18 The use of intraoperative microelectrode recording during surgery for movement disorders is controversial. Although many surgeons believe that MER is essential for the accurate localization of deep targets in the brain, others believe that techniques of preoperative high-quality imaging combined with intraoperative macrostimulation suffice. Proponents of the use of MER report that errors in target localization arise from multiple factors, including poor definition of the borders of deep brain structures and their subnuclei on magnetic resonance imaging, errors arising from MRI–related spatial shifts, methods of data deformation for digitized atlas images, variation in the anatomy between patients, errors associated with the mechanical accuracy of stereotactic frames, shifts in brain targets associated with intraoperative leakage of cerebrospinal fluid and patient positioning, and errors arising from lack of adequate clinical monitoring of stimulation-induced clinical effects or poor patient cooperation during surgery. Intraoperative MER gathers data that are independent of these factors and that can be helpful in localizing the target.1 Opponents of the use of MER believe that MER is not simple, requires the expertise of different disciplines, requires sophisticated and expensive operating room equipment, prolongs operative time, and may increase the risk of intracranial hemorrhage.2 They also claim that MER is unnecessary and that the use of techniques of preoperative high-definition imaging combined with intraoperative macrostimulation provides adequate clinical information for accurate target localization.3 Unless studies using one method or the other demonstrate different clinical outcomes, the controversy will most likely continue, and surgeons will keep using either technique based on their experience. This chapter reviews literature that correlates MER with clinical outcome as well as with the author’s clinical experience. Because the requirement for MER may be different for ablative and stimulation techniques and for different brain targets, the chapter reviews data on the use of MER for each of these surgeries. One may argue that the debate on whether MER during movement disorder surgery is essential or not could easily resolve if differences in clinical outcome using either method can be demonstrated. This is not as simple as it seems, however, for the following reasons. Movement disorder surgery comprises a multitude of various surgical procedures that target different brain structures.4 Methods that apply to one target (e.g., subthalamic nucleus) may not apply to another (e.g., globus pallidus). Even within the same brain structure (e.g., thalamus), different subnuclei have been targeted by different surgeons.5 Therefore, it is imperative that the use of MER in movement disorder surgery be studied for specific brain targets. Data pertinent to ablative surgery may not apply to neurostimulation surgery. Proponents of MER emphasize the importance of appropriately centering the lesion in the desired target to minimize complications and maximize effectiveness. Opponents of MER believe that the relatively large size of the lesion in ablative surgery negates the concept of a “micro” target and thus makes the use of MER unnecessary. A similar argument holds for neurostimulation therapy. Opponents of MER report that the relatively large size of stimulation and the multitude of lead contacts allow for some inaccuracy in lead placement. Proponents of MER report that lead contacts should still be centered on the target to maximize efficacy and minimize stimulation-induced side effects and that the multitude of lead contacts allows for errors in lead placement in the rostrocaudal direction but not in the anteroposterior or mediolateral directions.6 Different surgeons use different imaging techniques for target localization. The need for MER to refine the spatial coordinates of a target that is defined by a specific imaging technique may not apply to the same target when defined by a different imaging technique. Even when the same imaging technique is used, different institutions may have different results because of variations in machine accuracy such as that related to algorithms used to correct for spatial accuracy of MRI. Some surgeons use targeting techniques relative to the midcommissure, whereas others use direct anatomical targeting.7 Those who use direct anatomical targeting employ different targeting methods that may influence the accuracy of the spatial coordinates of the target.7–10 Surgeons who target brain structures relative to the anterior and posterior commissures (AC, PC) or the midcommissure may use different brain atlases.11,12 Those who use the same brain atlas may utilize different target coordinates and different atlas deformation techniques.13 Some surgeons scale the brain atlas to fit to the individual patient’s brain based on the length of the AC–PC line, whereas others scale it to fit to the size of the entire brain.14 Some digitized atlases use linear deformation algorithms, and others use nonlinear deformation algorithms.8 Some surgeons use MRI for targeting, whereas others use computed tomography, ventriculograms, or image fusion techniques.13 Scanning algorithms vary, resulting in differences in the quality of images used for preoperative planning. Some surgeons utilize complex mathematical computation to correct frame versus image tilts, others use computerized image tilt correction software, and still others do not use methods to correct for frame tilts.15 A discrepancy in the coordinates of targets has been reported when different methods of targeting are applied.16 Surgery is performed in various ways, using different equipment and techniques. Patient positioning, the type of stereotactic frame, the use of a burr hole versus a twist-drill hole, the amount of leakage of CSF allowed during surgery, lead trajectory, and the method of patient monitoring during surgery are variables that can affect clinical outcome. The method of macrostimulation, which is not standardized, can influence the data obtained by this method. Some surgeons monitor stimulation-induced electromyography changes during surgery, whereas others do not.17 Some surgeons rely on impedance measurements, whereas others do not.18 Different radio frequency generators measure impedances differently and may generate different impedance data for the same brain target.19 Macrostimulation using constant current yields different data than macrostimulation using constant voltage.19 The macroelectrode used for stimulation varies in size between different institutions, resulting in different current spread to adjacent structures. Stimulation parameters and stimulation thresholds that influence the decision to change the electrode trajectory vary among different investigators.13 Even when all applies, different patients may have different amounts of CSF pockets that surround the stimulation electrode and that can influence the data obtained by macrostimulation. The method of intraoperative microrecording is not standardized. Many surgeons claim the use of microrecording technique but instead perform field recordings.13 Those who use single-unit recordings do so differently. Whereas some surgeons use MER merely to verify passage through the target, others use it to define the inferior border of the target, to identify the longest trajectory through the targeted structure, or to map all the boundaries of a structure.20–28 Some surgeons map kinesthetic cells, whereas others record only spontaneous activity.29 Some surgeons utilize the technique of advancing several parallel-placed microelectrodes simultaneously, whereas others record from several subsequent trajectories.22 Some surgeons superimpose scaled reconstructions of the microelectrode tracks on outlines of parasagittal atlas-based brain slices, whereas others do not.29 There are no standard imaging techniques that properly localize the lesion or the implanted electrodes postoperatively. Although some series submit imaging techniques that may effectively localize lesions or implanted electrodes postoperatively, these techniques require sophisticated computer software that is hardly available, or their validity has not been adequately tested, or they have not been used by other investigators to allow valid comparison between published series.9,30,31 A critical review of most series that published postoperative images demonstrates that those images do not allow the accurate localization of lesions or implanted electrodes within a millimeter, accuracy that is claimed by those series.7,32–34 Spatial shifts on MRI of the implanted electrodes, the variability in the appearance of radio frequency–induced lesions on MRI at different postoperative time periods, poor definition on MRIs of the borders of deep brain structures, the difference in the plane of imaging and the patient’s head position between preoperative and postoperative MRIs, variability of pixel size, and the limited availability of software that can accurately and simultaneously reconstruct the lesion or the implanted electrode in more than one MR imaging plane are all factors that limit the ability to locate the lesion or the implanted electrodes in relationship to the three-dimensional structure of the brain and that can make comparison of preoperative and postoperative images difficult. A review of published postoperative MRIs demonstrated a discrepancy in the location of the electrode when assessed by the publishing surgeon relative to the location of the electrode when assessed by an independent reviewer.3 Postoperative images are infrequently obtained with the stereotactic frame still attached to the patient’s head, a method that allows accurate postoperative localization of the ablative lesion or of the implanted electrode relative to the midcommissure. Even under the best imaging circumstances, a patient-specific atlas may be required for the proper assessment of the location of the lesion or the implanted electrode. Several series report that MER improves targeting accuracy and thus improves the results of surgery. In most of these studies, investigators based their conclusions on clinical impressions that are not substantiated by comparative outcome analyses.22 Several series report that MER changes the final location of the electrode compared with that initially planned by preoperative imaging techniques; however, these series do not report whether such an adjustment of the target location could have been predicted simply by macrostimulation methods or whether such an adjustment did indeed change the clinical outcome. The author suggests that reviewing the cases of patients who fail to achieve good results may be more informative in determining the role of MER in movement disorder surgery than by reviewing overall results of movement disorder surgery. In summary, various surgical targets, different preoperative planning methods, various intraoperative surgical techniques and equipment, different methods of localization of targets postoperatively, and different methods of reporting clinical outcome make it difficult to compare the published results of surgery for movement disorders with or without the use of MER or to make definite recommendations regarding one technique over the other. A better approach may be to define the target for each surgery and study whether this target can be better defined by MER or by macro-stimulation techniques. Most surgeons consider the cerebellar receiving area of the thalamus, the ventral intermediate nucleus, to be the optimal target for thalamotomy.35,36 Different methods have been used to define this region of the thalamus. One method involves lesioning cells with tremor-synchronous activity.37 Such cells are clustered 2 mm anterior to the ventral caudal nucleus and 3 mm above the AC–PC line.37 Alternatively, targets have been placed anterior to the site at which evoked potentials can be recorded in response to cutaneous stimulation of the fingers.38 Lesions have been made in the region where intraoperative electrical stimulation produced tremor relief and anterior to the region where electrical stimulation evoked sensations.39 Lesions have been made in sites that contained kinesthetic and voluntary cells; however, not all kinesthetic cell sites where stimulation arrests tremor are uniformly effective as a target for thalamotomy.40 Finally, lesions were placed in the region where cells responded to contralateral passive or active muscle limb movements and where electrical stimulation effected tremor.36 In one series, the difference in the spatial coordinates between a thalamotomy target that is defined by MER and a thalamotomy target that is defined by its relative coordinates to AC–PC as visualized on CT imaging averaged 7.8 mm; however, clinical correlation was not attempted.41 There are several series that correlated clinical outcome with MER. The zones where kinesthetic or tremor-synchronous cells occur and where intraoperative electrical stimulation suppresses tremor overlap, but not all zones where kinesthetic or tremor cells are recorded or where intraoperative stimulation arrests tremor are uniformly effective in permanently suppressing tremor.39,42 The preferred thalamotomy site is selected out of a larger volume containing tremor cells and tremor arrest sites, anterior to the rostral border of the tactile cell.43 In one series, 60 mm3 lesions created within 2 mm of the center of a cluster of tremor-synchronous cells produced long-term tremor control, whereas similar lesions created greater than 2 mm away produced only transient tremor control.44 In another series, long-term tremor control and less postoperative complications were achieved when the thalamotomy target was not only located among kinesthetic or voluntary tremor cells in Vim or caudal ventral oral posterior at sites where acute stimulation most effectively arrested tremor, but also lay 2 to 3 mm above the AC–PC line.40 Other investigators share this opinion.44,45 The good outcome that was achieved by creating large thalamotomy lesions centered on an MER-defined target led Tasker’s group to create lesions that are similar in location but smaller in size, in an attempt to further reduce potential complications related to heat spreading to surrounding structures. Out of 40 patients who underwent such selective MER-guided thalamotomy, 75% had tremor subside or significantly improve but at the expense of having to repeat surgery on 11 (28%) patients. Tasker concluded that larger lesions are still required to achieve good clinical outcome, discrediting the concept of MER-guided “microthalamotomy” lesions.25 This conclusion is not shared by other investigators, who demonstrated that lesion location rather than lesion size is the decisive factor in long-term tremor control after thalamotomy.8 Several series that do not utilize MER report clinical results that are comparable to those reported by series that utilize MER. In one series of 43 thalamotomy procedures performed without MER for essential tremor, 60% of the patients achieved tremor relief, and another 14% retained mild residual tremor that did not interfere with daily activities; 12% of the patients had tremor recurrence 1 to 13 months later, and 14% of the patients developed permanent hemiparesis or speech difficulty.46 In another series of 57 patients who underwent thalamotomy without MER, tremor subsided in 79% of patients and significantly improved in another 22%; tremor recurred in 9% of the patients, and permanent complications occurred in 7%.47 Other series of thalamotomy without MER reported similar results.48,49 In general, similar rates for tremor control and complications following thalamotomy have been reported in series that utilize or do not utilize MER. Groups that used both techniques reported better outcomes when MER was used but not when smaller lesions were created. One may assume that MER allows the proper centering of the lesion to achieve long-term tremor control and reduce complications, but no prospective studies are available to allow definitive conclusions. A similar target to that of thalamotomy has been proposed for thalamic DBS.33,46,50,51 In Benabid et al’s series where the final electrode position was confirmed by ventriculography, the optimal tremor arrest site was located 4 to 8 mm anterior to the PC, 12 to 15 mm lateral to the midline, and 0 to 2 mm superior to the AC–PC.33 These coordinates correspond to Vim, although Vop is probably also affected by stimulation current spread. Capparros-Lefebvre et al52 reported a patient with Parkinson’s disease who died 43 months after successful thalamic DBS. Postmortem examination of the brain located the stimulator in the medial inferior part of Vim, at the entrance of the cerebel–lothalamic fibers. Other postmortem series reported similar findings.53 Although Vim seems to be the target for both thalamic DBS and thalamotomy, it seems that the anterior part of Vim is the preferred target for DBS, whereas the posterior part of Vim is the preferred target for thalamotomy.8,33 Benabid et al, who reported the largest experience with thalamic DBS, routinely used MER to properly place deep brain stimulator electrodes.33 One must note, however, that Benabid and colleagues implanted monopolar electrodes in the majority of their patients. To achieve best results, monopolar electrodes require strict accuracy in their positioning because of their relatively limited field of stimulation. Several investigators have successfully implanted quadripolar electrodes without intraoperative microrecordings, guided by digital imaging studies and intraoperative macrostimulation techniques.54,55 Some speculate that a certain margin of error in electrode placement can be allowed for quadripolar electrodes because of the relatively large field of stimulation that can be achieved with such electrodes.56 In one series of 94 patients who underwent DBS using intraoperative macrostimulation without MER and were prospectively followed for an average of 1 year, 89% of the patients reported excellent to marked improvement of tremor, with no persistent morbidity related to surgery.54 In another series of 11 patients who underwent DBS without MER, 10 patients achieved excellent tremor control.55 Other investigators adhere to the use of intraoperative microrecording for implanting thalamic deep brain stimulators even if quadripolar leads are used.23 It seems that with the use of quadripolar electrodes, MER during thalamic DBS may not add much to the overall long-term efficacy of the surgery. Because tremor is easily monitored clinically during surgery, several investigators rely solely on effective intraoperative tremor control using low parameter settings as the criterion for choosing final lead placement. Inadequate intraoperative assessment can arise if the patient experiences a microthalamotomy effect during surgery losing the ability to test stimulation-induced tremor arrest. Although microthalamotomy effect may be a sign of good electrode positioning,57 the relationship between intraoperative microthalamotomy effect and long-term tremor control has not been studied for DBS as well as it has been for thalamotomy. Currently there are no published series that compare long-term tremor control after DBS with or without the use of MER. For best results after pallidotomy it seems that the lesion should be placed in the sensorimotor part of the posteroventral component of the globus pallidus internalis.5 The lesion should be close enough to the optic tract without involving it and may extend to the ansa lenticularis.58,59 Within the posteroventral GPi, anteromedial lesions may be associated with greater improvement in rigidity and dyskinesia, and centrally located lesions may be associated with greater improvement in akinesia and gait.60 An accurately placed lesion in the posteroventral portion of the GPi is important to achieve good clinical results.61–63 Several series have been published that define methods for electrophysiologically defining the posteroventral region of the GPi during pallidotomy. These series demonstrate a significant variability in methodology, including that of MER. Some surgeons use MER to confirm passage through the GPi and to define the pallidal inferior border to avoid lesioning the optic tract.20 Other surgeons perform extensive examination of motor responses to define kinesthetic cells and to locate the sensorimotor component of the GPi.24 Other investigators use multiple simultaneous or sequential tracts during MER to define all borders of the GPi, then choose a site within the nucleus borders for lesioning.28,64 Investigators who do not use MER have utilized several other methods to define the pallidotomy target. Some surgeons perform field recordings to identify the GPi by its characteristic increased background noise activity and to identify the inferior pallidal border by the characteristic sudden drop of neuronal background activity.13 Other investigators record evoked potentials from the optic tract to create lesions close to it.65 Other surgeons perform impedance measurements to localize the pallidotomy target.66 In one series, impedance increased by 25 to 35 Ohms at 1 to 3 mm before the center of the pallidotomy target, then dropped abruptly when the electrode passed the target.67 Other investigators rely solely on macrostimulation techniques to define the surrounding structures, including the internal capsule and the optic tract.3 Numerous investigators support the use of MER in pallidotomy to localize the sensorimotor area of the internal pallidum and to avoid damage to adjacent structures, based on the following observations. Several investigators reported significant discrepancy between MRI-guided targets and MER-guided targets. In one series, MER-guided targets were more posterior and lateral to MRI-guided targets in the majority of patients. The difference between the two targets measured 2.3 ± 1.55 mm in the mediolateral coordinates and 3 ± 1.9 mm in the anteroposterior coordinates. The actual target overlapped the MRI-guided target in less than half of the patients.68 In another series, the first recording tract, as determined by MRI, was located outside the sensorimotor area of the GPi in around 60% of the patients and completely outside the GPi in less than half of the patients. In none of the patients was the center of the MER-guided target coincident with the center of the MRI-guided target. In more than a third of the patients, the difference in the centers of both targets measured > 3 mm in the anteroposterior direction; in 17% of the patients, the difference in the centers of both targets measured > 3 mm in the medial-lateral direction.69 One series compared the location of an MER-based pallidotomy lesion with the initial CT-based target. The MER-based lesion was consistently anterior and superior to the initial CT-based target by a vector that averaged 6.7 mm. A refined targeting algorithm based on these findings could only decrease the vector to 3.9 mm.41 Another series that compared the accuracy of MRI-guided versus CT-guided pallidotomy demonstrated that MRI-guided targets were closer to the final targets than CT-guided targets; however, MRI-guided targets still required as many changes in their position as did CT-guided targets.70 In another series, MER-guided targets varied in location with respect to the intercommissural plane and the midline depending on the variation in the width of the third ventricle and the obliquity of the internal capsule.62 Based on the latter information, one questions the ability of a deformed digitized atlas to accurately localize the GPi based on linear or even nonlinear algorithmic transformation. In yet another series, the use of ventriculography was also found not to be sufficient for proper targeting of pallidotomy lesions. In a study of 40 patients, differences between ventriculography-based and electrophysiologically based targets were significant, with ranges of 8 mm for laterality, 6.5 mm for depth, and 10 mm for posteriority.71 Proponents of MER reference the above-mentioned literature to emphasize the need for MER regardless of whether the target was planned by preoperative CT, MRI, CT–MRI fusion techniques, CT–MRI-digitized atlas techniques, or ventriculography. Proponents of MER, however, fail to demonstrate that a change in the target of a few millimeters actually translates into a significant change in the clinical outcome. They also fail to demonstrate that such a change could not have been predicted by macrostimulation techniques.72 Several series demonstrate worse cognitive outcome following pallidotomy when lesions extend into the anterior region of the GPi or into the external pallidum, structures that participate in the cognitive basal ganglia–thalamocortical circuits.73 The location of the lesion along an anteromedial-to-posterolateral axis was found to correlate with postoperative cognitive and motor outcomes. Memory was somewhat impaired in anteromedial lesions, improved in posterolateral lesions, and unaffected in lesions that lie in between. Motor function improved most when lesions were created in the center of the posteroventral GPi compared with more anteromedial or posterolateral locations.74 These data demonstrate that the cognitive effects of pallidotomy can be dissociated from the motor effects, depending on the placement of the lesions within the GPi. Advocates for MER stress the ability of the technique to better define the boundaries of the globus pallidus and the sensorimotor circuit of the GPi, thus improving the ability to achieve motor improvement without cognitive deficits. There are, however, no published series that compare postoperative cognitive outcome in patients who underwent pallidotomy with or without the use of MER. Several series relate improved motor outcome after pallidotomy to the use of MER, based on a comparison to a historical group. Several series reported decreased optical complications when MER was used related to the ability to better define the borders of the optic tract.26,28 In one series, tremor was better controlled in patients who had tremor cells recorded during pallidotomy than those who did not.75 Several surgeons who did not use MER reported good results after pallidotomy. Relying on impedance measurements and/or macrostimulation techniques, these investigators reported results that are in general comparable to those reported using MER.18,49,66,76,77 There are at least two series in the literature that performed retrospective meta-analysis of pallidotomy results with or without the use of MER, both reporting similar outcomes for either technique.2,3 The use of MER in pallidotomy remains controversial, but more investigators advocate the use of MER than not.78 Those who advocate MER stress the importance of restricting the lesion to the sensorimotor component of the GPi to reduce cognitive impairment and to maximize long-term benefit. It would be theoretically difficult to achieve that goal otherwise because of the inability of macrostimulation techniques to define the sensorimotor component of the GPi. Macrostimulation can only define the proximity of surrounding structures that should be avoided, such as the internal capsule and the optic tract. One can argue, however, that single-unit recording is not required and that recordings of field potentials may suffice to define kinesthetic cells. Although series that did not utilize MER reported comparable results to those that did, a head-to-head comparison is impossible, given different imaging and surgical techniques as well as different outcome analysis and follow-up periods. The optimal target for pallidal DBS, like that of pallidotomy, seems to be the posteroventral aspect of the GPi. Several investigators suggest that the effect of stimulation varies between lead electrodes that are located at the most ventral aspect of the GPi and lead electrodes that are located more dorsally, with the most inferior electrodes having the best effect on dyskinesia and the more superior electrodes having the best effect on bradykinesia.79 Nevertheless, the technique of targeting for pallidal DBS follows the same principles as that of pallidotomy. Several series reported good clinical results of pallidal DBS surgery with or without the use of MER.80–84 As in the case of pallidotomy, proponents of MER reported a significant discrepancy between the image-based target and the MER-based target. In one such series, more than 40% of 21 patients who underwent pallidal DBS had a mismatch of more than 3 mm between the image-based target and the MER-based target in both the anteroposterior and mediolateral planes, respectively.85 In another series of children who underwent pallidal DBS for dystonia, significant differences in the stereotactic spatial coordinates were found between targets that were determined by direct visualization on MRIs and those that were determined by referencing to a digitized atlas image.14 The same authors, in another series, accurately implanted pallidal DBS electrodes under general anesthesia in 12 children with dystonia, based on a sophisticated preoperative MRI-targeted protocols without the use of MER.10 Proponents of MER emphasize the importance of accurately centering the stimulation volume to the desired pallidal target to achieve the best improvement in parkinsonian motor symptoms and to minimize stimulation-induced side effects, such as postoperative cognitive decline.86 When the stimulation current is sufficient to excite large myelinated fibers near one of the quadripolar electrodes, an additional 1 mA current could activate similar fibers at an additional distance of 1.8 mm with bipolar stimulation and at a distance of 5.7 mm with monopolar stimulation.87 The use of MER may allow the accurate placement of the electrode by defining the borders of the GPi.4 There are, however, no published series that demonstrate that postoperative cognitive outcome is different in patients who underwent DBS with or without MER. The controversy regarding using MER for pallidal DBS probably exceeds that of using MER in pallidotomy because no gross lesioning is performed. Although several series reported a change in the image-guided target by the findings of MER, none of these studies demonstrated that the change in target location could not have been predicted by macrostimulation techniques or whether that change did indeed improve the clinical outcome. Although the subthalamic nucleus is the main target for this procedure, several investigators believe that clinical improvement associated with subthalamic DBS may not stem entirely from stimulation of the nucleus but may, at least in part, stem from stimulation of surrounding structures, such as the zona incerta, Forel’s fields, and the lowermost part of the anterior thalamus.88 It is not clear whether segregated sensorimotor and cognitive circuits exist in the subthalamic nucleus and whether somatotopic organization is clearly outlined. Several investigators claim that targeting the more dorsal and lateral aspects of the subthalamic nucleus yields better clinical results but do not provide clinical outcome studies that supports such a recommendation.89 Other investigators report that postoperative cognitive complications correlate with the location of the active electrode within the subthalamic region, but do not provide detailed analysis.90 Different surgeons use MER differently during subthalamic DBS. Some surgeons use MER simply to document the passage of the electrode through the subthalamic nucleus, whereas others use MER to define the longest axis of the subthalamic nucleus. Some surgeons use MER to identify kinesthetic cells within the subthalamic nucleus, whereas others map the boundaries of the entire subthalamic nucleus.91 Several investigators stress the importance of MER to target the subthalamic nucleus.22 In one series of 14 patients, the location of the clinically most effective electrode contact postoperatively was compared with both the initial target location, as determined by preoperative image fusion technique, and the intraoperative target location, as determined by MER. The initial image fusion–based target did not associate with typical MER neuronal activity of the subthalamic nucleus in more than a third of the cases. This finding emphasizes the lack of imaging techniques that can accurately localize the subthalamic nucleus and its boundaries. In the same series, the most active electrode postoperatively was found to lie 1.7 mm posterior, 1.7 mm inferior, and 12.3 mm lateral to the midcommissure, a location that lies 1 mm more anterior, 2.1 mm more dorsal, and 0.7 mm more lateral than the center of the subthalamic nucleus as defined by MER.89 This finding stresses the importance of mapping the boundaries of the subthalamic nucleus to place the active contact in the most effective location. In another series of 15 patients, the location of the final subthalamic target as determined by MER was compared with the initial target location as determined by preoperative MRI using direct anatomical targeting on coronal images, a digitized scaled computerized stereotactic atlas, or the relative coordinates to the midcommissure. All three preoperative imaging targeting techniques provided target locations that were significantly different from the final MER-defined target. The average distance error was 2.6 mm for targets that were defined by direct anatomic localization, 1.7 mm for targets that were defined by a digitized scaled atlas, and 1.5 mm for targets that were defined relative to the midcommissure.92 Other investigators also stress the importance of MER in targeting the subthalamic nucleus.21,93 Although several investigators support the use of MER for targeting the subthalamic nucleus, others report good results with subthalamic DBS without the use of MER.82 In one such series, only 1 of 17 patients who had bilateral subthalamic DBS required replacement of the implanted lead due to clinical inefficacy, whereas all the other patients achieved 50% reduction of daily off time.94 In another series, subthalamic DBS was performed successfully using macrostimulation techniques supplemented by EMG monitoring of tremor activity, without the use of MER.95 Other investigators reported good clinical outcome for five of seven patients who underwent STN DBS using bipolar recording of focal field potentials via the implanted stimulating electrodes for neurophysiological confirmation of the stimulation site, without the use of MER.96 Most investigators recommend the use of MER in targeting the subthalamic nucleus during subthalamic DBS. This recommendation is based on multiple factors that include the small size of the subthalamic nucleus, the inability to reproducibly identify the borders of the subthalamic nucleus on MRIs, the difficulty of macrostimulation in identifying the surrounding structures of the nucleus, and the occasional absence of intraoperative clinical signs and symptoms that can be tested by macrostimulation (e.g., tremor and rigidity). The delayed response of bradykinesia to subthalamic stimulation makes this symptom impractical to monitor repetitively during surgery. In primates, a three-dimensional map of the nigral complex has been constructed to infer the location of the substantia nigra pars compacta. The maps have been used to guide accurate intranigral placement of fetal dopaminergic cells. Based on the characteristic electrophysiological properties of SNc and surrounding structures in the parkinsonian state, MER may conceivably be used to place transplanted cells accurately in the intranigral region.97 Ablative surgery and deep brain stimulation of the thalamus, globus pallidus, or subthalamic nucleus have also been used to treat conditions other than Parkinson’s disease and essential tremor, including posttraumatic tremor,40 tremor of multiple sclerosis,98,99 dystonia,80,100–102 choreiform disorders,103,104 pain,40,105 and seizure disorders.106,107 Deep brain stimulation of brainstem structures has been used to treat pain108 and persistent vegetative states.104 The role of MER in these conditions is not defined. For the past 7 years, the author has used ablative surgery and deep brain stimulation to treat ~400 patients with various movement disorders, as well as various preoperative planning techniques, including CT, MRI, CT–MRI fusion, scaled digitized brain atlases, and various computer workstation techniques. The author has developed experience with single-unit microrecording techniques, field recording techniques, evoked potentials, impedance measurements, and macrostimulation techniques; has performed surgery with and without the use of MER to target various deep brain structures; and has used MER in various ways that include physiological verification of the target defining the base of the target, defining a minimum acceptable length of trajectory through the target, confirming the passage through the sensorimotor region within the target, identifying the borders of the target, and a combination of the above. The author has not completed detailed outcome analyses of his patients but has developed impressions based on clinical experience. The author’s impressions need to be confirmed by detailed retrospective analysis or by prospective studies before they can be recommended to others. The following is a summary of those impressions. Both thalamotomy and thalamic DBS can be performed safely and effectively without the use of MER. The effect of test lesion or test stimulation on tremor control can be monitored easily during surgery. Surrounding structures, such as the sensory region of the thalamus and the internal capsule, are usually easily identified using macrostimulation techniques. The trajectory that yields tremor arrest using the lowest test stimulation parameters is usually selected for lesioning or for implanting the DBS electrode. A microthalamotomy effect, although it renders interpretation of stimulation-induced tremor arrest during surgery difficult, is by itself a usually good indicator of proper target localization. The author has found MER necessary in some cases. The inferior thalamic border cannot be identified reliably using macrostimulation techniques or consistently using impedance measurements. Situations have arisen when macrostimulation was not reliable in identifying the surrounding structures and yielded conflicting results. Examples of such situations include the presence in some patients of large CSF pockets around the stimulation electrode that produced inconsistent spread of the electric current; slippage of a second-track stimulation electrode into the original trajectory despite changes on the stereotactic frame of the trajectory coordinates and angle, and the inability to test stimulation-induced clinical effects (e.g., sensory, cognitive, speech, and tremor) in patients who are confused or uncooperative during surgery. These situations more consistently arose in patients who had multiple sclerosis or previous trauma. In the latter group of patients, the author has consistently used MER during surgery. The author has developed the practice of localizing the thalamic target using macrostimulation techniques but was prepared to perform MER in cases where the data obtained by macrostimulation were inadequate. A preliminary review of the author’s cases of thalamic DBS with and without the use of MER suggests that patients who underwent MER-guided DBS achieved slightly better long-term tremor control, required less postoperative stimulation current, and needed fewer postoperative programming sessions than patients who underwent thalamic DBS without MER. Further analysis is required. MER is necessary to lesion or implant DBS electrodes in the pallidum for several reasons. First, proper intraoperative macrostimulation of adjacent structures does not necessarily imply proper placement of the electrode in the GPi. For example, the stimulating electrode may be located too anterior in the pallidum but medial enough to properly stimulate the internal capsule. Intraoperative MER helps identify the sensorimotor component of the GPi. Second, macrostimulation of the optic tract may not be reliable. The author has encountered occasions when the stimulating electrode was placed at the superior border of the optic tract, yet the patient did not consistently report stimulation-induced visual effects because the patient was too tired or confused or could not differentiate between visual stimulation and facial periocular motor contractions. Third, clinical monitoring of outcome measures is not identified easily during surgery. For example, a patient may not demonstrate rigidity during surgery, making it difficult to interpret the effect of stimulation. Fourth, improper centering of the target may result in undesirable lesioning or stimulation of adjacent structures outside the sensorimotor region of the pallidum. Intraoperative macrostimulation cannot exclude such a target. Full mapping of the pallidum using MER is not required for a successful pallidotomy or pallidal DBS. It may be necessary, however, to identify all borders of the GPi before depicting the desired target. The author has developed the practice of utilizing MER to confirm trajectory passage through the GPi, define its inferior border, and identify the optic tract. In most cases, this is achieved by using one or two trajectories. A preliminary review of the author’s cases of pallidotomy and pallidal DBS suggests that the best improvement in dyskinesia and bradykinesia occurred in patients who had a trajectory that passed through 5 cm of the GPi and very close to the optic tract. Further analysis is required. It is necessary to use MER in implanting electrodes in the subthalamic nucleus for several reasons. First, macrostimulation cannot adequately identify the location of the electrode in relation to its surrounding structures. For example, the electrode may lie too medial but anterior enough to adequately stimulate the internal capsule. Second, stimulation-induced changes in patients’ symptomatology may not be reliable. The author has encountered several occasions where the patient’s rigidity and tremor subsided before implanting the stimulating electrode. The effect of stimulation on bradykinesia lags in time, making this clinical outcome impractical to monitor consistently during surgery. Third, several adjacent structures can result in intraoperative clinical improvement of parkinsonian symptoms when stimulated. For example, a patient may achieve good stimulation-induced tremor control during surgery if the electrode is implanted in the zona incerta but will not exhibit improved bradykinesia postoperatively. The author has developed the practice of utilizing MER to confirm trajectory passage through the subthalamic nucleus and to identify its superior and inferior borders. This is usually achieved using one or two trajectories. When the patient does not exhibit clinical symptoms or signs that can be monitored easily during surgery, the author maps the borders of the subthalamic nucleus and its kinesthetic cells. A preliminary review of the author’s cases suggests that good results are achieved when 4 cm of the subthalamic nucleus is recorded, accompanied by appropriate stimulation-induced current thresholds for the surrounding structures. Further analysis is required. Good clinical results after thalamotomy, thalamic DBS, pallidotomy, pallidal DBS, and subthalamic DBS have been reported equally by investigators who used MER and investigators who did not. Variations in methodology, targeting, assessment of clinical outcome, and postoperative follow-up make a head-to-head comparison between both groups difficult. Currently, there are no published series that prospectively attempted such a comparison. Even if performed, the results of such a study will apply only to the investigators who performed it. The author believes that MER should be part of every surgeon’s armamentarium and has found it indispensable in localizing the target in a large number of cases.
Correlation of Intraoperative Microelectrode Recording with Clinical Outcome
JAMAL TAHA
Basis of Controversy
Various Surgical Targets
Ablative versus Neurostimulation Surgery
Various Preoperative Techniques for Target Localization
Various Intraoperative Surgical Techniques
Various Macrostimulation Techniques
Various Microrecording Techniques
Various Postoperative Imaging Techniques
Various Methods of Outcome Reporting
Thalamotomy
The Optimal Thalamotomy Target
The Validity of MER in Thalamotomy
Thalamic Deep Brain Stimulation
The Optimal Target for Thalamic DBS
The Validity of MER in Thalamic DBS
Pallidotomy
The Optimal Pallidotomy Target
The Value of MER in Pallidotomy
LESIONS BASED ON IMAGING ARE NOT ACCURATE
LESIONS THAT IIMPINGE ON SURROUNDING STRUCTURES CAUSE COMPLICATIONS
PALLIDOTOMY OUTCOME MAY IMPROVE WITH MER
Pallidal DBS
The Optimal Target for Pallidal DBS
The Value of MER in Pallidal DBS
Subthalamic DBS
The Optimal Subthalamic Nucleus Target
The Value of MER in Subthalamic DBS
Conclusions
Other Targets
Author’s Perspective
Thalamotomy and Thalamic DBS
Pallidotomy and Pallidal DBS
Subthalamic DBS
Conclusion
REFERENCES

Stay updated, free articles. Join our Telegram channel

Full access? Get Clinical Tree

