Efficacy and sleep impact of perampanel versus levetiracetam monotherapy for SeLECTS: A pilot prospective cohort...

Self-limited epilepsy with centrotemporal spikes (SeLECTS) is one of the most common childhood epilepsy syndromes. Its epileptiform discharges are closely linked to sleep, making sleep evaluation particularly important in its management. This study compared the effects of levetiracetam (LEV) and perampanel (PER) on seizure control and sleep outcomes to provide evidence for individualized treatment strategies in children with SeLECTS.

Thirty-five children with SeLECTS (aged 6-12 years) receiving monotherapy (PER, n = 19; LEV, n = 16) were enrolled and followed for at least 3 months. Clinical characteristics, seizure frequency, and adverse drug reactions were recorded. Electroencephalography (EEG) assessed epileptiform discharges and the spike-wave index (SWI). The Children's Sleep Habits Questionnaire (CSHQ) and the Epworth Sleepiness Scale for Children and Adolescents (ESS-CHAD) evaluated subjective sleep quality and daytime sleepiness, respectively. Polysomnography (PSG) measured objective sleep parameters: sleep efficiency (SE), sleep architecture, sleep latency (SL), wake time after sleep onset (WASO), and number of awakenings (NAW). The Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) and the Quality of Life in Children with Epilepsy Questionnaire-16 (QOLCE-16) assessed intelligence development and quality of life, respectively.

The median follow-up duration was 144.5 days (96.0, 209.5) in LEV group and 128.0 days (102.0, 196.0) in PER group, with no statistically significant difference between the two groups (P > 0.05). The baseline seizure frequency in LEV group and PER group were 0.11 (0.11, 0.39) and 0.22 (0.11, 0.33) per month, respectively. After treatment, it was 0.00 (0.00, 0.00) in LEV group and 0.00 (0.00, 0.11) in PER group at follow-up. All patients exhibited abnormal EEG findings at baseline, while 2 cases in the LEV group and 4 cases in the PER group showed normalization at follow-up. The PER group had significantly higher baseline SWI grades than the LEV group (P < 0.05). However, no significant between group differences were observed at follow-up (P > 0.05). The objective sleep parameters at follow-up: the PER group demonstrated a significantly lower proportion of N1 sleep (LEV vs. PER: 12.03 ± 2.75% vs. 10.15 ± 2.43%, P = 0.039) and a greater reduction in the number of NAW (LEV vs. PER: 1.13 ± 5.23 vs. - 2.68 ± 5.03, P = 0.036) than LEV group. There were no significant differences between the two groups in CSHQ, ESS-CHAD, or QOLCE-16 scores at baseline or during follow-up (P > 0.05).

Both LEV and PER monotherapy were tolerated and associated with short-term seizure control in children with SeLECTS. There are some indications that PER may offer small additional benefits for the sleep architecture, resulting in a decrease in the proportion of N1 and NAW. However, prospective studies with longer follow-up are needed to confirm the observations.