Authors :
Presenting Author: Ryan Verner, PhD – LivaNova PLC (or a subsidiary)
Riëm El Tahry, MD, PhD – UC Louvain
Firas Fahoum, MD, MS – Tel-Aviv University
Michal Tzadok, MD – Safra Children's Hospital, Sheba Medical Center
Francesca Beraldi, MS – LivaNova PLC (or a subsidiary)
Michael Gelfand, MD – University of Pennslyvania
Muhammad Zafar, MD, FACNS – Duke University Hospital
Gholam Motamedi, MD – Medstar Georgetown University Hospital
Massimiliano Boffini, PhD – LivaNova PLC (or a subsidiary)
Maxine Dibué, PhD – LivaNova PLC (or a subsidiary)
Charles Gordon, MS – LivaNova PLC (or a subsidiary)
George Morris, MD, MPH DIC – Ascension St Mary's Medical Center
Arjune Sen, MD – Oxford University
Rationale:
For nearly three decades, VNS has been an effective treatment to reduce the frequency of seizures in people with drug resistant epilepsy (DRE). Despite frequent use in this population, little evidence-based guidance has been made available to characterize appropriate dosing and titration practice. In 2022, investigators succeeded in leveraging a large database of existing VNS clinical evidence to identify a population-level target dose and titration method for VNS [1,2] which was aligned to the labeling-defined target dose range. The present work attempts to validate those retrospective findings using contemporary clinical data in a prospective observational study of VNS called CORE-VNS.
Methods:
CORE-VNS (NCT03529045) is a global, multicenter, observational study of VNS in patients with drug-resistant epilepsy. Participants in the study who were implanted with their first VNS device and attended at least one of either the 6- or 12-month follow-up were selected for this analysis. We employed a variety of statistical methods (generalized linear mixed models, weighted Cox regression, Kaplan-Meier, and Poisson regression) to assess the relationship between VNS titration and clinical response. In this work, participants who were predominantly manually titrated were separated from those who were predominantly titrated using automated programming (“Scheduled Programming (SP)”).
Results:
Of the enrolled population in CORE-VNS (First Implant: n=529), 499 participants met our inclusion criteria for this analysis. The majority of participants were titrated manually (n=385), and very few were titrated predominantly using the SP feature (n=46). The remaining participants (n=68) used a mix of manual and scheduled titration events such that the programming process did not appear to be heavily influenced by SP. There were no significant clinical or demographic differences between participants who were titrated manually versus those titrated using SP. Participants titrated with SP for 3 or more steps were able to achieve the target dose range (per product labeling: 1.5mA – 2.25mA dependent on pulse width) faster than those who used less SP or not at all (p< 0.001). These same SP-titrated patients had numerically shorter time-to-response (≥50% reduction from baseline in seizure frequency) compared to others, but, in this relatively small cohort, this effect was non-significant. Finally, participants titrated with SP achieved significantly higher rates of response following 12 months of VNS (63.2% vs 45.5%; p< 0.05).