# Medial pulvinar stereoelectroencephalographic biomarkers associated with deep brain stimulation response in focal drug‐resistant epilepsy

## TL;DR

This study identifies brain activity patterns in the medial pulvinar that predict successful deep brain stimulation for drug-resistant epilepsy.

## Contribution

The study introduces entropy-based SEEG biomarkers (PEI and ΔE) as potential predictors of DBS response in medial pulvinar.

## Key findings

- Entropy-based measures PEI and ΔE were significantly higher in patients who responded to PuM-DBS.
- Spike rates, connectivity, EI, and cEI did not differentiate responders from non-responders.
- Ictal PuM complexity disruption may serve as a candidate biomarker for DBS response.

## Abstract

Thalamic deep brain stimulation (DBS) represents an emerging therapeutic option for patients with focal drug‐resistant epilepsy who are ineligible for or have failed resective surgery. To optimize outcomes and guide DBS lead placement, thalamic stereoelectroencephalography (SEEG) has been proposed. This monocentric retrospective study aimed to identify interictal and ictal SEEG biomarkers of the medial pulvinar (PuM) associated with favorable PuM‐DBS response. Six patients (four female, two male) underwent SEEG including PuM sampling, were deemed unsuitable for resective surgery, and subsequently received bilateral PuM‐DBS. In total, eight PuMs were sampled in SEEG: four bilaterally (two patients) and four unilaterally (four patients). The SEEG exploration covered both the PuM and the ipsilateral epileptogenic zone network (EZN) in four patients, whereas in the other two the EZN was bilateral but PuM sampling was unilateral. All SEEG signal analyses were performed on the PuM sampling contacts available in each patient. Interictal SEEG analysis included spike rates and nonlinear functional connectivity (h2), whereas ictal analyses combined visual inspection with quantitative biomarkers: epileptogenicity index (EI), connectivity epileptogenicity index (cEI), permutation entropy index (PEI), and delta entropy (ΔE). Two patients were responders (≥50% seizure reduction at 1 year). PuM spike rates, connectivity, and EI and cEI values did not differentiate responders from nonresponders. In contrast, entropy‐based measures were significantly higher in responders: PEI (false discovery rate [FDR]‐p = .024) and ΔE (FDR‐p = .034). These findings suggest that ictal PuM complexity disruption, quantified through entropy‐based SEEG metrics (PEI and ΔE), may represent a candidate biomarker of response to medial pulvinar DBS and warrants validation in larger cohorts.