CT-Guided Asleep DBS

CHAPTER 23

CT-Guided Asleep DBS KIM J. BURCHIEL, MD, FACS

INTRODUCTION Deep brain stimulation (DBS) for Parkinson disease (PD), essential tremor (ET), and dystonia has become an established therapy for patients with medically intractable movement disorders.1 Indications for DBS for PD include frequent or unpredictable on/off motor fluctuations, disabling dyskinesia, or disabling tremor that is refractory to dopamine replacement medications. The indications for DBS for ET are less clear, but generally, DBS is reserved for patients with severe disabling, intentional or postural tremor. In the past 25 years, DBS has typically been performed with the patient awake in the operating room to allow for microelectrode recording (MER) to identify the targets for placement of the stimulating electrodes.2 This procedure involves placement of recording microelectrodes into the brain to identify the location and borders of the subthalamic nucleus (STN) and globus pallidus pars interna globus pallidus internus (GPi) for patients with PD or dystonia, and the ventralis intermedius (Vim) for patients with ET. MER often necessitates multiple passes into brain tissue to achieve a satisfactory result,3 a requirement that has been shown to be directly correlated with an increased incidence of intracerebral hemorrhage (ICH).4 The prospect of having an elective awake brain surgery is a barrier for many who are otherwise good candidates for this treatment, and there is no Class I or II evidence that MER improves the outcome of a DBS implant procedure. Often the MER procedure is accompanied by test stimulation with the patient awake to further verify target accuracy, further prolonging the duration of the procedure that might take up to 4 to 8 h.5,6 With the advent of advanced magnetic resonance and computed tomography imaging, particularly intraoperative imaging, the argument for the continued use of MER during DBS implantation has been substantially weakened. Furthermore, although the risk of a serious adverse event related to MER, such as ICH, remains low, it is not zero. Metaanalyses of surgical risk for DBS have reported ICH rates of 3.2%e5%.4,7,8 The

risk of ICH is related to the use of MER, the number of MER penetrations, and sulcal or transventricular course of the electrode trajectory.4 The bulk of the evidence indicates that multiple instrumented passes into the brain with the sharp tip of the recording microelectrode contributes to the risk of ICH, with the pertrajectory ICH rate estimated at 1.57%.9 ICH related to DBS may further be divided into asymptomatic, symptomatic, and that resulting in permanent deficit or death, at reported rates of 1.9%, 2.1%, and 1.1%, respectively.4 The incidence of hemorrhage in studies adopting an image-guided and imageverified approach without MER was significantly lower than that reported with other operative techniques (P < .001 for total number of hemorrhages, P < .001 for asymptomatic hemorrhage, P < .004 for symptomatic hemorrhage, and P ¼ .001 for hemorrhage leading to permanent deficit).4 Perhaps the most common argument for the use of MER is the sensitivity of this methodology in finding the precise target for implantation of DBS electrodes. Yet, a recent analysis of multiple databases between 2004 and 2013 revealed over 28,000 cases of DBS electrode placement, revision, and removal. Data from Medicare indicated that 15.2% of DBS procedures were for revision or removal. Similar analysis of the National Surgical Quality Improvement Program (NSQIP) dataset showed a 34.0% DBS removal or revision rate. The authors concluded that up to 48.5% of revisions may have been due to improper targeting or lack of therapeutic effect.10 Combining the Medicare and NSQIP datasets, the proportion of these procedures performed for a first DBS electrode implant with MER was 73.9%, compared with 10.8% without MER. That is, the overwhelming majority of cases in the analysis were related to patients in whom MER had been used during the initial implant. This result suggests that the presumed location of DBS electrode implant using MER may not be as predictable as has been previously assumed. Newer methods of performing DBS incorporate image guidance using intraoperative magnetic resonance 189

190

Functional Neurosurgery and Neuromodulation

image (iMRI) (Starr, 2010) or intraoperative computerized tomographic imaging (iCT)11e13 without MER. When iCT is used, iCT images are coregistered with a preoperative 3-T MRI to plan the trajectory to avoid sulci and ventricles, which are known to cause adverse events, visualize the planned target, and confirm lead placement of the target electrode contact before the patient leaving the operating room.11,13 This allows the patient to be asleep from the start to end of the entire procedure. Accuracy of lead placement using this method is excellent, with good outcomes.11e13

ASLEEP DEEP BRAIN STIMULATION PROCEDURE Imaging Before surgery, 3-T MRI is obtained for DBS targets in all patients using the following sequences: ventrointermediate nucleus (Vim; a standard 3D T1, TE: 4.61, TR: shortest, flip angle: 30, voxel size: 1.02, matrix 256  256), STN (a FLAIR sequence TE: 140, TR: 14,000, slice thickness 2.5 mm, voxel size: 1.02, matrix 256  256), and globus pallidus internus (GPi; an MPRAGE sequence TE: 80, TR: 8000, slice thickness: 2 mm, voxel size: 0.6, matrix 256  256).14 These digital imaging and communications in medicine (DICOM) images are downloaded into a StealthStation (Medtronic Inc., Minneapolis, Minnesota, United States) via a network connection.

Corp., Danvers, MA).16 Subsequently, these DICOM images are transferred to the StealthStation (Medtronic Inc.) surgical navigation system and merged with the preoperative MR images. Using a passive planar blunt probe, a nonsterile registration of the skull fiducials is then performed to link image and surgical spaces. The burr hole entry point of the predetermined electrode trajectory is then marked on the skin, and a small pilot hole was drilled to mark that point on the skull. After appropriate sterile preparation and draping, skin incisions are made and burr holes centered on the pilot hole are completed. The lead anchoring device (StimLoc, Medtronic, Inc.) and the NexFrame base are then attached to the skull, and a second sterile registration is performed using the implanted fiducials (target registration error <0.5 mm). The NexFrame tower is then attached and aligned to the corresponding target using FrameLink (Medtronic Inc.) software. Target depth is then calculated and set on the StarDrive (FHC Inc., Bowdoin, ME) positioning device. After opening the dura, a cannula is then placed to target, and fibrin glue is used to prevent egress of cerebrospinal fluid. A DBS lead (Medtronic lead model 3387) is then advanced through the cannula to the target position (Fig. 23.1). The cannula is then retracted, and the electrode is

Surgical Technique Implant procedures are carried out in two stages. The first stage is an inpatient procedure for implantation of the DBS electrodes, and the second stage is the implantation of the internal pulse generator (IPG), performed on an outpatient basis approximately 1 week later. The implantation of the DBS electrode(s) uses the extraoperative MR images aforementioned, and surgical planning of electrode trajectories is carried out preoperatively on the StealthStation (Medtronic Inc.). GPi and STN targets are visualized directly, and the Vim target is derived from the Schaltenbrand and Wahren atlas, using the intercommissural plane and midcommissural point for reference.15 All procedures are performed under general anesthesia. After intubation, the patient’s head is attached to the operating table using a Doro Halo Retractor System with carbon fiber extensions (Pro Med Instruments, Inc., Freiburg, Germany). After sterile preparation of the head, five skull-mounted fiducials (NexFrame, Medtronic Inc.) are placed. A stereotactic 1-mm thick, zero-gantry-angle CT image was then obtained using a CereTom CT scanner (NeuroLogica

FIG. 23.1 Placement of DBS leads using the NexFrame and Stardrive. DBS, deep brain stimulation.

CHAPTER 23 CT-Guided Asleep DBS

FIG. 23.2 An intraoperative CT scanner (Ceretom, Neuro-

Logica, Inc.) is used in conjunction with the StealthStation (Medronic, Inc.) to confirm accurate electrode location after the leads are placed and before completing the procedure. CT, computerized tomographic.

secured in the StimLoc system. The exposed electrode is then pulled back through the StarDrive, and in most cases, the process is repeated on the opposite side. With the NexFrame towers in place, a second iCT scan is obtained to ensure satisfactory electrode placement (Fig. 23.2). This postplacement scan is then merged with the preoperative planning scans, and any off-target error is recorded. Vector error is calculated as the linear distance between the center of the intended target contact on the DBS electrode and the planned target. The electrode is repositioned if the vector error is >2.0 mm. Once the accuracy of the electrode placement is confirmed by intraoperative CT, no further imaging is carried out on these patients intra- or postoperatively (Fig. 23.3A and B). After ensuring satisfactory electrode placement, the NexFrame towers are removed and the StimLoc caps are applied. The two leads are then capped and tunneled in the subgaleal plane to the side of the head just superior and posterior to the ear ipsilateral to the planned IPG subclavicular implant, after which the wounds are closed. Within 4e7 days, an outpatient procedure is performed with the patient under general anesthesia during which the DBS leads are connected to extension leads and tunneled to a subclavicular pocket for the IPG implant.

OUTCOMES OF ASLEEP DEEP BRAIN STIMULATION Electrode Accuracy We have used two measures to assess electrode placement accuracy: (1) vector error, defined as the distance

191

between the position of the center of the target DBS electrode contact (“1” for Vim and STN, and “0” for GPi) and the intended target location as determined by the Euclidean distance between these two points [(x2  x1)2 þ (y2  y1)2 þ (z2  z1)2]1/2 and (2) deviation off trajectory, defined as the perpendicular distance from the target electrode to the planned trajectory, reflecting only the radial deviation from plan (Fig. 23.4). Both values are felt to be useful in determining the accuracy of placement because although vector error reflected the true accuracy of the procedure, off trajectory (radial) errors theoretically would be less easily correctable by assigning different active contacts during DBS therapy. The accuracy of electrode placement has been compared among all three targets; GPi, STN, and Vim. The closest approach of the DBS lead trajectory to the lateral ventricle has also been measured and correlated with both the vector error and the deviation off trajectory. In all cases, the postimplantation CT scan is merged with the registration merged CT/MR used for planning the electrode trajectory and target to calculate both the vector error and the deviation off trajectory distances. The initial 60 patients (33 PD, 26 ET, and 1 dystonia) who underwent DBS electrode placement by this method at Oregon Health and Science University (OHSU) were analyzed11,13. Over 18 months, 119 electrodes were placed (all bilateral, except one). Mean patient age was 64  9.5 years, and electrode implant location was Vim (25 patients), GPi (23 patients), and STN (12 patients). Final intraoperative CT imaging showed no evidence of intracranial air or any other potential cause for brain shift in any case. The mean total operating room time (time in: time out), including anesthetic induction and recovery, was 190  9.8 min. In these cases, the closest approach of the electrode trajectories to the ventricle was 6.3  3.4 mm (range, 1e13.9 mm), and there was no statistically significant difference between sides. Mean overall vector error was 1.59  1.11 mm, and the trajectory deviation error was 1.24  0.87 mm. There was a significant correlation between (1) the distance from the ventricle and trajectory deviation error (r2 ¼ 0.325, P < .05, n ¼ 77) and (2) the distance from the ventricle and vector error (r2 ¼ 0.339, P < .05, n ¼ 76). Furthermore, when the distance from the electrode trajectory and the ventricular wall was <4 mm, the correlation of the ventricular distance to the deviation from the planned trajectory was stronger (r2 ¼ 0.419, P ¼ .05, n ¼ 19; Fig. 23.5). Electrodes placed in the GPi were significantly more accurate than those placed in the Vim (mean vector error of 1.29 vs.

192

Functional Neurosurgery and Neuromodulation

A

B FIG. 23.3 Final postimplant CT scan merged with preoperative MRI scan to confirm accurate electrode placement. Yellow (left) and green lines represent the preoperative planned trajectory and target points for bilateral placements in the (A) subthalamic nucleus and (B) globus pallidus internus. CT, computerized tomographic; MRI, magnetic resonance image.

CHAPTER 23 CT-Guided Asleep DBS

193

t tr rge Ta ry cto aje

de

tro lec

le

tua Ac

Axial plane

T V

FIG. 23.4 Accuracy of electrode placement is measured by (1) vector error (V), defined as the distance between the position of the center of the target DBS electrode contact (“1” for Vim and STN, and “0” for GPi) and the intended target location as determined by the Euclidean distance between these two points: [(x2  x1)2 þ (y2  y1)2 þ (z2  z1)2]1/2, and (2) deviation off trajectory (T), defined as the perpendicular distance from the target electrode to the planned trajectory, reflecting only the radial deviation from plan. DBS, deep brain stimulation; GPi, globus pallidus internus. (From the Burchiel KJ, McCartney S, Lee A, et al. Accuracy of deep brain stimulation electrode placement using intraoperative computed tomography without microelectrode recording. J Neurosurg. 2013;119(2):301e306 with permission.)

1.9 mm, P ¼ .01). There was no statistically significant difference in accuracy comparing the GPi with STN, the STN with Vim, or the GPi with STN and Vim combined. Mean distance from the ventricle for GPi electrodes was 9 mm compared with 4.2 mm for the Vim and 5.15 mm for the STN, and these differences were significant (P < .05). In this series, one implanted electrode had a vector error of >2 mm and was revised. There was one case of a late infection that required explant of the entire system. There were no hematomas, no new neurologic deficits, and there was no mortality. Subsequent to this early experience, we have now analyzed 169 leads in 94 patients. Targets were GPi (n ¼ 86), STN (n ¼ 31), and Vim (n ¼ 52); 85 leads were placed on the left and 84 on the right. Average

Euclidean error was 1.63 mm (SD: 0.87). Error magnitude is higher for Vim (1.95 mm) than for GPi (1.44 mm), whereas STN (1.65 mm) did not differ from either Vim or GPi (ANOVA: F ¼ 6.15, P ¼ 0.003). Electrodes targeting Vim and STN were significantly more likely to deviate medially compared with GPi (ANOVA: F ¼ 9.13, P < 0.001). Coronal approach angle affects error when targeting Vim (r ¼ 0.338, P ¼ 0.01). These findings were confirmed during multivariate analyses (Fig. 23.5).

Intracranial Air The introduction of intracranial air may be a consequence of prolonged DBS implantation procedures under local anesthesia. In fact, it is the potential brain shift attendant upon cerebrospinal fluid drainage and brain

194

Functional Neurosurgery and Neuromodulation

2mm

VIM

Target

A X Error (%)

STN

2mm

Posterior to Anterior

GPI

2mm medial

Target

2mm lateral

2mm medial

Target

2mm lateral

2mm medial

Target

2mm lateral

0.2 0.1 0

2mm medial

Target

2mm Lateral

2mm medial

Target

2mm Lateral

2mm medial

Target

2mm Lateral

Z Error (%)

Y Error (%)

0.4 0.2 0

0

2mm Posterior Target 2mm Anterior

Target 2mm Superior

2mm Inferior

Target 2mm Superior

2mm Inferior

Target 2mm Superior

Target

2mm

2mm Inferior

2mm

Inferior to Superior

2mm Posterior Target 2mm Anterior

0.2

B

C

2mm Posterior Target 2mm Anterior

0.4

2mm medial

Target

2mm lateral

2mm medial

Target

2mm lateral

2mm medial

Target

2mm lateral

FIG. 23.5 Stereotactic error by target. (A) Distribution of electrode locations by target in the axial plane

(average error in these directions is represented by the back arrow). Note that this error vector is greatest for Vim. Contour lines represent 50% (black) and 75% (gray) probability density estimates for electrode locations. (B) Histograms of error magnitudes in the X (top), Y (middle), and Z (bottom) directions for each target. No error (0 mm) is denoted with the dotted line. (C) Distribution of electrode locations in the coronal plane. Average error is again denoted with a black error, and probability density estimates for electrode locations are as above. GPi, globus pallidus internus; STN, subthalamic nucleus; VIM, ventralis intermedius.

“settling” that is often used as a rationale to perform procedures using iMRI as a means of accounting for this brain shift. To assess the incidence of intracranial air during iCT-guided DBS implantation procedures under general anesthesia, we conducted a retrospective review of our bilateral DBS cases performed from 2009 to 201317; a time spanning 2 years before and 2 years after the implementation of asleep DBS at OHSU. Post- or intraoperative imaging was examined to determine the presence and volume of intracranial air during these procedures. A total of 163 electrode implantations in

156 patients were reviewed. Intracranial air was noted in 27% of cases, with an incidence of 10.1% and 61.1% during asleep and awake DBS, respectively. The incidence of intracranial air was significantly higher in our awake DBS cases (P ¼ .00) (Fig. 23.6).

Clinical Outcomes DBS candidates with PD referred to OHSU underwent asleep DBS with imaging guidance. Six-month outcomes were compared with patients who previously underwent awake DBS by the same surgeon and center.18

CHAPTER 23 CT-Guided Asleep DBS Awake Asleep

Percent cases with air

80

60

40

20

0

50

100

150

Case number FIG. 23.6 Percentage of cases showing intracranial air after completion of the implant procedure. Gray bars reflect the incidence of intracranial air in 50 cases in which implants were performed with an awake patient, using MER. At case, 50 procedures were performed with the patient under general anesthesia without MER (white bars). MER, microelectrode recording.

Assessments included an off-levodopa Unified Parkinson’s Disease Rating Scale (UPDRS) II and III, the PDQ-39 questionnaire, motor diaries, and speech fluency. Thirty patients underwent asleep DBS, and thirty-nine patients underwent awake DBS. No significant difference was observed in improvement of UPDRS III (þ14.8  8.9 vs. þ17.6  12.3 points, P ¼ .19) or UPDRS II (þ9.3  2.7 vs. þ7.4  5.8 points, P ¼ .16). Improvement in on time without dyskinesia was superior in asleep DBS (þ6.4  3.0 h/day vs. þ1.7  1.2 h/day, P ¼ .002). Quality-of-life scores significantly improved in both groups (þ18.8  9.4 in awake, þ8.9  11.5 in asleep). Improvement in summary index (P ¼ .004) and subscores for cognition (P ¼ .011) and communication (P < .001) were superior in asleep DBS. Speech outcomes were superior in asleep DBS, both in category (þ2.77  4.3 points vs. 6.31  9.7 points, P ¼ .0012) and phonemic fluency (þ1.0  8.2 points vs. 5.5  9.6 points, P ¼ .038).18

Costs of Deep Brain Stimulation We have compared the cost of DBS performed awake versus asleep at OHSU, and compared costs across the University Health System Consortium (UHC) Clinical Database.19 Inpatient and outpatient demographics

195

and hospital financial data for patients receiving a DBS lead implant (from the first quarter of 2009 to the second quarter of 2014) were collected and analyzed. Inpatient charges included those associated with International Classification of Diseases, Ninth Revision (ICD-9) procedure code 0293 (implantation or replacement of intracranial neurostimulator lead). Outpatient charges included all preoperative charges 30 days before implant and all postoperative charges 30 days after implant. The cost of care based on reported charges and a cost-to-charge ratio were estimated. The UHC database was queried (January 2011 to March 2014) with the same ICD-9 code. Procedure cost data across like hospitals (27 UHC hospitals) conducting similar DBS procedures were compared. 211 DBS procedures (53 awake and 158 asleep) were performed at OHSU during the study period. The average patient age (SD) was 65  9 years, and 39% of patients were female. The most common primary diagnosis was PD (61.1%) followed by essential and other forms of tremor (36%). Overall average DBS procedure cost was $39,152  $5340. The total asleep DBS cost was $38,850  $4830, which was not significantly different than the awake DBS cost of $40,052  $6604. The standard deviation for asleep DBS was significantly lower (P  .05). In 2013, the median cost for a neurostimulator implant lead (Ninth Revision [ICD-9] procedure code 0293 [implantation or replacement of intracranial neurostimulator lead]) was $34,052 at UHC-affiliated hospitals that performed at least five procedures a year. At OHSU, the median cost was $17,150, and the observed single academic health center cost for a neurostimulator lead implant was less than the expected cost ([ratio 0.97]) (Figs. 23.7 and 23.8).

DISCUSSION Imaging technology has advanced to the degree that DBS electrode placement based entirely on imaging can be considered a feasible option. Several studies have demonstrated the reliability of image-defined STN and its dorsolateral subregion as location for movement-related cells during MER.15 We capitalized on previous reports that combined the use of a skullmounted frame (NexFrame), the use of intraoperative imaging, and general anesthesia.6 We have developed a method that uses an iCT scanner and a skull-mounted stereotactic system to place DBS electrodes. Our aim was to exploit the advantages of high-field extraoperative MRI, the availability of intraoperative CT scanning, and the capacity to merge images from both sources to provide the basis for target and trajectory planning for DBS surgery. Furthermore, this

196

Functional Neurosurgery and Neuromodulation

Deep brain stimulation cost

30 10

20

Mean total cost $ thousands

40

50

1/2011-3/2014

.5

1

1.5

2

2.5

Observed/expected cost FIG. 23.7 Scatterplot of cost benchmarking of the ICD-9 code 0293 (neurostimulator implant lead) and adult

inpatient cost data from 27 UHC hospitals from January 2011 to March 2014 compared with OHSU costs. ICD9, International Classification of Diseases, Ninth Revision; UHC, University Health System Consortium. (Data source: UHC.)

Cost per discharge

$80,000 5th percentile

$60,000

25th percentile $40,000

Median

$20,000

75th percentile 95th percentile

$2010

2011

2012

2013

OHSU

Year

FIG. 23.8 Percentile average total cost per discharge for ICD-9 code 0293, using adult inpatient cost data from 27 UHC hospitals (2010e13). ICD-9, International Classification of Diseases, ninth revision; UHC, University Health System Consortium. (Data source: UHC.)

system could be used to confirm target acquisition and any operative complications (hematoma) before leaving the operating room. Because this method relies only on anatomic target determination, the use of MER was obviated, and therefore general anesthesia was feasible.

Electrode Accuracy Using preoperative MRI, intraoperative CT, and a skullmounted stereotactic system, we were able to achieve an unprecedented degree of accuracy for DBS electrode placement. Our results (1.59 mm vector error) compare favorably with previously reported vector errors for iMRI (2.18 mm),13 intraoperative imaging (O-arm) combined with frame-stereotaxy (1.65 mm),11 and conventional frame-based stereotaxy without

intraoperative imaging (3 mm).12 The mean trajectory deviation error in our series was 1.24 mm, indicating that displacement was mostly radial, and not along the planned trajectory (depth). Our data indicate that the electrode trajectory distance from the ventricle plays a major role in the radial deviation of the trajectory. We attempted to correlate the vector error and trajectory deviation with factors that would impact location of the electrodes. The only significant correlation found was the distance of the electrode trajectory from the ventricle. Starr et al. previously reported on the correlation of the vector error and coronal approach angle, where the increased obliquity of the angle caused a higher positional error at the electrode tip.20 A negative correlation was found between the distance of the

CHAPTER 23 CT-Guided Asleep DBS electrode from the ventricle and both the vector error and trajectory deviation, i.e., both the vector error and the trajectory deviation were higher (less accuracy) when the electrode trajectory was closer to the ventricle. The strongest correlation was between the deviation error (radial displacement) and the distance from the ventricle in the range of 0e4 mm. Our analysis also showed that electrode trajectories approaching the GPi are significantly (P < .01) more distant from the ventricle in comparison with other targets (GPi ¼ 9 mm, STN ¼ 4.5 mm, and Vim ¼ 5.1 mm). Electrodes placed in the GPi were significantly more accurate than those placed in the Vim (mean vector error of 1.29 vs. 1.9 mm, P ¼ .01) but not more accurate than those placed in the STN (mean vector error of 1.29 vs. 1.52, P ¼ .3). Our data indicate that target trajectories that avoid a close approach to the ventricle are more likely to produce accurate placement. An alternative explanation is that traversing the internal capsule, e.g., approaching the Vim target from lateral to medial, may also cause distortion of the brain, with resultant electrode malposition. Trajectories that largely avoided the compact internal capsule, e.g., globus pallidus internus, were significantly more accurate. The reason behind this may well be the differential compliance of nuclei and white matter tracks to the penetrating cannula.

Intracranial Air We have shown that asleep DBS significantly reduces the incidence of intracranial air, which is likely correlated to shift in brain structures and hence accuracy of electrode placement. Our experience now with over 800 electrode implantations indicates that the appearance of intracranial air is rare (<5%) during asleep DBS implantation. This finding would obviate the need for special imaging procedures, such as iMRI, where the rationale for the intraoperative imaging is to account for brain shift.

Outcomes of Asleep Deep Brain Stimulation Asleep DBS for PD improved motor outcomes over 6 months on par with or better than awake DBS, was superior with regard to speech fluency and quality of life, and should be an option offered to all patients who are candidates for this treatment.

Cost DBS performed asleep was associated with a lower cost variation relative to the awake procedure. Furthermore, costs compared very favorably to UHC-affiliated hospitals (50% of median UHC cost).

197

CONCLUSIONS This method represents a clear departure from the use of MER and physiologic target confirmation during movement disorder surgery. It is our contention that MER has not been, nor will ever be, proven to add value to the implantation of DBS electrodes, or any other movement disorder procedure (Burchiel, 2004).13 Furthermore, there is, in fact, new evidence that MER adds risk to these procedures by way of increased chances of ICH and the production of new neurologic deficits.4 There is little question that MER requires additional expertise, added operative time, and the additional expense of the recording systems and electrodes. Effectively, MER also requires that procedures be performed under local anesthetic, a daunting prospect for most patients. We have demonstrated the feasibility and accuracy of image-guided DBS electrode placement. If this procedure can be shown to produce clinical benefits that are equivalent to the more traditional methods using MER, and by historical comparison, proves to be a safer procedure, then the continuance of routine MER-based DBS electrode implantation should be questioned. MER may well then be relegated to the position of a research technique, potentially requiring additional informed consent by the patient. As the demand for this procedure is expected to rise with the increasing prevalence of PD worldwide, the cost of this procedure is also an important issue. Prolonged duration of the DBS procedure and neurophysiologic assessments with MER may both add significantly to the cost of DBS. Costs associated with asleep DBS at our center were lower than comparable academic healthcare centers performing DBS19 likely because of shorter operative time and elimination of the need for MER. iMRI is another method that has been used to target DBS electrode placement in the asleep patient; however, iMRI is a relatively more expensive and generally less available guidance system at surgical centers performing DBS compared with iCT. Image-guided asleep DBS procedures with iCT afford an accurate, cost-effective, and clinically effective alternative to traditional MER-guided approaches.

REFERENCES 1. Weaver F, Follett K, Stern M, et al. Bilateral deep brain stimulation vs best medical therapy for patients with advanced Parkinson disease. JAMA. 2009;301(1):63e73. 2. Israel Z, Burchiel K, eds. Microelectrode Recording in Movement Disorder Surgery. New York: Thieme Medical Publishers; 2005.

198

Functional Neurosurgery and Neuromodulation

3. Starr PA, Christine CW, Theodosopoulos PV, et al. Implantation of deep brain stimulators into the subthalamic nucleus: technical approach and magnetic resonance imagingeverified lead locations. J Neurosurg. 2002;97: 370e387. 4. Zrinzo L, Foltynie T, Limousin P, et al. Reducing hemorrhagic complications in functional neurosurgery: a large case series and systematic literature review. J Neurosurg. 2012;116:84e94. 5. Umemura A, Jaggi JL, Hurtig HI, et al. Deep brain stimulation for movement disorders: morbidity and mortality in 109 patients. J Neurosurg. 2003;98:779e784. 6. Oh MY, Abosch A, Kim SH, et al. Long-term hardwarerelated complications of deep brain stimulation. Neurosurgery. 2002;50:1268e1276. 7. Kleiner-Fisman G, Herzog J, Fisman DN, et al. Subthalamic nucleus deep brain stimulation: summary and metaanalysis of outcomes. Mov Disord. 2006;21(suppl 14): S290eS304. 8. Videnovic A, Metman LV. Deep brain stimulation for Parkinson’s disease: prevalence of adverse events and need for standardized reporting. Mov Disord. 2008;23:343e349. 9. Kimmelman J, Duckworth K, Ramsay T, et al. Risk of surgical delivery to deep nuclei: a meta-analysis. Mov Disord. 2011;26(8):1415e1421. 10. Rolston JD, Englot DJ, Starr PA, et al. An unexpectedly high rate of revisions and removals in deep brain stimulation surgery: analysis of multiple databases. Parkinsonism Relat Disord. 2016;33:72e77. 11. Burchiel KJ, McCartney S, Lee A, et al. Accuracy of deep brain stimulation electrode placement using intraoperative computed tomography without microelectrode recording. J Neurosurg. 2013;119(2):301e306.

12. Mirzadeh Z, Chapple K, Lambert M, et al. Parkinson’s disease outcomes after intraoperative CT-guided “asleep” deep brain stimulation in the globus pallidus internus. J Neurosurg. 2016;124:902e907. 13. Burchiel K. The future of microelectrode recording. In: Israel Z, Burchiel KJ, eds. Microelectrode Recording in Movement Disorder Surgery. New York: Thieme; 2004: 209e210. 14. Sudhyadhom A, Haq IU, Foote KD, et al. A high resolution and high contrast MRI for differentiation of subcortical structures for DBS targeting: the fast gray matter acquisition T1 inversion recovery (FGATIR). Neuroimage. 2009; 47(suppl 2):T44eT52. 15. Schaltenbrand G, Hassler RG, Wahren W. Atlas for Stereotaxy of the Human Brain. 2nd ed. Stuttgart: Thieme; 1977. 16. NeuroLogica Corporation. Available at: http://www. neurologica.com/ceretom. 17. Ko A, Ozpinar A, Hamzaoglu V, et al. Asleep DBS reduces incidence of intracranial air during electrode implantation. J Funct Stereotact Neurosurg. 2018. https://doi.org/10.1159/ 000488150. 18. Brodsky MA, Anderson S, Seier M, et al. Clinical outcomes of asleep versus awake deep brain stimulation for Parkinson’s disease. Neurology. 2017;89:1e7. 19. Jacob RL, Geddes J, McCartney S, Burchiel K. Cost analysis of awake versus asleep deep brain stimulation: a single academic health center experience. J Neurosurg. 2016;124(5): 1517e1523. 20. Starr PA, Martin AJ, Ostrem JL, et al. Subthalamic nucleus deep brain stimulator placement using high-field interventional magnetic resonance imaging and a skull-mounted aiming device: technique and application accuracy. J Neurosurg. 2010;112:479e490.