Closed-loop neurostimulation for epilepsy leads to improved outcomes when stimulation episodes are delivered during periods with less epileptiform activity

Abstract

In patients with drug-resistant epilepsy, electrical stimulation of the brain in response to epileptiform activity can make seizures less frequent and debilitating. When effective, this therapy, known as closed-loop responsive neurostimulation (RNS), produces long-lasting changes in brain dynamics that correlate with clinical outcomes. Since periods with frequent epileptiform activity are less conducive to neuroplasticity, we hypothesize that stimulation timing, specifically stimulation during brain states with less epileptiform activity, is critical in driving long-term changes that restore healthy brain networks. To test this, we quantified stimulation episodes during low- and high-risk epochs--that is, stimulation during periods with a low or high risk of generating seizures and less or more epileptiform activity--in a cohort of 40 patients treated with RNS. Patients were categorized into three groups: super responders (>90% reduction, n=10), intermediate responders (>= 50% reduction and <= 90% reduction), n=19, and poor responders (<50% reduction, n=11). As hypothesized, in this retrospective study, seizure reduction (median 64.6% reduction at last follow-up) was correlated with more frequent stimulation during low-risk periods compared with high-risk periods. Additionally, stimulation events were more likely to be phase-locked to prolonged episodes of abnormal activity for intermediate and poor responders when compared to super responders, consistent with the hypothesis that improved outcomes are driven by stimulation during low-risk states. These results suggest that stimulation during low-risk periods may more readily induce plasticity that, in turn, facilitates network changes leading to long-term seizure reduction.

Unearthing the mechanisms of responsive neurostimulation for epilepsy

“…Indeed, a recent study found that the effects of changing RNS stimulation parameters (e.g., frequency, burst duration, and charge density) depend on the initial seizure risk state 47 , with parameter changes effective at reducing seizures in one seizure risk state being less effective during another risk state. Consistent with this, a recent study examining how stimuli were distributed across these low-and high-risk states found improved outcomes when stimulations were delivered preferentially in lowrisk states, i.e., those states less disrupted by ongoing epileptiform activity 48 . RNS stimulation parameters are typically adjusted every few months, remaining constant in the interim.…”

“…Patients with the best outcomes from RNS are those with the greatest ability to reorganize functional network connectivity. Stimuli inducing this plasticity may be more effective when delivered during periods with less epileptiform activity 48 , when endogenous neuroplasticity mechanisms are more active 52,53 .…”

Resting-state background features demonstrate multidien cycles in long-term EEG device recordings

“…In our retrospective cohort of 20 patients, we replicated analyses from previous studies [3], [5], [15] to compare the multidien dIEA cycle characteristics in our subjects with existing literature.…”

“…Extracting multidien cycles from dIEA counts. We followed a similar methodology as outlined in Baud et al to extract multidien cycles from dIEA activity (Figure 1A) [5], [15]. In brief, for each patient we applied a wavelet transform to the dIEA signal and identified the dominant modes in the average frequency spectrum (peaks in the periodogram) in the range of 3-60 days.…”

An embedded intracranial seizure monitor for objective outcome measurements and rhythm identification

“…For epilepsy, failure to deliver stimulation during seizure events may lead to worse therapeutic outcomes than falsely providing stimulation during non-seizure periods. In fact, there is increasing evidence that the therapeutic benefits of responsive neurostimulation are provided when stimulation is applied during non-seizure states, rather than seizure states [ 16 ]. Similar considerations should be taken when considering other neurological disorders where the therapeutic trade-off may differ in a disease-specific manner, potentially leading to worsened outcomes for patients’ daily activities.…”