deep brain stimulation for parkinson’s disease: surgical ...gaoy3/to...

Deep Brain Stimulation for Parkinson’s Disease: SurgicalTechnique and Perioperative Management

Andre Machado, MD, PhD,1* Ali R. Rezai, MD,1 Brian H. Kopell, MD,2 Robert E. Gross, MD, PhD,3

Ashwini D. Sharan, MD,4 and Alim-Louis Benabid, MD5

1Center for Neurological Restoration, Department of Neurosurgery, Cleveland Clinic Foundation, Cleveland, Ohio, USA2Department of Neurosurgery, Medical College of Wisconsin and Clement J. Zablocki VA Medical Center, Milwaukee,

Wisconsin, USA3Department of Neurology, Emory University, Atlanta, Georgia, USA

4Department of Neurosurgery, Thomas Jefferson University, Philadelphia, Pennsylvania, USA5Functional and Stereotactic Neurosurgery, Joseph Fourier University, Grenoble, France

Abstract: Deep brain stimulation (DBS) is a widely acceptedtherapy for medically refractory Parkinson’s disease (PD). Bothglobus pallidus internus (GPi) and subthalamic nucleus (STN)stimulation are safe and effective in improving the symptomsof PD and reducing dyskinesias. STN DBS is the most com-monly performed surgery for PD as compared to GPi DBS.Ventral intermediate nucleus (Vim) DBS is infrequently usedas an alternative for tremor predominant PD patients. Patientselection is critical in achieving good outcomes. Differentialdiagnosis should be emphasized as well as neurological andnonneurological comorbidities. Good response to a levodopachallenge is an important predictor of favorable long-termoutcomes. The DBS surgery is typically performed in an awake

patient and involves stereotactic frame application, CT/MRIimaging, anatomical targeting, physiological confirmation, andimplantation of the DBS lead and pulse generator. Anatomicaltargeting consists of direct visualization of the target in MRimages, formula-derived coordinates based on the anterior andposterior commissures, and reformatted anatomical stereotacticatlases. Physiological verification is achieved most commonlyvia microelectrode recording followed by implantation of theDBS lead and intraoperative test stimulation to assess benefitsand side effects. The various aspects of DBS surgery will bepresented. © 2006 Movement Disorder Society

Key words: deep brain stimulation (DBS); Parkinson’s dis-ease; stereotaxis; neuromodulation

Deep brain stimulation (DBS) has become a readilyaccepted treatment for medically refractory Parkinson’sdisease (PD). While ablative procedures such as pal-lidotomy and thalamotomy have been proven to be effi-cacious,1–8 they are currently reserved for patients withcontraindications to implantable hardware, those livingin areas without access to programming equipment/ex-pertise, and in countries with limited economicresources.

The three structures most commonly targeted for alle-viating the symptoms of PD are the ventralis intermediusnucleus of the thalamus (Vim), the posteroventral portion

of the internal segment of the globus pallidus (GPi), andthe subthalamic nucleus (STN). Vim is generally chosenfor treating tremor predominant PD with similar efficacyand lower rates of long-term neurological complicationscompared with thalamotomies.9–11 Alternatively, GPiand STN are chosen for treating not only tremor, but alsothe other cardinal symptoms of PD such as rigidity andbradykinesia as well as having benefits on reducing dys-kinesias. In this context, Vim DBS is the least commonsurgery for treating PD.

Both GPi and STN have been shown to have similarefficacy rates.12–14 However, the literature demonstratesa trend that STN may be more efficient in managing thesymptoms of PD.15 Choice of the STN over the GPi isoften based on institutional experiences, surgical andprogramming expertise and preferences, lower currentrequirements with resulting longer battery life, and a

*Correspondence to: Dr. Andre Machado, 9500 Euclid Avenue,Desk S31, Cleveland, OH 44195. E-mail: [email protected]

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mds.20959

Movement DisordersVol. 21, Suppl. 14, 2006, pp. S247–S258© 2006 Movement Disorder Society

S247

more consistent trend for significant reductions in dopa-minergic medication intake with the STN target.16–24

Stimulation of the GPi may be preferred for patients withborderline cognitive changes present preoperatively.STN DBS may have a greater impact than GPi DBS onexecutive frontal lobe functions.25 Neurosurgical exper-tise in stereotaxy, modern imaging, widespread use andease of image guidance, and consistent patient outcomeshave encouraged the popularity of STN, GPi, and VimDBS as a treatment for medically refractory PD. Thenumber of implants has risen dramatically since the firstreports of the therapy by the Grenoble group.26,27 Cur-rently, there are over 400 centers worldwide with over30,000 DBS implants to date. The operative approachvaries from one center to the other, particularly in thecriteria employed for microelectrode-guided recordingand angles of approach to the various subcortical targets.The authors recognize that there are various surgicalstrategies, tools, equipment, and procedures used forDBS surgery. This is because of the experience andpreference of the surgical team, individual center logis-tics, and healthcare system constraints. We provide ageneral overview of the surgical technique and periop-erative management realizing these inherent variations.

PATIENT SELECTION

Patient selection is a critical first step as poorly chosencandidates may not have optimal benefits and have in-creased morbidity. Several factors must be consideredbefore determining if a patient is an appropriate candi-date for DBS surgery. A multidisciplinary approach in-volving the neurosurgeon, neurologist, and neuropsy-chologist is important to determine the appropriatesurgical candidate. It is also important that the diagnosisof idiopathic PD be confirmed prior to proceeding withDBS surgery. Several neurological disorders can mimicthe signs and symptoms of idiopathic PD. Multiple-system atrophy28 and progressive supranuclear palsy arein the differential diagnosis that must be considered.29 Acomplete, thorough neurological examination with atten-tion to cognitive function, eye saccadic movements, pos-tural hypotension, and postural instability, and signs ofcerebellar or corticospinal involvement is important inthe preoperative evaluation.

Key to this assessment is evaluating the surgical can-didate in both the on and off medication states with acorroborating levodopa challenge. Perhaps the best prog-nostic indicator of a patient’s suitability for DBS surgeryis their response to levodopa.18 In general, a levodopachallenge following a 12-hour medication withdrawalshould provide at least a 33% improvement in the motorsection of the Unified Parkinson’s Disease Rating Scale

(UPDRS).30 Patients who do not improve significantlywith levodopa are unlikely to improve with surgery, asthe therapy is largely aimed at increasing on time anddecreasing on–off fluctuations. The levodopa challenge isalso useful to demonstrate to the patient and family theextent that the parkinsonian symptoms are likely to im-prove after surgery. Patients should be educated not toexpect an improvement in function greater than the beston medication period with the concomitant reduction indrug-induced dyskinesias. Tremor is an exception andcan improve with STN DBS even if not significantlyaffected by levodopa. Patients should be advised thataxial and gait symptoms are less likely to be impactedthan appendicular symptoms.

One of the most important presurgical assessments isthe neuropsychological evaluation. The team must becapable of identifying patients with elevated risk forsurgery and those incapable of adjusting to a lifestylecompatible with implanted neurological hardware. Sur-gical management may be temporarily deferred or out-right declined in these patients, depending on their re-sponse to adjunctive psychological counseling. Deepbrain stimulation for PD does not seem to cause perma-nent cognitive changes.31 However, patients who presentwith cognitive decline preoperatively tend to be poorcandidates and may worsen after surgery. A staged pro-cedure is a good alternative to simultaneous bilateralsurgery in elderly patients with borderline cognitivefunction.

Age is one of the major factors determining how apatient will cope with the surgical procedure and behavepostoperatively. The risks of cardiopulmonary events arecertainly higher, as in any other major surgery. Morespecifically to DBS, advanced age is linked to higherincidence of cognitive changes after surgery.32 Youngerpatients tend to recover faster and are usually fully ori-ented in the day following surgery. Older patients aremore prone to experience a period of confusion that maylast for several days. We do not recommend a specificcutoff for age, as the chronological age does not neces-sarily correlate with the biological aging among individ-uals. Older patients without many medical comorbiditieswho still present good mental and physical functionduring the on periods may be considered candidates.Staged or unilateral implantations may be wise alterna-tives for those in the eighth decade of life or older.Depression is common in the PD population. Ideally, thepatient should be counseled and treated prior to surgery,with the goal of maximizing the individual’s capacity tocope with the procedure and postoperative recovery pe-riod. Psychosis may also occur primarily but is often amanifestation of dopaminergic medication side effects.

S248 A. MACHADO ET AL.

Movement Disorders, Vol. 21, Suppl. 14, 2006

Brief periods of medication-induced psychosis may bemedically managed by altering the medication regimen.However, psychosis not related to medication may be areason to defer surgery in a particular patient. The oc-currence of significant mood changes after bilateral DBShas been reported33 but is not common in our experience.Severe anxiety may become a major impediment forawake stereotactic surgery, especially in prolonged op-erations. Counseling prior to surgery might be helpful(see Voon and colleagues in this issue for a more detaileddiscussion of these issues).

PREOPERATIVE MANAGEMENT

A full medical assessment is a necessary part of thepreoperative evaluation, as advanced PD patients tend tobe elderly with significant comorbidities. The risk ofdiscontinuing medications that affect anticoagulation andplatelet aggregation should be weighed against the po-tential benefits in the quality of life offered by DBSsurgery. However, timely discontinuation of these lattermedications is mandatory for stereotactic surgery sinceintracerebral hematomas are the most serious of all po-tential complications from DBS. Any anticlotting medi-cations, including aspirin, ticlopidine, clopidogrel, andall nonsteroidal anti-inflammatory drugs should be dis-continued at least 7 to 10 days preoperatively to ensurethe return of normal blood clotting function.

Arterial hypertension can also increase the risk ofintracranial bleeding during stereotactic procedures andmust be controlled in the weeks prior to surgery. Thenight prior to DBS surgery, the antiparkinsonian medi-cations are typically held to pronounce the Parkinson’ssymptoms at the time of surgery and the families must becounseled regarding their role in facilitating the patientthat day. Alternatively, some centers may chose to admitthe patient the night before to aid with rigidity, bradyki-nesia, and/or freezing that may be profound and dis-abling with the medications held.

Patients should be willing and able, both physicallyand cognitively, to undergo an awake surgical procedurerequiring several hours of immobilization in a stereotac-tic frame. Moreover, surgical candidates need to be ableto remain attentive and cooperative under stressful con-ditions. Realistic expectations of outcomes are also par-amount for both the patient and the family. The patientand family must be willing and able to bear the respon-sibilities for maintaining a chronically implanted hard-ware system.

A prolonged discussion on the short- and long-termeffects of DBS on Parkinson’s disease should be carriedout with the patient, family, and caregivers. DBS oftenresults in dramatic changes in the patient’s motor func-

tion and consequently in independence. Patients may nolonger require as much assistance in caring for them-selves or for the household, which may change the fam-ily’s routine and dynamics. The preoperative neuropsy-chological assessment may be useful in identifying theseissues and preventing them with counseling.

Furthermore, while the goal of the surgery is to restorethe patient’s quality of life, it is also important to discusssafety of indoor and outdoor activities. Releasing thepatient from the severe motor limitations imposed by themedically refractory disease may give the individual asense of aptitude that may be unrealistic due to therestrictions imposed by the concomitant aging vestibular,peripheral nervous, and musculoskeletal systems.

The DBS surgical candidate must be educated regard-ing environmental concerns with implanted hardware.Patients with DBS hardware should not be exposed tolarge magnetic fields. Compatibility exists with only alimited number of magnetic resonance (MR) sequencesin machines that have undergone safety testing.34–36 In-dividuals with other health issues that may require fre-quent MR imaging may not be ideal candidates for DBS.Body coils are particularly dangerous for DBS systemsand their need may pose a contraindication for patientswith spinal disease that require future MRIs. Unfortu-nately, the association of spinal problems with medicallyrefractory Parkinson’s disease is not uncommon and itmay be necessary to judge along with the patient andfamily which disease should be prioritized.

Finally, DBS implantation should be regarded the firststep in an extended management plan. The caretakersand the patient should be committed to treating thepatient within the specialized center for programming,and hardware management issues (such as the need forbattery change) in the near and long term.

UNILATERAL VS. BILATERAL SURGERY

There are various approaches to the surgical proceduredepending on the patient’s symptoms, tolerance of sur-gery, and team preferences. The patients can undergosimultaneous bilateral DBS lead and pulse generatorimplantation the same day, unilateral implantation of theDBS lead and pulse generator the same day, bilateralimplantation of the DBS lead in 1 day and subsequentstaged implantation of the pulse generators, or unilateralDBS lead implantation in 1 day and secondary stagedimplantation of the pulse generator another day. All theseapproaches have been used successfully and depend onvariables such as the patient symptoms, tolerance ofsurgery, team preference, and various other health sys-tem logistics that are individual center–dependent.

SURGICAL TECHNIQUE, PERIOPERATIVE MANAGEMENTE S249


SURGICAL PROCEDURE

Application of Head Frame and Imaging

Surgical strategies and details of procedure vary withthe surgical teams, the equipment they are used to, andthe local healthcare system constraints. Stereotactic sur-geons typically master the use of their institution’sequipment of choice, which are used for a variety ofstereotactic procedures. Among them, DBS is probablyone of the most technically challenging and should notbe considered a beginner’s procedure. A variety of headframes can be used such as the Leksell, CRW, Riechert-Mundinger, and other commercially available systems.Placement of the head frame should be done as close aspossible to surgery in order to minimize the duration ofdiscomfort and risk of displacement. The base of theframe should be placed parallel to a line extending fromthe lateral canthus/orbital floor to the tragus in order toparallel approximately the anterior commissure–poste-rior commissure (AC–PC) line. Ear bars (with ear plugsplaced for patient comfort) may be used to balance theideal position of the frame while applying the pins.Ideally, the head should be centered in the frame, so thatthe midline falls within the center point of the stereotac-tic space defined by the head frame system. Care must betaken such that the edges of the frame should never be incontact with the nose or the occipital/neck area due torisk of skin erosion, which could be encountered duringan extended stereotactic procedure. The frame should notobscure the patient’s eyes since this makes communica-tion with the patient and assessment of eye movementsmore difficult during surgery.

The scalp is shaved and prepared with betadine andthe pins are inserted under local anesthesia of preference,typically lidocaine and/or marcaine. A small dose ofintravenous midazolam or other sedation can be givenprior to insertion of the pins in order to minimize aware-ness and memory of frame placement. The anterior pinscan be placed two finger breadths above the orbital rim,taking care to avoid the supraorbital nerve. Posteriorly,pins should be located so as to avoid penetration of thecerebral venous sinuses. The height of the base of theframe must also be taken into consideration so thatprospective targets lie comfortably within the frame’scoordinate system, avoiding the extremes, which willmechanically hamper positioning the arc for targeting. Inaddition, since the pins can create significant artifactualdistortion of CT imaging, they should be placed such thatthey distort neither the AC, PC, nor the target.

The pins should firmly purchase the outer layer of theskull to prevent the frame from being displaced during aprolonged surgery. Alternatively, overtightening is unde-

sirable due to potential for causing distortions in theframe and consequently in the accuracy of the system.Finding the right balance may be challenging in patientswith severe tremor, who are more prone to cause dis-lodgment of the pins. After application of the frame, anattempt should be made to move the patient’s head andstress the pins. Any perception of pain by the patient atthis point might imply inadequate insertion of the pins.Patients with severe anxiety or low tolerance to pain maynot tolerate awake placement of the head frames. In thoseinstances, the head frame may be placed under heavysedation or general anesthesia.

Once the frame is placed, the patient is taken forpreoperative imaging. CT, MRI, and ventriculographycan all be used to construct the stereotactic imagingdatabase required for DBS implantation. The surgicalteam can utilize any of the above imaging modalities foranatomical targeting of the STN, GPi, and Vim. Manycommercially available image-guided systems offer themerging and image fusion capabilities that allow forintegration of multiple imaging data sets.

The CT scan should include the upper portion of theframe, the anterior and posterior limits of the frame, andvertex up to the skin edge in order to encompass all thefiducials. The slices are optimally obtained with thethinnest slice thickness with no gap between slices. Ifdislodgment of the frame occurs after the volumetricstereo CT, the entire procedure has to be restarted due tothe loss of relationship between the fiducials and intra-cranial structures.

Volumetric, gadolinium-enhanced, T1-weighted MRI,T2-weighted images, or inversion recovery images canbe obtained with the head frame on, or preoperatively toexpedite the process on the day of surgery. T2 andinversion recovery images are beneficial for direct tar-geting of the structure of interest and surrounding areas.Inversion recovery images can be problematic with cer-tain imaging software. Often, excessive tremor hampersthe interpretation of the preoperative MRI, and anotherone may be obtained with the head frame on, which helpsto stabilize the head during image acquisition. In order tooptimize the chances for a good outpatient MRI, neurol-ogists can prescribe levodopa to patients in such fashionas to match on medication time with the time of theexamination (i.e., tremor dominant patients should be inthe on medication state while dyskinetic patients shouldbe in the off medication state). MRI will be contraindi-cated in patients with cardiac pacemakers making reli-ance on CT and ventriculography if necessary. WhileMRI has superior soft-tissue resolution, CT is less sus-ceptible to the distortional artifacts produced by theinhomogeneities in the magnetic field and thus more



accurately represents the actual position of intracerebralstructures in space. Consequently, reliance on imagefusion and merging will allow for visualization of mul-tiple imaging data sets.

Frameless Technology

This is a method that is emerging as an alternative toframe-based stereotaxis. The so-called frameless ster-eotaxis has been used for brain tumors using skin fidu-cials for some time, transforming the process for local-ization of intracranial lesions.37,38 Recently, apreliminary report was released on frameless navigationfor deep brain stimulation utilizing a new technologywith promising results.39 Holloway and colleagues40

have compared the use of this new frameless methodusing skull-based fiducials with standard frame-basedstereotaxis. The authors assessed the actual location ofthe electrode with a postoperative CT scan and comparedthe location with the expected intraoperative targeting.There was no statistically significant difference compar-ing the two surgical methods, suggesting that framelessstereotaxis may have a future role in the field of deepbrain stimulation. Advantages of this method include aless restricted position during surgery and conveniencefor preoperative image acquisition and surgical planning.Disadvantages include the consequent risk of intraoper-ative dislodgment in patients with severe tremor or in-voluntary movements. A learning curve with this newsystem should also be expected for surgeons alreadyaccustomed with frame-based systems. At the currentstage of development, frameless stereotaxis is not thestandard choice for deep brain stimulation given thatonly limited experience exists with its accuracy. Figure 1shows a frameless apparatus and its microdrive mountedin surgery.

Target Selection

Centers use various combinations of anatomical tar-geting strategies. One can target using an indirect methodthat involves reformatted anatomical atlases as well asformula coordinates based on known distances from theAC and the PC landmarks. The STN and GPi can also bechosen on the basis of direct targeting by visualizing thestructure on T2 and inversion recovery images. Withcurrent imaging limitations, Vim can only be targetedindirectly and confirmed with intraoperative physiologi-cal data. Emphasis will be given to anatomical targetingof the STN since it is the most common target used forParkinson’s disease but a brief description for GPi andVim targeting is also included.

Direct Targeting

Direct targeting is based on MRI visualization of thestructures. A target within STN can be chosen at theventral part of the somatosensory STN, which is thelateral part of the nucleus. The axial and coronal T2images are particularly important for adequate visualiza-tion of the STN as a sharp contrast can often be seenbetween the nucleus and the surrounding white matter.The red nucleus can also be clearly seen and the STN liesanterior and lateral to the red nucleus. The anteriorborder of the red nucleus can be used as a landmark forthe anteroposterior localization of the STN target.41 Thisis best visualized in the axial T2 images, as shown inFigure 2.

The GPi and the adjacent optic tract and internalcapsule can be visualized via inversion recovery and T2images. Careful windowing of the images can also pro-vide proper visualization of the internal capsule as wellas the GPi and the pars externa of the globus pallidum(GPe). For GPi DBS, implantation target choice may bedifferent from that used for pallidotomies, which are

FIG. 1. Frameless stereotactic system. The cranial apparatus and mi-crodrive for frameless stereotaxis (Image Guided Neurologics). Thedisposable frame is attached to the skull with three self-tapping screws.The image guidance planning/neuronavigation station is used to alignthe entry point to the target. A microdrive is mounted on top of thehardware, which allows for microelectrode recording and DBS elec-trode implantation.



typically located at the posteroventral portion of thenucleus.1,4,42–44 Implantation at that position may not beoptimal for DBS since the DBS stimulation field mayrequire higher amplitudes and potential spread to theinternal capsule. In this regard, for DBS, a more anteriorand lateral target is preferable, which will allow thestimulation amplitude to be increased without significantinvolvement of the descending white matter fibers. Theoptic tract courses ventrally to the globus pallidus andcan be used as a landmark for the approach (Fig. 3).45,46

Indirect Targeting

Indirect targeting is based on a standardized stereotac-tic atlas and on a formula-derived method based on ACand PC landmarks. While atlas-based coordinates arebiased by the inherent inaccuracies of atlases, coordi-nates can be derived from the experience of teams andrepresent the average values of the coordinates of thesebest contacts, gathered in a rather large number of cases.Referring these coordinates to internal landmarks such asAC–PC length and height of the thalamus above theAC–PC plane provides normalized coordinates less de-pendent on individual variations. The formula uses themidcommissural point (MCP) or the PC as reference,which is calculated after selecting the AC and PC coor-dinates (Fig. 4). Typical initial anatomical coordinatesfor the ventral and sensorimotor STN are 11 to 13 mmlateral to the midline, 4 to 5 mm ventral to the intercom-

missural plane, and 3 to 4 mm posterior to the MCP.Coordinates for the sensorimotor GPi are selected as 19to 21 mm lateral to the midline, 2 to 3 mm anterior to theMCP, and 4 to 5 mm ventral to intercommissural plane.The Vim is targeted 11 to 12 mm lateral to the wall of theIII ventricle, at the level of the intercommissural plane.Most commonly, the goal is to target the topography ofthe upper extremity, which carries the greatest impact forquality of life. The upper extremity occupies a moreintermediate position between the more medially locatedface representation and the laterally located lower ex-tremity. In terms of anterior–posterior position, the Vim

FIG. 2. Direct STN targeting. Axial T2-weighed image showing thesubthalamic nucleus (white arrow) anterior and lateral to the rednucleus (black arrow). Note how the anterior border of the red nucleusaligns with the posterior third of the STN. The T2 images are crucial forthe direct targeting of the STN and should be windowed carefully tooptimize the accuracy of the nuclei boundaries.

FIG. 3. Direct GPi targeting. Direct targeting of GPi. A: GPi (arrow)is seen on an inversion recovery axial image. B: The optic tract (tip ofthe trajectory line) can be identified ventral to the GPi and used as alandmark for direct targeting. Careful windowing is critical to optimizethe resolution and allow accurate identification of the structures.



is typically located between 2/12 and 3/12 of the AC–PCdistance rostral to the posterior commissure.47 Thesecoordinates may not represent the ideal location for im-plantation but are used as initial anatomical targets,which are verified by physiological recordings. Figure 5demonstrates a typical digitized version of the Schalten-brand and Wahren atlas that is morphed, reformatted,and overlaid onto an MRI to conform to the patient’sanatomical structures such as the caudate head, thalamicheight, and brainstem. This atlas-and-formula–based tar-gets are then cross-correlated with the direct visualiza-tion target.

Indirect targeting has been the traditional method instereotactic neurosurgery. The limitation of this methodis that it does not take into account the discrete anatom-ical variations among individuals. Not all brains corre-spond to the topographical distribution depicted in a

given atlas. Even the reformattable atlases available incomputerized planning stations may not conform ade-quately to the patient’s anatomy. In this sense, directtargeting seems to have obvious benefits over indirect. Itis based on the imaging data set of the individual’s ownanatomy. Imaging techniques have advanced dramati-cally, allowing for adequate visualization of the STN inT2 images. However, several pitfalls exist, such as fail-ure of fusion between CT scans and MRI or magneticdistortions that may alter the geometry of the nucleus. Atthe current stage, it is best to perform all targetingmethods and correlate them prior to surgery.

Entry Point Selection

It is necessary to find an approach that will bestfacilitate interpretation of microelectrode recordingswhile concomitantly maximizing safety. From an elec-trophysiological mapping standpoint, a parasagittal ap-proach is preferable but not feasible in all cases. Entrypoints are chosen based on the patient’s anatomy toprovide a safe and optimal trajectory through the areas ofinterest. Typical angles of approach for the STN targetare 15 to 30 degrees from the sagittal plane and 50 to 70degrees in the anterior–posterior direction. The preciseentry point may be refined on the planning console suchthat the trajectory passes through the crown of a gyrusrather than into a sulcus. This avoids inadvertently dam-aging sulcal or pial vessels, which lie on the corticalsurface. The cortical vessels should be avoided as well asthe F–F2 sulcus. After selecting preliminary entry points,the images can be reformatted to the trajectory view,which provides three orthogonal planes positioned withrespect to the trajectory rather than the patient’s anat-omy. The approach then may be traced at millimeterintervals to ensure that no deep sulci are transgressed andthat the ventricular ependyma is not scythed. A typicalapproach will pass the electrode to course through the

FIG. 4. AC–PC coordinate localization. The anterior and posteriorcommissures can be identified on this axial T1-weighed images.

FIG. 5. Digital brain atlas and trajectory. Coronaland sagittal Schaltenbrand and Wahren digitized at-las is morphed, reformatted, and overlaid onto theMRI for the best fit to the individual patient’s anat-omy. Note that this reformatting is not always anoptimal match. The trajectory is targeted to the pos-terior portion of the STN nucleus.



anterior portion of the thalamus, zona incerta, subtha-lamic nucleus, and substantia nigra (Fig. 5). Any giventrajectory must be altered to avoid major visualized vas-cular structures. Venous tributaries draining into the su-perior sagittal sinus or branches of the anterior cerebralartery may limit a more medial approach. Enlarged ven-tricles and the presence of ependymal and intraventricu-lar veins may also force the entry point to be movedlaterally, at the expense of missing the thalamus andconsequently the transition to zona incerta in STN sur-gery. Gadolinium enhancement of the T1-weighted im-ages allows visualization of vessels that would otherwisebe unnoticed. It is clear that accounting for these vesselsminimizes the risk of hemorrhage, although other groupsreport a low rate of bleeding without this method.16 Withregards to the GPi, it is a more lateral structure than theSTN. As such, it is more often possible to find a moreparasagittal tract that will not approach any of the moremedially located vascular structures.

Positioning and Anesthesia

The patient is positioned supine on the operating roomtable, with the knees flexed and the head of the tableelevated to various degrees depending on surgeon pref-erence. The head frame is fixed to the table with anadapter to the table or the floor. The patient’s feedback isused to find a neck position that will best tolerate ex-tended immobility. Certain teams use intravenous seda-tion for the incisions and bony opening until electrophys-iological mapping begins. Although sedation can beused, there is a potential risk for interference with mi-croelectrode recording. Agents such as propofol arequickly reversible and may be preferred over otherlonger-lasting agents. Careful control of the arterialblood pressure is maintained to minimize the risk forintracranial bleeding. Patients with labile blood pressuremay benefit from invasive monitoring for enhanced ti-tration of the antihypertensive drugs. Preoperative anti-biotics should be administered before incision and re-peated thereafter. Draping should be performed to allowvisual access to the patient’s face, arms, and legs whilemaintaining a sterile surgical site.

Older patients or those with degenerative spinal dis-ease may not tolerate the extended awake immobilizationrequired for microelectrode recording (MER). In thesecases, epidural blocks or continuous epidural infusion ofopioids may be attempted to ease the pain.

Stereotactic Approach

The coordinates for the X, Y, and Z coordinates are setand the entry points and the midline are marked on theskin and the incisions are planned. The stereotactic arc

can be used to mark precisely the incision and the burrhole site, taking advantage of the trajectory planning.Depending on the thickness of the skull, the inner rim ofthe burr hole will limit the progression of the cannulas ifthe hole is drilled perpendicular to the skull surface. Inorder to prevent this problem, the holes should be drilledin the axis of the trajectory, at the same angle chosen forthe stereotactic approach. Some teams prefer to openboth incisions and burr holes before starting physiologyon the first side, in order to expedite the process, but thismay increase the time of exposure of the second side. Itis worthwhile to start implantation on the side contralat-eral to the patient’s worst symptoms in case the patientdoes not tolerate a bilateral procedure.

The dura mater, arachnoid, and pia mater on the firstside are coagulated and opened. Some teams do not openthe dura and insert the tube guide by puncture of thedura, thus preventing subdural air penetration and CSFleakage. At this point, sedation can be interrupted toallow the patient to be awakened for microelectroderecording. Depending on the equipment and proceduresof the various teams, different accessories are used tohold the microelectrodes and then the DBS lead. Thecannula is inserted for microelectrode recording to apredetermined dorsal offset to the chosen anatomicaltargeting. A hydraulic or electrical microdrive is used toadvance the microelectrode in submillimetric steps. Ex-cept when dura is not opened, either gelfoam and/orfibrin glue is placed around the cannula in the burr holeto provide a seal, thus minimizing CSF loss andpneumocephalus.

Electrophysiological Mapping

Intraoperative physiology including microelectroderecording and stimulation is discussed separately in thisissue (see Gross and colleagues).

DBS Electrode Implantation and Testing

Various implantation approaches are used, from plac-ing the active contacts in the center of the nucleus, at theventral border, beyond the ventral border, or spanningmultiple areas as determined by the team. After theoptimal location for implantation of the electrode hasbeen decided, the DBS lead (macroelectrode) can beimplanted. The two commercially available electrodeshave four contacts of 1.5 mm height and 1.27 mmdiameter and differ only in the spacing between contacts:1.5 mm in the 3387 model and 0.5 mm in the 3389 model(Medtronic, Minneapolis, MN). Contact 0 (the contact atthe electrode’s tip) can be positioned at the physiologi-cally defined ventral border of the nucleus and the re-maining contacts will span for 10.5 mm (3387) or 7.5



mm (3389) in the trajectory. When an inserting cannulais used, the chosen electrode is loaded in such fashionthat the first contact (and not the dead space at the tip ofthe electrode) is aligned with the tip of the cannula. Othersystems may have different loading structures, depend-ing on the distances between the electrode holder and theoffset to target. Fluoroscopy is used by many centers tomonitor DBS lead movement and migration. Some cen-ters perform AP and lateral X-rays while others performlateral only. In dedicated stereotactic operating rooms,the X-ray tubes and the frame are part of a solid-statesetup, which avoids the instability of portable X-raysystem alignments.

The patient is examined for baseline tremor, rigidity,and bradykinesia. The microdrive can be used to advancethe electrode to the desired target or, alternatively, man-ual advancement can be performed to the target. Fluo-roscopy can be used to confirm that the electrode isassuming a straight trajectory. A microsubthalamotomy,thalamotomy, or pallidotomy effect may occur secondaryto the mechanical insertion of the electrode, with im-provements in PD symptoms.

Testing the DBS electrode for efficacy and side effectsis most commonly performed by all teams. However,certain groups rely extensively on the microelectroderecording and stimulation. In STN surgery, reduction intremor, rigidity, and bradykinesia is expected duringintraoperative macrostimulation. During implantation atthe GPi, less changes in PD symptoms may be foundintraoperatively, but the thresholds to side effects mustbe determined. The electrode parameters for intraopera-tive testing is similar to those for final programming.Typically, a pulse width of 60 to 90 microseconds andrate of 130 Hz are used and bipolar stimulation is per-formed at various combinations in order to assess themost commonly used cathodes. Voltage should be in-creased stepwise and the effects recorded. Particularattention should be directed to determination of sideeffects, including corticospinal activation (upper extrem-ity, lower extremity, and face/tongue), speech changes,conjugate (corticospinal activation) and dysconjugate(activation of the third nerve) eye deviation. For STNtargeting, paresthesias may indicate a more posteriorimplantation of the lead and, if transitory, are not con-sidered a contraindication for implantation. A well-tar-geted electrode typically allows stimulation up to 4 voltswithout causing side effects. Lower thresholds indicatethat the lead may be close to other structures and mayhave to be repositioned. For the GPi, thresholds to sideeffects should be higher, since higher amplitudes may benecessary for therapeutic effect. Capsular activation withsubsequent motor contractions should not be achieved at

less than 1 or 2 volts above the voltage yielding signif-icant beneficial effects. For GPi, the deepest contact mayhave visual side effects from stimulation of the optictract.

If side effects are found at thresholds lower thanexpected, repositioning of the electrode will be neces-sary. It is advised to move by more than 2 mm, as the riskof the electrode falling into the previous tract is high. Itshould also be noted that multiple macroelectrode pene-trations may create fluid-filled spaces that may causestimulation to spread unpredictably and result in unreli-able macrostimulation. These problems are reduced byprocedures using simultaneous insertion of severalmicroelectrodes.

Once the DBS electrode is implanted at the finallocation, it must be secured in place while removing thecannula, carrier, and microdrive. Continuous fluoroscopyis helpful to monitor the potential of electrode displace-ment. Anchoring and securing the lead can be achievedby various techniques depending on the center prefer-ence and expertise. These include securing the lead to theskull with ligature embedded in dental cement,miniplates, and screws, the Medtronic burr hole ring andcap, as well as the Navigue Stim-Loc device (ImageGuided Neurologics, Melbourne, FL). Advantages anddisadvantages exist to each of these methods. Dentalcement and miniplates are not recommended by the leadmanufacturer. Their use may be necessary in peculiarscenarios such as when a craniectomy has to be added tothe burr hole due to unexpected findings and the burrhole rings will not conform to the opening in the skull.The burr hole ring that is provided by the lead manufac-turer can be safely used but a tendency exists to drive theelectrodes deeper when inserting the cap. This can beeasily observed on live fluoroscopy. The Stim-Loc de-vice provides good stabilization of the electrodes withouta tendency to push the electrodes deeper. It is a costlytool that may not be accessible to many centers. Oncesecured, the lead is attached to an extension wire that istunneled to the parietal/occipital region. The excess leadcan be coiled around the burr hole device or along thepath of tunneling to serve as strain relief. The burr holeand the wound should be thoroughly irrigated with rins-ing solution before closing the skin. Depending on thetime that will be allowed between implantation of theelectrode and the pulse generator in a staged procedure,scar tissue can form around the excess wire and burr holedevice. In manipulating the DBS lead, sharp instrumentsor instruments with “teeth” should always be avoidedsince insulation may be inadvertently damaged. Like-wise, when using metal instruments, care should be takento prevent accidental crushing of the wires. Rubber-



capped instruments can be used but the surgeon’s fin-gertips may be the best instruments to deal with thesedelicate devices.

INTRAOPERATIVE ADVERSE EVENTS

Clinical hypervigilance is imperative throughout theprocedure. If there is any intraoperative evidence ofhemorrhage, such as bleeding from the cannula, gentleirrigation down the cannula is performed until the efflu-ent is clear (Fig. 6). It should be noted that the only pathfor the hemorrhage to come out is via the cannula and itshould not be removed prematurely. It is crucial to con-tinue gentle irrigation until the hemorrhage is clearedwhile monitoring the patient for neurological status. Anyneurological deterioration such as the onset of lethargyor a new focal deficit is cause for stopping the procedureand proceeding immediately to the CT scanner. If thepatient cannot protect his or her airway, they should beintubated prior to leaving the operating room. Equipmentto perform an emergent craniotomy should be readilyavailable at all times.

In addition to neurological changes, changes in neu-ronal recordings may indicate the presence of an occultintraoperative hemorrhage. As there are no active neu-

rons within a hemorrhage cavity, a microelectrode thatencounters or produces a hemorrhage along its track willrecord only silence. Therefore, unexpected electrical si-lence not attributable to problems with the electrodeitself or its position (such as in the internal capsule)should raise the suspicion of an intraoperative hemor-rhage. In addition, deviation of the electrode tip onfluoroscopy after placement not only signals the presenceof the development of an intracranial hemorrhage, butalso provides some information as to the magnitude ofthe problem.

Implantation of Pulse Generator (IPG)

This is the last step of surgery and is performed undergeneral anesthesia. It can be performed the same day orin a delayed/staged fashion. Due to the long duration ofthe procedure and the need for general anesthesia, IPGscan be implanted several days after the DBS lead im-plantation when the patient has recovered from the in-tracranial procedure.

For IPG implantation, the patient is positioned supine,with the head turned to the opposite side of the intendedsite of IPG implantation. Preoperative antibiotics areagain administered 30 minutes prior to incision. A sub-cutaneous pocket is then created for the IPG and thispocket is connected to the DBS lead tunneled previouslyto the parietal/occipital region. The most common loca-tion for the IPG placement is infraclavicular and typi-cally marked 2 cm below the clavicle and 4 cm awayfrom the midline or 2 cm away from the lateral manu-brial border. However, certain patients may requireplacement in other locations due to body habitus (verythin patients), age (pediatric patients), a history of priorsurgery in the region, or vanity. In addition, certainactivities such as hunting require the use of the chest tostabilize equipment. The IPG should be placed in alocation that minimizes pressure and trauma to the unit.Anecdotal reports cite local trauma as an infection riskwith other implanted hardware systems. Other locationsinclude the subcostal area and the flank as well as thelumbar region and buttocks. All of these alternate sitesrequire the use of a longer extension lead. The standardextension is 51 cm in length, but a 66 cm version is alsoavailable.

Regardless of the location, it is useful to mark theincisions before the final patient position to achievemaximal symmetry and cosmetic appeal. Marking thesubclavicular incision in women deserves special atten-tion for cosmesis. It may be useful to mark the incisionsin the standing or sitting positions in order to best predictthe final location of the incisions and generators. In mostindividuals, it is possible to create a subclavicular sub-

FIG. 6. Intraoperative hemorrhage. Axial CT scan image of a hema-toma on the left side of a bilateral GPi implantation. Note the alreadyimplanted DBS electrode on the right side. Active bleeding was seenfrom the cannula during microelectrode recording and gentle irrigationwas performed with a spinal needle through the cannula. The cannulawas removed only after active bleeding stopped. The patient had nochange in neurological status. The procedure was aborted and thepatient taken to the CT scan.



cutaneous pocket that is deep enough for hardware im-plantation, just against the pectoral aponeurosis. In verythin patients, it may be necessary to create the pocket inbetween the pectoralis muscle and its fascia, althoughextra care must be taken as an increased risk for postop-erative hematoma formation may exist. Additionally, forimplantation of the IPG in a nonsubclavicular area, ut-most care should be taken to prevent approximation ofthe IPG to any bony prominence such as the rib or theiliac crest.

A small parietal/occipital incision allows identificationof the connector that protects the DBS electrode. Exter-nalization of the lead must be carefully performed sinceit can be tethered by scar tissue formation. Excessiveforce should be avoided to prevent breakage of the elec-trode or migration of the intracranial portion. A tunnel iscreated from the parietal region to the IPG pocket and the51 cm implantable extension wire is passed and con-nected to the distal aspect of the DBS lead and IPG. Theconnector is internalized and pulled distally but shouldnever be placed at the level of the neck below themastoid, where mobility tends to cause tethering andresultant hardware failure can occur. A good position forthe connector is the retroauricular region (Fig. 7). Whenlocated too medially, it can cause pain when sleepingsupine and may come in contact with the lesser or greateroccipital nerves. If located too laterally, it may causediscomfort while wearing glasses. Some teams place ithigher on the head, at the upper part of the temporalmuscle. The excess extension lead is coiled behind thegenerator and the IPG is anchored to the fascia withnonabsorbable sutures. Anchoring of the generator tosubcutaneous fat should be avoided since caudal migra-tion may occur due to the weight of the IPG. Closure isperformed in layers after copious irrigation. The patient

is re-draped for the contralateral side or when using aKinetra, both extension leads are passed under the skinon the same side. The Kinetra is a larger device that canbe used to control two DBS electrodes, thus allowing fora single-side IPG implantation. The disadvantages to thissystem include the greater volume that may impose a riskof erosion in very thin patients and the limitation to asingle set of pulse frequencies.

POSTOPERATIVE MANAGEMENT

In the immediate hours after surgery, it is important tokeep arterial blood pressure in the normal range. Inaddition, the patient’s preoperative PD regimen shouldbe restarted immediately after surgery to avoid problemswith dopaminergic withdrawal. Patients should undergopostoperative CT scans and/or MRI scans to assess theelectrode location and intracranial status. In addition,plain X-rays are obtained to assess the location andgeometry of the leads and hardware. Parkinson’s medi-cations may need to be adjusted depending on the pa-tient’s status. Cognitive and behavioral changes mayoccur in the postoperative period, particularly in olderpatients. Patients can be discharged as early as 24 hoursafter surgery, depending on their neurological and cog-nitive status.

REFERENCES

1. Bakay RA, Starr PA, Vitek JL, DeLong MR. Posterior ventralpallidotomy: techniques and theoretical considerations. Clin Neu-rosurg 1997;44:197–210.

2. de Bie RM, Schuurman PR, de Haan PS, Bosch DA, Speelman JD.Unilateral pallidotomy in advanced Parkinson’s disease: a retro-spective study of 26 patients. Mov Disord 1999;14:951–957.

3. de Bie RM, Schuurman PR, Bosch DA, de Haan RJ, Schmand B,Speelman JD. Outcome of unilateral pallidotomy in advancedParkinson’s disease: cohort study of 32 patients. J Neurol Neuro-surg Psychiatry 2001;71:375–382.

4. Vitek JL, Bakay RA, DeLong MR. Microelectrode-guided pal-lidotomy for medically intractable Parkinson’s disease. Adv Neu-rol 1997;74:183–198.

5. Vitek JL, Bakay RA. The role of pallidotomy in Parkinson’sdisease and dystonia. Curr Opin Neurol 1997;10:332–339.

6. Narabayashi H, Maeda T, Yokochi F. Long-term follow-up studyof nucleus ventralis intermedius and ventrolateralis thalamotomyusing a microelectrode technique in parkinsonism. Appl Neuro-physiol 1987;50:330–337.

7. Hitchcock E, Flint GA, Gutowski NJ. Thalamotomy for movementdisorders: a critical appraisal. Acta Neurochir (Wien) 1987;39(Suppl.):61–65.

8. Fox MW, Ahlskog JE, Kelly PJ. Stereotactic ventrolateralisthalamotomy for medically refractory tremor in post-levodopa eraParkinson’s disease patients. J Neurosurg 1991;75:723–730.

9. Alesch F, Pinter MM, Helscher RJ, Fertl L, Benabid AL, KoosWT. Stimulation of the ventral intermediate thalamic nucleus intremor dominated Parkinson’s disease and essential tremor. ActaNeurochir (Wien) 1995;136:75–81.

10. Tasker RR. Deep brain stimulation is preferable to thalamotomyfor tremor suppression. Surg Neurol 1998;49:145–153.

FIG. 7. Lateral skull X-ray. A lateral skull X-ray showing bilateralSTN DBS electrodes. Note the location of the connectors in theretroauricular region how the excess wires were coiled at the burr holesite.



11. Tasker RR, Munz M, Junn FS, et al. Deep brain stimulation andthalamotomy for tremor compared. Acta Neurochir 1997;68(Suppl.):49–53.

12. Burchiel KJ, Anderson VC, Favre J, Hammerstad JP. Comparisonof pallidal and subthalamic nucleus deep brain stimulation foradvanced Parkinson’s disease: results of a randomized, blindedpilot study. Neurosurgery 1999;45:1375–1382.

13. Vitek JL. Deep brain stimulation for Parkinson’s disease: a criticalre-evaluation of STN versus GPi DBS. Stereotact Funct Neurosurg2002;78:119–131.

14. Deep-Brain Stimulation for Parkinson’s Disease Study Group.Deep-brain stimulation of the subthalamic nucleus or the parsinterna of the globus pallidus in Parkinson’s disease. N Engl J Med2001;345:956–963.

15. Peppe A, Pierantozzi M, Bassi A, et al. Stimulation of the subtha-lamic nucleus compared with the globus pallidus internus in pa-tients with Parkinson disease. J Neurosurg 2004;101:195–200.

16. Krack P, Batir A, Van Blercom N, et al. Five-year follow-up ofbilateral stimulation of the subthalamic nucleus in advanced Par-kinson’s disease. N Engl J Med 2003;349:1925–1934.

17. Benabid AL, Krack PP, Benazzouz A, Limousin P, Koudsie A,Pollak P. Deep brain stimulation of the subthalamic nucleus forParkinson’s disease: methodologic aspects and clinical criteria.Neurology 2000;55(12 Suppl. 6):S40–S44.

18. Charles PD, Van Blercom N, Krack P, et al. Predictors of effectivebilateral subthalamic nucleus stimulation for PD. Neurology 2002;59:932–934.

19. Herzog J, Volkmann J, Krack P, et al. Two-year follow-up ofsubthalamic deep brain stimulation in Parkinson’s disease. MovDisord 2003;18:1332–1337.

20. Kleiner-Fisman G, Fisman DN, Sime E, Saint-Cyr JA, LozanoAM, Lang AE. Long-term follow up of bilateral deep brain stim-ulation of the subthalamic nucleus in patients with advanced Par-kinson disease. J Neurosurg 2003;99:489–495.

21. Krack P, Limousin P, Benabid AL, Pollak P. Chronic stimulationof subthalamic nucleus improves levodopa-induced dyskinesias inParkinson’s disease. Lancet 1997;350:1676.

22. Limousin P, Pollak P, Benazzouz A, et al. Bilateral subthalamicnucleus stimulation for severe Parkinson’s disease. Mov Disord1995;10:672–674.

23. Limousin P, Krack P, Pollak P, et al. Electrical stimulation of thesubthalamic nucleus in advanced Parkinson’s disease. N EnglJ Med 1998;339:1105–1111.

24. Krack P, Pollak P, Limousin P, et al. Subthalamic nucleus orinternal pallidal stimulation in young onset Parkinson’s disease.Brain 1998;121(Pt. 3):451–457.

25. Jahanshahi M, Ardouin CM, Brown RG, et al. The impact of deepbrain stimulation on executive function in Parkinson’s disease.Brain 2000;123(Pt. 6):1142–1154.

26. Limousin P, Pollak P, Benazzouz A, et al. Effect of parkinsoniansigns and symptoms of bilateral subthalamic nucleus stimulation.Lancet 1995;345:91–95.

27. Benabid AL, Pollak P, Gross C, et al. Acute and long-term effectsof subthalamic nucleus stimulation in Parkinson’s disease. Ster-eotact Funct Neurosurg 1994;62:76–84.

28. Consensus Committee of the American Autonomic Society and theAmerican Academy of Neurology. Consensus statement on thedefinition of orthostatic hypotension, pure autonomic failure, andmultiple system atrophy. Neurology 1996;46:1470.

29. Litvan I, Agid Y, Calne D, et al. Clinical research criteria for thediagnosis of progressive supranuclear palsy (Steele–Richardson–Olszewski syndrome): report of the NINDS–SPSP internationalworkshop. Neurology 1996;47:1–9.

30. Lang A, Widner H. Deep brain stimulation for Parkinson’s disease:patient selection and evaluation. Mov Disord 2002;17:S94–S101.

31. Morrison CE, Borod JC, Perrine K, et al. Neuropsychologicalfunctioning following bilateral subthalamic nucleus stimulation inParkinson’s disease. Arch Clin Neuropsychol 2004;19:165–181.

32. Saint-Cyr JA, Trepanier LL, Kumar R, Lozano AM, Lang AE.Neuropsychological consequences of chronic bilateral stimulationof the subthalamic nucleus in Parkinson’s disease. Brain 2000;123(Pt. 10):2091–2108.

33. Berney A, Vingerhoets F, Perrin A, et al. Effect on mood ofsubthalamic DBS for Parkinson’s disease: a consecutive series of24 patients. Neurology 2002;59:1427–1429.

34. Sharan A, Rezai AR, Nyenhuis JA, et al. MR safety in patientswith implanted deep brain stimulation systems (DBS). Acta Neu-rochir 2003;87(Suppl.):141–145.

35. Rezai AR, Phillips M, Baker KB, et al. Neurostimulation systemused for deep brain stimulation (DBS): MR safety issues andimplications of failing to follow safety recommendations. InvestRadiol 2004;39:300–303.

36. Baker KB, Tkach JA, Nyenhuis JA, et al. Evaluation of specificabsorption rate as a dosimeter of MRI-related implant heating. JMagn Reson Imaging 2004;20:315–320.

37. Barnett GH, Nathoo N. The modern brain tumor operating room:from standard essentials to current state-of-the-art. J Neurooncol2004;69:25–33.

38. Gumprecht HK, Widenka DC, Lumenta CB. BrainLab VectorVi-sion Neuronavigation system: technology and clinical experiencesin 131 cases. Neurosurgery 1999;44:97–104.

39. Henderson JM. Frameless localization for functional neurosurgicalprocedures: a preliminary accuracy study. Stereotact Funct Neu-rosurg 2004;82:135–141.

40. Holloway KL, Gaede SE, Starr PA, Rosenow JM, RamakrishnanV, Henderson JM. Frameless stereotaxy using bone fiducial mark-ers for deep brain stimulation. J Neurosurg 2005;103:403–413.

41. Andrade-Souza YM, Schwalb JM, Hamani C, et al. Comparison ofthree methods of targeting the subthalamic nucleus for chronicstimulation in Parkinson’s disease. Neurosurgery 2005;56(Suppl.2):360–368.

42. Vitek JL, Bakay RA, Hashimoto T, et al. Microelectrode-guidedpallidotomy: technical approach and its application in medicallyintractable Parkinson’s disease. J Neurosurg 1998;88:1027–1043.

43. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventralpallidotomy in the treatment of Parkinson’s disease. J Neurosurg1992;76:53–61.

44. Laitinen LV. Pallidotomy for Parkinson’s disease. Neurosurg ClinN Am 1995;6:105–112.

45. Starr PA, Turner RS, Rau G, et al. Microelectrode-guided implan-tation of deep brain stimulators into the globus pallidus internus fordystonia: techniques, electrode locations, and outcomes. Neuro-surg Focus 2004;17:E4.

46. Starr PA. Placement of deep brain stimulators into the subthalamicnucleus or globus pallidus internus: technical approach. StereotactFunct Neurosurg 2002;79:118–145.

47. Guiot G, Derome P. The principles of stereotaxic thalamotomy. In:Schneider R, Kahn E, Crosby E, editors. Correlative neurosurgery.Springfield, IL: Charles C. Thomas; 1982. p 481–505.



deep brain stimulation for parkinson’s disease: surgical ...gaoy3/to...

Documents