Authors :
Presenting Author: Alexandre Berger, MSc – Institute of Neuroscience (IoNS), Catholic University of Louvain
Elise Beckers, PhD – GIGA-Cyclotron Research Center-In vivo Imaging, University of Liège; Vincent Joris, MD – Institute of Neuroscience (IoNS), Catholic University of Louvain; Venethia Danthine, MD – Institute of Neuroscience (IoNS), Catholic University of Louvain; Nicolas Delinte, PhD – Institute of Information and Communication Technologies, Electronics and Applied Mathematics (ICTEAM), Catholic University of Louvain; Inci Cakiroglu, PhD – Institute of Neuroscience (IoNS), Catholic University of Louvain; Enrique Germany, PhD – Institute of Neuroscience (IoNS), Catholic University of Louvain; Andres Torres Sanchez, PhD – Institute of Neuroscience (IoNS), Catholic University of Louvain; Benoit Macq, Prof – Institute of Information and Communication Technologies, Electronics and Applied Mathematics (ICTEAM), Catholic University of Louvain; Laurence Dricot, Prof – Institute of Neuroscience (IoNS), Catholic University of Louvain; Gilles Vandewalle, Prof – GIGA-Cyclotron Research Center-In vivo Imaging, University of Liège; Riëm El Tahry, MD – Institute of Neuroscience (IoNS), Catholic University of Louvain
Rationale:
Vagus Nerve Stimulation (VNS) has been used for over 30 years as an adjunctive treatment for patients with drug-resistant epilepsy (DRE). However, response to the therapy is variable across patients, and the requisites to become responder (R : ≥ 50% reduction in seizure frequency) as well as a complete understanding of the mechanisms of action are still lacking. Previous animal and human studies demonstrated the involvement of the brainstem locus coeruleus (LC), the main source of norepinephrine (NE) in the brain, in the clinical effects of VNS by preventing seizure development. The development of specific Magnetic Resonance Imaging (MRI) sequences made the
in vivo visualization of the LC possible in humans. However, no study assessed the structural and functional characteristics of the LC in VNS patients with DRE, so far.
Methods:
Twenty-three DRE patients (13 F – 10 M, mean age 37.26 ± 12.92) implanted with VNS (10 Non-Responders - NR : < 30% reduction in seizure frequency, 8 R and 5 Partial Responders – PR : 30-50% reduction in seizure frequency) were recruited for the study. Data were acquired with a 3T GE SIGNA Premier MRI system using a transmit-receive 48-channel head coil, with the VNS turned off. A structural MRI sequence sensitive to a neuromelanin-related feature of the LC was acquired to localize the LC. Functional MRI data were also acquired while patients completed an auditory oddball task - which mimics novelty detection and recruits the LC, in order to extract task-related estimates of LC reactivity. Moreover, using a multi-shell diffusion MRI sequence, tractography of LC-hippocampus connections was performed (Figure 1), and diffusion metrics (Diffusion Tensor Imaging and Microstructure Fingerprinting [Delinte et al., 2021] metrics) were extracted within the tracts. Finally, multiple linear regressions were used to model these features in terms of VNS response, controlling for age, sex, antiseizure medication intake, benzodiazepine intake and epilepsy duration.
Results:
A significant lower LC reactivity was found in the medial portion of the left LC in R/PR compared to NR (p = 0.01) – (boxplots shown in Figure 2). A lower mean diffusivity (MD, p = 0.001), lower axial diffusivity (AD, p = 0.02) and lower radial diffusivity (RD, p < 0.001), as well as a higher weighted fiber volume fraction (wFVF, p = 0.02) were found in the left LC-hippocampus connections in R/PR compared to NR. A lower MD (p = 0.01) and lower RD (p = 0.003) were also observed in the right LC-hippocampus connections in R/PR compared to NR.