Keywords
dystonia - deep brain stimulation - globus pallidus internus - visual evoked potential
- microelectrode recording
Introduction
Deep brain stimulation (DBS) of the globus pallidus internus (GPi) is an effective
treatment for patients with medically refractory dystonia, offering substantial improvement
in motor symptoms and quality of life. Accurate targeting is essential, especially
near critical structures like the optic radiations.[1]
[2] Visual evoked potentials (VEPs) can aid in identifying the optic tract during such
procedures. Although DBS is typically performed awake, patients with severe dystonia
and involuntary movements may require general anesthesia, complicating intraoperative
neurophysiological monitoring. We present a case of bilateral GPi DBS under general
anesthesia with intraoperative VEP monitoring, highlighting key anesthetic and neurophysiological
strategies.
Case Report
A 62-year-old female with a 1.5-year history of medically refractory oromandibular
dystonia, dependent on a Ryle's tube for feeding, was scheduled for bilateral GPi
DBS. Electrode placement was planned under general anesthesia with intraoperative
VEP monitoring to ensure accurate targeting.
The patient was counseled regarding the anesthetic plan, including the possibility
of reducing anesthetic depth to facilitate VEP acquisition, and the associated risk
of intraoperative awareness. Reassurance was provided that analgesia and comfort would
be prioritized. Antidystonic medications were continued up to the day of surgery.
On the morning of surgery, general anesthesia was induced with fentanyl (2 mcg/kg),
propofol (2 mg/kg), and cisatracurium (0.3 mg/kg). Following tracheal intubation,
a computed tomography scan was obtained after Leksell frame placement under pin-site
infiltration with local anesthetic. Anesthesia was subsequently maintained with dexmedetomidine
infusion, desflurane (minimum alveolar concentration (MAC) < 0.4), and a propofol
infusion (50–100 µg/kg/min), titrated to maintain a bispectral index (BIS) value of
40 to 60. No additional neuromuscular blockade was given, allowing assessment of motor
side effects during internal capsule stimulation.
For VEP monitoring, bilateral light-emitting diode (LED) goggles delivered photic
stimulation, and electrodes were placed at O1, O2, and Oz, with Fz, A1, and A2 as
references. As the stereotactic frame precluded goggle straps, the LED goggles were
secured with adhesive tapes. A cloth cover was additionally applied as a precaution
against any possible ambient light interference. Cortical VEPs (cVEPs) were recorded
using the NIM-Eclipse system at 3 Hz with 50 ms stimulation, generating the N75–P100
waveform after averaging. Optimal amplitudes were obtained when BIS was between 70
and 80, necessitating adjustment of anesthetic depth while providing intermittent
fentanyl boluses for analgesia.
A 2-mm active-tip DBS electrode was advanced along a trajectory extending from 4 mm
above to 2 mm below the planned GPi target and used for direct optic tract stimulation.
Stimulation parameters (3 Hz, 50 ms pulses, 4–5 mA) reliably evoked optic tract VEPs
(oVEPs), which were recorded from the same scalp electrodes while photic stimulation
was suspended. This was followed by microelectrode recordings (MERs), which demonstrated
characteristic dystonic neuronal firing. Subsequent macroelectrode stimulation (MES)
was performed with stepwise current increments (1–7 mA) to elicit contralateral motor
responses and further refine electrode positioning with respect to the internal capsule.
Final lead placement was achieved at a site 1 mm from the point of maximal oVEP amplitude
and 2 mm lateral to the capsule. Correct positioning was confirmed intraoperatively
with O-Arm imaging, after which the implantable pulse generator was successfully inserted.
The patient was extubated uneventfully and transferred to the intensive care unit.
She remained neurologically stable and had no recall of intraoperative events at 24-hour
follow-up.
Discussion
DBS of the GPi is an established treatment for medically refractory dystonia, with
proven efficacy in improving motor symptoms and reducing disability. Precise targeting
is crucial to optimize outcomes and avoid MES-related side effects, especially given
the GPi's proximity to the internal capsule.[3] However, stimulation responses may originate from adjacent structures such as the
putamen, globus pallidus externus, or ansa lenticularis, limiting their value for
confirming accurate localization.[2] While modern neuronavigation systems assist in targeting, optic tract stimulation,
due to its close anatomical relationship just above the GPi, offers a reliable functional
marker to further refine lead placement and prevent off-target positioning.
In this case, both cVEP and oVEPs were employed to enhance the safety and precision
of electrode placement near the optic tracts. Although oVEPs provides more targeted
information, initial cVEPs obtained via photic stimulation offered key preparatory
benefits. First, they served as a functional check of the recording setup, signal
integrity, and the baseline functional integrity of the visual pathways. Finally,
cVEPs helped optimize anesthetic depth, allowing the team to adjust agents to maintain
BIS values (70–80) that preserved evoked responses without compromising immobility.
These preparatory steps ensured reliable intraoperative monitoring and contributed
to the successful application of oVEPs during targeting. VEPs are near-field potentials
and their amplitude reflects the proximity between the target site and the optic tract.
In our case, VEP amplitude increased progressively as the stimulating electrode approached
the optic tract, from 4 to 1 mm ([Fig. 1]).
Fig. 1 Direct optic tract stimulation (optic tract visual evoked potential [oVEP]) potentials
of right optic tract. (A) oVEP waveforms at depth of 4 mm from optic tract. (B) oVEP waveforms at the depth of 1 mm. Note the increase in amplitude p100 waveform
in O2-A2 and Oz-A2 on the right side.
Both MER and VEP are highly sensitive to anesthetic agents. Thus, the primary anesthetic
challenge is to maintain an optimal depth that preserves neurophysiological signals
while ensuring patient immobility. In this case, we chose a combination of low-dose
desflurane with titrated propofol infusion rather than propofol alone. This strategy
allowed greater flexibility in titrating anesthetic depth during prolonged neurophysiological
monitoring, reduced cumulative propofol requirements, and facilitated smoother and
faster postoperative recovery. Although we maintained a BIS of 70 to 80, which is
slightly lighter than the typical surgical depth,[3]
[4] adequate immobility was achieved using intermittent fentanyl boluses. To further
suppress respiratory efforts and prevent patient movement, mild hypocarbia was maintained.
This strategy was planned and appropriate consent regarding the possibility of intraoperative
awareness was obtained. However, no awareness was reported postoperatively in this
case. Timing of opioid administration is crucial, as bolus dosing can transiently
suppress neurophysiological signals.[5] A low-dose fentanyl infusion offers more stable effects. Although remifentanil,
with its rapid onset and neurophysiological compatibility, would have been ideal,
it was unavailable. Dexmedetomidine was used instead as it provided analgesia, lowered
the need for additional anesthetics, and had minimal effect on MER.
Finally, successful execution of such a procedure relies not only on tailored anesthetic
strategies but also on seamless multidisciplinary collaboration. Close coordination
among the neuroanesthetist, neurophysiologist, neurologist, and neurosurgeon was essential
by contributing their expertise to address the unique challenges of DBS in dystonia.
