The Influence of Etiology on the Effectiveness of the Ketogenic Diet in a Cohort of Children with Epileptic Encephalopathy

Objective: To investigate the effectiveness of the ketogenic diet (KD) in epileptic encephalopathies during childhood with regards to the new ILAE criteria of classification and terminology of 2017.

Methods: This study was planned as an explorative descriptive data analysis of a prospective longitudinal database from March 1999 to September 2018, which includes consecutive patients treated with the KD at the Medical University Vienna. The new classification of epilepsy was applied including the new range of etiologic groups: known (structural, genetic, metabolic, infectious, and immune-mediated), as well as unknown causes. The frequencies of patients showing seizure freedom (i.e. 100% seizure reduction) seizure reduction > 50% were analyzed at 3, 6 12 months and last follow-up observation.

Results: 201 patients were included into the cohort. Drop-outs due to compliance occurred in 17 patients (8.5%) during the first weeks of evaluation. In one child the KD was established due to a complex I deficiency without seizures and was excluded from analysis. From the remaining 183 patients, 46% (85) were females. 26% (48%) achieved long term-seizure freedom at last follow-up. Seizure reduction > 50% at last FU on diet occurred in 84 participants (46%). Etiology was known in 70% (129/183); 22% of the participants (n = 28) achieved seizure freedom at last follow-up compared to 37% (n = 20) in unknown etiologies (p =0.042). Seizure reduction > 50% at last FU on diet was observed in 43% participants (n = 55) with known and in 54% participants with unknown (n = 29) etiology. Comparing groups of etiology to unknown etiology (37% seizure free at last FU), 19% (13/67) became seizure free in the structural group (p = 0.040), 30% (13/43) participants became seizure free in the genetic subgroup (n.s.), 8% (1/13) participants became seizure free in the metabolic subgroup became seizure free (p = 0.049) and no one in the infectious subgroup (n.s.), and immune-mediated group. Considering electro-clinical syndromes, seizure freedom at last follow-up was observed in 34% (34/100) of West syndrome, in 7% (1/14) of Dravet syndrome, in 40% (2/5) of MAE, in 18% (4/22) of LGS, in 16% (5/32) of cortical malformations, in 80% (4/5) of Down syndrome, in 33% (1/3) of SCN2A, in 33% (1/3) of Glut- 1 deficiency) and in 0% of TSC (0/4).

Significance: Unknown etiologies showed a higher proportion of seizure free infants, especially compared to the structural and metabolic subgroup. Electro-clinical syndromes and etiologies responding best to the KD included Westsyndrome, Dravetsyndrome, MAE, and LGS. Glut-1-deficiency, SCN2A, TSC, and cortical malformations responded less frequently as described in literature, as well as CDKL5 mutations and Ohtahara- Syndrome.

No conflict of interest has been declared by the author(s).