Efficacy and tolerability of the ketogenic diet versus high-dose adrenocorticotropic hormone for infantile spasms: A single-center parallel-cohort randomized controlled trial.

Objective: To compare the efficacy and safety of the ketogenic diet (KD) with standard adrenocorticotropic hormone (ACTH) treatment in infants with West syndrome.
Methods: In this parallel-cohort (PC) randomized controlled trial (RCT), infants were randomly allocated to KD or high-dose ACTH. Those who could not be randomized were followed in a PC. Primary end point was electroclinical remission at day 28. Secondary end points were time to electroclinical remission, relapse after initial response, seizure freedom at last follow-up, adverse effects, and developmental progress.
Results: One hundred one infants were included: 32 in the RCT (16 KD; 16 ACTH) and 69 in the PC (37 KD; 32 ACTH). Electroclinical remission at day 28 was similar between KD and ACTH (RCT: 62% vs 69%; PC: 41% vs 38%; combined cohort: 47% vs 48%; KD vs ACTH, respectively). In the combined cohort, time to electroclinical remission was similar between both treatments (14 days for KD, 16 days for ACTH). However, relapse rates were 16% (KD) and 43% (ACTH, P = 0.09), and seizure freedom at last follow-up was 40% (KD) and 27% (ACTH, P = 0.18). Adverse effects needing acute medical intervention occurred more often with ACTH (30% with KD, 94% with ACTH, P < 0.001). Age-appropriate psychomotor development and adaptive behavior were similar. Without prior vigabatrin (VGB) treatment, remission at day 28 was 47% (KD) and 80% (ACTH, P = 0.02); relapse rates were 29% (KD) and 56% (ACTH, P = 0.13). Consequently, seizure freedom at last follow-up was similar. In infants with prior VGB, seizure freedom at last follow-up was 48% (KD) and 21% (ACTH, P = 0.05).
Significance: The study is underpowered; therefore, its results should be interpreted with caution. KD is as effective as ACTH in the long term but is better tolerated. Without prior VGB treatment, ACTH remains the first choice to achieve short-term remission. However, with prior VGB, KD was at least as effective as ACTH in the short term and was associated with lower relapse rates in the long term; therefore, it represents an appropriate second-line treatment after VGB.
(Wiley Periodicals, Inc. © 2019 International League Against Epilepsy.)