Abstracts

Rationale: Several recent studies suggest that the ketogenic diet is highly effective in the treatment of infantile spasms. However, none of these studies used complete resolution of both spasms and hypsarrhythmia as the measure of efficacy, despite a recent consensus statement recommending 24-hour video-EEG to verify clinical response (to any treatment) and rule out the persistence of subtle spasms or hypsarrhythmia (Pellock et al, 2010). As these efficacy reports were at odds with our impressions from clinical practice, we sought to systematically describe the UCLA experience with the ketogenic diet for the treatment of infantile spasms. Methods: We retrospectively identified patients with video-EEG confirmed infantile spasms who were treated with the ketogenic diet at Mattel Children's Hospital UCLA, with diet initiation between April, 2010 and April, 2013. Relevant clinical and demographic factors were abstracted from medical records, including video-EEG confirmation of spasms and hypsarrhythmia (when present) before and during diet therapy. Results: During the study period, 19 patients with infantile spasms were treated with the ketogenic diet. 11 patients were female and the median age of onset of spasms was 5.4 mo (interquartile range 3.0-7.7). Etiology of spasms was identified in 14 cases, and the remaining 5 patients exhibited developmental delay prior to spasms onset; there were no classically "cryptogenic" patients. The median number of drug failures was 7 (6-7), 16 patients each had failed hormonal therapy (prednisolone and/or ACTH) or vigabatrin, and 14 patients failed both hormonal therapy and vigabatrin. Median age of diet initiation was 35.2 mo (25.8-47.2) and median delay from onset of spasms to diet initiation was 27.1 mo (18.4-39.6). The median peak fat:carbohydrate+protein ratio was 3.5 (3.0-3.9), and median duration of diet therapy was 10.1 mo (7.2-23.0). None of the patients became spasm-free during ketogenic diet therapy, except for one patient who responded immediately after placement of a vagal nerve stimulator, and another patient who became spasm-free 2 days after initiation of therapy with vigabatrin. Median duration of follow-up after diet initiation was 25.6 mo (15.6-32.4). Two patients have died, and 7 patients continue on the diet having achieved at least subjective reductions in either the frequency of spasms or other comorbid seizures. The diet was generally well-tolerated in this population. Conclusions: Among our highly refractory infantile spasms cohort, the ketogenic diet exhibited only modest efficacy, with no patient having achieved a complete response to therapy. Our lack of efficacy likely reflects both the rigorous criteria for response as well as the refractory character of our patient population. A randomized trial with blinded outcome assessment (including video-EEG) is of great need to determine whether or not the ketogenic is effective in quickly yielding clinical and electrographic response. Until then, there is insufficient data to guide clinicians as to when and if the ketogenic diet should be recommended to patients with infantile spasms.