Background: Despite widespread use of vagus nerve stimulation (VNS), evidence regarding optimal stimulation titration and dosing remains limited. Significant inter-practice variability and deviation from guidelines, results in sub-optimal dosing.
Objective: Identify optimal stimulation parameters and establish target dosing for VNS therapy in epilepsy.
Methods: We linked long-term outcomes with dosing parameters in 415 patients from a VNS registry (median follow-up = 79.0 months). A generalized linear mixed model (GLMM) assessed relationships between stimulation parameters (Output Current (OC), Frequency, Pulse Width (PW), Duty cycle [DC]) and very good clinical response (≥75 % seizure frequency reduction from baseline).
Results: Output current was the strongest predictor of seizure reduction (p = 0.001). Increasing OC by 1 mA more than doubled the probability of ≥75 % seizure reduction, peaking at 2.70 mA. Greatest likelihood of ≥50 % seizure reduction was at 2.1 mA. Post-hoc analysis supported GLMM findings, showing significantly more patients (46.3 %) receiving >2.0 mA OC achieved ≥75 % seizure reduction compared with 30.7 % (p = 0.01) among patients using lower OC. Although changes in DC were not significantly associated with seizure reduction, post-hoc analysis suggested benefits from other duty cycles. Standard 250 μs PW and 20 Hz frequency were associated with best effect. Longer follow-up improved seizure control (p < 0.001); intellectual disability negatively impacted response (p = 0.005). Seizure-free patients had significantly lower total charge (122.4 vs. 250.1 mC, p = 0.014) than patients with persistent seizures.
Conclusion: Output current should be the primary focus when titrating VNS patients to optimal dose. We suggest a target range of 1.5-2.25 mA with 250 μs PW and 20 Hz frequency. Higher OC should be attempted if tolerated.
Keywords: Drug resistant epilepsy; Output current; Titration; Vagus nerve stimulation.
Copyright © 2025 The Authors. Published by Elsevier Inc. All rights reserved.