou ∗ Department of Pediatric Neurology, Children’s Hospital of Fudan University, China
a r t i c l e
i n f o
Article history: Received 21 May 2016 Received in revised form 18 October 2016 Accepted 6 November 2016 Available online 9 November 2016 Keywords: Ketogenic diet Lennox-Gastaut syndrome EEG Children
a b s t r a c t Objective: The aim of this study was to evaluate the efficacy of the ketogenic diet (KD) on the clinical and electroencephalographic (EEG) features of Lennox-Gastaut syndrome (LGS) and explore the relationships between EEG changes and clinical efficacy. Methods: We retrospectively studied 47 patients with LGS who accepted KD therapy between May 2011 and May 2015. Clinical efficacy and EEG features such as background activity, abnormal interictal epileptic discharges (IEDs) and the discharge location were evaluated prior to and at 3 and 6 months after therapy. Responders were defined as ≥50% seizure reduction. Results: At 3 months of treatment, 23 patients (48.9%) had ≥50% seizure reduction. Seven patients (14.9%) discontinued treatment between 3 and 6 months because of lack of efficacy or inability to adhere to the diet. At 6 months of treatment, 4 patients (10%) were seizure free, 5 (12.5%) had ≥90% seizure reduction, 12 (30%) had a reduction of 50–89%, and 19 (47.5%) had <50% reduction. Patients with improved EEG background and reduced IEDs had an improved seizure reduction rate compared with patients without change in EEG background or IEDs (p < 0.01). Conclusions: The results show that the KD is effective in LGS. It can control seizures and improve EEG abnormalities. Early improvement in the EEG background and a reduction in IEDs may be predictors of a patient’s response to diet. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Lennox–Gastaut syndrome (LGS) is a severe childhood epileptic encephalopathy characterised by multiple seizure types, intellectual disability and the presence of diffuse fast activity and slow (<3 Hz) spike-wave complexes on electroencephalogram (EEG) (AlBanji et al., 2015). A literature review of LGS found that because a precise definition of this syndrome is lacking, the literature contains contradictory data on many aspects (van Rijckevorsel, 2008). It is a rare disease, with an estimated incidence of 0.28 per 1000 live births (Rantala and Putkonen, 1999). The prevalence may be higher because of its refractory characteristics, and it perhaps accounts for about 1–10% of all cases of epilepsy in children (Camfield, 2011). The aetiology of LGS is extensive and diverse. About 30% of LGS patients are cryptogenic. Perinatal brain damage, central nervous system infection sequelae, and malformations of cortical development are some of the most common causes, and probably account for 70% of all cases (van Rijckevorsel, 2008). There is only lim-
∗ Corresponding author at: 399 Wanyuan Road, Minhang District, Shanghai, China. E-mail address: [email protected] (S. Zhou). 1 Co-first author. http://dx.doi.org/10.1016/j.eplepsyres.2016.11.003 0920-1211/© 2016 Elsevier B.V. All rights reserved.
ited evidence for a genetic component. Half of West syndrome cases and other infantile epileptic encephalopathies will progress to LGS, and 20–30% of LGS cases develop from West syndrome or other epileptic encephalopathies (Trevathan, 2002; You et al., 2009). Intractable seizures of LGS include tonic, atypical absence, atonic and myoclonic seizures. All seizure types are resistant to antiepileptic drugs (AEDs) (van Rijckevorsel, 2008). An alternative therapeutic option is the ketogenic diet (KD), which may be useful, and according to some authors should be used as early as possible (Caraballo et al., 2014; Kim et al., 2015). The KD is a high-fat, low carbohydrate diet with 90% of calories derived from fat. It has been used in the treatment of intractable childhood epilepsy since the 1920s (Winesett et al., 2015). The KD has been shown to be beneficial for the treatment of drug-resistant epilepsy. It has been reported that more than approximately 50% of intractable epileptic children had at least a ≥50% seizure reduction after KD therapy (Henderson et al., 2006). Some patients achieved >90% reduction (Hallböök et al., 2015). It has been observed that the efficacy of the KD differs among different seizure types or epileptic syndromes (Caraballo et al., 2011). According to the International Ketogenic Diet Study Group recommendations, the KD may be particularly effective in Dravet syndrome, Doose syndrome, and some hereditary epilepsies, with
Y. Zhang et al. / Epilepsy Research 128 (2016) 176–180
larger than expected rates of seizure freedom in these patients (Kossoff et al., 2009). However, there is a paucity of published data on the efficacy of the KD in LGS. In a retrospective study, the records of 71 children with LGS who were treated with the KD were reviewed (Lemmon et al., 2012). Intent-to-treat analysis was used, and at the end of 6 months there were 36 (51%) children who had >50% reduction in seizures and 16 (23%) who had >90% reduction in seizures. One child was seizure free. In a prospective study, 20 children with LGS received KD therapy (Caraballo et al., 2014). Fifteen of the children continued to receive the KD for at least 18 months. Eight of these 15 children had a ≥50% decrease in seizures, including three who became seizure free. Although a number of published studies have attempted to identify predictors of the response to the KD, the studies often used a group of medically refractory epilepsy patients (including patients with LGS), and therefore the results obtained may not be specific to LGS. In particular, the relationship between EEG features and the clinical efficacy of the KD is unknown in patients with LGS. Therefore, the purpose of this study was to examine the effects of the KD on clinical and EEG features in children with LGS and to evaluate the correlation between them. 2. Methods 2.1. Participants A total of 47 LGS patients who experienced daily or weekly seizures despite therapy with at least two appropriate AEDs were enrolled in our study and received KD therapy between May 2011 and May 2015. To be enrolled, patients had to meet the following criteria: (1) multiple seizure types, delay of intelligence development, and an abnormal EEG pattern of fast activity and slow spike-wave complexes; (2) have tonic seizures during sleep; (3) intractable epilepsy defined as two or more drugs used without achieving control; and (4) signed informed consent provided by the parents or legal guardians. Patients were excluded if they were not suitable for the KD; had any deficiencies of the transportation or oxidation of fatty acids or ketobodies; had deficiency of pyruvate carboxylase; had a mitochondrial disease; or their parents did not adhere to providing the diet. A paediatric epileptologist and a dietician managed all patients on the KD. The patients were started on the classic 4:1 ratio KD (fat:protein plus carbohydrate). KD therapy was given for 6 months. All patients continued to take the same drugs they were taking before KD therapy during the 6-month period; no changes in drug prescription were permitted. The patients were given potassium citrate supplementation to prevent hypokalemia or lithiasis. Calcium and vitamin D were also given. Patients stayed in the hospital for 1 week and were closely monitored for efficacy and potential side effects. Seizure frequency was documented daily using seizure diaries kept by the families, which were reviewed at each clinic visit. After discharge, the children were reviewed as outpatients monthly for an initial 3 months and then followed up once every 3 months. Adherence to the KD was evaluated during these visits. The efficacy, side effects, laboratory tests and video EEG recordings were assessed at regular intervals. 2.2. MRI and EEG All patients underwent magnetic resonance imaging (MRI). Imaging abnormalities were classified as lesional (e.g., tuberous sclerosis, subdural effusion, hippocampal atrophy) or other (e.g., atrophy, developmental abnormalities). Electrographic patterns were measured through digital EEG monitoring. EEG was performed when patients were both awake
177
and asleep. Electrocardiography (ECG) and electromyography (EMG) were performed during the awake, drowsy and sleep states. The results from each patient were reviewed by two of the authors; if no consensus was reached, a third author reviewed the results and made the final decision. Hyperventilation and photic stimulation were performed in patients who were cooperative. The EEG markers such as background activity (rhythm, frequency) (Dressler et al., 2010), abnormal interictal epileptic discharges (IEDs), and discharge location were recorded. Improvement of EEG background included an increase of wave frequencies in the occipital region or amelioration of disorganisation. More specifically, after therapy, regular rhythm was present or the rhythm was increased in frequency, that is, the frequency of the background wave on the EEG increased by at least 1 Hz compared with before therapy. The spike-wave index of IEDs was expressed as the number of seconds containing spikes or sharp waves divided by the total number of seconds times 100. Kessler et al. (2011) reported that those patients with an individual IED decline of ≥10% from the baseline level were more likely to be KD responders. Therefore, we adopted ≥10% as the IED reduction criteria for this study. A total of three 5-min wakeful EEG samples without artefacts were selected for evaluation, and the mean spike-wave index was calculated as described previously. 2.3. Seizure frequency Seizure frequency was expressed as the percentage change compared with the baseline level after 1, 3 and 6 months of treatment. All types of seizures were included. The outcome was classified as <50%, 50–89%, 90–99%, or seizure free. Treatment response was defined as ≥50% reduction in seizure frequency. A seizure frequency reduction between 90% and 99% was considered to be highly effective therapy. Patients who had a <50% seizure frequency reduction were defined as non-responders. EEG features were examined prior to and at 3 and 6 months after the initiation of KD therapy, and the EEG characteristics and clinical efficacy were evaluated. 2.4. Statistical analysis General characteristics of patients were represented as mean ± standard deviation (SD) with range (minimum to maximum) for continuous data and n (%) for categorical data. Binary logistic regression was applied to identify the association of successful seizure reduction rate (≥50%) at 6 months after treatment with consideration of patients’ characteristics and EEG characteristics. Results are shown as odds ratio (OR) with corresponding 95% confidence intervals (95%CI). Variables with significant association in univariate analysis (p < 0.05) were selected and put in a multivariate logistic regression model. A bar graph was also used to represent the distribution of patients with regard to seizure reduction for given follow-up times after treatment. All statistical assessments were two-tailed and considered significant at p < 0.05. All statistical analyses were carried out with IBM SPSS statistical software version 22 for Windows (IBM Corp., Armonk, NY, USA). 3. Results 3.1. General characteristics of patients A total of 47 patients (30 males, 17 females) with LGS were treated with the KD between May 2011 and May 2015 and monitored at the Children’s Hospital of Fudan University. The mean age at the initiation of the KD was 4.4 years (range: 1–16 years). The time interval of treatment delay from seizure onset to KD therapy ranged from 0.5 to 15 years. The seizure types other than tonic seizures that were most frequently detected were atypical absence, myoclonic, epileptic spasms and atonic. All patients had received
178
Y. Zhang et al. / Epilepsy Research 128 (2016) 176–180
Table 1 Patients’ demographics and baseline characteristics (n = 47). Characteristics
(n = 47)
Age at initiation of ketogenic diet, years Age at seizure onset, years Time delay to ketogenic diet therapy, years
4.4 ± 3.2 (1–16) 1.9 ± 2.0 (0.08–10) 2.5 ± 2.4 (0.5–15)
Aetiology Cerebral dysplasia/atrophy Post-infectious sequelae Hypoxic Other structural Unknown
17 (36.2) 7 (14.9) 2 (4.2) 8 (17.0) 13 (27.7)
Seizure types Tonic Atypical absence Myoclonic Epileptic spasms Atonic Focal GTCS (Generalised tonic–clonic)
47 (100) 35 (74.5) 17 (36) 15 (32) 12 (26) 8 (17) 5 (10.6)
Number of antiepileptic drugs used before diet Two Three Four
18 (38.3) 22 (46.8) 7 (14.9)
Antiepileptic drugs used before diet (only top five listed) Valproate Topiramate Levetiracetam Clonazepam Lamotrigine History of infantile spasms
37 (78.7) 30 (63.8) 27(57.4) 24 (51.1) 6 (12.8) 18 (38)
MRI abnormality Lesional Developmental Normal
17 (36.2) 15 (31.9) 15 (31.9)
EEG characteristics Discharge location at baseline Generalised Multifocal
9 (19.1) 38 (80.9)
Table 2 Patients’ EEG characteristics, long-term outcomes and side effects after treatment. (n = 47). Characteristics
(n = 47)
EEG background after treatment (3 months later) No change Improved
25 (53.2) 22 (46.8)
IEDs change after treatment (3 months later) No change Decreased Duration of continuation of the diet on the last follow-up of this study after the 6 months plan, months
26 (55.3) 21 (44.7) 9.45 ± 6.71 (3–24)
Effects evaluation <50% seizure reduction ≥50% seizure reduction Early termination
19 (40.4) 21 (44.7) 7 (14.9)
Side effects Hyperlipidaemia Gastrointestinal Fatigue Drowsiness
15 (31.9) 8 (17.0) 7 (14.9) 5 (10.6)
Data are represented as mean ± SD (range: min. to max.) for continuous data and n (%) for categorical data.
Data are represented as mean ± SD (range: min. to max.) for continuous data and n (%) for categorical data.
at least two types of AEDs before being given the KD. The most commonly used drugs were valproate, topiramate, levetiracetam and clonazepam. Eighteen patients (38%) had a history of infantile spasms. Thirty-two patients (68.1%) had abnormalities noted on MRI scans; 17 had lesional and 15 developmental abnormalities. All the patients’ characteristics and post-treatment outcomes are summarised in Tables 1 and 2. The long-term treatment duration was recorded as a mean of 9.45 months (range: 3–24 months), and the therapy was effective for 21 patients (44.7%) and not effective for 19 (40.4%). Seven patients (14.9%) had early termination of therapy before the 6-month follow-up after the start of treatment. Among these seven patients, two had successful seizure reduction (≥50%), but the patients refused to continue the therapy. Five patients did not have effective reduction of seizures, and two of these five patients refused to continue the therapy and the other three stopped the KD because the therapy was ineffective. After 3 months of treatment, 22 (46.8%) of the 47 patients had an improvement in EEG background (increased background frequency), and 21 (44.7%) had decreased IEDs. The side effects that were noted were hyperlipidaemia (n = 15), gastrointestinal (n = 8), fatigue (n = 7) and drowsiness (n = 5) (Table 2). 3.2. Clinical efficacy of ketogenic diet therapy The clinical efficacies at 1, 3 and 6 months after the initiation of KD therapy are shown in Fig. 1. Seven patients who left the
Fig 1. The distribution of patients according to the stratification of seizure percent reduction rate at the first, third and sixth month after treatment.
study were not included for the efficacy evaluation at the sixth month after start of treatment. The proportion of patients who were seizure free at the first, third and sixth month was 4.3%, 4.3% and 10%, respectively. The seizure frequency reduction rate was as follows: at 1 month, 63.8% of patients had <50% reduction in seizure frequency, 25.5% had 50% to 89%, and 6.4% of patients had 90% to 99% reduction in seizure frequency; at 3 months, 51.1% of patients had <50% reduction in seizure frequency, 36.2% had 50% to 89%, and 8.5% of patients had 90% to 99% reduction in seizure frequency; and at 6 months, 47.5% of patients with a reduction rate <50%, 30% from 50% to 89%, 12.5% from 90% to 99%. There was no significant difference in the seizure reduction rate among the time points according to the McNemar test (Fig. 1). The aetiologies of LGS in the 47 patients are presented in Table 1. Table 3 shows the number of patients in seizure reduction groups at the sixth month after treatment with regard to aetiology. Table 4 shows that there might be an association between successful seizure reduction (≥50%) and change in improved EEG background (p = 0.007) or IEDs reduction (p = 0.025) after 3 months of treatment. Fig. 2 illustrates the distribution of patients with successful seizure reduction rate at the first, third and sixth month after
Y. Zhang et al. / Epilepsy Research 128 (2016) 176–180
179
Fig. 2. The distribution of patients with successful seizure reduction rate (≥50%) at the first, third and sixth month after treatment with combining the change of EEG background and IED reduction at the third month after treatment. EEG (−) and (+) mean no change and improved EEG background, and IED (−) and (+) mean no change and reduced IEDs. Table 3 Summary of the patients for given seizure reduction at sixth month after treatment according to the aetiology. Seizure reduction at sixth month after treatment Aetiology Cerebral dysplasia/atrophy Post-infectious sequelae Hypoxic Other structural Unknown
Seizure free 2 (15.4%) 0 (0%) 0 (0%) 0 (0%) 2 (15.4%)
≥90% 1 (7.7%) 0 (0%) 0 (0%) 0 (0%) 4 (30.8%)
50%–89% 4 (30.8%) 2 (40%) 0 (0%) 3 (42.9%) 3 (23.1%)
<50% 6 (46.2%) 3 (60%) 2 (100%) 4 (57.1%) 4 (30.8%)
Data were summarised as n (%).
treatment combined with the change in EEG background or IEDs reduction at the third month after treatment. The combined effect also showed that patients with both improved EEG background and reduced IEDs had an improved seizure reduction rate ≥50% compared with patients without improvement in EEG background and IEDs (OR = 98.0, 95%CI = 9.75–1208.5, p < 0.001) (results not shown in tables). 4. Discussion Our results suggest that the KD is effective for the treatment of LGS. We found that 52% of the children had ≥50% seizure reduction 6 months after KD therapy compared to the baseline level, and four children achieved complete freedom from seizures. Most of
the responders to the diet exhibited a marked reduction in seizure frequency during the first 3 months of KD therapy. Our results also suggest that improved EEG background change and reduction in IEDs are associated with a better seizure reduction rate, and therefore the EEG might be useful as a predictor of response. Our findings on clinical efficacy of the KD are similar to those of two previous studies of children with LGS (Caraballo et al., 2014; Lemmon et al., 2012). The KD has also been shown to be effective in children with various types of epileptic conditions. In a large study of 315 children with epilepsy resistant to drug therapy, Hallböök et al. found that based on intention-to-treat analysis >50% seizure reduction at 6, 12 and 24 months was achieved in 50%, 46% and 28% of the children, respectively (Hallböök et al., 2015). In another large study of 216 children with refractory epilepsy who received KD therapy, Caraballo et al. reported that 56.5% of the patients had >75% seizure reduction by the end of the study period (Caraballo et al., 2011). It was observed that LGS was one of three epileptic conditions in which the KD produced the best results. None of the routine clinical variables including gender, age at seizure onset, age at initiation of the KD, or MRI abnormalities have been found to be predictors of the response to KD therapy (Lemmon et al., 2012). It has been reported that the treatment delay in responders was significantly shorter than in non-responders (Dressler et al., 2010), but in our study, no significant difference between responders and non-responders was found. Thirty-eight percent of our patients had a history of infantile spasms, but it was
Table 4 Association of the patients with successful seizure reduction rate (≥50%) at sixth month after treatment considering patients’ characteristics and EEG characteristics. ≥50% seizure reduction (n = 21)
OR (95%CI)
EEG background after treatment (3 months later) 3 (15) Improved No change 16 (80)
17 (85) 4 (20)
11.820 (1.969, 70.955)* Reference
IEDs change after treatment (3 months later) 3 (15.8) Decreased No change 16 (76.2)
16 (84.2) 5 (23.8)
7.986 (1.305, 48.883)* Reference
Characteristics
<50% seizure reduction (n = 19)
Data are represented as n (%) for patients with and without poor seizure reduction rate (<50% and ≥50%) at sixth month after treatment for given patients’ characteristics and EEG characteristics. Binary logistic regression was applied to identify the association of successful seizure reduction rate (≥50%) at sixth month after treatment considering patients’ characteristics and EEG characteristics. Variables with significant association in univariate analysis (p < 0.05) were selected and put in multivariate logistic regression model. Results are shown as OR with corresponding 95%CI in multivariate logistic regression model. * p < 0.05.
180
Y. Zhang et al. / Epilepsy Research 128 (2016) 176–180
not predictive of a response to the KD. In our sample, differences in aetiologies between responders and non-responders were not statistically significant. Few studies have examined EEG characteristics between KD responders and non-responders. A study of 37 patients undergoing KD therapy who underwent routine EEGs prior to, 1 month, and 3 months after starting the diet showed that patients with ≥10% improvement in IED frequency at 1 month were more than six times as likely to be KD responders (Kessler et al., 2011). In our retrospective study, there were significant differences between responders and non-responders with respect to the characteristics of the EEG. For 21 patients, a significant reduction of IEDs after 3 months of treatment was identified. Sixteen of these patients were clinical responders. After 3 months of treatment, the EEG background of 22 patients was improved; 17 of those were responders. Four patients were seizure free, and three of these patients had a normal EEG. Patients who experienced a reduction in IEDs and improvement in EEG background after KD initiation were more likely to be KD responders. The EEG background rhythms of some patients before therapy were difficult to differentiate and were disordered. After therapy, regular rhythms were present. The rhythms of other patients increased in frequency after therapy, which meant that the frequency of the background wave on the EEG increased by at least 1 Hz compared with before therapy. Thus, EEG background and IEDs can be predictors of the KD clinical outcome. The findings in the present study will help in the recognition of early clinical improvement according to EEG characteristics. For infants who showed seizure reduction after 3 months of treatment, the improvement of EEG also supported the therapeutic effect of the KD. For some infants who had no significant clinical improvement after 3 months of therapy, the improvement shown by EEG could encourage KD therapy to be continued. For some patients who had no improvement clinically or on EEG, the long-term efficacy of KD would be predicted to be not good, and therefore other therapies should be recommended. Thus, the EEG examination at 3 months of KD therapy may provide guidance on continuation of diet therapy. Few side effects were seen in our study population. None of the side effects led to cessation of KD therapy. There were some limitations in this study. Data on seizure frequency was based only on seizures diaries. The EEG reviewers were not blinded to case disposition. The diagnosis of LGS lacked specificity and therefore a mixture of aetiologies was represented in the study group. There was no comparison or control group and no randomisation. Also, data on ketone levels were not available for comparison between responders and non-responders. Of course, more studies, especially multicenter studies with large sample size, are required to confirm our findings because the sample size was small in this study. 5. Conclusion In general, the KD is an effective and safe treatment for children
with LGS. It can not only effectively control clinical seizures, but it can also improve abnormal EEG background activity and reduce IEDs. Early improvement in the EEG background and a reduction in IEDs may be predictors of a patient’s response to diet. Conflict of interest statement None of the authors has any conflicts of interest. References Al-Banji, M.H., Zahr, D.K., Jan, M.M., 2015. Lennox-Gastaut syndrome. Management update. Neurosciences (Riyadh) 20, 207–212. Camfield, P.R., 2011. Definition and natural history of Lennox-Gastaut syndrome. Epilepsia 52 (Suppl. 5), 3–9. Caraballo, R., Vaccarezza, M., Cersosimo, R., Rios, V., Soraru, A., Arroyo, H., Agosta, G., Escobal, N., Demartini, M., Maxit, C., Cresta, A., Marchione, D., Carniello, M., Panico, L., 2011. Long-term follow-up of the ketogenic diet for refractory epilepsy: multicenter Argentinean experience in 216 pediatric patients. Seizure 20, 640–645. Caraballo, R.H., Fortini, S., Fresler, S., Armeno, M., Ariela, A., Cresta, A., Mestre, G., Escobal, N., 2014. Ketogenic diet in patients with Lennox-Gastaut syndrome. Seizure 23, 751–755. Dressler, A., Stocklin, B., Reithofer, E., Benninger, F., Freilinger, M., Hauser, E., Reiter-Fink, E., Seidl, R., Trimmel-Schwahofer, P., Feucht, M., 2010. Long-term outcome and tolerability of the ketogenic diet in drug-resistant childhood epilepsy—the Austrian experience. Seizure 19, 404–408. Hallböök, T., Sjolander, A., Amark, P., Miranda, M., Bjurulf, B., Dahlin, M., 2015. Effectiveness of the ketogenic diet used to treat resistant childhood epilepsy in Scandinavia. Eur. J. Paediatr. Neurol. 19, 29–36. Henderson, C.B., Filloux, F.M., Alder, S.C., Lyon, J.L., Caplin, D.A., 2006. Efficacy of the ketogenic diet as a treatment option for epilepsy: meta-analysis. J. Child Neurol. 21, 193–198. Kessler, S.K., Gallagher, P.R., Shellhaas, R.A., Clancy, R.R., Bergqvist, A.G., 2011. Early EEG improvement after ketogenic diet initiation. Epilepsy Res. 94, 94–101. Kim, H.J., Kim, H.D., Lee, J.S., Heo, K., Kim, D.S., Kang, H.C., 2015. Long-term prognosis of patients with Lennox–Gastaut syndrome in recent decades. Epilepsy Res. 110, 10–19. Kossoff, E.H., Zupec-Kania, B.A., Amark, P.E., Ballaban-Gil, K.R., Christina Bergqvist, A.G., Blackford, R., Buchhalter, J.R., Caraballo, R.H., Helen Cross, J., Dahlin, M.G., Donner, E.J., Klepper, J., Jehle, R.S., Kim, H.D., Christiana Liu, Y.M., Nation, J., Nordli Jr., D.R., Pfeifer, H.H., Rho, J.M., Stafstrom, C.E., Thiele, E.A., Turner, Z., Wirrell, E.C., Wheless, J.W., Veggiotti, P., Vining, E.P., 2009. Optimal clinical management of children receiving the ketogenic diet: recommendations of the International Ketogenic Diet Study Group. Epilepsia 50, 304–317. Lemmon, M.E., Terao, N.N., Ng, Y.T., Reisig, W., Rubenstein, J.E., Kossoff, E.H., 2012. Efficacy of the ketogenic diet in Lennox-Gastaut syndrome: a retrospective review of one institution’s experience and summary of the literature. Dev. Med. Child Neurol. 54, 464–468. Rantala, H., Putkonen, T., 1999. Occurrence, outcome, and prognostic factors of infantile spasms and Lennox-Gastaut syndrome. Epilepsia 40, 286–289. Trevathan, E., 2002. Infantile spasms and Lennox-Gastaut syndrome. J. Child Neurol. 17 (Suppl. 2), 2S9–2S22. Winesett, S.P., Bessone, S.K., Kossoff, E.H., 2015. The ketogenic diet in pharmacoresistant childhood epilepsy. Expert Rev. Neurother. 15, 621–628. You, S.J., Kim, H.D., Kang, H.C., 2009. Factors influencing the evolution of West syndrome to Lennox-Gastaut syndrome. Pediatr. Neurol. 41, 111–113. van Rijckevorsel, K., 2008. Treatment of Lennox-Gastaut syndrome: overview and recent findings. Neuropsychiatr. Dis. Treat. 4, 1001–1019.