Pulse Width and Implantable Pulse Generator Longevity in Pallidal Deep Brain Stimulation for Dystonia: A Population-Based Comparative Effectiveness Study.

ABSTRACT

INTRODUCTION: A wide range of pulse widths (PWs) has been used in globus pallidus internus (GPi) deep brain stimulation (DBS) for dystonia. However, no specific PW has demonstrated clinical superiority, and the paradigm may differ among DBS centers. OBJECTIVE: To investigate how different paradigms of PWs in GPi DBS for dystonia affect implantable pulse generator (IPG) longevities and energy consumption. METHODS: Thirty-nine patients with dystonia treated with bilateral GPi DBS at 2 Swedish DBS centers from 2005 to 2015 were included. Different PW paradigms were used at the 2 centers, 60-90 µs (short PWs) and 450 µs (long PW), respectively. The frequency of IPG replacements, pulse effective voltage (PEV), IPG model, pre-/postoperative imaging, and clinical outcome based on the clinical global impression (CGI) scale were collected from the medical charts and compared between the 2 groups. RESULTS: The average IPG longevity was extended for the short PWs (1,129 ± 50 days) compared to the long PW (925 ± 32 days; χ2 = 12.31, p = 0.0005, log-rank test). IPG longevity correlated inversely with PEV (Pearson's r = -0.667, p < 0.0001). IPG longevities did not differ between Kinetra® and Activa® PC in the short (p = 0.319) or long PW group (p = 0.858). Electrode distances to the central sensorimotor region of the GPi did not differ between the short or long PW groups (p = 0.595). Pre- and postoperative CGI did not differ between groups. CONCLUSIONS: Short PWs were associated with decreased energy consumption and increased IPG longevity. These effects were not dependent on the IPG model or the anatomic location of the electrodes. PWs did not correlate with symptom severities or clinical outcomes. The results suggest that the use of short PWs might be more energy efficient and could therefore be preferred initially when programming patients with GPi DBS for dystonia.