Abstracts

Rationale: Ketogenic diets and responsive neurostimulation (RNS) are both nonpharmacologic treatments for patients with drug-resistant epilepsy who are not candidates for resective surgery. The classic ketogenic diet and the modified Atkins diet (MAD) are high fat, low carbohydrate diets which are increasingly used in patients with epilepsy. The RNS device is implanted in the skull and connected to one or two intracranial depth and/or strip leads targeting a seizure focus or foci. The device continuously monitors electrocorticographic (ECoG) signals and delivers stimulation when epileptiform patterns are detected (RNS detections). While many patients experience seizure reduction with RNS, those who continue to have seizures provide an opportunity to assess the response to new anti-seizure treatments using longitudinal continuous ECoG data collected from RNS devices. This case demonstrates the use of RNS continuous ECoG data to examine the efficacy of a ketogenic diet. Methods: A patient with a history of seizures since age 14 years with intractable focal left frontal seizures and two prior surgical resections underwent RNS device implantation in 7/2006 during enrollment in a clinical trial. Due to frequent seizures despite the RNS, she started a 20 gram per day carbohydrate limit MAD in 3/2013. The patient kept a calendar of seizures and ketones (measured using urine dipsticks). The RNS device was explanted in 1/2014. There were no RNS parameter changes in the 8 months prior or 9 months after starting. Seizure counts and RNS detections (total detections and long events) were compared before and after MAD initiation and these were also correlated with ketone levels. Results: The patient was in sustained ketosis (?- 40 mg/dL urine acetoacetate) within 2 days of starting MAD. She reported a subjective reduction in seizures after starting MAD in her follow-up clinic visit interview, however, there was no clear reduction in seizure counts on her calendars. Although it is possible that seizures were shorter or less intense, there was also no clear or sustained reduction in RNS total daily detections or long events after starting the MAD compared to baseline. In addition, degree of ketosis did not correlate with RNS detections, including during a timeframe when ketosis was lost due to diet noncompliance. Conclusions: In this case, there was not a clear reduction in RNS detections after starting MAD despite subjective report of seizure reduction. Although this patient did not benefit from adjunctive ketogenic diet therapy, there is potential for synergy between RNS and diet similar to vagus nerve stimulation and diet. Valuable data could be obtained from patients in this unique situation, including quantitative measures of diet therapy response and whether correlations exist between ketone levels and RNS detections. In addition, comparisons can be made between objective ECoG data and subjective patient reports of seizure frequency. This case serves as a proof of concept for information that can be collected and analyses that can be performed using the longitudinal ECoG data in RNS patients on a new anti-seizure treatment such as a ketogenic diet. Funding: N/A