Abstracts

Rationale: The majority of children receiving the ketogenic diet (KD) have attempted at least two anticonvulsant drugs (ACDs) previously. Many parents choose to start the KD not only for seizure reduction, but also in order to potentially wean ACDs. Although there have been several publications regarding both positive and negative potential interactions between the KD and ACDs, as well as results from ACD reduction on the KD, it is unknown how often complete medication withdrawal occurs on the KD. Additionally, there is little data how long this drug-free period lasts, if there are certain characteristics describing patients who are more likely to achieve drug-free diet (DFD) therapy, and how often children restart ACD after achieving medication freedom. Methods: We reviewed the charts of all children started on the KD at Johns Hopkins Hospital (JHH) and Johns Hopkins All Children’s Hospital (ACH) from January 2011 to April 2018. Children were excluded if they were started on the KD without any ACD (n=9). Patients were defined as achieving DFD status if they started the KD on at least 1 ACD and achieved a period of time (at least 1 week) where they were on the KD alone without any ACD. Information was also obtained regarding patient demographics, time on the KD when DFD status occurred, restart of ACD, and why they wished to achieve drug freedom Results: Over the time period, 232 children (mean age at KD onset: 4.7 years) were evaluated, with a mean of 2.4 ACD at onset (range 1-6) and 21.7 seizures per day (range 0.05-240).  DFD status occurred in 43 (18.5%), of which 32 (74%) remained off ACDs for the remainder of their KD treatment course.  The most common last drug weaned was levetiracetam, followed by valproate and clobazam. During DFD therapy, 63% were seizure free, 28% had 90-99% seizure reduction, and 9% had 50-90% seizure reduction. DFD status was achieved at a mean of 7.5 months after KD onset (range 1.5-21 months), for a mean duration of 22.6 months (range 1-66 months). Eleven children restarted AED after a mean of 7 months (range 1-50 months) and subsequently remained on the KD for a mean of 11.7 months. Children achieving DFD therapy were more likely to be younger (3.8 vs. 4.9 years, p =0.02), have fewer ACDs at KD onset (1.7 vs 2.6, p <0.001), have Glut1 deficiency (12% vs 1%, p = 0.003) or epilepsy with myoclonic-atonic seizures (26% vs 37%, p < 0.001), but were less likely to have Lennox-Gastaut syndrome (5% vs 24%, p = 0.004).   Conclusions: Achieving complete anticonvulsant medication freedom while on the KD occurred in 1 in 5 children in our series, sometimes for extended periods of time.  Younger children on fewer medications, with excellent seizure reduction, and especially those with Glut1 deficiency and epilepsy with myoclonic-atonic seizures, were most likely to become drug-free. Further studies developing a protocolized, algorithm-based method for stopping all drugs while on the KD may be valuable. Funding: Carson Harris Foundation