Abstracts

Rationale: Prior uncontrolled studies have reported seizure reductions following Deep Brain Stimulation (DBS) in Lennox-Gastaut syndrome (LGS), but data from randomised controlled studies are lacking. We aimed to assess the safety and efficacy of DBS to the centromedian nucleus of thalamus (CM) for treatment of LGS.

Methods: We performed a prospective, double-blind, randomised study of continuous, cycling stimulation of CM-DBS, in patients with LGS. All participants had tonic seizures, generalised paroxysmal fast activity and slow spike-wave on EEG. Following 3-month pre- and post-implantation periods, half received 3-months stimulation (blinded phase), then all received 3-months stimulation (unblinded phase). The primary outcome was the proportion of participants with a ≥50% reduction in seizures recorded in a diary (stimulated versus control groups), measured at the end of the blinded phase. Secondary outcomes were: proportion of participants with a ≥50% reduction in electrographic seizures on 24-hour ambulatory scalp EEG at key timepoints, clinical seizure frequency at study exit relative to baseline, burden of scalp EEG interictal generalised paroxysmal fast activity pre- and post-stimulation, and changes in cognitive assessments (GADS, ABAS-III). Safety was assessed throughout.

Results: Between Nov 2017 and Dec 2019, 20 young adults with LGS (aged 17-37 years; 13 females) underwent bilateral CM-DBS at a single centre in Australia, with 19 randomised (treatment, n=10; control, n=9; one removed due to infection). 50% (5/10) of the stimulation group achieved a ≥50% reduction in seizures, compared with 22% (2/9) of controls (OR 3.1; 95% CI 0.44-21.45; p=0.25). The stimulation group showed a significantly greater reduction in absolute seizure counts (p=0.025).

For electrographic seizures, 59% assigned early stimulation had ≥50% reduction compared with none of the controls (OR 23.25; 95% CI 1.0 to 538.4; p=0.05). Median relative change in electrographic seizures was different between treatment and control groups (p=0.006), but not for diary-recorded seizures (p=0.124) (Figure 1). Across all patients (early and delayed stimulation), median (IQR) reduction in seizures (baseline vs study exit) was 46.7% (38.98) (Figure 2). There were no observable changes in disability (GADS score), assessed after the blinded and unblinded periods. After 3-months of stimulation, functional ability (median ABAS-III score) improved by 0.2%, but this was not significant (p=0.75). Serious adverse events occurred in seven (35%) participants over the course of the study. Adverse events in the blinded phase were reported by three participants only.

Conclusions: Bilateral CM-DBS in LGS patients reduces electrographic seizures after 3 months of stimulation. Seizure diaries showed a non-significant reduction in seizures, potentially reflecting the known unreliability of this seizure counting method. Overall, median reduction in clinical seizures was 46.7% at study exit. No adverse effects on cognition were seen.

Funding: Please list any funding that was received in support of this abstract.: NHMRC project grant #110881.